Viewing Issue Advanced Details
ID Category [?] Severity [?] Reproducibility Date Submitted Last Update
03919 Sound Minor Always Jul 3, 2010, 18:38 Jan 3, 2022, 14:50
Tester Ace View Status Public Platform MAMEUI
Assigned To Resolution Open OS
Status [?] Confirmed Driver
Version 0.138u2 Fixed in Version Build
Fixed in Git Commit Github Pull Request #
Summary 03919: salamand & clones, other Konami sets with VLM5030: VLM5030 sound is unfiltered
Description There appears to be a filter of some kind applied to sound coming out of the VLM5030 in Konami games that is missing in MAME. A particular example is Salamander and Life Force's speech. I've attached 2 sound clips each containing 2 voice samples, one from MAME and one from GameStone's recording of his Salamander arcade board.
Steps To Reproduce
Additional Information
Github Commit
Flags Verified with Original
Regression Version
Affected Sets / Systems salamand & clones, other Konami sets with VLM5030
Attached Files
ogg file icon Salamander_voice_samples_MAME.ogg (32,707 bytes) Jul 3, 2010, 18:38
ogg file icon Salamander_voice_samples_original_PCB.ogg (26,392 bytes) Jul 3, 2010, 18:38
related to 03538ResolvedFujix  salamand, lifefrce, lifefrcej, salamandj: Music is too quiet in Salamander and Lifeforce 
User avatar
Jul 3, 2010, 21:27
I'm pretty sure the fact that audio isn't perfect in these Konami games has been made abundantly clear with your reports over the last few months. When will it be fixed.. god only knows. I hate to sound snippy or upset, but I don't see the point in completely nitpicking stuff which is already known to need work.
User avatar
Jul 4, 2010, 02:42
Sorry. I just can't stand inaccuracies in emulators and feel the need to nitpick every little flaw I find.

Although I can propose some fixes for some of the Konami games that have sound issues(not this particular one, sadly), but I don't know how accurate my fixes are to the original sound chips and sound processors(and the original arcade board, for that matter). Since anything that's not 100% exactly like the original hardware appears to be frowned upon on MAME Testers, I doubt these fixes will be applied in future MAME releases, but I don't think it would hurt to give it a shot.
User avatar
Jul 4, 2010, 03:03
I can understand the devotion. Apologizes again if it came across too harshly - I don't want to discourage anyone from posting legitimate bugs related to emulation. 'Filtering' is a thing that most of the games in MAME are missing and are most commonly discrete components which simply haven't been implemented. It's just that if we file bugs for every single game which does not have his hooked up, we'd double our bug total overnight. That is where my irritation was coming from. It's my problem, I'll deal with it.
User avatar
Jul 4, 2010, 08:02
edited on: Jul 4, 2010, 08:06
Probably it is not a simple problem whether it is filtered or not.

According to the chip driver author, the current VLM5030 was made by brute force I/O signal comparison of the real chip, although, it is still a guess work and imperfect largely.

Regarding VLM5030, ultimately we have to dump the internal chip program for perfect emulation. Maybe we need to add an imperfect flag to VLM5030.
User avatar
Jul 6, 2010, 04:08
Documentation on it was horrible:
User avatar
Jul 23, 2017, 16:10
User avatar
Senior Tester
Jan 3, 2022, 14:50
quote from Ace:

Bugs 03738 and 03919 deal with emulation issues pertaining to the VLM5030 and are completely unrelated to the Yamaha FM timers. 03919, in particular, is due to emulation inaccuracies with the VLM5030 in general, which I would suggest to rebuild based on this gate-level VHDL model made by Arnim Laeuger of FPGAArcade: