| layout | title | permalink | category |
|---|---|---|---|
question |
Frequently Asked Questions |
/FAQ/ |
help |
{::options auto_ids="true" /}
- Will be replaced with the ToC {:toc}
See Also
Operating system: 64-bit Windows 10 - 11, Apple macOS 11 - 14., or 64-bit Linux with at least glibc 2.31.
CPU: x86-64 Intel or AMD; at least one 2 GHz core for SD, 4 cores for HD, and 8 cores for 4K.
GPU: OpenGL 2.0 that works correctly and is compatible on Linux, Direct3D 11 on Windows.
RAM: At least 4 GB for SD, 8 GB for HD, and 16 GB for 4K.
NOTE: If your computer is at the lower end of CPU and RAM requirements, you should use both the Preview Scaling and Proxy features to help reduce preview lag.
Hard drive: yes, get one; the bigger, the better :-)
Network: Shotcut does NOT require access to the network to activate, check a subscription, or send usage analytics. However, some links in the Help menu do link out to this web site. If you have files on a fast (at least 1 Gb/s) network share you can access them from there through your operating system.
Is this a new - for the first time ever - installation of Shotcut? If NOT, and it used to work, then you may need to delete your Shotcut registry entries and try again. The registry keys are stored at HKEY_CURRENT_USER\Software\Meltytech\Shotcut.
One Windows 8.1 user reported he needed to change the following in the Compatibility tab of the Properties for the Shotcut icon/exe: Run in compatibility mode for "Windows XP (Service Pack 2)", and "Run this program as administrator." However, not every one needs to do that, and we are not sure why he needed it.
If you just want to save your changes to re-open it later in Shotcut, you can save your project as a MLT XML file by clicking Save on the toolbar at the top of the window. If you want to upload the video to a web site or somehow share the result with someone as separate video file, then click Export on the toolbar, which opens or raises the Export panel. In the Export panel, there are 3 basic steps:
- choose and click a Preset,
- click Reload to load the current video settings, and
- click Export File at the bottom of the panel.
By default, if you have made a playlist Export uses the Playlist unless you have put something into the Timeline, in which case, it uses the timeline. Otherwise, it will Export the clip or live source (stream, device, screen, etc.). However, you can control what is used for the source of the export using the From control at the top of the Export panel.
It all depends; that is why there are so many of them! But here are some suggestions:
- Upload to a video-sharing web site: H.264 Main Profile
- Extract the audio: Ogg Vorbis or MP3
- Save to an intermediate file to use with another tool: DNxHD or lossless/ProRes - really depends on what the other tool accepts
- Put onto my own web site: WebM and/or H.264 Main Profile
- Screen recording (Linux only): lossless/MJPEG
- Capture from SDI/HDMI: lossless/ProRes if your system can handle it, otherwise lossless/MPEG-2
If a job appears to be stuck or hung (please allow ample time for percentage to change), right-click the job in the Jobs panel and choose Stop. Try export again with Export > Video > Parallel processing disabled (unchecked). If there is still a problem see what percentage it reaches before failing or hanging. Then, apply the percentage to your project duration to find the rough time of where it seems to be having trouble. Is there something complicated or different there? You can temporarily remove some things in this area and try again to see if export goes further.
There are 2 text filters at this time: Text: Simple and Text: Rich. You can apply the text filters to a video clip, an image file or a whole track. You can also apply a text filter to a transparent clip to act as its own text clip by choosing File > Open Other > Text from the menu or Open Other > Text from the main toolbar. The default background color in Open Other > Text is transparent which as well can be changed to a solid color by clicking on the Background color button.
You can also create text with an external program as an image with an alpha channel and composite it. To composite, you add a video track to the Timeline to use as a layer. Then, you open an image, set its Properties to adjust duration or enable an image sequence, add it to the new video track, and further adjust its position and duration as-needed. You might also want to apply the Size, Position & Rotate filter to the image clip. The image file formats that support an alpha channel are PNG, SVG, and TGA. You can also use Quicktime Animation format. There are many tools that can create images with alpha channels for this purpose. Some of them include GIMP, Inkscape, Krita, Paint.NET, etc. For animation, consider Blender and Synfig.
Some formats and compression methods simply make it take longer. In the Settings menu set Interpolation to Nearest Neighbor. This setting not only affects the quality of image scaling but also the accuracy of seeking. Please be aware that this setting may cause seeking to become less accurate resulting in some frames repeating when stepping frame-by-frame backwards or the first several frames in the forward direction immediately after a seek.
In the Settings menu set Interpolation to something other than Nearest Neighbor - Bilinear is recommended. When the interpolation level is set to nearest neighbor it relaxes the accuracy of seeking to make the responsiveness of the video player faster.
The best way to see the shortcuts is in the Actions and Shortcuts dialog. This dialog appears by navigating "Help->Actions and Shortcuts..." or by pressing the "?" key.
There is also a [page with all of the keyboard shortcuts]( {{ "/howtos/keyboard-shortcuts/" | prepend: site.baseurl }} ).
