HCS Forum - What's going on in other xsf formats

by marioman at 8:58 AM EDT on July 19, 2007: @UF: Just to let you know, there are currently two Genesis music formats and neither of them support the Pulse Width Modulation channel for the Sega 32X. If you are going to make a new format for the Genesis, then PLEASE make sure that you support PWM.

Additionally, have you consulted CS2 about your 2SF format? I am sure that with CS2's GSF experience, he should be able to figure out 2SF.
by unknownfile at 9:44 AM EDT on July 19, 2007: First: Yes, it will support the 32X, but only if someone rips 32X games.

Second: Nah, I doubt he cares.
by marioman at 1:17 PM EDT on July 19, 2007: Nevermind, I got the SNSFs to work.

Is there any chance that a Winamp plugin for SNSFs will be available soon? Thanks.

edited 1:21 PM EDT July 19, 2007
by unknownfile at 7:38 PM EDT on July 20, 2007: In other news, Sonic the Hedgehog is pretty much ripped, I can say that I can't get it patched down below 48 KB... Expect a modified Gens soon.
by Richter X at 11:47 PM EDT on July 20, 2007: Oooh! Does this mean I can finally rip the music to SegaCD Snatcher?
by unknownfile at 12:35 AM EDT on July 21, 2007: if you know how, yes

by the way, i will not host any mp3 rips in the mdsf format for the sega cd so don't try it
by agu fungus at 8:17 PM EDT on July 21, 2007: So, is there a way to rip the SNSFs out of the game rom, like an automated ripper?
by marioman at 9:10 PM EDT on July 21, 2007: I don't know, but I would think that it would be after CS2 makes a SNSF plugin for Winamp. (And after he rips Wario's Woods. :P )
by Richter X at 9:34 PM EDT on July 21, 2007: It'll be nice to finally hear the music from the SNES Lord of the Rings.
by unknownfile at 12:29 AM EDT on July 22, 2007: bah, rip it yourself
by Knurek at 4:40 AM EDT on July 22, 2007: >So, is there a way to rip the SNSFs out of the
>game rom, like an automated ripper?

Chances are about as high as automated NSF or GBS ripper (ie very slim).
warning, rant ahead by unknownfile at 10:21 AM EDT on July 22, 2007: Why is it that people can't take the time to look at a game's code and figure out how it works? Patience is one issue, but you're making it seem like it's impossible.

For one thing, ripping *sfs is an educational experience in that it teaches you how a system works, not to mention all the assembly operands that you need in order to work on rips. What knowledge does VGM and SPC give you? Just click on a command, wait a bit, then stop logging, and then maybe loop it later. Automation takes away from the fun and the challenge that manual ripping gives.

Another thing that pisses me off is that people are willing to rip something but they don't even bother trying. A good example of this case is Firedivine, who constantly wants to rip something, but never ends up doing it. Maybe if he tried, he'd get somewhere.

So if you're willing to learn how to rip something and have a nice feeling inside after you've conquered the game, I suggest you give it a try. Otherwise, you can keep living in your world where you think ripping can be done with a click of a mouse.

</rant>
by Knurek at 10:43 AM EDT on July 22, 2007: >Why is it that people can't take the time to
>look at a game's code and figure out how it
>works? Patience is one issue, but you're making
>it seem like it's impossible.

Bad tutorials for most part.
Listen, I've done loads of SID rips. Including some 'semi hard to do, rewrite the playroutine, multispeed' rips.
And I can't really understand even the simplest rip example from that one NSF ripping tutorial.
Don't get me started on GBS rips, there's no imformation anywhere.
Really, if you want to do SID rips, there's ton's of (step by step!) tutorials, there are helpful people from HVSC, there are tools for checking/optimizing rips. If you want NSF/GBS/GSF/etc you're required some arcane knowledge no one's willing to teach you.
by marioman at 1:30 PM EDT on July 22, 2007: Another thing it a lack of free time. As is the situation with me. Automation makes the process faster and therefore possible for people who just don't have as much free time as you do.
by unknownfile at 10:03 PM EDT on July 22, 2007: ripping goes something like, "look for writes to insert sound register here"
by kingshriek at 10:14 PM EDT on July 22, 2007: The problem with step-by-step tutorials is that they can really only target a specific sound driver, and in the case of NSF/GBS where you run across hundreds of different drivers, the ripper needs to possess more general knowledge, which can be found in various tech docs, on how the system works. And really, other than general system-specific knowledge (memory map, memory management, interrupts, etc.) and assembly language experience, the only other things a ripper really needs to know is the specs of the format (sometimes this information isn't really clear - with HES I had to look in the source code of NEZplug to answer some questions I had) and what tools are available that can help out (any emulator with a built-in debugger is a godsend)

