Viewing Issue Advanced Details
ID Category [?] Severity [?] Reproducibility Date Submitted Last Update
05262 Sound Minor Always Aug 11, 2013, 20:27 Aug 13, 2013, 15:32
Tester Ace View Status Public Platform MAME (Self-compiled)
Assigned To Resolution Open OS Windows Vista/7/8 (64-bit)
Status [?] Acknowledged Driver
Version 0.149 Fixed in Version Build
Fixed in Git Commit Github Pull Request #
Summary 05262: All sets using SegaPCM: [possible] SegaPCM emulation issues
Description I've been looking through some videos of a few Sega games using the SegaPCM chip, more specifically Galaxy Force II and After Burner II, and noticed the SegaPCM audio is noticeably different from what it sounds like in MAME. I have two videos here taken from the original boards of each of these games (not my videos):

Galaxy Force II (audio is a bit distorted in this video):



After Burner II (there are times where the audio gets severely garbled - also, be sure to only click the Play button once otherwise you will be redirected to NicoNico): http://nicoviewer.net/sm20566807

However, I'm not sure whether the issue has to do with filtering or inaccuracies in the SegaPCM emulation, hence why I noted this bug as "possible." I have found schematics of Galaxy Force here (After Burner II uses the same audio setup, so the audio section of the schematic also applies to After Burner II. I did find a schematic for that game, but the scan is of really poor quality and the audio circuit is illegible): http://tamdb.net/G/Galaxy%20Force%20Schematics.pdf

Page 9 shows the audio circuit on the right side of the page. Looking at the audio circuit, there's an active low-pass filter with a cutoff frequency of 32883.25Hz followed by a passive high-pass filter with a cutoff frequency of 0.07Hz. That does not alter the audio in any perceivable way.

Then, there's the After Burner owner's manual: http://dweeb.net/afterburner/owners_manual/afterburner_owners_manual.pdf

Page 25 has a wiring diagram of all the major electronic components in the cabinet, and there's an extra amplifier board labeled 839-0035-01/PWR AMP BD between the audio out of the X-Board and the speakers. Other Sega games have this board as well like OutRun (not sure about Galaxy Force).

Going by these, the audio differences here could be either caused by that board or emulation issues with the SegaPCM chip (I don't have any schematics of the 839-0035-01, so I don't know if it applies any extra filtering or not). If the audio differences do turn out to be caused by filtering on the 839-0035-01 board, please disregard this bug report.

EDIT: Apparently, I missed something in the audio circuit. There are two MF6CN-50 chips (IC114 and IC118 respectively) after the AD7520 DAC hooked up to the Sega PCM chip and those are 6th-order switched Butterworth low-pass filters. Okay, my mistake, I thought there was actually something wrong with the Sega PCM emulation, but it's not the source of the problem. Please disregard.
Steps To Reproduce
Additional Information As I don't have access to any games using the SegaPCM chip (yet, I will have to speak to the owner of a local arcade shop about getting any X-Board, Y-Board, etc. games for a few days to verify this), an audio sample taken straight from the SegaPCM chip before being mixed with the YM2151 would be of great help.
Github Commit
Flags Possible
Regression Version
Affected Sets / Systems All sets using SegaPCM
Attached Files
 
Relationships
There are no relationship linked to this issue.
Notes
2
User avatar
No.09731
Fujix
Administrator
Aug 13, 2013, 13:11
SegaPCM games have their PCM sound data in a unsigned 8-bit format.
For example in After Burner 2,

epr-11102.13
mpr-10930.12
mpr-10931.11

have raw PCM data. Please open them with a proper wave editor such like Cool Edit 2000, you can play PCM sounds as they are.
So if you notice the difference between MAME and arcade, I believe it is by the amp board.
User avatar
No.09733
Ace
Tester
Aug 13, 2013, 15:32
edited on: Aug 13, 2013, 15:33
I did just that with Audacity and the raw PCM data is a lot more muffled than what MAME outputs. Maybe even the raw audio out of the AD7520 DAC sounds the same as in MAME, but before jumping to this conclusion, I will see if I can get a game with the SegaPCM chip and get a sample from before the MF6CN-50 chips. Although I'm fairly certain that's what makes the audio sound so much different (that would be an issue to resolve for another time), I have my doubts as to whether the raw audio output from the SegaPCM chip is correct because even when trying to simulate a 6th-order low-pass filter, the audio still sounded a bit different than the original boards. So I'll see what I can do about obtaining any games with the SegaPCM chip (hopefully After Burner II) and getting audio off the AD7520 or if someone else has an After Burner II board on-hand before I can obtain one, they can do the same.