Just like j, k, and l for playback transport control, the bare i, o, x, v, b are very common shortcuts use by other professional video editing software from Apple, Avid, Lightworks, and others.
P.S. While it is rather obvious to use cursor left and right keys for single frame stepping, there is another technique so you do not have to remove your fingers from the JKL: while holding down K, tap J to step left or tap L to step right.
Sometimes the keyboard "focus" might be captured by something in the GUI causing the shortcuts to not function. In that case, click the video preview region to return focus to the player.
Some file managers do not like the launcher icon provided with the binary download from this site. The launcher icon was tested successfully on GNOME Nautilus and KDE Dolphin.
- open a terminal/console window
- cd to the location where you have extracted Shotcut
- enter
Shotcut.app/shotcut\ Please do not try to run bin/shotcut; always use the wrapper script in the Shotcut.app folder! If it still does not start, then it should report that some libraries could not be loaded.
First, use Properties to see if the Video tab is disabled. If it is disabled, then Shotcut is not compatible with this format or codec. If the video tab is enabled, more than likely OpenGL (or also Direct3D on Windows) is not working on your system, or it is too old. First, make sure GPU Effect is disabled in Settings. GPU effects requires OpenGL version 3.2. When it is disabled, you only need OpenGL version 2.0 on Linux, Direct3D on Windows, or Metal on macOS.
Yes. Simply rename your existing program folder to put the version number in it or move it out-of-the-way to another location.
On Windows, the installer is mostly just a fancy zip extractor that also adds a start menu item. So, you can install the new version to a different location, or rename the existing folder to prevent it from being overwritten. Then, you can just navigate to whichever program folder you want in Explorer and run shotcut.exe.
On macOS, you do not need to copy Shotcut to the /Applications folder - that is merely a suggestion. Simply drag Shotcut out of the .dmg to wherever you like and rename the app bundle to put the version number into it. Or, rename the existing version to move it out of the way before copying Shotcut from the .dmg.
The same concepts apply to Linux, where Shotcut is simply delivered as a compressed tar archive. However, on Linux, it is important to understand that the launch icon always looks in Shotcut.app; so, either version the folder containing the launch icon or have multiple, versioned Shotcut.app folders and run the launch script that is inside of it.
This is available through Properties > menu > Extract Sub-clip...; however, it is not frame accurate and there are some pitfalls. We recommend LosslessCut, VirtualDub, or Avidemux as alternatives for this. However, those usually have the same limitations and problems. In Shotcut you can try the extract, and instead export if that did not work good enough.
- Some filters are not compatible with GPU Processing.
- Some filters cannot be applied to a track or the entire timeline (i.e. Output). Please check what is currently selected. The top of the Filters panel shows the name of what is selected.
- By design, we do not make all LADSPA and frei0r filters available through the UI. We do not want to present a confusing generic interface to all available plugins and overwhelm users with so many, often redundant, options. Also, the Shotcut team wants to better understand the filters we are providing and ensure each one is operating as intended with MLT. If you are manually authoring XML, then you can add any filter you want.
- Sometimes a MLT plugin may fail to load due to a missing or incompatible library. Shotcut downloads provide all dependencies for Windows and macOS, but it may rely upon some libraries provided by distribution packages on Linux.
Yes, because the software is made available under the GPLv3 license, it supports The Free Software Definition, which includes:
The freedom to run the program as you wish, for any purpose (freedom 0).
The Shotcut software does not collect, store, or share any personal information. It does not even collect any usage activity to send to our server. The exception to that is the Microsoft Store, and you can read the Microsoft Privacy Statement. Of course, there might be other app stores and distribution systems, as permitted by the license, with their own personal information collection that is not under our control.
The Shotcut forum has a privacy policy.
Currently, the cache is used to store thumbnail and waveform data.
Windows
The log, cache, and presets are stored in %LOCALAPPDATA%\Meltytech\Shotcut\. In the Explorer location bar, enter "%LOCALAPPDATA%" and press Enter. Then look for Meltytech\Shotcut.
The settings are stored in the registry at key HKEY_CURRENT_USER\Software\Meltytech\Shotcut\.
Linux
The log, cache, and presets are stored in ~/.local/share/Meltytech/Shotcut/ where '~' is your home directory, of course.
The settings are stored in ~/.config/Meltytech/Shotcut.conf, which is a text file in INI format.
macOS
The log, cache, and presets are stored in ~/Library/Application Support/Meltytech/Shotcut/, where '~' is your home directory, of course.
The settings are stored in ~/Library/Preferences/com.meltytech.Shotcut.plist, which is a binary plist file.
Shotcut was originally announced in November, 2004! You can read more about it from a backup of its original website. The current version of Shotcut is a complete rewrite with none of the original requirements in mind. The original Shotcut was created by Charlie Yates, a MLT co-founder and original lead developer. Since Dan Dennedy, another MLT co-founder and its current lead, wanted to create a new editor based on MLT, he simply chose to reuse the name since he liked it so much. The current Shotcut had its origins as the MLT BuildOnMe project. BuildOnMe was created in January, 2011 as a minimal, example project of how to use MLT in a cross-platform Qt 4 project. Eventually, Dan sought to replace the aging Melted GTK+ client, Rugen, with a cross-platform replacement and needed an app on which to easily test the cross-platform compatibility of new MLT features such as WebVfx and Movit. That led him to fork BuildOnMe and start Shotcut.