As for automation, one case where I was hoping for a way of automatic ripping is PC-Engine CD games, since most games use the System Card supplied PSG driver. Unfortunately, there is no standard way in which the music data is stored. To initialize the sound routines, you basically have to pass in a bunch of pointers to various data (track data, waveform data, envelope data, etc.), and these initializations are pretty much different for each game (some even use indirect addressing modes to do this). Another problem is actually isolating the music data out of the games, as there really isn't anything you can reliably search for.

edited 10:16 PM EDT July 22, 2007
by unknownfile at 10:47 PM EDT on July 22, 2007: In other news, MDSF page up here

Yep, expansion sound's supported.
by Knurek at 1:08 AM EDT on July 23, 2007: Guess I'll stick with PC-98 music ripping.
With the speed the Hoot guys are releasing music driver patches I recon we'll have every game ripped this year. :)
by marioman at 11:48 AM EDT on July 23, 2007: I have found a bug in the latest version of Highly Advanced. If you play a GSF and have more than one GSF into the Winamp playlist, the GSF entries in the playlist window flash repeatedly as if their tags are being updated multiple times per second. If a number of GSF files are visible in the playlist, it can cause the music to skip.

Any idea what's going on with this?
by unknownfile at 12:15 PM EDT on July 23, 2007: I don't know what causes this to happen, but there are multiple bugs in HA that I know of:

- Equalizer settings don't work in Winamp.
- There's a major bug in XMplay that causes the thing to slow the entire machine down if a song is played more than once.
- Timers are not updated correctly in XMplay.
by marioman at 10:06 PM EDT on August 2, 2007: Highly Competitive v0.1 (CS2's SNSF plugin for Winamp) is available at the SNSF page. This just makes me want the Wario's Woods SNSF set even more.

Good job CS2.

--EDIT--

He has also put up what looks like the source to the plugin.

edited 10:15 PM EDT August 2, 2007
by Richter X at 11:20 PM EDT on August 2, 2007: in_snsf is hellishly choppy-sounding. Then again, it's a first release. :)
by Lunar at 5:53 AM EDT on August 3, 2007: Cool, I'm checking out the Chrono Trigger and Tales of Phantasia sets right now. You weren't kidding about choppy. Still, I'm interested in seeing how this progresses. It'll be useful for all those sets that weren't rippable before.

music-0002.minisnsf - vocals :D

I don't see this as a potential replacement for SPC.
by marioman at 9:14 AM EDT on August 3, 2007: Yet it is choppy, but it is after all v0.1. I do not see it as a replacement for SPC either, but rather a complementary format that will fill the shortcomings of SPC.

I think that the best way for CS2 to utilize SNSF is to take snesmusic.org's list of undumpable games and rip them first. It seems to be a little redundant to rip games that already have tagged/timed SPC soundtracks.

Nonetheless, the format is shaping up very well and I am excited to see a real release of in_snsf.

edited 9:15 AM EDT August 3, 2007
by unknownfile at 9:44 AM EDT on August 3, 2007: It could be a replacement for SPC as SPCs usually have redundant data which could be stored in a ROM... but that's just my guess.
by agu fungus at 11:10 AM EDT on August 3, 2007: Can you give me the link to the snsf page?
by marioman at 11:15 AM EDT on August 3, 2007: Sure.
by agu fungus at 11:28 AM EDT on August 3, 2007: Thank you. Now if I could understand how to rip the music out of the games.

