- v5.0.0 'Live Loop', To be released
- v4.5.1 '8oh8', 26th April, 2024
- v4.5.0 '808', 18th Oct, 2023
- v4.4.0 'Sawtooth', 28th June, 2023
- v4.3.0 'Pathfinder', 30th Sept, 2022
- v4.2.0 'Batch', 14th Sept, 2022
- v4.1.0 'Phase', 30th August 2022
- v4.0.3 'Dovetail', 18th July 2022
- v4.0.2 'Relink', 15th July 2022
- v4.0.1 'Verknüpfen', 12th July 2022
- v4.0.0 'Link', 6th July, 2022
- v3.3.1 'Beamer', 1st Feb, 2021
- v3.3 'Beam', 28th Jan, 2021
- v3.2.2 'Tau3', 5th April, 2020
- v3.2.1 'Take Tau', 3rd April, 2020
- v3.2.0 'Tau', 28th Feb, 2020
- v3.1.0 'Sauna', 23rd Jan, 2018
- v3.0.1 'IOIO', 27th July, 2017
- v3.0 'IO', 18th July, 2017
- v2.11.1 'Hack', 16th Dec, 2016
- v2.11 'Time Warp', 3rd Nov, 2016
- v2.10 'Cowbell', 15th April, 2016
- v2.9 'Venster', 31st Dec, 2015
- v2.8 'Karlsruhe', 20th Nov, 2015
- v2.7 'Rerezzed', 10th Sept, 2015
- v2.6 'Algorave', 30th July, 2015
- v2.5 'Craft', 13th April, 2015
- v2.4 'Defrost', 11th Feb, 2015
- v2.3 'Bitcrush', 28th Jan, 2015
- v2.2 'Slicer', 18th Dec, 2014
- v2.1.1 'Firewall', 25th Nov, 2014
- v2.1 'Core', 21st Nov, 2014
- v2.0.1, 9th Sept, 2014
- v2.0 'Phoenix', 2nd Sept, 2014
This new release of Sonic Pi introduces a breaking change in the interaction between live_loop
s and with_fx
that significantly improve using them together in a live performance. There has also been a significant overhaul of the shortcuts in the GUI to allow you to change them. You can now choose between the default (called Emacs Live), Windows or Mac modes as well as customise them yourself. There's also a wonderful new function for working with tuplets designed by Dago Sondervan.
Before we get to the full breakdown list of all the changes, let's take a moment to explore the breaking change with live_loop
and with_f
:
Consider this code:
with_fx :reverb do
live_audio :moog
end
When you run this code, the audio from the sound card is played through reverb. Now, what if you're in the middle of a performance and wanted to change it to distortion? Easy, you just change the code to this and hit run again:
with_fx :distortion do
live_audio :moog
end
The live_audio magically moves out of the Reverb context and into the Distortion. There's no need to stop and start the code - it all happens live and is completely seemless. As an added bonus, behind the scenes the Reverb figures that the live_audio has moved on and automatically garbage collects itself freeing the resources. The crowd dances to the distorted Moog synth.
Now consider this:
with_fx :reverb do
live_loop :foo do
play 70
sleep 0.5
end
end
As with the live_audio - this works as expected, the sound from the live loop is played through Reverb. However, if you changed the reverb to Distortion and hit run again, it did not change the FX over to the Distortion. The live loop wass stuck in the Reverb FX it was originally created in. Forever. This wasn't ideal and confused a lot of people.
It unfortunately stayed like this for many years because it was an incredibly hard problem to fix due to the architecture of Sonic Pi's internals. It required deep changes to the thread management system, the live loop implementation, and with_fx and its auto FX garbage collection system. Each of these is tricky enough to work on independently - but the intersection of them all is a thing of nightmares.
Well, now more because the work is done and this behaves like live loop. Swap the code to this, hit the run whilst it's still playing and you'll hear the :foo
live_loop
played through distortion:
with_fx :distortion do
live_loop :foo do
play 70
sleep 0.5
end
end
You can even change the FX to one with a control parameter and it will still just work:
with_fx :lpf do |fx|
live_loop :foo do
control fx, cutoff: rrand(70, 120)
synth :square, note: :e3, release: 0.2
sleep 0.25
end
end
When you're working with fast spinning live_loop
s it's now generally better to put your FX outside the live_loop
as it's much more efficient. You just have one FX running the whole time, rather than creating a new FX every time the loop spins round.
- Complete overhaul of the shortcut system.
live_loop
now moves to new FX contexts similar tolive_audio
. You can therefore now usewith_fx
both within and outside oflive_loop
and be able to change things live in a performance.- Where supported, you can now indepedently specify different input and output sound cards in the
audio-settings.toml
config file. This is done with the new optionsinput_sound_card_name = ""
andoutput_sound_card_name = ""
. Note, you still have to ensure that the sample rate is the same for input and output. - Incoming OSC bundles are now supported. Timestamps are ignored (if OSC scheduling is a commonly requested feature this could be added in the future). This increases compatability with software which exlusively sends OSC in budle format such as TouchDesigner.
tuplets
- Run block with tuplet timing and optional swing. Lets you supply a nested list such as[70, 72, [73, 74], 76]
and a block. It will run the block passing in each element of the list and the sleep time between block calls will be automatically calculated depending on the length of the tuplet. Non-nested values e.g.70
,72
and76
will take one beat and nested values73
, and74
will change the time density depending on the length of the nested list. So, a nested[73, 74]
will be with density 2 (all time values are halved) and a nested[73, 74, 75]
will be with a density of 3 which creates triplets. There's also options to add swing and set the duration of one tuplet beat (defaults to 1).
- Two new ride cymbals
:ride_tri
andride_via
. - New hi-hat
:hat_len
.
- Improvements to the French, German, Korean, Polish, Portuguese, Portguguese (Brazil), Russian and Turkish, translations.
This is a maintanence and bugfix release which includes updates to all the major internal dependencies. The main goal of this release is to fix boot issues on Windows with specific locale settings.
- Improvements to the Chinese, Dutch, German, Indonesian, Italian, Japanese and Russian translations.
- Only remove one level of comments when uncommenting
This release comes packed with a number of exciting new synths for you to play with. We have a new hard-hitting kick drum :gabberkick
for that gabber sound and a new vintage electric piano :rhodey
synth. There are also 16 new percussive synths inspired by the infamous TR-808 drum synthesiser. All the synths come with many opts for you to play with to manipulate and change the default timbres.
There are also a number of minor improvements and bug fixes as detailed below.
Have fun and get your live coded 808 on!
- New synth:
:rhodey
- The sound of an electric piano from the 60s and 70s, producing a characteristic metallic sound for notes below:g2
. Adapted for Sonic Pi from SuperCollider Code. - New synth
:gabberkick
- An aggressive Gabber synth sound adapted for Sonic Pi from SuperCollider Code - 16 new 808 inspired synths based on Yoshinosuke Horiuchi's:
:sc808_bassdrum
,:sc808_snare
,:sc808_clap
,:sc808_tomlo
,:sc808_tommid
,:sc808_tomhi
,:sc808_congalo
,:sc808_congamid
,:sc808_congahi
,:sc808_rimshot
,:sc808_claves
,:sc808_maracas
,:sc808_cowbell
,:sc808_closed_hihat
,:sc808_open_hihat
,:sc808_cymbal
,
- Improve error message for invalid
midi_to_hz
args
- Improvements to the Chinese, Danish, Portuguese, and Spanish translations.
- Add native language name for Arabic
- Fixed a regression in the error reporting - "did you mean" suggestions now work again.
- Fix system language detection when the language is generic English
This release comes chock-full of new samples to play with. Firstly we have 5 new samples donated to Sonic Pi by the wonderful electric artist Arovane. These are a selection of glitchy beats recorded at 130 BPM. To best use these either use_bpm 130
or use the beat_stretch
opt to make them fit seamlessly into your current BPM. They're also wonderful to use with the onset:
option for extracting individual sounds to use on their own. Thank-you so much to Arovane for letting us play with these amazing new sounds.
Continuing on the sample theme, there are also 20 new hi-hat samples. These all have a :hat_
prefix and there's a wide variety of sounds to play with ranging from acoustic to electro to metal. They're perfect for adding a sense of drive or groove to your tracks.
This version of Sonic Pi is also the first to support a fully native build for Apple Macs running Apple Silicon CPUs such as the M1 and M2. This means that Sonic Pi can now run on modern Mac computers without needing Rosetta to run things under emulation.
- New looping samples contributed by the wonderful electronic artist Arovane. The new samples are:
:arovane_beat_a
,:arovane_beat_b
,:arovane_beat_c
,:arovane_beat_d
,:arovane_beat_e
. - 20 new hi-hat samples:
:hat_snap
,:hat_zap
,:hat_cats
,:hat_tap
,:hat_bdu
,:hat_psych
,:hat_zild
,:hat_zan
,:hat_zgump
,:hat_noiz
,:hat_sci
,:hat_star
,:hat_gem
,:hat_raw
,:hat_yosh
,:hat_mess
,:hat_cab
,:hat_gnu
,:hat_hier
,:hat_metal
.
- Improve text formatting of the information window that pops up with a boot error.
- Improvements to section 8.2 of the tutorial where the code has been updated to better match the behaviour of
play_pattern_timed
. - Improved
play_pattern_timed
function documentation.
- Improvements to the Catalan, Chinese, Croatian, French, Georgian, Hebrew, Indonesian, Latvian, German, Portuguese, Russian, Slovak, Spanish, Swahili and Turkish translations.
- Syntax and runtime errors are now correctly coloured once again - pink for runtime and blue for syntax.
- Fix
==
operator when comparing maps withnil
. Now returnsfalse
as expected.
This is another maintenance and bugfix release - primarily addressing a critical issue with configuring external sound cards for audio input. This was accidentally introduced in v4.2. An issue with the auto-indentation preference in the GUI has also been addressed.
The GUI is now restricted to having one instance open at any one time. Having multiple instances of Sonic Pi running simultaneously can cause conflicts. This is now no longer possible, which should reduce any issues users may have observed when accidentally opening the app multiple times in quick succession. Now, only the first instance will open and all the subsequent others will exit before starting if they detect another instance already running.
- Restrict running instances to only allow one instance of Sonic Pi to be running at any one time on a given machine.
- Improvements to the Dutch, Estonian, Japanese and Polish translations.
- The behaviour of external audio card configuration should now be as it was in v4.1.
- The auto-indent preference in the GUI now works as expected. Unchecking will stop the GUI from automatically reformatting the code.
14th September 2022
The main purpose of this release is to address a booting issue discovered on a school Windows cluster with common security restrictions. This issue was reported by Adam Cooper from Moor Park High School & Sixth Form in the UK. Adam was wonderfully helpful in finding the specific issue and testing potential solutions. Thanks Adam!
If you happen to have had any trouble running v4 of Sonic Pi on your school clusters, please give this release a go! (Note, this issue didn't affect v3 of Sonic Pi).
In addition, there have been some translation improvements, Global Time Warp now has a greater range and a couple of time_warp
related bugs were found and dealt with. Have fun!
- Change Global Time Warp range from +-250 to -250 -> 999. This lets users access the full phase down to a BPM of 48.
- Booting process on Windows has been modified to enable it to run when executing in an environment which restricts access to CMD (which prevents terminal usage and running .bat files). This is typical in schools.
- Sonic Pi stores all its history, configuration and logs in a folder called
.sonic-pi
which is typically placed in the user's home directory. If the user happens to not have permission to write to their home directory, Sonic Pi now raises a descriptive error on boot and explains that the fix is to set an environment variable calledSONIC_PI_HOME
to point to a directory the user does have permission to write to.
- Improvements to the Dutch, Hungarian and Indonesian translations.
- Teach timing safety system that
time_warp
doesn't sleep (even if it contains calls tosleep
internally). This now means that wrapping all code within alive_loop
withtime_warp
no longer causes the thread to spin out of control. It instead reports that the live loop did not sleep as expected. - Throw error when attempting to call
sync
within atime_warp
. This is because time warps are effectively timeless andsync
asks the running thread to wait for an unknown period (until a matching event appears). Attempting to wait within a timeless context is clearly nonsensical!
