Arpeggiator - Note repeats

[zonkmachine]

I finally got around to finish the arpeggiator note repeats function that I started working on here #687.

The first attempt modified the direction logic but I found a simpler way to do away with that.

[LmmsBot]

🤖 Hey, I'm @LmmsBot from github.com/lmms/bot and I made downloads for this pull request, click me to make them magically appear! 🎩

Linux

• Linux (AppImage): lmms-1.3.0-alpha.1.60%2Bg075e608-linux-x86_64.AppImage (build link)

Windows

• Windows 32-bit: lmms-1.3.0-alpha.1.60%2Bg075e608df-mingw-win32.exe (build link)
• Windows 64-bit: lmms-1.3.0-alpha.1.60%2Bg075e608df-mingw-win64.exe (build link)

🤖

{"platform_name_to_artifacts": {"Linux": [{"artifact": {"title": {"title": "(AppImage)", "platform_name": "Linux"}, "link": {"link": "https://11790-15778896-gh.circle-artifacts.com/0/lmms-1.3.0-alpha.1.60%2Bg075e608-linux-x86_64.AppImage"}}, "build_link": "https://circleci.com/gh/LMMS/lmms/11790?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link"}], "Windows": [{"artifact": {"title": {"title": "32-bit", "platform_name": "Windows"}, "link": {"link": "https://11794-15778896-gh.circle-artifacts.com/0/lmms-1.3.0-alpha.1.60%2Bg075e608df-mingw-win32.exe"}}, "build_link": "https://circleci.com/gh/LMMS/lmms/11794?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link"}, {"artifact": {"title": {"title": "64-bit", "platform_name": "Windows"}, "link": {"link": "https://11791-15778896-gh.circle-artifacts.com/0/lmms-1.3.0-alpha.1.60%2Bg075e608df-mingw-win64.exe"}}, "build_link": "https://circleci.com/gh/LMMS/lmms/11791?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link"}]}, "commit_sha": "b474a9c1694dabd7a13b2338aa24ddc0598a6f4d"}

[IanCaio]

First review just goes through code style, which I know sounds pointless considering the current state of the InstrumentFunctionArpeggio::processNote method (IMHO it needs a refactor on the code style and improvements on comments, but that's outside the scope of this PR), but I guess the general consensus is to at least keep new additions compliant, even if the surrounding code isn't.

I'm still learning the logic of the method itself to further review, but testing the build I noticed a misbehavior that I'm still trying to figure out the source of in the code: The repeating notes overlap with the previous note. Probably something to do with the calculation of the frames per note, but I suppose the repeating notes should be played alone not on top of the previous one.

[zonkmachine]

The repeating notes overlap with the previous note.

Right. The envelope I used on my test case hide that well.

Thanks for the review, I'll look into it!

[superpaik]

This is a very tricky one to test. Something weird happens but I'm not sure it's due to this PR, though the repeat function make it clearer.
Anyway. I try to explain it.
If you set a project like the one in the image, where REP is set to one (no repeat), mode FREE and TIME synced to 16th note, everything seem to work fine. But if you set REP to two, you'll notice a silence before the thrid beat (just before E4 plays), and again at the end of 4th beat.
If you set mode to SYNC, some notes of the arpegiator (I think are those of the E4 note) are not sync at all.
I said that maybe is unrelated to this PR because in a similar situation (without no REP, of course) on the stable version, though notes are sync, some notes of the E4 arpeggio are not playing.

[zonkmachine]

@superpaik Thanks for testing the PR!
I don't think the problems you mention are related to this PR. There are known issues with disappearing notes. #2606 #3880

[IanCaio]

@zonkmachine The overlapping notes issue seems to be fixed!

I noticed some notes skipping or odd behavior (even on other modes like Sort), but as you mentioned those are unrelated to this PR but a problem with the current state of the Arpeggiator. I was going to suggest a refactor of this method (on another PR ofc), but seems like there are some issues opened already regarding this and other deeper problems that apparently causes those glitches on the notes.

Everything seems to be fine on the code, but since I'm still getting familiar with this method I'll review it again.

[zonkmachine]

I'm thinking we can give the ComboBoxes a bit more space.

Changes to the layout on the left instrument window

-       mainLayout->setHorizontalSpacing( 20 );
+       mainLayout->setHorizontalSpacing( 14 );

[IanCaio]

Sorry I resolved the comments and wrote all over again, but I think I couldn't reuse the comments for a last review. It has been brought up recently again the argument in favor of us being more strict with code style on lines that are touched by the PR, which I agree with since if we take the surrounding code into account we will probably never move away from the previous non-conforming styles used.

As for the logic, I reviewed again after the gate issue was fixed and it all seems good! 😃

[zonkmachine]

@IanCaio @Veratil I find the interface a bit confusing. Can you please commit your changes or just push them directly?

[zonkmachine]

I have now gone over your suggestions and merged two of them. Much appreciated! In regards to the clang format changes the pendulum now seem to have changed, once more, to waiting for a script to do the job.

[superpaik]

If there is only one note playing at a time, repeat function works well in every mode (free, sort, sync)
But if there are two notes playing at a time, in sync mode some notes, at random, are not repeated, they "desappear" (you can use the same use-case I detailed above)

[zonkmachine]

I think it's this issue you're seeing: #2606

[superpaik]

Yes. I see, it's related to that other PR.

[zonkmachine]

I think this has been reviewed for both style and code and has been tested thoroughly (I've been trying out most of these code changes for well over six years).

[zonkmachine]

Merge? :)