Edit: Another set needs to be reripped into this format: Clayfighter (reason: the Clayfighter song cannot be ripped to its fullest in SPC)

edited 11:52 AM EDT August 3, 2007
by marioman at 8:51 AM EDT on August 7, 2007: in_snsf has been updated, and it appears that CS2 has fixed the issue with the music being choppy. Go get it.

By the way, when will we be seeing some MDSF music/players? Just wondering.
by unknownfile at 9:10 AM EDT on August 7, 2007: when I feel like it
by marioman at 5:37 PM EDT on August 7, 2007: The Lost Vikings SNSF set has been tagged/timed.
by Lunar at 8:02 AM EDT on August 9, 2007: Glad to see the plugin is at least usable now :D now to wait patiently for a Wario's Woods rip!
by marioman at 9:40 AM EDT on August 9, 2007: now to wait patiently for a Wario's Woods rip!

Seconded. I can't even find a MP3 rip of that game.

Before snesmusic.org took undumpable games off of their requests list, it was one of the top 15 requested SPC sets. Hopefully CS2 will rip it soon.
by ugetab at 2:55 AM EDT on August 10, 2007: With Wario's Woods, I've gotten as far as reducing the compressed rip to 95k, and making the music play fairly accurately. All of my attempts to make this thing run as an snsf have failed though. The ROM version runs fine, but after being converted, it gives me a lot of trouble.

Here's the rom, and something for other people to mess with:
http://www.angelfire.com/nc/ugetab/WariosWoods_SNSF.zip

Use the code 80FF54?? to select alternate themes.

---

I got the SNSF working(HiROM=00FF, LoROM=007F), but now it plays too fast in the winamp plugin, but not in the custom snes9x build.

It must have something to do with the vblank timing for NMIs, because that's the only thing that's too severely different about this rip that I can think up.

edited 5:03 AM EDT August 10, 2007
by marioman at 9:36 AM EDT on August 10, 2007: Thank you ugetab! If you figure anything out about how to fix the speed of the music when it plays in Winamp, please let us know. You have made myself and quite a few other people very happy. Thanks.
by Lunar at 10:46 AM EDT on August 10, 2007: ugetab does it again. genius! hopefully in time you'll figure out the speed issues.
by unknownfile at 11:31 AM EDT on August 10, 2007: this is uf here, i request you do the dkc series kthxbye.
by ugetab at 2:15 PM EDT on August 10, 2007: The winamp plugin needs fixing, not the set.

As for the DKC series, that was made by RARE. After ripping a few of their games, I've learned to make a check-list of ways not to program a game, and check off which ones they didn't do that time. I'll give it a shot, but no way will I promise anything with a game RARE had a hand in. Every past RARE game I've done has had a catch when ripping, something just big enough to stub your toe on if you're not looking. I can't imagine them avoiding this tendancy on the SNES when they've been at it on the NES(3 music init/play routines? A pause theme disintegrator? Direct PCM?), Gameboy(they perfected the art in the later games, like perfect dark), and N64(Split-apart tracks, anyone?)
by marioman at 3:27 PM EDT on August 10, 2007: The winamp plugin needs fixing, not the set.

OK. Thank you so much for ripping the set.

Hopefully CS2 will fix the plugin soon. Maybe someone who can easily contact CS2 (like UF or Mouser) can get in touch with him and alert him of this problem.

Thanks again.
by Lunar at 3:44 PM EDT on August 10, 2007: ^^what marioman said
by ugetab at 5:18 PM EDT on August 10, 2007: On the subject of SNSFs, consider looping the main CPU with CB80FD(WAI, BRA to WAI), instead of just 80FE (Infinite Loop). I think that the sets could be more stable with this bit of coding forcing the CPU to wait for the next interrupt to occur, instead of wasting time going through an uninterruptable loop. Tests with the SNES9X debugging version confirm a lower amount of CPU being used by about doubling the turbo speed of the music in Wario's Woods, with the only difference being the use of a NOP in the infinite loop vs the use of a WAI command. The results of this command are even more dramatic when used on a game that has interrupts disabled, like Chrono Trigger. In that game, the main CPU never recovers from the command. The sets must sometimes crash winamp for other people too, and this change may make the player more stable. I can only guess that it's related to the level of activity on the emulated CPUs.
by marioman at 6:29 PM EDT on August 10, 2007: In_snsf crashes for me all the time. The most notable/reproducible occurrences are as follows:

- Seeking using the right arrow key.
- Loading a SNSF file in the Winamp Playlist, restarting Winamp, and then double-clicking the entry in the playlist.

