Chat

Author Topic: bug reports and probably more  (Read 10648 times)

HertzDevil

  • Guest
bug reports and probably more
« on: March 12, 2012, 12:53:20 AM »
Annoyances
1. [GB] Is the volume output calculated by linearly offsetting the volume envelope with the channel volume? For example, an envelope {E} with channel volume B gives E-(F-B)=A, but {8} with channel volume 7 yields 8-(F-7)=0. 0. How is that possible?
2. [GB] [SMS] These sound chips only have 16 volume levels as opposed to 128 for YMU759 and YM2612. When Axy changes the channel volume by x or y every frame, the result is disastrous, creating an effect that is nearly unusable.

Bugs
1. [GB] [SMS] Omitting the channel volume or putting an F for the first note of any channel will mute all notes until the channel meets a channel volume that is not F.
2. [All, unconfirmed] Axy doesn't change the volume if doing so will cause the final volume to "clip".

Offline Delek

  • Salí para ver
  • Administrator
  • Hero Member
  • Posts: 1809
    • Delek's Website
    • Email
bug reports and probably more
« Reply #1 on: March 12, 2012, 02:50:24 AM »
Wow, thanks a lot!. This types of bugs are the harder to find. :)

They will be fixed on next release!.

Offline Delek

  • Salí para ver
  • Administrator
  • Hero Member
  • Posts: 1809
    • Delek's Website
    • Email
bug reports and probably more
« Reply #2 on: March 21, 2012, 05:05:50 AM »
I am fixing some bugs before the release so:
1. [GB] Is the volume output calculated by linearly offsetting the volume envelope with the channel volume? For example, an envelope {E} with channel volume B gives E-(F-B)=A, but {8} with channel volume 7 yields 8-(F-7)=0. 0. How is that possible?
That is logical, because F-7 = 8, so 8-(F-7)=8-8, then the zero appears. How is that possible?, math. :P
2. [GB] [SMS] These sound chips only have 16 volume levels as opposed to 128 for YMU759 and YM2612. When Axy changes the channel volume by x or y every frame, the result is disastrous, creating an effect that is nearly unusable.
This is true, the volume goes down rapidly isn'it?.

1. [GB] [SMS] Omitting the channel volume or putting an F for the first note of any channel will mute all notes until the channel meets a channel volume that is not F.
That issue exists only if you do not use an envelope, and actually, the bug is THE SOUNDING note. Because DefleMask always needs a volume envelope, at least of size 1.
2. [All, unconfirmed] Axy doesn't change the volume if doing so will cause the final volume to "clip".
I checked this over and over again, and I cannot replicate the bug. Seems that everything is working correctly. Remember that the Axy effect is: A10 -> DOWN WITH SPEED OF 1, A01 -> UP WITH SPEED OF 1. A11, A21, A23, A41, ETC = NO EFFECT. A ZERO VALUE IS REQUIRED.

Offline Ralnir

  • Newbie
  • Posts: 8
bug reports and probably more
« Reply #3 on: April 03, 2012, 11:43:35 PM »
I hate to hijack this thread, but I figured it'd be the best place to mention my woes regarding DefleMask. Having only recently stumbled upon this tracker while looking for something to fill my FM synth needs, I rather enjoy it. That said, there's a few things that doesn't sit well with me. It could very well be just on my end (I'm running Windows 7 64-bit, which is known to cause a few problems with programs here and there), but nevertheless here is my admittedly short list of problems that I've encountered.

1) The sound stutters like mad. It only happens sometimes and it eventually goes away, but there's no real fix for it or a method to prevent it as far as I've experimented. I've tried setting the core affinity for DefleMask to just one CPU and that doesn't seem to help at all. I don't have an audio card or anything like that so it's hard to record the sound when it acts like that. Perhaps this is the fault of the buffer? It's been like this since I started, which would be 5.1b or so?
2) The samples screen is very tempermental when in Genesis mode. I'm not sure how it was originally imagined, but if you click the X next to a sample, it's hard to load a sample into a new slot without reloading the song. That could be how it's meant to act, so I don't know.
3) Exporting a song to .vgm doesn't seem to produce anything that runs. I've tried exporting to both version 1.60 and 1.61 and neither work in foobar2000 with the Game Emu Player plugin or XMplay with in_vgm.dll.
4) In the changelog for 5.2, it says that a bug was fixed regarding the AR slider in YMU759 mode. I didn't think much of it at the time, but later on I was messing around in YMU759 mode and the AR slider for operator 2 does absolutely nothing.

