More NSF Requests

Discuss NSF files, FamiTracker, MML tools, or anything else related to NES music.

Moderator: Moderators

User avatar
MrNorbert1994
Posts: 987
Joined: Tue May 06, 2014 1:50 pm
Location: Szeged, Hungary
Contact:

Re: More NSF Requests

Post by MrNorbert1994 »

Reripped Bayou Billy with a new lookuptable.
Attachments
Mad City (SFX).nsf
(31.3 KiB) Downloaded 178 times
The Adventures Of Bayou Billy (NTSC) (SFX).nsf
(29.97 KiB) Downloaded 181 times
The Adventures Of Bayou Billy (PAL) (SFX).nsf
(29.97 KiB) Downloaded 178 times
NewRisingSun
Posts: 1510
Joined: Thu May 19, 2005 11:30 am

Re: More NSF Requests

Post by NewRisingSun »

rainwarrior wrote:
BlackGuyRX wrote:NSFPlay ... still glitchy with NSFE track names.
Here is a beta version where that problem is fixed:
https://www.dropbox.com/s/xh09cae6g9myg ... 3.zip?dl=0
Playlists themselves do not seem to be followed even in that beta, case in point attached.
Attachments
Ai Senshi Nicol.nsfe
(10.02 KiB) Downloaded 176 times
User avatar
rainwarrior
Posts: 8732
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: More NSF Requests

Post by rainwarrior »

NewRisingSun wrote:Playlists themselves do not seem to be followed even in that beta, case in point attached.
The NSFe file you attached has no 'plst' chunk, so it does not have a playlist.

There is also an option to ignore NSFe playlists if desired (in the General tab of Settings), which should be off by default but if you aren't getting valid NSFe playlists to work that could be turned on.
NewRisingSun
Posts: 1510
Joined: Thu May 19, 2005 11:30 am

Re: More NSF Requests

Post by NewRisingSun »

Then why does every other NSFE player play Stage 1's music at the track labelled "Stage 1", while NSFPlay plays a sound effect labelled "Stage 1"? The option you mentioned unfortunately does not change that.

Edit: Extracting the NSF data out of the NSFE file produces a NSF file that still plays the wrong songs in NSFPlay. Since every other emulator I have tried plays that NSF correctly (and I could try more than I could try the NSFE), I don't think it's the NSF. But you're right that it's not an NSFE-specific problem.
Attachments
Ai Senshi Nicol from NSFE.nsf
(9.92 KiB) Downloaded 177 times
User avatar
MrNorbert1994
Posts: 987
Joined: Tue May 06, 2014 1:50 pm
Location: Szeged, Hungary
Contact:

Re: More NSF Requests

Post by MrNorbert1994 »

That's a broken rip, what you are experimenting with.
Attachments
Ai Senshi Nicol (SFX).nsf
(20.13 KiB) Downloaded 183 times
NewRisingSun
Posts: 1510
Joined: Thu May 19, 2005 11:30 am

Re: More NSF Requests

Post by NewRisingSun »

Still, it would be good to understand in what way it is broken that it only fails in one player.
User avatar
rainwarrior
Posts: 8732
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: More NSF Requests

Post by rainwarrior »

MrNorbert1994 wrote:That's a broken rip, what you are experimenting with.
Ah, yes I forgot this was one of those problem FDS rips.

Thread here: https://forums.nesdev.com/viewtopic.php?p=98299#p98299

Basically the problem is that the rip was banking the same pages into different locations, which caused mirrored writes to corrupt the program. It was relying on a "copy" semantic for banking that was impossible on the real FDS.

You can actually play the bad rip by turning on the "write protect" option on the FDS tab of Settings (though this breaks other FDS NSFs), but the better idea would be to fix the NSFe; I believe all it requires is some new numbers in the BANK section, but I'm going by memory here. (Or you could just take the working NSF and add the track names yourself in NotSoFatso's editor.)
NewRisingSun wrote:Still, it would be good to understand in what way it is broken that it only fails in one player.
It fails on PowerPak as well. (Hardware players are incapable of doing "copy" banking; it's really a kludge that players like FCEUX do this, IMO.)