30th August 2022
The release of v4 of Sonic Pi introduced the ability to change the tempo of live running code to an external beat. This is possible either by setting the tempo manually to a specific BPM (Beats Per Minute) or by using the Tap Tempo button to tap along to an external rhythm. However, even when you do this, it's quite likely that things still don't sound quite right as there's a good chance that the two rhythms are out of phase despite being at the same tempo. This release introduces a new feature - Global Time Warp which lets you shift the phase of Sonic Pi's timing back or forward up to 250 milliseconds to let you align it with an external rhythm.
Taking a quick step back, what does it mean for two rhythms to be out of phase and why is it a problem? Imagine you had two metronomes ticking away. If one has a higher tempo than the other, then one will be ticking faster than the other. This can be fixed by setting the tempos to be the same. However, unless they were both started at exactly the same time, they both wont be ticking at the same time. Instead, one will tick, then the other, then the first, then the second, etc. The second will always be a little later than the first and as they are at the same tempo, they will never catch up with each other, they will always tick at different times. This is called being out of phase. When two rhythms are out of phase it can sound very messy. Therefore, being able to get rhythms to start at exactly the same time is really important to being able to jam together and for it to sound good.
The new Global Time Warp control lets you shift the timing of all triggered synths, MIDI and OSC messages forwards or backwards by a specified time in milliseconds. You can either slide a slider or type in a specific time to adjust the phase of Sonic Pi's timing to match the phase of an external rhythm.
This means you can now easily jam to an external track - first by tapping out the tempo and then by shifting the phase using the new Global Time Warp controls until it sounds right. This is essentially what DJs do when they mix tracks together in a live set - only now you can layer coded rhythms, basslines and riffs on top of external music with code. This is not only a great way to perform live, but it's also a wonderful way to practice - by jamming along to music you love.
In addition to this, the accuracy of the Tap Tempo functionality has been improved in addition to UI tweaks to all the metronome controls.
Have fun jamming in phase to external music and take your live coding collaborations to the next level!
- New Global Time Warp controls. This lets you adjust the timing of all scheduled events (synths, samples, MIDI and OSC) forwards or backwards up to 250 milliseconds. This is important for matching the phase of an external rhythm.
- Improved Tap Tempo logic aimed towards increased BPM detection accuracy and speed.
- BPM scrubber and new Global Time Warp scrubber now respond to mouse wheel changes.
- Clearly specify in the tutorial that user-defined function names should start with a lower case letter. Therefore use
define :foo
rather thandefine :Foo
.
- Minor improvements to the Indonesian, Korean Portuguese (Brazil) and Russian translations.
- The
duration:
opt formidi_clock_beat
has been fixed. - Fix
scsynth_opts
andscsynth_opts_override
options in the~/.sonic-pi/config/audio-settings.toml
config file. - Audio output via external soundcards on macOS should now work without requiring audio inputs to be enabled in the GUI.
current_debug
now correctly returnstrue
if logging/debug is enabled (the default). Previously it reported the inverse.- Minor improvements to documentation layout.
18th July 2022
This is a minor bugfix and maintenance release.
- The docs for
beat
have been updated to reflect the beat value can no longer be assumed to start at0
. This changed since Ableton Link integration in v4 which now defines the initial beat value based on its internal state.
- Minor improvements to the Vietnamese translation.
- A regression in the function
midi_sysex
introduced in v4 has been resolved.midi_sysex
can now be called as expected.
15th July 2022
Hot on the heals of v4.0.1 here's another bugfix release addressing an issue with the translations.
During the last year there has been significant upheaval of the process used to build Sonic Pi. Somehow, during this upheaval, the translations never got added into the final release builds, despite being part of the dev builds. This has been true since v4.0 BETA 1 back in August 2021 - yet it somehow went unnoticed until now. Apologies! This release fixes this and re-introduces the translations which are so important for lowering the barrier to entry across the world.
As a bonus, a couple of nice improvements also managed to sneak in for good measure. Have fun!
- All the translations are now available again - including the vast amount of new translation effort since v3.3.1 was released.
- Loading a file no longer erases the undo history. This means that you can now "undo" a file load in the current buffer and return it to the state prior to the load.
- The BPM-scrubbing behaviour (clicking your mouse and dragging to change the BPM) has been tweaked to make it easier to use.
- Improvements for the Portuguese translation.
12th July 2022
This is a bugfix release addressing a startup issue for a number of users with non-ASCII characters within audio hardware information on macOS.
We also sadly say goodbye to Core Team member Hanno Zulla who did amazing work with Linux packaging - especially getting Sonic Pi into Debian. Hanno also spearheaded the translation work setting up the crowdsourced translation system which has now enabled Sonic Pi to be translated into over 50 languages. Thank-you so much Hanno - we wish you success and happiness in all you do going forwards.
This release also includes Italian and Polish translation updates.
6th July, 2022
Get ready to Jam with v4 of Sonic Pi.
In this release, the internal timing algorithms have been completely refactored to enable them to link to a new global metronome. The GUI now features a new metronome panel which will let you dynamically change the BPM of your code whilst it is running. You can also use the new "Tap Tempo" button to click out a tempo and Sonic Pi will automatically shift to play in time with your taps. Additionally, you can link the metronome with other versions of Sonic Pi running on other computers connected to the same network (wifi or ethernet). This means it is now possible to gather some laptops together to form an ensemble or turn a whole classroom of computers into an orchestra of live coders - and for everyone to be in time with each other. You can even dynamically change the BPM for everyone on the network whilst code is running on everyone's computers independently and everything maintains synchronised and in time. It's a real jam.
Finally, you're not limited to syncing tempo with other copies of Sonic Pi. The new global metronome is built on top of a wonderful piece of technology called Link which was developed and open sourced by Ableton who also make Live and Push, both widely used by professional musicians and also automatically sync using Link. Additionally there are also over 200 other apps or hardware products that also sync with Link. Sonic Pi can therefore now automatically jam with any of these with just a click of a button. It's so exciting to imagine the new kinds of bands and collaborations that are going to be possible with this new timing system.
There's also a large number of other improvements and fixes within this release that together combine to form the most powerful and exciting release of Sonic Pi yet. See below for a full breakdown of visible changes.
This release would not have been possible without the huge number of fabulous contributors that have been involved. Key contributions have been a huge refactoring of our GUI by Chris Maughan, the wrapping of Link as an Erlang NIF by Luis Lloret and the tireless and constant attention to detail by Ethan Crawford. Thanks so much to you all. Thanks also to José Valim for financially supporting development during a tricky period. José is the creator of the programming language Elixir which is now integrated into the internals of Sonic Pi. Elixir opens up a rich world of exciting opportunities within Sonic Pi which we have already started exploring with earnest. Expect exciting things for the future and consider becoming a Patreon supporter to automatically get access to all upcoming BETAs.....
Finally, we would love for you all to extend a warm welcome to the latest member of the Sonic Pi Core Team: Lily Foster. She brings a huge depth of Linux knowledge to the team and we're terrifically excited to have her on board.
Now, go and get your Live Coded Jam On!
- The default BPM is now set to the new global Link metronome. This is set to 60 by default, so on the surface the behaviour will appear identical. However, if you change the Link metronome BPM using the GUI or the new fn
set_link_bpm!
it will change the BPM for all threads dynamically. Previous behaviour (a static BPM of 60 unlinked to the global metronome) can be obtained by starting your code withuse_bpm 60
. - Previously it was possible that the
onset:
option forsample
silently ignored the last onset of a given sample. This has now been fixed. Some samples may therefore have an additional onset index which won't affect any code using earlier indexes but will affect code which uses indexes larger than the number of onsets (and therefore relying on the index wrapping behaviour). - The Minecraft Pi Edition API has been removed (all
mc_
fns). Minecraft Pi Edition appears to no longer ship on Raspberry Pi OS and the Pi Edition API is not the same as the standard Minecraft API. - The main mixer settings have been modified to reduce load on the limiter in common scenarios with the GUI volume slider set to a high position. This caused distortion and other unwanted audio issues. This change should be an improvement for most users - however if you wish to return to the old behaviour, you can run the following at the start of your session:
set_mixer_control! pre_amp: 1, amp: 1
. - The value returned by
beat
no longer starts at0
. Instead it is now initialised with Ableton Link's current beat value. Other than the initial value not being0
,beat
's behaviour remains unchanged.
- Support for Ableton Link. This enables you to synchronise the tempo of Sonic Pi running on multiple computers connected on the same network. It will also enable automatic BPM synchronisation with music production tools such as Ableton Live, VJ tools such as Resolume, DJ hardware such as the MPC and many compatible iPad music apps. For a full list see: https://www.ableton.com/link/products/,,
- New
:link
option to fnuse_bpm
. This enables Link mode for the current thread which automatically syncs the BPM to the Link metronome (which also syncs it with all other Link-capable apps running on any computer connected to the local (wired or wifi) network. - New fn
link
which sets the BPM to a new:link
mode and also waits until the start of the next bar (as determined by Link) before continuing. This lets you automatically sync tempo and beat phase in one command. - New fn
link_sync
which sets the BPM to a new:link
mode, waits for the Link session to be playing and also waits until the start of the next bar (as determined by Link) before continuing. This lets you automatically "arm" Sonic Pi to sync tempo and beat phase and wait for an external "play" command from another Link device - such as Ableton Live. - New fn
set_link_bpm!
to change the BPM/tempo of the Link metronome (and simultaneously change the tempo of all connected Link-capable apps on the network). - New fn
current_random_source
which returns the current random number source kind (seeuse_random_source
). - New fn
load_synthdef
which lets you load a single synthdef file. load_synthdefs
now loads both directories and single files (by dispatching toload_synthdef
where necessary).
- New synth
:winwood_lead
- a lead synth inspired by the Winwood songs from the early 80s. - New synth
:bass_foundation
- a soft bass synth inspired by the sounds of the 80s. - New synth
:bass_highend
- an addition to the :bass_foundation synth inspired by the sounds of the 80s.
- New Algomancer example - Blockgame coded by DJ_Dave.
- New Sorcerer example - Lorezzed.
- Preference pane is now an overlay which hovers over the main window. This means that opening and closing it does not inadvertently modify a carefully chosen layout e.g. for a performance.
- New preference option to show and hide the pane titles such as Scope, Log, Cues, Context, Help, etc.
- New preference option to hide the menubar when in full-screen mode (Windows and Linux only as this behaviour is standard on macOS).
- New preference option for enabling audio inputs (now disabled by default). Modifying this setting requires a reboot to take effect.
- New preference option for selecting the default language translation. Modifying this setting requires a reboot to take effect.
- Preferences now show audio hardware information such as connected input/out devices, sample rate, block size etc. The amount of information displayed varies by platform.
- New GUI controls for interacting with the new Link Metronome. You can connect/disconnect to the network (to share tempo with others), change the tempo (in BPM) and tap out a new tempo.
- Increase width of panel dividers and highlight on mouse hover.
- Highlight scrollbars and preference checkbox descriptions on mouse hover.
- Scrollbars now have rounded edges.
- Teach autocompletion about random source choices:
:white
,:light_pink
,:pink
, etc. - Improve syntax indentation.
- Improvements for Arabic, Catalan, Chinese (Simplified), Dutch, Estonian, French, German, Italian, Japanese, Korean, Persian, Polish, Portuguese (Brazil), Russian, Sinhala, Spanish, Swedish, Ukrainian.
- Introduced new translations for Basque, Gaelic.
- Many minor documentation fixes and improvements.
- When running on Raspberry Pi, Sonic Pi connects to PulseAudio by default.
- The scheduling accuracy of outgoing OSC and MIDI messages is improved on Windows.
- Optimise
midi_clock_beat
. note_range
can now handle both increasing and decreasing note ranges.- No longer connect to the input sound device on macOS by default - this also means audio input/output rate mismatches no longer cause boot issues on macOS by default.
