VOGONS


Reply 561 of 660, by root42

User metadata
Rank Oldbie
Rank
Oldbie

Found another bug: I have two VGM files in the folder and make a playlist of them. The 2nd VGM is simply played twice.

https://youtu.be/-ba_9ZuUAiQ

YouTube and Bonus
80486DX@33 MHz, 16 MiB RAM, Tseng ET4000 1 MiB, SnarkBarker & GUSar Lite, PC MIDI Card+X2+SC55+MT32, OSSC

Reply 562 of 660, by FreddyV

User metadata
Rank Member
Rank
Member
root42 wrote on 2020-09-15, 20:20:

By the way: do you support vgz? Probably not...

No, I need the uncompress code for this.
With 7zip, right click extract file to do a .vgm from a vgz.
For .vgz, mod master start sbvgm if you copy it in the same folder as mod master.

Last edited by FreddyV on 2020-09-16, 08:55. Edited 1 time in total.

Reply 563 of 660, by FreddyV

User metadata
Rank Member
Rank
Member
root42 wrote on 2020-09-15, 20:18:

EDIT: The change is to null 0x5c to 0x5E, which contains the OPL3 clock, and instead to set 0x50 to 0x52 to 99 9E 36, which is the OPL2 clock.

Yes, only this 😀

Reply 564 of 660, by MJay99

User metadata
Rank Member
Rank
Member
FreddyV wrote on 2020-09-13, 14:58:

- S3M: Bug corrected in the pattern loader (MJay99, PASSTIME.S3M)

That's much better now, a big thank you for fixing it for one of my favorite tracks 😁

Last edited by MJay99 on 2020-09-16, 19:39. Edited 1 time in total.

Reply 567 of 660, by matze79

User metadata
Rank l33t
Rank
l33t

Not yet 😒

https://dosreloaded.de - The German Retro DOS PC Community
https://www.retroianer.de - under constructing since ever

Co2 - for a endless Summer

Reply 568 of 660, by genesis8bit

User metadata
Rank Newbie
Rank
Newbie
FreddyV wrote on 2020-09-19, 17:17:

Anybody tested with a Tandy LPT ? and OPL2LPT ?

I just tested it on my Amstrad PPC 640 with a mod of the Mystic Towers (Apogee) music, it worked.

First sound wasnt perfect, it was correct after I erased the .DIR and .CFG, I suppose because I was using a covox 4 before ?

But, I only have sound with the internal amplifier, strangely no sound via the jack cable which is connected to Logitech speakers.

It is the first time I am using the OPL3LPT since I bought it a few months ago. Any advice will be welcome.

Reply 569 of 660, by FreddyV

User metadata
Rank Member
Rank
Member
genesis8bit wrote on 2020-09-21, 20:04:

I just tested it on my Amstrad PPC 640 with a mod of the Mystic Towers (Apogee) music, it worked.

Thanks,

Note I added OPL2LPT support only, Access to the OPL3 registers (Channels 10 to 18) is not programmed yet as I needed to have confirmation for OPL2LPT first.

Reply 570 of 660, by zyga64

User metadata
Rank Member
Rank
Member

Thank you for new release. VGM support works excellent !!!
However, I think I found a bug. Darkness2.s3m by Purple Motion is completely silent !
I can see song is playing (notes and effects on tracks), but can't hear nothing.
Same situation in Cubic Player, and OCP (even newest version 0.1.20 for Linux). No sound at all.
XTC-Play and Inertia Player are able to play it correctly.

What is even more strange - Scream Tracker V3.2 plays it silently (but not Fasttracker II).

1) VLSI SCAMP /286@20 /4MB /TVGA9000C /CMI8330
2) i420EX /486DX33 /16MB /Trio64V+ /AZT2316+GUS
3) i430HX /P233MMX /64MB /VirgeDX+3DFX /YMF701
4) i440BX /P II 400 /256MB /FX5500/AWE64
5) i865G /E5800 /2GB /Ti4200 /YMF724

Reply 571 of 660, by FreddyV

User metadata
Rank Member
Rank
Member
zyga64 wrote on 2020-09-22, 19:21:

Thank you for new release. VGM support works excellent !!!
However, I think I found a bug. Darkness2.s3m by Purple Motion is completely silent !
I can see song is playing (notes and effects on tracks), but can't hear nothing.

Thanks,

For the module with no sound, its global volume is set to 0 by default and is not changed during the partition.
Simply edit the file in ST3 to change its global volume.
The players able to play it are the players having a bug 😀

As it was made for a game, it is probably the game engine that change the global volume.
It was then not supposed to be ripped and released as is.