So, all in all, it's mostly just 3 and 4 that's really bringing me down and it may simply be a result of my choice of players. Sorry if this little list of mine isn't terribly descriptive, but that's about the most I can glean from what I've encountered.
« Last Edit: April 03, 2012, 11:46:47 PM by Ralnir »

Offline Delek

  • Salí para ver
  • Administrator
  • Hero Member
  • Posts: 1809
    • Delek's Website
    • Email
bug reports and probably more
« Reply #4 on: April 04, 2012, 03:01:35 AM »
1) The sound stutters like mad. It only happens sometimes and it eventually goes away, but there's no real fix for it or a method to prevent it as far as I've experimented. I've tried setting the core affinity for DefleMask to just one CPU and that doesn't seem to help at all. I don't have an audio card or anything like that so it's hard to record the sound when it acts like that. Perhaps this is the fault of the buffer? It's been like this since I started, which would be 5.1b or so?
Have you tried change the buffer's size?, under options menu.
2) The samples screen is very tempermental when in Genesis mode. I'm not sure how it was originally imagined, but if you click the X next to a sample, it's hard to load a sample into a new slot without reloading the song. That could be how it's meant to act, so I don't know.
Wow, strange bug, I will fix it now.
3) Exporting a song to .vgm doesn't seem to produce anything that runs. I've tried exporting to both version 1.60 and 1.61 and neither work in foobar2000 with the Game Emu Player plugin or XMplay with in_vgm.dll.
What?, I am testing the vgm export right now and works perfectly. Seems to be a problem of your vgm player.
4) In the changelog for 5.2, it says that a bug was fixed regarding the AR slider in YMU759 mode. I didn't think much of it at the time, but later on I was messing around in YMU759 mode and the AR slider for operator 2 does absolutely nothing.
Actually it does everything but the envelope is from the operator 3!!, my mistake (remember that the YMU759 have both 2 or 4operators modes). I fixed it right now, and DefleMask was Re-Uploaded. Could you please re-download the file and check if the operators of the YMU759 now are fine and the samples on Genesis mode are right?.
So, all in all, it's mostly just 3 and 4 that's really bringing me down and it may simply be a result of my choice of players. Sorry if this little list of mine isn't terribly descriptive, but that's about the most I can glean from what I've encountered.
This type of feedback helps me a lot, many many thanks!. :)
« Last Edit: April 04, 2012, 03:10:41 AM by Delek »

HertzDevil

  • Guest
bug reports and probably more
« Reply #5 on: April 06, 2012, 09:16:18 AM »
1. 0Dxx does not work when it is located in the last pattern.
2. The background box in the Gameboy waveform editor does not align with the macro editing area.
3. Gameboy duty cycle has 5 settings.
4. The volume table is still subtractive instead of multiplicative like FamiTracker's current volume matrix.
5. Choosing the WAVE Channel instrument and inputting notes into the wave channel does not actually write the correct instrument number to the instrument column.
6. EDxx does not delay volume column values.
7. Dragging the mouse over the pattern editor cannot select more than one column.
8. Chain of 0Axx effects does not work like the same effect in almost every other tracker.

Offline Delek

  • Salí para ver
  • Administrator
  • Hero Member
  • Posts: 1809
    • Delek's Website
    • Email
bug reports and probably more
« Reply #6 on: April 06, 2012, 07:05:22 PM »
1. 0Dxx does not work when it is located in the last pattern.
So you want to jump to the next pattern where there does not exist a next pattern?. The 0Dxx effect WILL NOT WORK on the latest pattern, because there is NO NEXT PATTERN TO JUMP.
From the manual:
Quote from: manual
Dxx - Pattern Break

This command will stop playing the current pattern and will jump to the next one in the pattern matrix. You can also select the row where to start the next pattern. Note that the specified row xx is in Hex.

...
2. The background box in the Gameboy waveform editor does not align with the macro editing area.
I checked this over and over again and I can not find any "not alignment" in the editing areas. Could you please be more specific?.
3. Gameboy duty cycle has 5 settings.
That's true, my bad. However, the fifth value will be the 4, because DefleMask has a parameter limiter on every command. Btw, I fixed this and DefleMask was re-uploaded. Thanks.
4. The volume table is still subtractive instead of multiplicative like FamiTracker's current volume matrix.
Why should DefleMask be equal to FT in every way?. DefleMask is a multi-system tracker, a multiplicative volume is something not desirable for systems with a high resolution of volumes.
5. Choosing the WAVE Channel instrument and inputting notes into the wave channel does not actually write the correct instrument number to the instrument column.
The "WAVE Channel" is NOT an instrument, you should set the Waves and THEN select a STD instrument to use the envelopes with that desired waveform, in other way the instrument will be "0". For that reason the "WAVE Channel" does not have indexes and the other instruments yes.
7. Dragging the mouse over the pattern editor cannot select more than one column.
This is intentional, however the selection way will be changed in future releases.
8. Chain of 0Axx effects does not work like the same effect in almost every other tracker.
I do not know how work like the Axx effect in almost every other tracker, could you please be more specific?.