- SuperCollider audio server boot issues are now immediately detected and reported as a specific issue to the user alongside the full scsynth log file.
- Improve robustness of outgoing OSC messages in the case where an outgoing hostname is malformed and can't be resolved.
- Synths
:dull_bell
and:pretty_bell
now properly free themselves when they finish playing, which now means the resources they consumed are also properly freed. - Indexing into an empty ring no longer causes a divide by zero error.
- No longer attempt to increase audio server priority on Windows which causes booting errors in some cases.
- Fixed encoding issues when saving/loading files containing non-ascii characters on Windows.
range
no longer loops infinitely with a step size of 0. Instead it now throws an error.- In some circumstances having the lissajous visualiser visible caused the GUI to crash on startup. This has now been addressed.
- Stop thread hanging when passing
0
as the first argument tospread
. Now just returns a ring of false values.
1st Feb, 2021 (view commits):
This is a compatibility release addressing two issues with macOS Big Sur:
- Correct language translations now used based on user's current locale
- The language server is no longer killed by macOS Gatekeeper when a sample's onset times are queried.
This release also includes minor translation updates.
28th Jan, 2021 (view commits):
Ten months of development, over 700 individual commits, loving contributions from many many people have gone into this new version of Sonic Pi. This release is also the result of a 3 month long community beta process which took place over on Patreon where supporters got access to early releases and really helped polish and shape development. Huge thanks to everyone that has supported continued development. We're really excited and proud to share this new release with you all.
In terms of significant changes, there have been many internal improvements, including a full rewrite of our MIDI subsystem. This shouldn't result in much change as a user (although some MIDI cue names might have changed slightly) but it really sets us up for some exciting developments in the future. We're essentially moving much more of the core input/output functionality of Sonic Pi into a special system called the BEAM and a language called Erlang. This is a system that has a strong emphasis on low latency, low jitter and high concurrency which is exactly what the IO of a music system requires! We're so excited to see what we'll be able to achieve with this new approach that the BEAM features as the codename of this release.
Another area that has seen a lot of love and attention is our strong focus on accessibility. We now have a much improved menu bar which essentially duplicates much of the functionality found in the preferences but in a format that's extremely easy to work with via a screen reader. We've also added a new context pane that displays the current line and character position as a direct result of a collaboration with Leiden University's accessibility group. There have already been numerous studies that are reporting success for blind and low-vision users of Sonic Pi - so we're confident we're on the right track with respect to accessibility. We're also not stopping here and are always looking for ways to lower the many barriers to entry for creative experiences with code for everyone.
This is also the first release where we officially welcome Chris Maughan as the latest member of the Sonic Pi core team. Chris has made significant contributions over the last year including improvements to the scope and a complete reworking of our build process. I hope you join us in welcoming Chris to the team.
Finally, we have official support for macOS Big Sur, some lovely new synths and FX, improved translations, support for advanced configuration (audio and colour theming), new Turkish scales and many, many bug fixes and improvements.
We really hope you love this version of Sonic Pi as much as we do.
Happy Live Coding!
- MIDI port description names may be slightly different from previous versions. Precise matches on MIDI ports via
sync
andget
will have to be updated accordingly. - Incoming MIDI cues now no longer include clock tick events. Optionally re-enabling these and new ways of working with incoming timing systems will feature in a future release.
- Completely new MIDI subsystem.
use/with_random_source
- change the current random stream used when selecting random values. We now have:white
,:light_pink
,:pink
,:dark_pink
and:perlin
. Default is:white
which is the same stream as previous releases to preserve compatibility.- Many Turkish scales (makams) have been added.
- New user config directory for fine tuning aspects of the app such as advanced audio settings. See
~/.sonic-pi/config/README.md
for more information.
- New synth
:rodeo
- emulating an electric piano. - New synth
:kalimba
- an African thumb piano. - Add sliding to
:autotuner
FX optsnote:
andformant_ratio:
. mix
,pre_mix
andpre_amp
opts are now affected byslide:
.
- New context pane which displays the current line number and position making this information accessible to screen readers. This may be shown/hidden with a preference setting.
- New keyboard shortcuts for switching directly to a buffer. S-M-1 (hold shift and Meta and the number 1) will jump to buffer 1. Meta is Cmd on macOS and alt on Windows/Linux.
- Autocompletion system can now be enabled or disabled. This is intended to help those using screen readers for which the autocompletion system is currently incompatible.
- Sample listings in the help system now have a handy play button that allows you to preview them by clicking rather than having to write any code.
- Improvements for Catalan, Chinese, Danish, Dutch, Finnish, French, German, Hebrew, Hungarian, Icelandic, Italian, Korean, Norwegian, Polish, Portuguese, Portuguese (Brazil), Russian, Slovak, Spanish, Swedish, Thai, Turkish and Vietnamese translations.
- Sinhalese translation added.
- GUI now has a much more complete menu bar which duplicates the functionality of the preferences pane whilst making it easily accessible to screen readers.
- MIDI port names are now autocompleted.
- The GUI now correctly saves and restores the visibility preferences for the buttons, tabs and log pane.
- Many minor improvements to the documentation.
- Unsent external MIDI messages are now flushed and no longer sent after hitting the Stop button.
- The common mistake of calling both
play
andsample
on the same line such asplay sample :loop_amen
is now an error with an appropriately helpful description to explain thatplay
should be removed.
- Fix
pulse_width_curve:
andpulse_width_slide:
opts which were incorrectly swapped onpulse
,subpulse
anddpulse
synths. - Improve boot on Windows systems that have default locale code pages currently unsupported by Ruby. We now force UTF-8.
- Fix issue causing the incorrect error line to be highlighted in the GUI.
- Fix issue when syncing with cues with
-
symbols in their paths. - Fix bug with
use_merged_synth_defaults
after a call touse_synth_defaults
. This bug was triggered in the "Sonic Dreams" example which now plays through correctly again. - MIDI device hotswapping (detecting whether devices were connected or disconnected) now works on macOS.
- Fix
.take_last
to no longer returnnil
for min and max of the range. - Fix off-by-one error in
line
fn. Now should always return rings of the correct length.
5th April, 2020 (view commits):
This is a minor bugfix and maintenance release.
- Significant improvements to the Danish and Swedish translations.
- Fixed a build issue on macOS that stopped recording from working.
- Fixed issues rendering images in the documentation for non-English translations.
3rd April, 2020 (view commits):
This release is another maintenance release with a primary focus on increasing stability and reliability. There are therefore no major new features, but instead there are significant improvements to the internals and the packaging of the Windows installer.
Whilst there are no major new features, we have still been working hard engaging with users with accessibility requirements. We have added a number of modifications to the GUI to improve keyboard navigation which should improve things for users that do not use a mouse (for example, users working with a screen reader). As mentioned in the previous release, we will continue to work to improve accessibility support wherever possible in each release.
One core internal focus has been improving memory utilisation of the app. A number of memory leaks have been hunted down and fixed. A memory leak is where the application requests increasingly more memory from the computer and never gives it back. Over time this means that the app would waste a lot of the system memory. This might have not been too noticeable using Sonic Pi on modern computers that typically have incredibly large amounts of memory. However, it was noticeable and problematic on low-powered systems with limited memory capacity. Sonic Pi now uses a lot less memory and after a warm-up period, the memory usage stays much more constant than before. We will continue to look for and fix memory leaks going forward.
Finally, this release is the first to ship with a fully signed Windows installer. This is a major milestone as we now ship with signed versions for both macOS and Windows which should help more users and organisations (especially schools) trust the installation and running of Sonic Pi.
Now, go and get live coding!
- The
ping_pong
FX now merges the dry sound with the result of the echo to match the behaviour ofecho
. - In the last release, the
piano
synth had its default amplitude accidentally altered when being updated to work with arbitrary frequencies. This has been restored. - Incoming MIDI Active Sensing messages are now ignored
-
New shortcuts and menu items for explicitly placing focus on specific areas of the app. This is primarily intended to improve accessibility for screen readers or those not using a mouse or trackpad. The new shortcuts are:
Control-shift-e
- move focus to editorControl-shift-l
- move focus to logsControl-shift-c
- move focus to cuesControl-shift-p
- move focus to preferencesControl-shift-h
- move focus to help listing (e.g. the list of available help section)Control-shift-d
- move focus to help details (e.g. the contents of the chosen help section)Control-shift-w
- move focus to syntax/runtime error warnings
-
Once focus is in the help listing, the up and down arrow keys now navigate up and down in the list and the left and right arrow keys switch tabs to work through the different help categories.
-
When an error occurs, focus is now placed on the error pop-up window. Focus can be easily returned back to the editor by pressing either the Escape key or
Control-g
. -
Minor improvements for Catalan, Greek, Chinese, Russian, French and Romanian translations.
- Upgraded the Ruby process library which improves/fixes process shutdown behaviour on some Linux systems.
- Fixed a number of memory leaks which resulted in the memory requirements of the app increasing over time until it was closed and opened again. This was particularly noticeable and potentially problematic on machines with small memory footprints.
- General memory usage has been significantly reduced due to the introduction of further caching in hot areas of the code that created many unnecessary object allocations.
- The Windows installer now works on more Windows 10 systems thanks to additional libraries now being included in the installer.
28th Feb, 2020 (view commits):
After a long 2 years without updates, we are threaded with excitement to bring you Sonic Pi's first fully independent release. Sam Aaron's work on this release has been 100% supported by kind and generous donations from people on Patreon in addition to funds raised by delivering talks, workshops and performances. Sam's work has also been hugely assisted by many wonderful code and translation contributions both by members of the Core Team and by our growing international community of live coders. This release has truly been a community effort.
So, what's new? This time there are no radical new features, instead we have a much more polished, accessible and stable version than ever before. (This release has been heavily battle-tested in many performances worldwide).
This release is codenamed Tau - (an improved Pi).
One of the core focusses of this release is accessibility. Tau introduces support for screen readers. Previously the code editor was invisible to the main screen readers on both macOS and Windows and this is now fixed. There is also an accessible menu bar and a new high contrast colour theme which is WCAG 2 level AAA compliant. Accessibility is very important to us - we acknowledge that this is only the beginning and we value and encourage feedback from users who have specific accessibility requirements that we are not directly meeting. We plan to improve our accessibility support where possible in every release going forward.
Another aspect of accessibility is support for Windows devices with high resolution displays. Previously on certain displays font and icon sizes could be far too small, unreadable and out of proportion with the app in general. This has all been addressed with significant work on ensuring Sonic Pi looks great on a variety of monitors on both Windows and macOS. We believe that this release looks better and is more usable than any previous release.
Tau features two beautiful new audio visualiser scopes - a spectrum scope showing the frequency content of the audio and a beautiful mirror stereo scope. These scopes have been contributed by Chris Maughan who has also been pivotal in improving and simplifying the Windows build process. We are hugely grateful for his contributions - thanks Chris!
We have two fabulous new FX - a ping pong delay contributed by Ethan Crawford and an autotuner contributed by Xavier Riley. We hope you have as much fun playing with these as we have.
Finally, this release has seen a significant improvement in translation coverage - both of the GUI and the tutorial. We truly believe that this helps us achieve our mission of lowering the barrier to entry for a creative experience with code by helping more non-English speakers start their live coding journeys.
Now, go and make some noise with code!
spread
now produces identical patterns as in the Toussaint paper. Previously, some of the patterns had been shifted. Use therotate:
opt to match prior behaviour if required.- OSC cues now include the IP address and port number of incoming messages
- MIDI cues now include the source information in the first part of the
path with
:
separators to match the new OSC cue format. - The
osc
fn now forces all outgoing args to either be numbers or strings (binary blobs and timestamps are not supported at this point). If the value is neither a number or string, it is 'inspected' and the resulting description string is sent instead. - The default incoming OSC cues port is now 4560 as the previous value was registered on the Iana Service Name and Transport Protocol Port Number Registry. However, 4560 is currently free.
