On SD2SNES, transitions between SPC music/soundfx (such as the Wind Sound) and MSU music can cause buzzing/silence.

Issue #15 resolved
Dylan Morrison repo owner created an issue

"Two things I noticed, though. First, the transition between Narshe Town and the Mine Entrance will ALWAYS cause the wind SFX to play for a brief instant and then buzz until a new audio track is loaded (like from a battle).

Secondly, if the opening intro is allowed to play all the way through and then a 'new game' is started (which will put the player just entering Narshe, as the cutscene has already played), then the wind SFX will not load properly and buzz (though much quieter).

Both of these are 100% reproducible. I played up to the second save point, so there may be other issues, but these are the ones I notice right away." - Retro Dan

Comments (14)

  1. Dylan Morrison reporter

    Just a note that I seem to be unable to reproduce this bug in Higan, thus the title.

    For me in Higan there is a bit of a crackle/stutter when it transitions from wind SFX to the Narshe Mines theme initially, but nothing major, and the Narshe Mines theme does play as normal.

    I am suspecting this issue may be due to the "diagnostic code" so I'm committing a version to experimental that will not have the diagnostic code.

  2. Dylan Morrison reporter

    Alright, Retro Dan please try the ips from commit 965ce09 and see if this fixes this issue. This commit is without the diagnostic code, and it also contains a fix to the unnecessary requests to "track 0" from the MSU-1 (Issue #14). One or the other might fix your issue simply due to simplifying the code.

  3. Maximilian Rehkopf

    Reposting this from #13:

    Yeah, the buzz is another bug I encountered yesterday. When requesting a non-existing track while another is being played back, output to the DAC isn't stopped; after a bit the FPGA audio output buffer gets filled with new data for some reason (probably opening+preparing track 0), but is never updated subsequently - thus creating a static buzz noise. This is now fixed as well.

  4. Dylan Morrison reporter

    So we can probably assume this is for the most part fixed on my end by explicitly skipping Track 0 on the application side. I won't close this one resolved quite yet, but good to know.

  5. Maximilian Rehkopf

    It should be fine even with track 0 as it needs to be fixed in the sd2snes firmware. If you / @RetroDan are interested I can prepare a current development build for testing. It should work fine even with 6cbec33 or earlier.

  6. Dan Yetman

    That would be nice, ikari. I'm more than willing to be available for testing whatever is needed.

  7. Dylan Morrison reporter

    Awesome. I don't have an SD2SNES so a development build of the firmware wouldn't be of much use to me, but I'm really glad that my project can be of help to yours.

    For now I'm offering a stopgap IPS on the master branch containing both current workarounds, but I'll probably revert back to the current mainline "beta" code (commit 6cbec33) if it can be verified that bugs #13 and #15 are resolved with the new firmware and said firmware gets a public release.

  8. Dan Yetman

    Sorry, I've been super busy recently. I'll see what's what with this newest commit over the weekend.

  9. Dylan Morrison reporter

    Doing some cleanup in here in preparation for moving this project to GitHub to hopefully get some more eyeballs since I just haven't had time for this project in months. I'm marking this issue closed for now under the assumption ikari's firmware fix linked above handled the problem, even though I haven't gotten any feedback from Dan on whether it worked yet.

  10. Log in to comment