There are other crashing/CPU usage problems, but I haven't nailed them down yet. Overall it is very unstable (not as much as the first version though), but I am just glad that it is available. However, if you can make any improvements it would be appreciated.
by ugetab at 7:30 PM EDT on August 10, 2007: Reduced CPU usage seems to generally increase the stability. I managed to rather quickly scan about 25 tracks with the WAI fix in place in a set, where I'm usually lucky to get 8 before a crash
by marioman at 7:38 PM EDT on August 10, 2007: Great! If get around to it, could you please post a bugfix version of in_snsf? It would at least make my life a little easier. Thanks in advance.
by ugetab at 9:59 PM EDT on August 10, 2007: This isn't a fix for in_snsf, this is a programming trick that makes sets use less CPU power. I don't know enough about emulator design to start changing the plugin, but I'm familiar with some *SF format storage schemes, and I believe I'm adequately familiar with SNES programming to find some tricks and info for making or optimizing some sets.

The trick is to turn off the main SNES CPU when it's not in use using the WAI command, which essentially puts the main CPU into a comatose loop instead of an active one, while the sound CPU continues. It's a bit technical, but I've tested the effect, and it definately reduces the amount of processing required using only 1 instruction put into the right place just before the main CPU is no longer required.

Caitsith is still the plugin developer, but a source for simple solutions to SNSF ripping doesn't exist yet, there's no standard decompressor for the sets to help you learn from them(I built my own power tool for that purpose), and the WAI solution would be better put into place with 9 sets around than when there's 50. I may update the documents with some notes, and release a friendlier compression/decompression tool for the snsflibs.
by marioman at 10:43 PM EDT on August 10, 2007: I seem to have been a little confused, but I understand now. Sorry.

This info should be relayed to CS2 before any more SNSF rips are made. I am not sure of the best way to reach him, but there is an E-mail address for him in some of his GSF rips. He may also be available on IRC, but I am not sure.

Thank you for all of your work on this. It will definitely help the format get established faster.
by ugetab at 1:42 AM EDT on August 11, 2007: I'm releasing some minor info about the SNSF format, along with a compression/decompression program. This should also be something you could modify to work with other zlib compressed formats. Right now, only SNSF is structurally supported.

http://www.angelfire.com/nc/ugetab/SNSF_ZLIB.zip
by unknownfile at 5:14 PM EDT on August 31, 2007: Bumping to see how the RSS feed reacts
by Josh W at 11:49 PM EDT on August 31, 2007: ive had an RSS feed for ages:

http://joshw.hcs64.com/rss.php
by unknownfile at 12:45 AM EDT on September 1, 2007: lol wrong thread
by Josh W at 3:55 PM EDT on September 1, 2007: :p
by F3582 at 6:57 AM EDT on September 4, 2007: In other news: I recently read Oddigy saying something about the music in Tri-Ace PS2 games being un-ripable. Does anyone know details?
by Mouser X at 8:09 AM EDT on September 4, 2007: All PS2 games that use a custom driver are unrippable because to rip them requires tools that only Neill and Zoopd have. At least, they're the only 2 that have ever ripped custom drivers off the PS2. I know that Daivironica (however you spell that) has the song data for Radiata Stories, and I think for Star Ocean 3 as well. However, s/he doesn't have the drivers to play them. So, if I understand correctly, they're unrippable because no one currently has the ability (tools [specifically, I think you need a PS2 developer kit, since you need to rewrite/recompile some stuff, but I'm not sure on that] and such) necessary to rip them.

If I'm mistaken, then hopefully someone else will be able to clarify. Mouser X over and out.
by Lunar at 7:16 AM EDT on September 6, 2007: There was a load of PSF rips added to Lawrence Lin's mirror recently. Mostly WIPs.

