Denon MC7000: Add optional jog wheel acceleration to the controller mapping

[fwcd]

Based on #4757

While toying around with the jog wheels on the MC7000, I have noticed that Serato performs a form of accelerated scrubbing when scratching in vinyl mode, i.e. the jog sensitivity increases the faster you turn the wheel. Acceleration can be useful to both precisely navigate the track and scrub quickly without having to press shift.

This PR adds a simple, configurable and entirely optional acceleration whose curve is a simple monomial function. In pseudocode:

accelerationFactor = pow(baseSpeed, accelerationExponent) * accelerationCoefficient
scratchSpeed = baseSpeed * accelerationFactor

When acceleration is disabled (which it is by default, or the exponent is 0 and the coefficient 1), we get the current behavior:

scratchSpeed = baseSpeed

Thoughts?

[fwcd]

cc @toszlanyi

[toszlanyi]

@fwcd Thanks a lot for your improvement, this looks interesting. Unfortunately I cannot test it myself as i don't have the MC7000 to hand. To get rid of the code style issues you may check the wiki page here

[fwcd]

Many of these code style issues are unrelated to this patch, therefore I've created a separate PR #4754 to tackle those. Once merged, I would rebase this branch.

[Swiftb0y]

Hey there @fwcd, thank you for your addition. I'm sorry that I forgot about this PR. In regards to the codestyle issues: I'd suggest you rebase your bug fixes and feature additions (this PR and #4755) onto the 2.3 branch, wait until those two are merged into 2.3 (and main) and then rebase #4754 ontop of that new state of main. This way, the bug fixes and feature additions get merged into the next 2.3.x release, while the changes related to the newer javascript environment stay in 2.4. What do you think?

[fwcd]

That sounds like a good plan. I have updated both this and #4755 to be based on 2.3. Since there was still a linting issue, I didn't base them directly on 2.3, but on #4757 (which in turn is based on 2.3), to separate the concerns there. To summarize, the branching now looks as follows:

                                           +--- #4755 (inverted shift bug)
                                          /
2.3 <--- #4757 (2.3 code style issues) <-+
                                          \
                                           +--- #4686 (jog wheel acceleration, i.e. this PR)

I'll keep #4754 based on main and will update it once the 2.3 PRs are merged.

[Swiftb0y]

I have noticed that Serato performs a form of accelerated scrubbing when scratching in vinyl mode, i.e. the jog sensitivity increases the faster you turn the wheel.

Can you elaborate a bit (linking to an explanation would also be fine). If interpret this correctly, this seems to alter the scratch behavior significantly (ontop of mixxx's own scratch emulation). Is this behavior desired from DJs?

[ronso0]

I have noticed that Serato performs a form of accelerated scrubbing when scratching in vinyl mode

This would lead to so-called "sticker drift", right? Like, scratching forward fast und slowly moving the same distance back would not bring you to the scratch starting point in the track.

[fwcd]

Yes, it does lead to 'sticker drift' as you've described it (this would be unavoidable with any kind of acceleration), making it unsuitable for certain kinds of scratching, therefore I would definitely leave it disabled by default.

The primary value in such a feature, as I see it, would be a very convenient way to seek through a track, both on a fine-grained and a larger scale. Whether implementing it on top of Mixxx's own scratch emulation is a good idea or not is something I have pondered over too, though. Since this wouldn't be a feature specific to the MC7000, perhaps it would be better to include it in Mixxx directly?

[Swiftb0y]

Can you link me to a manual or a video where this acceleration behavior is described / demonstrated? I don't know how popular this feature is. Many controllers already feature a "touch strip" for seeking within a track (even the MC7000 has this).

[daschuer]

Maybe we can start the acceleration only after a full round of the jog wheel. Normally scratch samples are only ~ 1/2 rotation = 1 s and searching for a beat when cuing a track is also only a meter of a 1/2 rotation. This probably gives enough distance to distinguished between scratching and seeking.

[fwcd]

Hm, I am having a hard time finding a video that explicitly demonstrates this feature. https://youtu.be/72hSiPS3yfQ?t=379 shows it very briefly, but I'll see if I can find anything specifically on this. Traktor seems to have a setting named "Rotary acceleration" that lets users map arbitrary MIDI controls with acceleration.

[Swiftb0y]

Mhmm, the video you sent is a standalone controller though. Its not even related to Serato. Do they not mention this feature anywhere in their manual?

[github-actions]

This PR is marked as stale because it has been open 90 days with no activity.

[JoergAtGithub]

Mhmm, the video you sent is a standalone controller though. Its not even related to Serato. Do they not mention this feature anywhere in their manual?

The Numark NV II is a Serato controller with two embedded displays, not a standalone unit.

[github-actions]

This PR is marked as stale because it has been open 90 days with no activity.

[JoergAtGithub]

What's the state of this PR? At least you need to retarget it to 2.4.

[JoergAtGithub]

Controller settings are ony available in main, not in 2.4. I suggest a follow up PR for main.

[fwcd]

I'll probably just retarget this to main, since I don't think this will be very useful without controller settings (digging into the JS is likely too inconvenient for most users). Since this is not a large change, I don't expect this to be a big risk for merge conflicts.

[JoergAtGithub]

Retargeting to 2.5 is enough. My comment about Main was from before we branched 2.5.

[fwcd]

Fair enough, thanks for noticing that.

[fwcd]

The jogwheel options are now fully user-configurable:

[JoergAtGithub]

I guess for the jog the input-latency matters. Because for the output-latency we have an undocumented CO: engine.getValue("[App]", "output_latency_ms").

[JoergAtGithub]

LGTM! Could you please open a manual PR, to describe the new settings, curently we have only: https://manual.mixxx.org/2.5/en/hardware/controllers/denon_mc7000#user-variables

[fwcd]

Could you please open a manual PR, to describe the new settings

Done:

• Denon MC7000: Document new jogwheel settings manual#690

[fwcd]

Can this be merged? I'm planning on adding more options to the mapping, but to avoid conflicts, I'd base them on this PR.

[JoergAtGithub]

Thank you!