Different character of sound of N106 in Rolling Thunder.

This is an archive of a topic from NESdev BBS, taken in mid-October 2019 before a server upgrade.
View original topic
Different character of sound of N106 in Rolling Thunder.
by on (#147619)
Why? Interesting to know. In other games with N106 such did not hear.
https://yadi.sk/d/GFz30J7HgfzX5
Recording from emulator (Nestopia), N106 sounds only.
Re: Different character of sound of N106 in Rolling Thunder.
by on (#147620)
Nestopia's emulation of the Namco 163's wavetable sample length was wrong until quite recently: https://github.com/rdanbrook/nestopia/issues/118
Rolling Thunder is one of the games that actually uses long enough samples for this to be audible: viewtopic.php?t=12326
Re: Different character of sound of N106 in Rolling Thunder.
by on (#147621)
Thanks! Already I read.
P.S. Somehow I heard, that in N163 there are samples in length more than 32. And Nestopia, means, cannot simply play such samples...
Re: Different character of sound of N106 in Rolling Thunder.
by on (#147631)
Yes, you'll have to use a newer release of Nestopia UE in order for longer samples on the 163 to work correctly. (Anything newer than 2015-02-09, when the change was committed)
Re: Different character of sound of N106 in Rolling Thunder.
by on (#147640)
I can't find a newer Nestopia with fixed bug of N163...
Re: Different character of sound of N106 in Rolling Thunder.
by on (#147647)
Looks like you'll have to make it / get someone to make it for you... (see viewtopic.php?p=145550#p145550 )
Re: Different character of sound of N106 in Rolling Thunder.
by on (#148818)
I implemented N106 sample length properly in my emulator (HalfNES) from the beginning but had to add heuristics to trim samples down to 32 bytes maximum when the remaining length was all zeroes, to cater to the expectations of certain NSFs by Robokabuto. Not sure I should have done that but someone did complain and it wasn't like Robokabuto would issue a fixed version.
Re: Different character of sound of N106 in Rolling Thunder.
by on (#148819)
"when the remaining length was all zeroes" ?

You could fix the NSFs yourself, FWIW.
Re: Different character of sound of N106 in Rolling Thunder.
by on (#148820)
I mean that if the sample length is set to more than than 32 but the program hasn't written anything into at least 3 consecutive bytes of that sample area, then I shorten the length.

Also, interestingly enough, someone else already fixed the NSFs in question and posted them right here on Nesdev six months before I ever encountered the issue and I never knew about it until just now, because the versions in the Famicompo archive were never fixed.

viewtopic.php?f=6&t=9268&start=105#p115201

(That Dropbox link is dead too now.)
Re: Different character of sound of N106 in Rolling Thunder.
by on (#148821)
I found this lying on my hard drive, with the right download date:
Attachment:
robokabuto.rar [355.95 KiB]
Downloaded 126 times
(tell me if I should remove it)
Re: Different character of sound of N106 in Rolling Thunder.
by on (#148822)
Well I did go ahead and patch the NSFs I had to deal with this anyway. There's one track that wasn't in the other set, and all of them have different metadata because they were from the Famicompo archives.

Now if only I could get the fixed versions back in those archives so this never comes up again.
Re: Different character of sound of N106 in Rolling Thunder.
by on (#151561)
Whether exist a player which correctly playback a wavetable sample in length more than 32?
Re: Different character of sound of N106 in Rolling Thunder.
by on (#151569)
At least NSFPlay does.
Re: Different character of sound of N106 in Rolling Thunder.
by on (#151601)
Thank you!
P.S. In Final Lap samples too are longer, than 32. Sound on NSFPlay and in the emulator different.