- Constrain
:flanger
FX'sfeedback:
opt to range 0->1. Values greater than 1 had the effect of increasing the amplitude which should be achieved via theamp:
opt. - Add arg checker for
sample
'shpf:
opt which now ensures the value is not higher than 119 as values higher than this can cause bad noise artefacts. - Improve
sample
'sstart:
/finish:
/onset:
/slice:
opt semantics. Thestart:
andfinish:
opts forsample
now work within the specified slice or onset. This allows you to easily play, say the first half of a given slice or onset by specifying astart:
of0
and afinish:
of0.5
. Where both theslice:
andonset:
opts are specified, then the slice operates within the onset. Where theslice:
,onset:
andstart:
/finish:
opts are all used, the start/finish operate within the slice which in turn operates within the onset. See the updated sample examples for more information.
eval_file
evaluates the contents of the specified file inline in the current thread as if it was a function.midi_sysex
for sending MIDI sysex messages.- Add new chain fn
.notes
which takes a ring of values and returns a new ring with each element treated as a note. This will leave numbers unchanged but will turn note names like:C1
to their corresponding MIDI note.
- New FX
ping_pong
- delayed echo alternating between left and right channels. - New FX
autotuner
- Autotune/phase vocoder effect. - The
gverb
FX now checks to ensure that theroom:
opt is greater than or equal to 1. - The
piano
synth now supports fractional notes such as70.3
- New WCAG 2 level AAA compliant high contrast colour theme. This is especially suited for those with low vision.
- New application menu and shortcuts.
- New default icons which match the pro icon feel.
- 2 new visualiser scopes - mirror stereo and spectrum.
- Move to simple numbered labels for buffers.
- Improve scaling of icons, text etc. on high DPI displays on Windows.
- The version number is no longer placed in the initial comment of new empty buffers. This felt like a friendly thing to do, but can be confusing if a given buffer hasn't been used and the version was updated. This results in the buffer reporting the old version number that was used to create the buffer not the current version used to display it.
- Many, many translation improvements. Thanks to all the wonderful volunteers contributing to the translation effort: https://hosted.weblate.org/projects/sonic-pi/
- Error pane height is now higher by default on all platforms and its height is correctly scaled on high DPI monitors.
- Preference checkboxes are now similarly styled on all platforms.
- Allow scale degrees over an octave and augmented/diminished degrees.
- Calling
.tick
and.look
on a normal array such as[1, 2, 3]
will now automatically convert it to a ring first. - Add docstring for the
set
fn. - Add new article A.20 - Creative Coding in the Classroom as appeared in issue #9 of the Hello World magazine.
- OSC messages are now sent from the external cue listening port. This allows OSC servers that reply directly back to incoming messages (such as SuperCollider) to be able to communicate back to messages sent from Sonic Pi.
- Random fns now understand notes. For example
rrand(:e1, :e2)
will return a random frequency between E octave 1 and 2. - Windows can now detect sample onsets from samples stored in
.flac
and.ogg
format in addition to arbitrary sample rates such as 48khz.
- Further improve boot stability on Mac in the cases where audio input/output sample rates do not match (typically due to the use of bluetooth headsets). Audio inputs on macOS are now disabled by default unless we can definitely determine the audio rates are the same.
- Revert synthdefs to original bytecode version. This fixes a regression
in at least the
:tb303
synth and possibly others. All synths should sound and behave as they did inv3.0.1
. - Fix issue with scaling default opts. Previously it was possible that unspecified opts correctly fell back to the default value - however that default value wasn't scaled. Default values are now always scaled whether or not explicitly specified.
with_sched_ahead_time
now correctly sets the schedule ahead time before running the block and returns the result of the block itself- When "Enforce Timing Guarantees" is selected, Sonic Pi wil no longer
throw out of time warnings in 'real time' threads (specified using
use_real_time
). set
andcue
no longer print duplicate messages which was possible in some cases.- Fix
time_warp
examples which were incorrect and misleading. - Improve pid handling on Linux. This should reduce the number of zombie processes accidentally generated.
- The
offset:
opt forwith_swing
now matches documented behaviour. - Teach
sample_duration
to scale correctly. - Fix GUI bug which could result in a buffer stuck in the highlighted state after hitting Run.
- Fix the
on:
opt when playing chords to work similarly across all notes in a given chord. - Stop
defonce
from being executed multiple times - Calling
.take
on an empty ring no longer results in infinite recursion but instead returns an empty ring. - Scope now has a minimum height.
- Windows now supports sending and receiving OSC messages over the network to and from external machines.
23rd Jan, 2018 (view commits):
The major feature of this release is that it brings v3 functionality to
Windows. Windows now supports all the good stuff listed alongside
v3.0.x
including live_audio
, midi
, osc
and much more. See
v3.0.1
and v3.0.0
release notes below for more information.
Of course, no new release comes without new features, and there are a
number of additions and improvements on all levels. Firstly, we are
honoured to include a number of wonderful new samples kindly recorded
and released into the public domain by our friends Mehackit. If you
aren't already aware, Mehackit organise and run creative coding
workshops all around Finland and Europe and were a core part behind the
CodeBus Africa project which used Sonic Pi to engage almost 2000 African
students with creative code. Our favourite of these samples is the new
:ambi_sauna
which has given us the name for this release.
We have added support for Ogg Vorbis encoded audio files .ogg
and
.oga
. These are lossy encoders similar to .MP3
but free from license
restrictions. See: https://en.wikipedia.org/wiki/Vorbis
Another notable addition is the increase in translation coverage across a number of languages. If you still feel the translation for your language could be improved, please do consider joining in the crowd-sourced effort here: https://hosted.weblate.org/projects/sonic-pi/
We have also included some styling tweaks to the GUI - to make it look even cleaner and also work better when projecting code with visuals underneath.
Finally, this release also gives me the opportunity to welcome Ethan Crawford to the Core Team. He has been a prolific contributor to the project over the years and his keen eye for detail has helped polish Sonic Pi from the rough stone it once was to the shining gem it is today.
Have fun and enjoy the new features - especially you lovely Windows users!
- The synth
chip_noise
now has standard default envelope opts (sustain:
andrelease:
are now 0 and 1 respectively (as opposed to 1 and 0))
set_audio_latency!
Apply positive or negative timing offset for audio events to allow for any timing differences between OSC/MIDI events caused by external latencies (such as wireless speakers).midi_pc
for sending MIDI Program Change messages
- Add support for Ogg Vorbis encoded audio files
.ogg
and.oga
. - Add new samples (kindly recorded and released into the public domain by Mehackit).
- new
glitch_
sample group containing sounds with a glitchy texture. - new
mehackit_
sample group with sounds of old toys - new samples:
:perc_bell2
,:perc_door
,:perc_impact1
,:perc_impact2
,:perc_swoosh
,:ambi_sauna
,:bd_mehackit
,:sn_generic
,:loop_3d_printer
,:loop_drone_g_97
,:loop_electric
,:loop_mehackit1
,:loop_mehackit2
,:loop_perc1
,:loop_perc2
,:loop_weirdo
,
- new
- Add new in_thread forum to the community listings (and update other entries).
- Many, many translation additions and improvements for a multitude of languages. The tutorial is now available in German, Spanish, French, Italian, Japanese, Dutch, Polish, Portuguese, Romanian, and Russian.
- GUI now lets the user know when the buffer capacity has been exceeded on macOS and Linux (Windows behaviour is currently unchanged). This capacity limitation will be addressed more thoroughly in a future release.
- Stop the internal sample pattern matcher from duplicating matches in some cases.
- Stop pro icon preference from resetting to 'off' when switching from dark to light mode
28th July, 2017 (view commits):
This is a minor release addressing a few bugs and includes a small number of modifications.
- Raspberry Pi only - based on a request from Raspberry Pi we have removed the ability to switch between HDMI and headphone output as this is now duplicate functionality to what is now available in PIXEL. Now, to change audio output, you need to right-click on the audio logo in the menu at the top right of the screen.
- The IO menu has been slightly tweaked to improve amount of space used.
- The Time State system no longer maintains a history of all events - instead it retains at least 20 events for each path and beyond that culls events 30s in the past.
- Fix minor issues with docs (grammar and rendering).
- Incoming OSC now correctly handles OSC paths which contain capital letters.
18th July, 2017 (view commits):
This release is our most ambitious to date. The goal is to open up the code within Sonic Pi to the outside world. We want you to be able to both manipulate the real world to change the code and change the real world with the code. Input Output. IO.
We have therefore focussed on getting events and audio in and out of Sonic Pi in new ways whilst keeping to our philosophy of simple code, live manipulation, and strict, powerful timing. This release introduces a number of brand-new components:
- Time State - a powerful new time-based deterministic shared memory system
- MIDI - support for input and output of MIDI
- OSC - support for sending and receiving OSC messages on the network.
- Live Audio - for getting multiple streams of audio into Sonic Pi
- Multi-channel audio out - for outputting multiple streams of audio.
- Audio Buffers - for internal recording of audio enabling the creation of loopers.
It's important to mention that MIDI and OSC output functionality is possible due to a new well-timed scheduler. This was built in a remarkable programming language called Erlang which is now part of the Sonic Pi distribution. We have had the great honour of one of Erlang's co-creators, Joe Armstrong working directly with us on the implementation of this scheduler and we look forward immensely to continuing to work with him on new functionality.
Another very important part of this release is the addition of Robin Newman to the core team. We have always been consistently impressed with his fearless and experimental attitude - pushing Sonic Pi into new areas with ease. We're excited that he's decided to join our core team and hope you're as excited by his future work with us as we are.
Finally, we're really very excited about what new kinds of instruments people will be able to create with this new technology both in the classroom and on stage at musical festivals. Sonic Pi has now become a fully programmable music studio. Have fun live coding!
- Ring's
.pick
now returns 1 element by default. Previously calling.pick
on a ring would pick n elements randomly from the ring (including duplicate picks) where n would be the size of the ring. With this change,.pick
only returns a single element. This makes it similar to choose.
-
midi_*
- many new MIDI-specific fns such asmidi_note_on
,midi_pitch_bend
,midi_cc
,midi_clock_tick
. See new tutorial section for more information. These fns send MIDI messages to connected MIDI devices. Incoming MIDI is received via the new event log. -
with_swing
- add swing to successive calls to do/end block. -
get
- get a named value from the Time State at the current time. This will return the last value entered. Previous values can be read when within atime_warp
. Has full support for OSC -
set
- set a named value in the Time State at the current time. Future values can be set when within atime_warp
. -
use_real_time
- convenience fn for setting the schedule ahead time to 0 for the current thread. Very useful for removing latency from live loops that are working with external cues (such as MIDI or OSC). -
use_midi_defaults
- set defaults to be used for all subsequent MIDI calls. Similar touse_synth_defaults
. Also available:with_midi_defaults
. -
use_osc
- set the default hostname and port number for subsequent outgoing OSC messages. See alsowith_osc
. -
osc
- send Open Sound Control messages in time with the music to default hostname and port -
osc_send
- similar toosc
but requires you to specify the hostname and port -
use_sched_ahead_time
- set the schedule ahead time specifically for the current thread. Also available -with_sched_ahead_time
. -
current_time
- return the current logical time. -
assert_error
- An assertion to ensure the specified block of code raises the specified error.
-
New synth
live_audio
- directly stream audio from your sound-card as a synth. -
New FX
record
- enables you to internally record any audio into named buffers. Perfect for building looper systems. -
New FX
:sound_out
- stream out audio to a specific output on your sound card. This enables multi-channel audio out. -
New FX
:sound_out_stereo
- similar tosound_out
but streams out to a pair of consecutive audio card (left and right) output channels. -
New FX
eq
- Parametric EQ with three centre freqs - low, mid & high - all with Q values and gain (-1 -> 1). Also has low and high shelves with centre freqs and slope adjustment. -
New FX
tremolo
- simple tremolo effect which modulates the volume within thedo/end
block.
- New 'pro' icon set for performances.
- New GUI translations for the following languages: (BS) Bosnian, (CA) Catalan, (CS) Czech, (DA) Danish, (EL) Greek, (ET) Estonian, (HI) Hindi, (ID) Indonesian, (KO) Korean, (PT) Portuguese, (TR) Turkish, (ZH) Chinese
- Added new pane for displaying new cue events (including incoming OSC and MIDI)
- Added new IO preferences tab for configuring MIDI and network settings.
- Automatically autocomplete
sync
,cue
andget
orset
- Increase width of autocompletion popup.
- New articles on additive and subtractive synthesis techniques.
- New example piece 'Cloud Beat' by Pit Noack
- Teach
time_warp
about input ranges. It now works similar toat
in that it can now take two lists of args - times and values - which represent a list of time destinations to be visited in turn. - Ensure any unprinted messages are displayed if an exception occurs.
- Teach
range
to work as expected with both floats and integers. - Teach rings a new chain method -
.scale
which will return a new ring with all elements multiplied by the scale value. - The fn
control
now returns the node you're controlling. - Add many new chords
- Fix randomisation aspects of
:slicer
,:wobble
and:panslicer
FX (i.e. via theprobability:
opt). - Fix file path drag and drop on Windows to not accidentally prefix path with /.
- Teach
chord_invert
andsample
to work with floating point args.
16th Dec, 2016 (view commits)
This is primarily a maintenance release containing a number of bug
fixes and minor tweaks. The feature set remains unchanged from v2.11
.
One of the core bugs fixed in this release is an issue with the return key within some Japanese layouts. Another important change is the unification of the font on all platforms to Hack. This is from the same family as Menlo and Bitstream Vera (the previous fonts for Mac and Linux respectively) and a dramatic improvement to the Courier font previously used for Windows.
- Move all platforms to the same font - Hack.
- Improve initial log information and messages.
- The scope axes are no longer shown by default for a cleaner look and feel.
- Add new MagPi article on sample stretching.
- Ensure
chord_invert
's shift value is a whole number - otherwise it is possible to get into an infinite recursion. - Ensure thread locals have correct default values on
clear
. When a run is created, the thread is given a set of default thread locals. Previouslyclear
removed all thread locals. Now, we reset the thread locals to the defaults for a brand new thread. - Fix line reported on syntax error.
- Fix unknown synth error message.
sync
can once again handle multiple cues correctly.- Fix boot issues for users that don't have a standard /etc/hosts file that contains an entry for localhost.
sample_duration
now handles onsets, slices and sustain.- Fix issue with return key not being recognised with Japanese keyboards.
- Fix sporadic flickering of current line on Raspberry Pi.
- Fix errors in buffer 0 being reported as being from buffer 3.
- Fix scrollbar background colour on Windows in dark mode.
- Improve error message reported when required ports are not available at boot.
- Fix issue with calling
control
on a chord group. - Fix
rand
andrand_i
to honour their arguments.
3rd Nov, 2016 (view commits)
This release is the biggest and most adventurous release yet. There are as many (invisible) modifications and improvements to the internal systems as there are new external features that you can see and play with. The aim was to create a solid foundation for new and exciting features both in this release and in preparation for future releases.
We also open our arms to welcome two new Core Team members - Luis Lloret and Adrian Cheater. Both have made generous and substantial contributions to this release. Thank-you. Sadly we also say farewell to Jeremy Weatherford. Please extend your kind thoughts and gratitude to Jeremy for all of his contributions - in particular for turning the Windows release from a possibility into a reality. Luckily Luis has kindly stepped in to maintain the Windows installer.
The main visible feature is the new scope visualisers. The overall audio output can now be visually monitored by one of three wave form visualisers. Firstly there is the individual left and right channels, next is a single mono scope which is mixed down from the stereo channels using RMS and finally there is a Lissajous scope which displays phase differences between the left and right channels. Typically the mono output will be most useful. Use the preferences pane to hide and show each of these scopes. All of them may be viewed at the same time if necessary. Thanks to Adrian Cheater for the core work behind this feature.
We now have support for multi-channel input (up to 16 channels) via the
new sound_in*
synths for systems that have audio in. This opens up the
possibility to use Sonic Pi as an FX unit for vocals, guitars and any
other audio source.
Another exciting new feature is the sample opt onset:
- which lets you
play a specific percussive part of a sample. This uses an automatic
onset detection algorithm to determine all the points in the sample that
go from quiet to loud quickly such as a drum, synth or bass hit. For
example, this allows you to take a complex drum sample and trigger each
of the individual drums in your own order and to your own timing.
Finally, translations are now crowd-sourced and small or large
contributions for any language can be made here:
https://hosted.weblate.org/projects/sonic-pi/
. If your language isn't
yet available or you would like to improve things, please join in the
effort. Thanks to Hanno Zulla for making this possible.
sample
now supports the optpath:
which enables the sample's path to be overridden.use_sample_pack
is now deprecated and no longer available. Consider using the new filter system. See documentation forsample
for more details.current_sample_pack
is now deprecated and no longer available.inspect
has been removed. (Standard printing now callsObject#inspect
by default)load_sample
now only loads the first matched sample.load_samples
now loads all matched samples.- Remove SuperCollider server automatic reboot system as it was badly
conflicting with machines that went into a 'sleep state' (for example,
when a laptop is closed). The fn
reboot
is still supported and may still be triggered manually if required. - Calls to
play
,synth
andsample
now consume all their arguments before testing to see if the synth should be triggered. This ensures all declared rands are consumed. This change might therefore potentially modify your random stream consumption. Consider usingrand_back
orrand_skip
to re-align the stream if necessary. - New threads now start with a fresh set of tick counters and a new random stream.
- It is no longer possible to use lambdas as values for synth defaults. This is because synth defaults are shared across thread boundaries and there is now a new safety system that only allows immutable/serialisable values to be used. Unfortunately Ruby has no notion of a 'pure' function and each lambda captures over its environment and therefore may contain free variables pointing to mutable data. A replacement system for describing a simple set of pure functions is being designed.
reset
- resets the user's thread locals (such as ticks and rand stream index) to the snapshot of the state as recorded at the start of the current thread.clear
- clears all the user's thread locals to a blank state.time_warp
- allows whole blocks of code to be shifted forward or backwards in time up to the value ofcurrent_sched_ahead_time
.rand_look
- generate a random number without consuming a rand by looking ahead in the random stream.rand_i_look
- generate a random integer without consuming a rand by looking ahead in the random stream.run_file
- Runs the contents of file at path as if it was in the current buffer.run_code
- Runs the contents of the specified string as if it was in the current buffer.Numeric#clamp
- max and minimum bound (will clamp self to a value <= other and >= -1*otherset_recording_bit_depth!
- set the bit depth for WAV files generated by recording the audio. Default is 16 bits, and can now be set to one of 8, 16, 24 and 32. Larger bit depths will result in better quality audio files but also much larger file sizes for the same duration.scsynth_info
- obtain information about the running audio synthesis server SuperCollider such as the number of available busses and buffers.
- New synth
:tech_saws
- an implementation of a slightly modified supersaw. - New synth
:sound_in
- reads audio from the sound card. - New synth
:sound_in_stereo
- reads audio from the sound card. - All FX now have a
pre_mix:
opt. This allows the audio flow to completely bypass a given FX (unlikemix:
which passes the audio through the FX but modifies the amplitude afterwards). - Teach
control
to manipulate the last triggered synth by default. For example,control amp: 3
will set theamp:
opt of the last triggered synth to 3. However,control foo, amp: 3
will still specifically control synthfoo
.
- New opt
slice:
- lets you play a specific slice of a sample. The default number of slices is 16 which may be changed with thenum_slices:
opt. Sample is divided equally into the number of slices without regard for audio content and onset points. Theslice:
opt also works withpick
for triggering random sample slices:sample :loop_amen, slice: pick
. - New opt
onset:
- lets you play a specific percussive part of a sample. Uses automatic onset detection to determine the points in the sample that go from quiet to loud quickly. Unlikeslice:
,onset:
does not necessarily divide a sample into equal onsets - some onsets will be smaller or bigger than others and the number of onsets is determined by the algorithm and isn't known in advance.
- New scope visualisers.
- Allow files to be dragged from the OS into the text area. This inserts the file/folder name as a string.
- GUI now remembers the last directory you saved or opened a file to/from as the default location for next time.
- Swap align button for a scope button. Given that alignment now happens automatically, a specific button seems somewhat redundant. Instead we now have a button for toggling the visibility of the scope(s).
- Loading multiple samples simultaneously is now much faster.
- Preferences have been slightly re-organised.
- Preferences now has a Master volume slider which controls Sonic Pi's audio amplitude independently from the system volume.
- All buttons now display status message + shortcut where available.
- Enable app transparency slider for Windows.
- Dark and light theme colours have been slightly polished and unified to use the same logic.
- On multi-screen systems, full-screen mode now defaults to the app's current screen.
- Translations are now crowd-sourced. See:
https://hosted.weblate.org/projects/sonic-pi/
- Improve docstring for
live_loop
. - Add 3 new MagPi articles on amplitude modulation, performance and practice techniques.
- Add missing
pulse_width:
opt to flanger FX doc.
- Improve log messages written to
~/.sonic-pi/log
- Improve booting on Mac in the case that the audio card's rate can't be determined.
- Massively improve boot stability on Windows.
- Improve error message for
play_chord
when notes isn't list like. - The number of samples that may be loaded at any one time has been increased from 1000 to 4000. However, memory limitations still apply (4000 1MB samples will require 4000MB of free system memory)
pick
now returns a lambda if no list is given as the first argument (which makes it useful for using with sample'sonset:
andslice:
opts.- Audio server is now paused when app is not in use - reducing CPU load and battery consumption.
- Error messages now report names matching the editor tabs such as
buffer 0
.
- Decrease duration of
:loop_tabla
so that it correctly loops. (Length reduced to 10.674 seconds). - Enforce UTF-8 encoding of all incoming text.
- Fix
:reverb
FX'smix:
opt to ensure it's in the range 0 to 1. sample nil
now no longer plays a sample - it was incorrectly defaulting to the first built-in sample (:ambi_choir
)pick
'sskip:
opt now works as expected:pick(5).drop(1) == pick(5, skip: 1)
sample
now prints a 'no sample found' message with bothsample nil
andsample []
rather than incorrectly playing the first built-in wav file.- Limit
:piano
synth to notes less than 231 as higher values crash the audio server.
Friday 15th April, 2016 (view commits)
"I gotta have more cowbell!" - The Bruce Dickinson
The main feature in this release is the completely new sample system for
working with large folders of samples. You can now easily index into and
filter over whole folders of samples with a simple yet powerful filter
system. This system has been in development for over a year now and has
been tuned and tweaked over many performances and practice sessions. See
the documentation and examples for sample
as well as the new tutorial
section 3.7 for details.
We also have a number of exciting new sounds to play with including some beautifully sounding chiptune synths, fun retro FX and new drum samples including a full tabla set and a cowbell.
Finally, even more boot issues on both OS X and Windows have been fixed, making this the most polished and stable release to date.
Now go and get your live code on!
use_sample_pack_as
andwith_sample_pack_as
have been removed.- A synth opt value of
nil
now no longer resolves to 0.0. Instead it masks any defaults and ensures the downstream default (for the actual synthdef itself) is honoured. This allows you to override any existing synth defaults specified viause_synth_defaults
for a given synth trigger. - Default schedule ahead time is now 0.5s on all platforms for consistency (except for Raspberry Pi 1 which remains at 1s).
current_random_seed
- Get the current seed value of the random generator.set_cent_tuning!
- global tuning.on
- similar toif
but behaves the same as theon:
opt for synths and samples.halves
- create a ring of successive halves.doubles
- create a ring of successive doubles.pick
- similar to shuffle but allows duplicates. You may also specify how many items to pick.fx_names
- create a ring of all available FX.
- New synth
:dtri
- detuned triangle waves. - New synth
:pluck
- a lovely synthesised plucked string. - New synth
:chiplead
- retro NES style lead synth. - New synth
:chipbass
- retro NES style bass synth. - New synth
:chipnoise
- retro NES style noise synth. - New FX
:whammy
- low-tech transposition effect similar to the Digitech Whammy guitar pedal. - New FX
:octaver
- low-tech octave effect similar to early guitar pedals. - New FX
:vowel
- modifies incoming signal to match human vowel sounds. - New FX
:mono
- mono effect for summing stereo channels. :tanh
FX is now more crunchy by default.:compressor
and:krush
FX now correctly honour themix:
opt.
- Samples in FLAC format (Free Lossless Audio Codec) are now supported!
- The
sample
fn gained source and filter pre-opts. You may now specify a number of parameters directly before the opts which describe both folders containing samples and filters over those folders to allow you to easily and quickly work with large sample sets. See Section 3.7 of the tutorial for more information. - Samplers now have
hpf:
andlpf:
opts. Anycutoff:
opts are automatically switched tolpf:
and any errors will be reported with anlpf:
prefix. - The sampler synth gained a compressor enabled via the
compress:
opt. This means you can now compress any triggered sample directly without the need to reach for an FX. - Samplers gained the
pre_amp:
opt which applies the amp at the beginning of its internal signal chain. You can use this to overload the new compressor. - Samplers now have both high pass and low pass filter envelopes which
behave similarly to the amplitude envelope but control internal hpf
and lpf FX. These are available via new opts which mirror the standard
envelope opts but with a
hpf_
andlpf_
prefix (such ashpf_attack
,lpf_decay_level
and friends). - Passing a directory path to
load_samples
will now load all the samples within that directory. - Passing a directory path to
free_samples
will now free all the loaded samples within that directory. - Samples are now loaded asynchronously in a separate thread. This therefore doesn't block the current thread and cause any subsequent synth/sample triggers to be late.
- Sample trigger logging now includes the directory of the contained sample.
- Samples are now reloaded asynchronously after a server reboot (and therefore no longer block execution).
- Add new
:tabla_
sample group with a range of tabla drum sounds. - Add new
:vinyl_
sample group with a range of vinyl scratches and hisses. - Add new samples:
:drum_cowbell
,:drum_roll
,:misc_crow
,:misc_cineboom
,:perc_swash
,:perc_till
,:loop_safari
,:loop_tabla
.
- Add new preference to enforce strict synth/FX timing. When enabled synths and samples no longer trigger if it is too late - instead they are silent and print a warning message. This behaviour ensures samples or synths are never triggered out of time.
- New load button which will load the contents of a file into the current buffer.
- The vertical bars which help visualise nesting now render in a contiguous fashion over blank lines.
C-k
now nukes over trailing whitespace.load_sample
now has sample autocompletion.- GUI now correctly reports if the host is a Raspberry Pi 3.
- New editor preference - Log Auto Scroll. When enabled will always scroll log to the bottom after printing every new message.
- Whitespace at top and bottom of buffer is no longer trimmed.
- Hitting
RETURN
now auto-indents the buffer - ensuring that the cursor is moved to the correct indentation on the following line. - Added Chinese Simplified GUI translation.
- Log visibility now correctly matches GUI preference.
- New tutorial section 3.7 on Sample Pack Filters.
- New appendix sections.
- Examples for
sample
fn have been completely rewritten and extended.
scale
andchord
can now handle being passed no tonic such as:(chord :minor)
which will return a ring of offsets from 0.chord
learned new variants:add2
,add4
,add9
,add11
,add13
,madd2
,madd4
,madd9
,madd11
,madd13
- The ring's
.take
method can now take more elements than the original ring by wrapping around:(ring 1, 2, 3).take(5) #=> (ring 1, 2, 3, 1, 2)
- Rings may now be added or subtracted from each other e.g.
(ring 1, 2, 3) + (ring 4) #=> (ring 1, 2, 3, 4)
- Adding or subtracting a number from a ring will create a new ring with
the number added or subtracted from each of the original ring's
elements:
(ring 1, 2, 3) - 1 #=> (ring 0.0, 1.0, 2.0)
- Calling
(ring 1, 2, 3).take(0)
will now return an empty ring. density
now complains if it isn't passed a do/end block.- Improve error messages for
use/with_synth
when accidentally passed opts.
- On OS X only raise an error on boot if it's clear the sound card's in and out rates are different.
- Improve robustness of boot process on Windows.
- Rest notes are no longer printed if synth logging is disabled.
- No longer apply synth defaults to FX.
- You may now control opts that have no associated info (previously it would raise a 'not modulatable' error).
- Fix index lookup of Vectors.
- Fix
C-i
doc lookup shortcut to work with:tb303
synth docs. C-i
now always displays docs where available (previously it was possible for docs not to be displayed).- Sliding between chords now works correctly
- Windows version will now boot on multiple networked machines logged in with the same account.
Thursday 31st December, 2015 (view commits)
Hot on the heels of the previous release comes v2.9
codenamed
Venster
(Dutch for window). This release has a specific focus on
fixing all the known issues with Sonic Pi running on Windows. If you've
tried Sonic Pi on Windows before and had issues, make sure to try it
again with v2.9
. If you're still having issues on Windows please do
let us know so we can fix it as soon as possible.
For all you Raspberry Pi and Mac OS X users - you're not left out. Sonic
Pi is just as stable as it was before (if not more stable) and v2.9
comes with a surprising amount of tweaks and features for its short 1
month development cycle. There's two new FX to play with: tanh
and
gverb
as well as a heavily revamped logging system which is much
clearer about printing which opts are actually being used for your
synths and samples. Finally, we now include of all the published MagPi
magazine articles in the tutorial. We also now have Hungarian and
Norwegian translations of the GUI.
Happy Live Coding!
- Rename fn
invert_chord
tochord_invert
- Sampler no longer mixes
rate:
andpitch_stretch:
opts. It's now possible to set therate:
independent of thepitch:
modification caused bypitch_stretch
.
block_duration
- returns the duration of the given block in seconds (not scaled with the BPM).block_slept?
- returns true if the contents of the block slept or synced.synth_names
- returns a list of all the available synthsreset_mixer!
- resets the main mixer back to its default values.sample_free
- unload a sample from memory to free up memory usage.sample_free_all
- unload all samples from memory.use_octave
- similar touse_transpose
but for whole octaves.with_octave
- similar towith_transpose
but for whole octaves.use_merged_sample_defaults
- similar touse_merged_synth_defaults
but for sampleswith_merged_sample_defaults
- similar towith_merged_synth_defaults
but for samplesuse_cent_tuning
- uniformly tune all subsequent notes in centswith_cent_tuning
- uniformly tune all notes in cents within the block
- New FX
tanh
- for more distortion goodness. - New FX
gverb
- a different reverb FX to play with.
- Synths and samples now also log defaults set by
use_synth_defaults
and friends. - Opts are logged in the order they were defined with local opts first followed by inherited opts.
- BPM scaling is now accounted for in logging values.
- Log metadata such as run number and time is now printed in a more code-like way: as a hash of key value pairs.
C-k
will now kill multiple lines if lines end with,
.- When saving a buffer a
.txt
extension is automatically added to the file name if none specified. - Add Hungarian and Norwegian translations of GUI.
- Add Spanish translation of tutorial.
- Add title to main window - enables Sonic Pi to be selected in screen-sharing app lists such as Google Hangouts and OBS.
- Add autocompletion for tuning systems.
- Add 8 complete MagPi magazine articles to the tutorial in appendix A.
- Add new example 'ambient experiment' by Darin Wilson.
- Add new example 'chord inversions' by Adrian Cheater.
- Change tutorial license to CC-BY-SA 4.0.
- Add instructions for compiling and building on Windows.
- Many, many minor tweaks and improvements.
- Add
sync:
opt tolive_loop
andin_thread
. This now syncs the live loop once on creation. Similar to thedelay:
opt. If bothsync:
anddelay:
are specified then the delay happens before the sync. - The
synth
fn can now play chords with thenotes:
opt such as:synth :dsaw, notes: (chord :e3, :minor)
. This will return a single controllable synth node representing all the notes similar toplay_chord
. - BPM scaling and other normalisation is now honoured when controlling nodes
- The
on:
opt is now honoured when playing chords. - Samplers sound signal now bypasses filter envelope when not used.
- It is now possible to use externally defined FX synths by passing a
string as the FX name:
with_fx "my_shiny_effect"
. This needs to be loaded manually viaload_synthdefs
. - OS X now supports rates other than 44.1k provided they are similar for both audio in and out.
- Run code in
~/.sonic-pi/init.rb
on launch if it exists. - If environment variable
SONIC_PI_HOME
is set it will be used over~
as the root of the user's.sonic-pi
directory (used to auto-save files and store logs). - Default sound card selection is now honoured on Raspberry Pi - so you may now use your IQaudIO hat (or similar) out of the box.
- Fix number of synth validation errors.
- Fix sporadically failing boot issues on Windows
- Add auto-reboot functionality for audio server. This now detects errors in the server (such as a killed or zombied process) and automatically reboots it to enable Sonic Pi to continue without a full restart. Reboots automatically reload all loaded samples and synthdefs.
sample_duration
now correctly takes account of TL defaults set byuse_sample_defaults
.- Sampler opts
pitch_stretch
,beat_stretch
andrpitch
can now be used in TL defaults.
Friday 20th November, 2015 (view commits)
This release, named after Karlsruhe, one of the home cities of live coding, is mainly a maintenance release with a strong focus on both performance, stability and documentation. This is therefore the fastest and most stable release of Sonic Pi ever with a massive 10% performance improvement on the original Raspberry Pi 1. It also ships with new translations in Polish, Japanese and French. Many of these improvements (such as the complete rewrite of the OSC stack) are not documented in this release list but can instead be found in the commit logs over on Github.
However, not to go unnoticed are a number of exciting new features. For
example we now have a new Band EQ FX, the ability to use MIDI note names
such as :e5
as values for opts such as cutoff:
, and new powerful
cutoff envelopes on the sampler.
- Shortcuts for switching buffers have changed. They are now
M-{
andM-}
for switching left and right respectively. sync
no longer inherits BPM by default. Set thebpm_sync:
opt to true to enable BPM inheritance.- Random seed generation for new threads is now reset on calls to
use_random_seed
.
octs
- returns a ring of successive octaves.assert
- for raising an exception if the argument is not true.assert_equal
- for raising an exception if the two arguments are not equal.bt
- converts the specified time w.r.t. the current BPM.inspect
- similar toprint
but prints the inspected version of the argument.
- New translations for Polish, Japanese and French.
- Improve efficiency of logging panel.
M-RET
is now a duplicate shortcut for running the code.- Log title bar is now hidden in full-screen mode.
- Log - don't display └ or ├ if the line is blank, instead display │
- Add sample name autocompletion to more fns such as
sample_duration
.
- New tutorial section on ring chains (chainable functions for modifying rings)
- Tilburg 2 example slightly tweaked for Raspberry Pi 1 compatibility.
- Many minor tweaks and improvements in all areas.
- New FX - Band EQ for attenuating or boosting a specific frequency band.
- New synth - DPulse - a detuned pulse wave.
- Sampler now has a cutoff envelope which can be accessed via new opts
which mirror the standard envelope opts but with a
cutoff_
prefix (such ascutoff_attack
,cutoff_decay_level
and friends). - Sampler now correctly handles samples with different sample rates.
- Bitcrusher FX now has an internal low pass filter modifiable via a new
cutoff
opt. - Panslicer now correctly honours min and max pan values.
- New default opt
on:
for bothsample
andsynth
. This acts likeif
but ensures all the opt values are evaluated (useful if wanting to keep the consumption of random streams or ticks consistent even when not triggering a synth. - MIDI opts such as
cutoff:
can now accept note names such as:c4
. - FX learned the global
slide:
opt to match synths.
- Massive performance improvements.