What are the other NSFe players? I only know of NotSoFatso.
NewRisingSun
Posts: 1510
Joined: Thu May 19, 2005 11:30 am

Re: More NSF Requests

Post by NewRisingSun »

I meant the extracted NSF. But the problem is solved with the FDS copy versus banking explanation.

It's funny: I asked about that issue about twelve years ago, and the answer I got suggested the "copy" method being correct. (The actual Yuushi no Monshou rip I ended up producing uses no bankswitching whatsoever, since I rearranged the data, so in the end it didn't matter to me. Also, the emulator behavior of today's versions does not match what I described about then contemporaneous emulator versions. I didn't create that Ai Senshi Nicol rip, to be clear. ;))
User avatar
Karmic
Posts: 308
Joined: Sun Jun 05, 2016 2:10 am
Location: Canada
Contact:

Re: More NSF Requests

Post by Karmic »

Fixed noise error in Deadly Towers.
Attachments
Deadly Towers.nsf
(8.26 KiB) Downloaded 179 times
User avatar
Karmic
Posts: 308
Joined: Sun Jun 05, 2016 2:10 am
Location: Canada
Contact:

Re: More NSF Requests

Post by Karmic »

Added missing stuff to Monster Party.
Attachments
Monster Party.nsf
(21.08 KiB) Downloaded 172 times
User avatar
MrNorbert1994
Posts: 987
Joined: Tue May 06, 2014 1:50 pm
Location: Szeged, Hungary
Contact:

Re: More NSF Requests

Post by MrNorbert1994 »

Monster Party US track arranged.
Boy, this was a nigthmare.
Attachments
Monster Party (SFX).nsf
(21.08 KiB) Downloaded 171 times
NewRisingSun
Posts: 1510
Joined: Thu May 19, 2005 11:30 am

Re: More NSF Requests

Post by NewRisingSun »

By the way, MrNorbert1994, your NSF archive contains two modified versions of my Yuushi no Monshou rip from 2006. Apart from updating the composer credits (which is fine), both add an 18th song, which is just a repeat of song 6, and one version moves the entire rip from $6000 to $8000. What's up with that? $6000-$7FFF should be perfectly valid memory for FDS rips.
User avatar
MrNorbert1994
Posts: 987
Joined: Tue May 06, 2014 1:50 pm
Location: Szeged, Hungary
Contact:

Re: More NSF Requests

Post by MrNorbert1994 »

Ah... those are just test modifications when I had pretty much no experience with NSF ripping, and started to learn about the registers.

You know, I use 1 player to test almost all of my rips which is located here:

http://user.tninet.se/~zxy965r/nsf.zip

This player has a full memory viewer feature which is VERY useful.
A lot of FDS games does not play in this player, and I was curios what causes this problem.

Since then I learned that FDS games loads from $6000, and that a lot of players don't really like when a game is loaded below than $8000.

Even my "new rip" is broken as all hell, since only the pulse channels seems to play.
I might try to rerip it in the future, plus to have all the SFXs as well in it.
NewRisingSun
Posts: 1510
Joined: Thu May 19, 2005 11:30 am

Re: More NSF Requests

Post by NewRisingSun »

What about the additional song?
Even my "new rip" is broken as all hell, since only the pulse channels seems to play.
I think that's a problem of that particular player, not of a broken rip. While I agree that the memory usage window of NSFLive seems neat, it doesn't seem useful to me for optimizing rips. I usually optimize by setting read memory breakpoints in FCEUltra, then going through all songs while holding down the fast forward hotkey, modifying the minimum and maximum addresses until no breakpoint is invoked anymore.
I might try to rerip it in the future, plus to have all the SFXs as well in it.
That game was a nightmare to rip, and no-one should have to go through that again. I think I'll just look into adding sound effects myself and update my own rip that way.
User avatar
Karmic
Posts: 308
Joined: Sun Jun 05, 2016 2:10 am
Location: Canada
Contact:

Re: More NSF Requests

Post by Karmic »

NewRisingSun wrote:With all due respect, I think that's a problem of that particular player, not of a broken rip.
Yes, I agree. I've talked to Norbert outside of this board, and unfortunately he does not abandon NSF Live purely due to some of the features it offers, despite it having some major problems following the NSF spec.

It's kind of insulting to have a rip called "broken" just because it doesn't work in an inaccurate player.
Post Reply