All of the videos that I have been posting since 1.7 and calling "1.8" are fixes leading up to 1.7.5. I have had testers doing extensive testing since 1.7, but of course any changes whatsoever could totally invalidate that testing. The testers help me to make progress to a perfect sound, and I try no to mess up that course.
The last change to code was done on September 18. Most of these changes were related to getting latency down, so did not change the sound as seen in the videos, perhaps with the exception of the time it takes for the sound to ramp up (not the sustain phase).
One of the guys doing testing had a vaguely described problem that sounded like it might be the all purples up and red all the way down setting, at full volume on a half-stack. Then one of the comments states that 1.7 was better at this setting. I am *really* puzzled by these comments, because I think the harmonic mix is completely unchanged for 1.7.5. One of the fixes that I had to do to 1.7 was that at this setting which supposedly sounded so "amazing" was, when you played a note down an "impulse" (audible clicking noise) happened... In fact, this setting was the one particular setting that I hated in 1.7! So the only thing I know of that is different at that setting was a bug. Is anybody seriously saying that they loved that? That makes no sense at all.
I thought that exactly this setting was completely useless in 1.7.
If you feel that I broke 1.7's organ sound, I would be curious to get a concrete explanation. I suppose that it's possible that I could have goofed and changed the harmonic mix, especially when I was reducing latency. Latency reduction changed a little bit of the code everywhere. One thing about Mugician that makes it interesting is that absolutely everything is inter-related. I can't fix the clean sound without changing the distorted sound. Visual doodads can create audible latency. There is next to nothing that doesn't affect the sound in some way. So I can't fix a particular sound in isolation. Even the stupidest little change invalidates a lot of testing, even though it takes testers days to a week to get back to me with feedback since the last line of code has changed.
As you can imagine, that's maddening from the perspective of trying to fix sound bugs without changing the sound. :-)
No comments:
Post a Comment