- Teach
play_pattern_timed
to handle rings. current_transpose
now returns 0 if there is no current transposition.- BPM scaling is now honoured when controlling synths and FX
- All
with_fx*
fns now return the result of their block. spark
now handles rings correctly.spark
now treats booleans as 1s and 0s so you can now spark rings of bools.puts
,print
andmc_message
now handle multiple message arguments
- Ensure
with_fx
doesn't swallow TL modifications such as transposition, current synth etc. - Ensure
with_fx
doesn't affect random seed. - Improve reliability of boot process on Mac and Windows.
- The FX
pre_amp:
opt is no longer scaled w.r.t. the current BPM. - Fixed GUI side of update checking system.
Thursday 10th September, 2015 (view commits)
This release brings a substantial change to the random number
generator. This has the unfortunate side effect of breaking backwards
compatibility. If you have been using rand
, choose
, shuffle
and
friends to create predictable patterns for your riffs, your code will
produce different results in this release. Please let me apologise and
say it's for a good cause. So what is this good cause? Well, you can
now jump backwards and forwards through the random stream giving you way
more creative control than before! The random stream is now also unified
with the random stream on the synthesis server allowing you to sync
behaviour between synths and code. Exciting times.
The sampler has also been super charged. We can now easily change the
rate via MIDI note intervals with rpitch:
and stretch the sample whilst
preserving pitch via pitch_stretch:
(although with variable results
;-)
).
Finally you can now control the global mixer with set_mixer_control!
for those full filter sweeps over the entire sound...
Have fun and happy live coding!
- Complete rewrite of random number system. This means if you've been
combining calls to
use_random_seed
with randomisation to create predictable melodies/rhythms/timbral manipulations you will unfortunately get different results inv2.7
. This change is to synchronise both the Ruby rand stream with the one in SuperCollider as well as enabling the reversal of calls to rand viarand_back
. sync
now causes the BPM to be inherited from the thread calling the matchingcue
. This may be disabled with the newbpm_sync:
opt.rrand
andrand
now return 0 if called with 0.invert_chord
now handles negative inversions in a more musically appropriate manner..
ratio_to_pitch
which provides the inverse ofpitch_to_ratio
midi_notes
- returns a ring of numbers (mapping the source ring/array through the fnnote
).rand_back
- reverse the random stream and 'undo' calls torand
rand_skip
- skip forward through the random stream.rand_reset
- reset the random stream to the last seed.
- It is now possible to toggle the commenting of whole selections or
individual lines with the shortcut
M-/
. - Added Icelandic translation.
- All synths learned the
decay_level
opt to allow the sustain phase to shift between two distinct values. The default value fordecay_level
is to mirrorsustain_level:
thus preserving current behaviour. play
andsynth
have now learned thepitch:
opt to matchsample
. This just increments or decrements the final note.sample
now correctly validates opts.sample
learned thepitch_stretch:
opt which combinesrate:
modification (similar tobeat_stretch:
withpitch:
. Attempts to maintain the pitch whilst simultaneously changing the rate. The underlying implementation is very basic and can easily destroy the sound.sample
learned therpitch:
opt for simple rate pitch modulation. This modifies the rate to match the specified number of MIDI notes relative from the current pitch. For example, arpitch:
of 12 would double the rate.- The unit of the FX
:echo
'sdecay:
opt is now beats and the value is scaled with the BPM.
- Most examples have been tweaked to sound good with the new random generator.
- Tilburg has been replaced with Tilburg 2. Play it and get your Algorave on!
- Auto-align code on Run.
live_loop
learned theseed:
opt which will set the new thread with the specified seed prior to initial run.- Add check to ensure BPM is a positive value.
density
has now been taught to handle values between 0 and 1 which will now stretch time for the specified block.- Errors now no longer print out unusual print version of context object i.e. #SonicPiSpiderUser1:0x007fc82e1f79a0
- Both
in_thread
andlive_loop
have now learned thedelay:
opt which will delay the initial execution by the specified number of beats. - Buffer and thread name are now printed on error.
sample_duration
now understands all the opts that you can pass tosample
- It is now possible to do basic arithmetic on symbols representing
rests:
:r + 3
returns:r
(a rest plus any MIDI note shift is still a rest).
- Fixed crash when synth args were specified as Rationals.
note_info
now correctly handles octaves.- Fix windows paste shortcut
C-v
. - Teach
invert_chord
how to properly handle out of range index ranges.
Thursday 30th July, 2015 (view commits)
The laser beams sliced through the wafts of smoke as the subwoofer pumped bass deep into the bodies of the crowd. The atmosphere was ripe with a heady mix of synths and dancing. However something wasn't quite right in this nightclub. Projected in bright colours above the DJ booth was futuristic text, moving, dancing, flashing. This wasn't fancy visuals, it was merely a projection of Sonic Pi running on a Raspberry Pi. The occupant of the DJ booth wasn't spinning disks or playing MP3s, she was writing, editing and evaluating code. She was live coding and this was an Algorave.
This release is codenamed Algorave to
celebrate that Sonic Pi is now ready to be performed within nightclubs
as well as still being a fantastic tool for learning how to code
creatively. There are many improvements as detailed below. However,
let's just take a brief look at some of the most fun. Firstly we have
the new error reporting system to make it easier to find and debug your
errors. Syntax errors are now blue and runtime errors pink. We also have
a new look and feel including a new dark mode for performing in dark
places. We also have some fantastic new synths, FX and have even
improved the standard synths. For example, sample
now lets you stretch
to match the beat with the beat_stretch:
opt and change pitch with
pitch:
. The slicer
FX now sports a fantastic deterministic
probability:
opt for creating and manipulating rhythmic structures
with ease. Finally there's the new thread local tick
/look
system
which will revolutionise the way you work with ring
s within
live_loop
s. Of course there's so much more too!
Enjoy this release and happy Algoraving!
- The
res:
opt for all synths and FX now has range 0->1 rather than 1->0. This means that a higher res value results in more resonance. This will hopefully be more intuitive to beginners and less surprising for people with existing synth knowledge. - The fn
stop
has been renamed tokill
for killing specific synths. In its place a new fnstop
has been added to stop a given thread orlive_loop
. invert_wave
opts are now inverted. The default is now 0 which has the same behaviour as the old 1. This means that it's more intuitive to use the opt as to invert the current wave, you now specify:invert_wave: true
, rather thaninvert_wave: false
. This shouldn't affect any code which doesn't explicitly set theinvert_wave:
opt. Pieces which have explicit inversion need to swap all 0s for 1s and vice versa.- The
res:
opt forrrand
andrdist
has been renamed tostep:
to avoid confusion with the resonance opt for cutoff filters. - Rename
pitch_ratio
topitch_to_ratio
to keep in line with other similar fns such asmidi_to_hz
.
- New thread-local (i.e. live_loop local) counter system via fns
tick
andlook
. - New fn
vector
which creates a new kind of Array - an immutable vector (SPVector
) which is now the default base class for all rings. - New fns
use_sample_defaults
andwith_sample_defaults
which act similarly as their*_synth_defaults
counterparts but for samples not synths. - New fns
use_tuning
andwith_tuning
for exploring other tuning systems such as:just
,:pythagorean
, and:meantone
. - New fn
invert_chord
for chord inversions. - New fn
current_beat_duration
for returning the duration of the current beat in seconds. - New fn
note_range
for returning a range of notes between two notes with apitches:
opt for constraining the choice of filler notes. - New fns
scale_names
andchord_names
for returning a ring of all chords and scales. - New example
rerezzed
- strongly influenced by Daft Punk's trackderezzed
. - New example
reich phase
- a nice way of combininglive_loop
s andtick
to create sophisticated polyrhythms. - New fns
use_cue_logging
andwith_cue_logging
for enabling and disabling the logging ofcue
messages.
- New visual look and feel including a new Dark Mode for live coding in night clubs.
- New preferences for hiding/showing aspects of the GUI such as the buttons, log, tabs etc.
- New preference for full screen mode.
- Improve error message reporting. Syntax errors are now made distinct from runtime errors with colour-coded messages. Also, the line number of the error is much more visible, and the line of the error is highlighted with an arrow in the left-hand margin.
- Workspaces are now named buffers. This is a smaller word which works better on lower res screens and is also a lovely term used by a number of wonderful programming editors such as Emacs.
- Print friendly messages to the log on boot.
- Add pref option to check for updates now.
- New FX -
krush
for krushing the sound. - New FX -
panslicer
similar toslicer
andwobble
but modulates the stereo panning of the audio. - New synth
subpulse
for a full range pulse with extra bass. - New synth
blade
- a moody Blade Runner-esque synth violin - New synth
piano
- a basic piano emulation. Only capable of whole notes. - FXs
slicer
andwobble
now have a wonderful newprobability:
opt which will only slice on (or off depending on wave inversion) with the specified probability. The behaviour is deterministic, so repeated calls with the sameseed:
andprobability:
opts will result in the same behaviour. Great for adding interesting rhythmic variation to sound. - FXs
slicer
andwobble
now have smoothing opts for even more control over the resulting wave form. - Teach
sample
the optbeat_stretch:
for modifying the rate of the sample to make sure the duration of the sample is n beats long (with respect to the current bpm). Note: stretching the beat does change the pitch. - Teach
sample
the optpitch
to enable pitch shifting on any sample. - FX
flanger
's feedback mixing is now more fair and is less likely to hike up the amplitude.
- Teach
note_info
to also handle a number as its param. - Teach
factor?
to handle division by 0. - Teach
load_sample
to throw exception when passed an empty path. - Now throws an exception when you attempt to create an empty ring.
- Rings are now immutable (in the Clojure sense) which means they can be safely passed to multiple threads/live_loops without any issues.
- Teach
use_sample_bpm
the optnum_beats:
to indicate that a given sample consists of a specific number of beats. - Teach
comment
anduncomment
to require blocks. - Teach synth chord groups to allow their notes to be controlled individually to allow transitions between chords.
- Throw nicer exception when unable to normalise synth args
- Teach
chord
the new optinvert:
as a shortcut to the newinvert_chord
fn. - Teach
sample_duration
about the optsstart:
andfinish:
and envelope opts such asattack:
andrelease:
. This allows you to replace any call tosample
withsample_duration
to get the exact duration of that call. - Teach
chord
the optnum_octaves
to enable the easy creation of arpeggios. - It is now possible to set the block type in the Minecraft API.
- Fix bug in
with_sample_pack_as
to now correctly accept a block. mc_surface_teleport
no longer throws an error.Array#shuffle
now works correctly with the random seeds for deterministic behaviour.- Fix broken behaviour with multiple nested calls to
*_sample_bpm
.
Monday 13th April, 2015 (view commits)
This release comes with support for Minecraft: Pi Edition installed on the Raspberry Pi. You can now create music with Minecraft visuals or even code up a synth score in Minecraft blocks and read and play the score from Sonic Pi! Another exciting aspect of this release is much improved editor functionality for navigating around and manipulating code via keyboard shortcuts. This means that live coding just got a lot more fun. The keyboard shortcuts are based on the standard shortcuts provided by GNU Emacs - the oldest and most powerful text editor in use by wizard programmers today.
invert_wave
now defaults to 1 everywhere. I found I always inverted the wave every time I used a synth/fx where wave inversion was key. This seemed like such a better default I've broken compatibility for it. Apologies if this has affected you.- The
flanger
FX now defaults the optional argstereo_invert_wave
to 1. - Renamed FX
ring
toring_mod
to reduce the potential for confusion with thering
data-structure. Tab
now indents current line or selection rather than inserting a useless tab character.
- Support for programming Minecraft Pi Edition.
sync
now accepts multiple cue ids and will sync on the first matching id.- New fn
pitch_ratio
for converting a midi note to a frequency ratio. Useful for tuning samples. - New fn
line
for creating a line from start to finish with a specific number of slices. - New fn
spark
for displaying lists of numbers in a fancy text-graph (▁▃▅▇
) in the log. - On stop, amplitude of output slides down over 1s to silence for a smoother transition.
sample_duration
now scales result based on current BPM.range
now accepts optional args:inclusive:
andstep:
.
- German translation of GUI and tutorial. Simply open Sonic Pi with a machine with a German localisation setting.
- Display GUI fully maximised when opening for first ever time.