Reply 572 of 660, by zyga64

User metadata
Rank Member
Rank
Member
FreddyV wrote on 2020-09-24, 07:06:

Simply edit the file in ST3 to change its global volume.
The players able to play it are the players having a bug 😀

Just did it. Very useful tip ! Thank you !

1) VLSI SCAMP /286@20 /4MB /TVGA9000C /CMI8330
2) i420EX /486DX33 /16MB /Trio64V+ /AZT2316+GUS
3) i430HX /P233MMX /64MB /VirgeDX+3DFX /YMF701
4) i440BX /P II 400 /256MB /FX5500/AWE64
5) i865G /E5800 /2GB /Ti4200 /YMF724

Reply 574 of 660, by FreddyV

User metadata
Rank Member
Rank
Member

The New Tandy stuff is working and as usual, need some adjustment before I can release it.
(You can see it on my Twitter account)

The Sound quality of this thing is excellent, what a pity that too few software support it.

Reply 576 of 660, by FreddyV

User metadata
Rank Member
Rank
Member
MobyGamer wrote on 2020-10-05, 15:32:

Tandy DAC was released a year before Sound Blaster, so even more of a shame. I'm delighted that your TL/2 (an 8MHz 286) could run modm xt all the way up to 48 KHz.

Yep, When I bought the Tandy TL and I did look on internet for the DAC Speed, I saw 22KHz almost everywhere, but I also saw 44kHz somewhere.
Looking at the DAC Documentation, it can go much higher, even more than 48K.

When we check the games supporting the DAC, the samples are bad. The Deskmate samples are also not good...

I did not expect a sound quality like this 😀 (No ISA Bus noise like with the Sound Blaster)
The DAC quality seems to be better than a SB2.0, for something released one year before the SB 1.0

The problem with the Tandy DAC is that it use the DMA Signal (TC) that is not on the ISA Bus, so not possible to use in something else.

I still have to improve it: The sound crack when it is too slow as my code reducing the "quality" when it is slower is not finished, so the 48KHz in the video is "Real"
Then, we can go higher 😀

I spent 2 weeks being convinced that the DAC used 2Kb fixed size buffer. (But in reality, Tantrack mix 2kb Buffers..)
Then, Looking at the DOSBox code, I saw the Tandy DMA generate the IRQ at the end of the DMA Buffer (Only possible if the chip is on the motherboard)
I had then to redo the code "like" in tantrack and other software : Do nothing in the DAC IRQ and Read the DMA offset in a 50Hz Timer to know what is the next buffer to mix.

I found a bug in DOSBOX, so tantrack does not work it and I "Corrected" this in Mod Master: The Bit telling that the IRQ was a DAC IRQ is wrong.
After working under DOSBOX, I checked on the real hardware it did not work....
I spent 4 hours to realize it was a variable used for the SB1.0 (Not more supported) that was causing the problem, I needed to remove it or delete MODMXT.CFG...

Reply 577 of 660, by MobyGamer

User metadata
Rank Member
Rank
Member

I wondered about the 2K buffer thing, because when I was looking at the docs I couldn't see that as a limit. The BIOS extensions for the DAC don't have that limit either, so glad that we were both mistaken on that. But it is a shame it does IRQ only at the end of a buffer, so we have to resort to SB 1.0 player tricks which are ugly.

Is there a need to go higher than 48 KHz? Even 8-bit, that frequency response (including the distortion added by 8-bit) is already above the range of human hearing...

Reply 578 of 660, by Cloudschatze

User metadata
Rank Oldbie
Rank
Oldbie
MobyGamer wrote on 2020-10-08, 01:34:

Is there a need to go higher than 48 KHz? Even 8-bit, that frequency response (including the distortion added by 8-bit) is already above the range of human hearing...

Not to mention the 5kHz roll-off and ~11kHz cutoff filtering affecting the output of most PSSJ implementations.

Reply 579 of 660, by root42

User metadata
Rank Oldbie
Rank
Oldbie
MobyGamer wrote on 2020-10-08, 01:34:

Is there a need to go higher than 48 KHz? Even 8-bit, that frequency response (including the distortion added by 8-bit) is already above the range of human hearing...

We need to make an experiment to figure that out, if anyone can hear the difference. But although the sampling theorem says that with 48kHz you can reproduce up to 24kHz signals, that is of course only true only to an extent. You obviously gain resolution in reproducing the waveform when having a higher sampling rate. Granted, having this with only 8 bit resolution is probably overkill. But yeah, I propose Freddy and whoever has a Tandy with DAC make an experiment! 😀

YouTube and Bonus
80486DX@33 MHz, 16 MiB RAM, Tseng ET4000 1 MiB, SnarkBarker & GUSar Lite, PC MIDI Card+X2+SC55+MT32, OSSC