If you are strictly asking about subtitles or closed captions, Shotcut does not read, make, edit, or pass-through subtitles. There is planned the ability to read, show, pass-through, and burn-in subtitles, but there is no estimated time of arrival. We recommend that you try the free, open source, cross-platform subtitle editor Aegisub.
However, if you are just asking about the ability to put/overlay text in your video, use the Text: Simple or Text: Rich filter. Since it is a filter, that means you need something to which to apply it. If you just want a solid color, choose File > Open Other > Color. You can also use a picture/photograph just like a video clip by opening it and adding it to your playlist or timeline. If you just want a text clip (transparent background) choose Open Other > Text and add it to your playlist or timeline.
This is implemented as of version 16.01. With a clip open in the source player or selected in the timeline, choose Properties and look for the Speed field. Shotcut only provides simple frame dropping or duplicating. However, if the frame rate of your source footage is higher than the Video Mode (under Settings menu), then you can achieve a fairly smooth slow motion. If you are looking for more sophisticated results using more advanced optical flow techniques, we recommend you try the free, open source, cross-platform tool slowMoVideo.
Shotcut does not offer that, but we recommend to try the free, open source, cross-platform tool Audacity.
Shotcut uses the GPU in three ways:
- For drawing parts of the user interface and showing video: Direct3D on Windows, Metal on macOS, and OpenGL on Linux.
- hardware encoding (where available and enabled)
- OpenGL for GPU Effects (filters and transitions) mode
Shotcut does NOT use the GPU or hardware acceleration for the following:
- decoding and pixel format conversion
- automatic (as-needed) filters to deinterlace, scale, and pad video and to resample or downmix audio
- filters that you add
- transitions
- compositing/blending video tracks
- mixing audio tracks
Thus, you cannot expect Shotcut to use close to 0% CPU and much % of GPU when exporting using the hardware encoder because the reading of files and decoding alone becomes a bottleneck to feed the hardware encoder. Also, if you have any decent amount of image processing, you should expect a significant amount of CPU usage especially if parallel processing is enabled (it is by default). Software from other companies may limit itself to one GPU vendor API such as CUDA in order to provide almost entirely GPU-based pipeline. Shotcut has not chosen to go that route because it is a cross-platform solution.
The video stream always originates in system (CPU) RAM, and CPU-based video decoding is highly optimized and fast. Meanwhile, transferring full, uncompressed video from the GPU RAM to system RAM is a relatively slow. Thus, in the context of a video editor (not simply a player or transcoder), hardware-accelerated decoding should only be done when all video processing can also be done on the GPU. That alone is non-trivial. Shotcut does have an OpenGL-based effects system that is disabled and hidden currently due to instability. Even when enabled it is a small subset of all effects and does not include a deinterlacer. Next, assuming you do not need to deinterlace and agree to limit oneself to the GPU effects, there is a major technical hurdle to transfer the decoded video in GPU RAM to OpenGL textures due to multiple APIs for multiple operating systems. Likewise, the complexity to convert OpenGL textures to hardware encoder frames for the various hardware encoding APIs. Any tool that claims to do all of these but does not ensure the video stays in GPU RAM is going to have limited performance gain if any.
Even if made available (integrated) there are major hurdles to handle resource limitations (number of simultaneous decodes) in a robust fashion and to handle incompatible video streams with many permutations of encoding profiles/parameters, APIs, and devices. That would result in a huge source of unreliability and support issues.
If you want to help with this, please feel free to [contribute]( {{ "/howtos/contribute/" | prepend: site.baseurl }} ). We have not made much progress here due to higher priorities: fixing bugs, rework on some features, adding basic expected UI features, upgrading dependencies, providing support, making documentation, and stabilizing GPU Effects.
Shotcut's engine (MLT atop FFmpeg and other libraries) uses multiple CPU cores/threads for:
- decoding video on many (most?) video codecs
- image slice-based multi-threaded processing in some processes
- frame-based multi-threaded image processing in many processes
- encoding video when not using the hardware encoder on most video codecs
When any of the above is not enabled, a bottleneck is introduced. Some of these are minor and others major depending on the weight of the operation.
Shotcut's interface - in addition to the main UI thread - uses multiple background CPU cores/threads for:
- generating video thumbnails
- generating audio levels for waveform display in the timeline
- the engine itself (see above)
- sending video to OpenGL for display
- exporting
Use the QT_SCALE_FACTOR command line options.
On Windows, you can also create a shortcut to shotcut.exe. Then, edit the
shortcut and modify the target field in the shortcut to include the
QT_SCALE_FACTOR. E.g. C:\Program Files\Shotcut\shotcut.exe --QT_SCALE_FACTOR 2
Shotcut does not properly display or output HDR video. However, it can convert HDR video with tone-mapping into a Rec. 709 SDR video. In the UI use either Properties > Convert or View > Resources... in the menu.