I'm hoping Airgrave and Battle Bugs PSF rips will appear sometime down the line.
by marioman at 7:36 AM EDT on September 6, 2007: Wow! I had no idea that there were so many PSFs that have been ripped recently. There are quite a few PSF sets there that I have been looking for. Thanks for pointing that out.
by Lunar at 8:09 AM EDT on September 6, 2007: Yeah this is what I mean, it's so hard to keep track of these ripped formats :P I don't even know if there's a place where discussion about PSF ripping goes on (actual "I'm working on a rip of (X) right now" etc.) The ripped music scene is very convoluted indeed.

Lawrence Lin's mirror seems to be the most comprehensive and regularly updated anyway, so it's a good place to keep bookmarked.
by Knurek at 9:31 AM EDT on September 6, 2007: As most of recent PSF rips are done by japanese 2CHN users, you might try this link:

http://pc11.2ch.net/test/read.cgi/dtm/1174838154/l50
by Lunar at 2:37 PM EDT on September 6, 2007: Ooh, that's kinda cool. Damn, I wonder if stuff has been ripped and I just don't know about it. Though then... I assume that many rips appearing on Lawrence Lin come from 2ch anyway. Good to have, nonetheless~
by Knurek at 5:10 PM EDT on September 6, 2007: Well, they have recollection packages available on Share I guess.
http://pc11.2ch.net/test/read.cgi/dtm/1174838154/66

About 2.4 GB in all, can't say how much is available elsewhere, as I can't get Share to work here.
by snakemeat at 11:23 AM EDT on September 7, 2007: @Knurek - Thanks Knurek, I didn't know anything about that Jp stuff until these two posts. Now I know where to find links for 2ch and have learned to use Share (something I never knew existed). Even better, are the thousands of MDX files I found with it! In any case, I'm not sure what clusters to use, but I couldn't get those PSF recollection packages to even show up as search results.

@Lunar - I grabbed all the PSFs I could from the 2ch posts, and from what I saw, they were all on LLin's mirror.
by Knurek at 11:27 AM EDT on September 7, 2007: /me want those MDX files (heck, everything you get from share). Rapidshare or similar, if you may?

edited 11:27 AM EDT September 7, 2007
by snakemeat at 1:26 PM EDT on September 7, 2007: No problem, I'll upload them tonight or tomorrow. I had a few more coming in as I left for work, so when I return home I'll do a minor sort (removing viruii, etc...) and throw 'em up somewhere.
by agu fungus at 4:58 PM EDT on September 7, 2007: I finally finished updating the Pokemon Emerald GSF set, and it's currently available at PSF Rippers. I also uploaded some untagged GSF sets as well:

Crayon Shin-Chan: Arashi no Yobu Cinema-Land no Daibouken!
Bomberman Tournament
The Magical Quest
The Magical Quest 3
by snakemeat at 10:16 PM EDT on September 7, 2007: I have uploaded the MDX files I found on Share here. I didn't sort them, so I am sure there are plenty of duplicates.

edited 10:16 PM EDT September 7, 2007
by Knurek at 4:08 AM EDT on September 8, 2007: Woah. Thanks for those, gonna take a while to sort out but I see new stuff already (System Sacom's Yami no Ketsuzoku for instance).

If you manage to find some FMP/PMD files. I know the pack I put on the FM forums is vastly incomplete.
by Richter X at 12:05 AM EDT on September 9, 2007: So many repeats in that 7z! @.@'

Hope it gets sorted/translated soon. :(
by snakemeat at 1:04 AM EDT on September 9, 2007: Yeah, I didn't do any cleanup on it. I believe the first 3 sets are actually the same, maybe with slightly different additional files added. They all had different sizes and checksums. One thing to note, however, is that many of these are covers or re-rips, I not entirely sure how it works. So some duplicate sets actually contain different versions or covers.
by kingshriek at 4:25 PM EDT on September 9, 2007: I've uploaded my FMPMD collection here:

http://rapidshare.com/files/54535651/FMPMD.7z.html

Has some stuff that doesn't appears to be in Knurek's pack (Metal Force, Misty Blue, A Ressha de Ikou 4, Genocide, etc.), though most of the collection is duplicated (too lazy to sort out the differences).