- Workspace indexing now starts at 0 to match standard programming indexes.
- New shortcuts -
M-<
andM->
for switching workspaces. - Many new Emacs-based code navigation and editing shortcuts. See the shortcut cheatsheet in the built-in tutorial for more information.
- Increase height of doc and error panes.
- Improve error pane colour scheme.
- Auto-align now trims whitespace from start and end of buffer.
- Add preference toggle to hide/show line numbers.
- Documentation now supports semantic formatting and highlighting.
- Docsystem tabs are now positioned at the bottom for better navigation.
- New preference to hide/show line numbers in editor.
- New FX -
:pitch_shift
- Fix OSC lib to properly encode multi-byte chars such as UTF8
- Fix sporadic issue on some platforms when trigging percussive sounds
within a
reverb
FX caused a serious audio overload. - Add missing fn metadata for
*_sample_bpm
- Fix synth metadata for FX
:bpf
. - Fix arg metadata for
use_sample_pack_as
- Rings now pretty print themselves as
(ring 1, 2, 3)
rather than[1, 2, 3]
. C-k
keyboard shortcut now copies text into the clipboard.- Scales and chords now return actual
ring
s rather thanring
-like things. - Improve Ring Mod FX arguments
- Exceptions created within
with_fx
are now raised correctly.
Wednesday 11th February, 2015 (view commits)
A quick release following v2.3
to address an issue with the GUI
freezing on specific CPUs. However, although this release has had a
small development cycle, it ships with three fantastic features. Firstly
we now have the spread
fn which provides an amazing way to create
interesting rhythms with very little code. Secondly we can now use
cutoff:
on any sample massively increasing their timbral range and
finally we have three exciting new synths for you to play with. Have
fun!
- Unfortunately 5 pre-release synths accidentally slipped into v2.3. Three of them have been polished up and are in this release (one with major changes including a name change). However, the other two have been removed.
- New fn
spread
for creating rings of Euclidean distributions. Great for quickly creating interesting rhythms. - GUI now automatically appends a
:
to the FX opt autocomplete list - Synths and FX now raise an exception if any of their non-modulatable opts are modulated. This is disabled when the pref 'check synth args' is unchecked.
- GUI now renders pretty UTF-8
└─
├─
characters when printing in the log on RP. - Improve docstrings for sample player.
- New Synth
:dark_ambience
, an ambient bass trying to escape the darkness. - New Synth
:hollow
, a hollow breathy sound. - New Synth
:growl
, a deep rumbling growl. - Sampler synths now sport built-in
rlpf
andnormaliser
FX. These are disabled by default (i.e. won't affect sound of the sample) and can by enabled via the newcutoff:
,res:
andnorm:
opts.
- Fix insanely obscure bug which caused the GUI to freeze on certain platforms (32 bit Windows and RP2 with 2G/2G kernel).
- Remove DC Bias offset from Prophet synth (see http://en.wikipedia.org/wiki/DC_bias)
Wednesday 28th January, 2015 (view commits)
- Playing chords with the fn
chord
now divides the amplitude of each resulting synth by the number of notes in the chord. This ensures the resulting amplitude isn't excessive and is normalised. - Chords now evaluate their args once and those args are used for all synth triggers. This means random values are only generated once and are similar across all notes in the chord. Previous behaviour can be obtained by calling play multiple times with no interleaved sleeps.
- Ensure each new thread's random number generator is unique yet seeded
in a deterministic manner. This stops random values across
at
from being identical. range
is now exclusive:(range 1, 5) #=> (ring 1, 2, 3, 4)
- New fn
density
for compressing and repeating time Dr Who style. For example, wrapping some code with a call to density of 2 will double the bpm for that block as well as repeating it twice. This ensures the block takes the same amount of time to execute while doing double the work. - New fns
with_bpm_mul
anduse_bpm_mul
which will multiply the current bpm by a specified amount. Useful for slowing down and speeding up the execution of a specific thread or live_loop. - New fn
rdist
- generate a random number with a centred distribution - New examples:
square skit
,shufflit
andtilburg
- Teach control to ignore nil nodes i.e.
control nil, amp: 3
will do nothing. - Teach Float#times to yield a float to the block. For example,
3.4.times {|v| puts v}
will yield0.0
,1.0
and2.0
. - Synth, Sample and FX args now handle bools and nil correctly.
true
resolves to1.0
andfalse
,nil
resolve to0.0
. This allows you to write code such as:play :e3, invert_wave: true
- Teach
at
to handle varying block arities differently. See docs for more detail. Original behaviour is preserved and only extended. - App now checks for updates (at most once every 2 weeks). This may be disabled in the prefs.
- Teach
:reverb
FX to extend its kill delay time with larger room sizes to reduce the chance of clipping.
- New FX
bitcrusher
- for crunching and destroying those hi-fi sounds. - New FX
flanger
- a classic swhooshing effect typically used with vocals and guitars. - New FX
ring
- ring modulation for that friendly Dalek sound - New FX
bpf
- a band pass filter - New FX
rbpf
- a resonant band pass filter - New FX
nbpf
- a normalised band pass filter - New FX
nrbpf
- a normalised resonant band pass filter
perc_snap
- a finger snapperc_snap2
- another finger snapbd_ada
- a bass drumguit_em9
- a lovely guitar arpeggio over Em9
- Namespace
live_loop
fn and thread names to stop them clashing with standard user defined threads and fns. - GUI no longer crashes when you start a line with a symbol.
with_fx
now returns the result of the block- Kill zombie scsynth servers on startup for better crash recovery.
- Handle paths with UTF8 characters gracefully
- Force sample rate for output and input to 44k on OS X. This stops scsynth from crashing when output and input sample rates are different.
Thursday 18th December, 2014 (view commits)
This release brings a number of nice enhancements. However the main feature is the accurate timing for triggering FX. This means you can now reliably use FX for accurate rhythmic purposes such as wobbling, slicing and echoes.
use_sample_pack_as
now uses a double underscore__
as a separator between the user-specified alias and the sample name.
- Teach synth args to take prefixed maps:
play 50, {amp: 0.5}, {release: 2}, amp: 2
- Don't round Floats when user specifically prints them to log with puts
with_fx
FX synths are now triggered using virtual time rather than real time. This means that FX can now be used for rhythmical purposes.- Work on new
RingArray
data-structure. This is essentially an array that wraps its indexes so you can use indexes larger than the array size. - New fn
ring
-(ring 1, 2, 3)
creates a new ring array. - New fn
knit
-(knit :a1, 2, :c1, 1)
returns(ring :a1, :a1, :c1)
- New fn
bools
-(bools 1, 0, 1)
returns(ring true, false, true)
- New fn
range
-(range 70, 100, 10)
returns(ring 70, 80, 90, 100)
- New fn
sample_loaded?
- to detect whether a specific sample has been loaded
- Fixed regression in
:tb303
synth - sound is reverted to v2.0 behaviour - New Synth -
:square
- Pure square wave
- Help system now auto-docks on close
- Preferences are now remembered across sessions
- On Raspberry Pi, previous volume and audio output options are forced on boot.
bd_tek
- Bass drum
one_in
now returns false if num is < 1- Ensure
live_loop
's no-sleep detector works within nestedwith_fx
blocks chord
now returns a ring.
Tuesday 25th November, 2014 (view commits)
- Windows version no longer needs special firewall exceptions to run
- Added license information to info window
- Minor grammar and spelling tweaks to tutorial
Friday 21st November, 2014 (view commits)
The focus of release is very much on technical improvements, efficiency and general polish.
The most obvious and exciting change is the introduction of the
live_loop
which will change the way you work with Sonic Pi. For more
information on live_loop
take a look at the new section in the
tutorial. Another very exciting development is that v2.1 marks the
official support for Windows thanks to the excellent work by Jeremy
Weatherford. Finally, this release is also the first release where Sonic
Pi has a Core Team of developers. Please give a warm welcome to Xavier
Riley, Jeremy Weatherford and Joseph Wilk.
- New fn
live_loop
- A loop for live coding - New fn
inc
- increment - New fn
dec
- decrement - New fn
quantise
- quantise a value to resolution - New fn
factor?
- Factor test - New fn
at
- Run a block at the given times - New fn
degree
- for resolving a note in a scale by degree such as:i
,:iv
- New fn
chord_degree
- Construct chords based on scale degrees - New TL fn
use_sample_bpm
- for changing the BPM based on a sample's duration - New fn
rest?
- Determine if note or args is a rest - New fn
vt
- Get virtual time - New fn
set_control_delta!
- Set control delta globally wait
now handles bothsleep
andsync
functionality- Allow first arg to
play
to be a proc or lambda. In which case simple call it and use the result as the note - Teach
play
to accept a single map argument (in which case it will extract:note
key out if it exists. - Fns
play
andsynth
now treat 'notes'nil
,:r
and:rest
as rests and don't trigger any synths.
- Updated and improved tutorial
- Much improved autocompletion
- Add HPF, LPF, mono forcer and stereo swapping preferences to new studio section for use when performing with Sonic Pi through an external PA.
- Shortcuts overhauled - now supports basic Emacs-style Ctrl-* navigation.
- Shortcuts Alt-[ and Alt-] now cycle through workspaces
- Shortcuts now work when toolbar is hidden
- Font sizes for individual workspaces are now stored between sessions
- Ctl-Mouse-wheel zooms font on Windows
- Links are now clickable (opening external browser)
- Entries in docsystem's synth arg table are now clickable and will take focus down to arg documentation
- Stop users accidentally clearing entire workspace if they type quickly after hitting run
- Hitting F1 or
C-i
over a function name now opens up the doc system at the relevant place
- Reworked examples.
- Much improved efficiency in many areas - especially for Raspberry Pi.
- Avoid occasional clicking sound when stopping runs
- Note Cb is now correctly resolved to be a semitone lower than C
- Non RP systems now start with more audio busses (1024)
- Array#sample and Array#shuffle are now correctly seeded with thread local random generator
- Log files are now placed into ~/.sonic-pi/log
- Chords and scales now wrap around when accessed from indexes outside of their range.
rand_i
andrrand_i
now correctly return integers rather than floats- rrand arguments now correctly handle a range of 0 (i.e. min == max)
- Line offset in error messages is now correct
- When saving files on Windows, CRLF line endings are used
- Stop users defining functions with same name as core API fns
- New samples (bass drums, snares and loops)
- Allow
mod_range:
opt to have negative values (for oscillating with lower notes) - Change slide mechanism to default to linear sliding with support for changing the curve type. All modifiable args now have corresponding
_slide_shape
and_slide_curve
args. - Improve TB303 synth - now supports separate cutoff ADSR envelopes. New opts:
cutoff_attack:
,cutoff_sustain:
,cutoff_decay:
,cutoff_release:
,cutoff_min_slide:
,cutoff_attack_level:
,cutoff_sustain_level:
,cutoff_env_curve:
Tuesday 9th September, 2014 (view commits)
- Fix recording functionality
- Improve documentation content and layout
- Close off OSC server from external clients
- Add History, Contributors and Community pages to info window
- Improve startup speed on OS X
- Re-work and add to shortcuts for key actions:
- on RP they are all
alt-*
prefixed - on OS X they are all
cmd-*
prefixed
- on RP they are all
- Improve highlighting of log messages (
cue
/sync
messages are more clearly highlighted) - Log now communicates when a run has completed executing
- Fix bug encountered when stopping threads in super fast loops (stopped comms with server)
Tuesday 2nd September, 2014 (view commits)
- Complete rewrite since v1.0
- Support for Live Coding - redefining behaviour whilst music is playing
- New timing system - timing now guaranteed to be accurate
- Many new synths
- New chainable studio FX system
- Support for sample playback
- Inclusion of over 70 CC0 licensed samples from http://freesound.org
- Support for controlling and modifying synth, fx and sample playback arguments after they have been triggered
- Completely re-designed GUI
- Help system with full documentation, tutorial and many examples
- Record functionality (for recording performances/pieces)
- Support for controlling system audio settings on RP