rtoy@ can you help triage this?

I tried the repro case on Win10, Chrome 54, the current stable release.  The notes play out correctly for me.  The notes are the same as when I use Firefox to play.

Works fine in Chrome canary (57).

Could you try a newer version of Chrome?

I am using Chromium Version 53.0.2785.143 Built on Ubuntu , running on Ubuntu 16.04 (64-bit), which is the version in the stable repos for Ubuntu.

I just now ran an update, but the only Chromium related update is: chromium-browser-l10n

Here are two links to videos I created of the issue:
https://owncloud.libretools.com/index.php/s/xS42Pm96x3uVlhl
https://owncloud.libretools.com/index.php/s/tWyXwAENYiu951M
^^Perhaps the second video would be more helpful. In the second video, I am using the tone.js code so that you can see the synth log side-by-side our log. The issue is that the both synth AND our software report the correct note, but the note that is played is the wrong note. You have to listen to notice the bug, hence the video.

On another machine, I have version 54.0.2820.0 (64 bit) and I experience the same issue. As stated before, the issue was noticed on Windows, Mac, and GNU/Linux machines running Chrome/ium 53 and higher.

If there is a newer stable version of Chromium that does not have the bug, would you please try to add it to the Ubuntu Repos? 

Thanks again!
Devin

Ok. Thanks for the videos. I do hear it, but you're also moving blocks around so I don't know what that means.

I tried this with Chrome 55 (stable) on Ubuntu.  I do hear the first note being played with the wrong pitch once in a while.  I have been unable to reproduce this with Chrome beta or chrome dev.

So some fixes maybe have come in between what you're using and Chrome beta, but without a smaller repro case that doesn't use tone.js, finding the underlying cause likely won't happen.  Especially since I can't reproduce this with Chrome beta.

> Ok. Thanks for the videos. I do hear it, but you're also moving blocks around so I don't know what that means.

Moving the blocks just helped me isolate pitches. As you can see from the log, the intended output is unaffected.

> I tried this with Chrome 55 (stable) on Ubuntu.  I do hear the first note being played with the wrong pitch once in a while.  I have been unable to reproduce this with Chrome beta or chrome dev. So some fixes maybe have come in between what you're using and Chrome beta...

Do you mean that the newest versions after version 55 (stable) may have fixed this issue?

Excuse my ignorance, but are these the newest beta versions?
https://www.google.com/chrome/browser/beta.html?platform=linux
https://www.google.com/chrome/browser/desktop/index.html?platform=linux&extra=devchannel

> but without a smaller repro case that doesn't use tone.js, finding the underlying cause likely won't happen.  Especially since I can't reproduce this with Chrome beta.

"repro case" means "reproducible case"?

I will try and test with latest versions. If I still experience the issue, I will try to experiment with other synths.

Thanks!

I assume the logs are just telling you what the output was supposed to be without actually measuring the actual output.

Yes, those links are the correct ones for downloading beta and dev channel builds of chrome.

Yes, "repro case" means reproducible case.  Sorry for the jargon.

Thanks for testing.  It's much easier to find and fix bugs if the test case is small; having tone.js involved makes it much more difficult to do and I couldn't even find where you were creating the tones. (Because I don't know tone.js).

We do not seem to be experiencing the issue since we updated tone.js

Please see https://github.com/walterbender/musicblocks/issues/295#issuecomment-275160656 and https://github.com/walterbender/musicblocks/commit/343c1b28d4013ef27240ab8a789379ab671c3a87

If this is the case, I apologize for raising the alarm. Thank you!

devinulibarri@

Which version of Chrome did you test with?

Version 54.0.2820.0 (64-bit), which is one of the versions that I had tested previously and reproduced the bug. I am asking for our team (SugarLabs) to test on various versions and report back to us. I will post their findings once they come in.

I am especially curious about the results for Chrome/ium 53 - Chrome/ium 55 (stable).

Based on c#9 and the fact that no new information has been given, I'm closing this bug as Fixed since it seems Chrome 54 has fixed it.

If this issues shows up again, please re-open this issue or file a new one.

Thanks!