Thanks a lot. :)
« Last Edit: April 06, 2012, 07:08:26 PM by Delek »

Offline kinkinkijkin

  • Newbie
  • Posts: 41
  • Delek FTW
bug reports and probably more
« Reply #7 on: May 20, 2012, 07:14:25 PM »
Sorry to add this in here and not somewhere else, but 04xy doesn't work on D-4 - G(?)-4 on SMS. The exact range I don't know, I just know it works on G#4 and A#4 and C-5.

Actually it seems 04FF works... but it lowers the pitch of the D-4 and F-4 by half a note... y=6 makes F-4 lose the vibrato... and it seems y=4 - y=9 read as y=4.

Very weird stuff here. Very annoying. I also know on D-5 the pitch bend is applied instead as really fine volume changes. Weird 04xy bug is weird.

Offline Delek

  • Salí para ver
  • Administrator
  • Hero Member
  • Posts: 1809
    • Delek's Website
    • Email
bug reports and probably more
« Reply #8 on: May 20, 2012, 09:28:58 PM »
Could you please give me a module example?, maybe a vibrato that does not work could help. Thanks.

Offline Delek

  • Salí para ver
  • Administrator
  • Hero Member
  • Posts: 1809
    • Delek's Website
    • Email
bug reports and probably more
« Reply #9 on: May 21, 2012, 02:14:50 AM »
I found a mistake in the vibrato code!, thanks!, I fixed it and everything works fine. Check it out and test it kinkinkijkin, maybe the speed and depth could change a little because the math behind the vibrato changed for a more precise one.

Offline Ralnir

  • Newbie
  • Posts: 8
bug reports and probably more
« Reply #10 on: June 30, 2012, 12:51:15 AM »
I don't really want to flood the forums with a new topic every time something miniscule crops up, so I'll use this thread again.
Okay, so. I'm not sure how much of this is supposed to function as intended and I'm simply misusing it, but I have a rather bothersome issue with the latest version of DefleMask. If you don't understand what I'm trying to describe, reply and I'll try to clarify. I was messing around with a Genesis song as a means of passing the time when I noticed that the instruments seem to stack upon each other as they play. I started with a rudimentry bass hit and then just started toying with the values. When I use the second instrument and then the bass instrument right afterwards, it sounds like it's been run through a grinder. It's the same if you use the second instrument after the bass instrument though it is not as distorted. Hitting stop seems to get it to reset, though I'm worried about what effect this will have on existing songs. Is it merely something wrong with the envelopes ... oh balls. I just attempted to load up a premade Genesis instrument and the folder is shown as empty when accessed within DefleMask.
Though I have doubts as to its usefulness, I have included the DMF as an attachment.
Not related to the topic but as an aside, is there some indication of what operator # is what on the algorithm? I'm used to VOPM and what have you so I kind of have a general idea, but still, the potential for confusion is there.

Offline Delek

  • Salí para ver
  • Administrator
  • Hero Member
  • Posts: 1809
    • Delek's Website
    • Email
bug reports and probably more
« Reply #11 on: June 30, 2012, 09:59:15 PM »
About the empty folder, check if that is happening now, re-download DefleMask.

Also I'm working on configurable keys and GB Rom exporting. So, check that stuff too. :)

I will release DefleMask 7f in a few days btw.

Offline Ralnir

  • Newbie
  • Posts: 8
bug reports and probably more
« Reply #12 on: July 01, 2012, 01:56:13 AM »
Well, the folders work now which is a great help. It's just mindblowing how the instruments seem to feed into each other. Yet, it only affects new songs. I loaded up a song from a few versions ago and it worked perfectly. Has any changes been made as to how DM creates new songs since version 6? I've yet to check the other systems, though, so it may only be a glitch in the handling of the FM synth.

Or am I just doing it wrong? Am I not handling the operators properly? That's a possibility, as I've only started messing around with this sort of thing a month or two ago.