Here's the listing:

http://h1.ripway.com/kingshriek/FMPMD.txt

edited 4:30 PM EDT September 9, 2007
by Knurek at 6:38 PM EDT on September 9, 2007: Quite a lot of them I simply forgot to copy from my Hoot dir.

But, oh wow, you had data from some PC-98 games I couldn't find anywhere like (Cuisse) More and More or (You-en-tai) You-en-tai Kuusou Yugi Vol1 Yozakura Dennou Ninpouchou...
This will kinda delay Hoot Archive update, but for a good reason. Thank you very much!
by snakemeat at 1:22 AM EDT on September 10, 2007: Many thanks to the both of you, the links are always appreciated. I've got some coming soon, once I finish leeching them, I'll be sure to put em up for all :D.
by Knurek at 11:58 AM EDT on September 10, 2007: Oh, one more question Kingshriek.

I've noticed that someone on 2CHN did a rerip of Famicom Tantei Club (available here: http://vgmmusic.66ghz.com/)

It's different than the fix on your site, and I can make out something about FDS in the rip description. Could you tell which of the files is better?
by kingshriek at 10:48 PM EDT on September 10, 2007: I can't tell any difference between the two other than the fix that I did containing more tracks (short jingles).
by kingshriek at 10:12 PM EDT on September 12, 2007: Hmm, SSF ripping wasn't as bad as I thought it would be. For the standard Sega sound driver, it didn't even require me to do any 68K hacking since complete documentation is available (see Saturn Sound Driver Implementation Manual at http://www.antime.org/sega/docs.html).

All you really need to do is map the right files (from the game CD) to the right place in the 68K's RAM, setup the correct sound area map at 0x500-0x57F and setup 0x700-0x77F so that the right commands are issued to the SCSP (really just setting up the mixer and DSP, initializing the music, and setting up the volume). Both the area map data and emulator save states are very useful in setting up the 68K's ram correctly. One thing to watch for is when setting up the sound area map, make sure to set the transfer complete bit in all the area map blocks (don't want the driver waiting for a nonexistant SH2 to feed it data).

Of course, I'm only talking about getting something playing at all, not necessarily getting it playing correctly. I surely don't claim that I completely know what I'm doing. Here's what I managed to come up with tinkering with Sakura Taisen:

http://www.megaupload.com/?d=5IRX493J

Doesn't sound very accurate, but it's a start. The sound problems are probably a combination of both inaccuracies in the in-development Highly Theoretical and any errors I made (highly likely).

edited 10:14 PM EDT September 12, 2007
by snakemeat at 10:55 PM EDT on September 12, 2007: Many thanks Kingshriek.

After many days of auto hotkey leeching, here's ~21,000 FMPPMD files . There are plenty of dupes, but way too many for me to sort in a short amount of time. Onto MDX now.
by Knurek at 1:58 AM EDT on September 13, 2007: @kingshriek: Ya, I recall Neill Corlett saying something along the lines of SSF being as easy to rip as SPCs.

@snakemeat: Thank you, I'll get to sorting them right away. Should've a new pack with all those kingshriek, yours and my new rips soonish.
by marioman at 7:43 AM EDT on September 13, 2007: Nice work kingshriek. I am amazed that anyone is giving any effort to SSF. Keep it up!
by anewuser at 1:45 PM EDT on September 13, 2007: Hi.

kingshriek, could you use use hosting servicies that are not megaupload? -_- Unable to grab those :( (international user stuff) (sorry for asking too much stuff)

Also, *SWEET* about trying SSF files!!!

edited 2:05 PM EDT September 13, 2007
by kingshriek at 5:12 AM EDT on September 14, 2007: Ok, I've made a rapidshare link as well.

http://rapidshare.com/files/55636773/sakutai_ssf.zip.html

I won't be doing anymore SSF-related work until an official player release.

edited 5:56 AM EDT September 14, 2007
by marioman at 8:14 AM EDT on September 14, 2007: Awww. And here I had my hopes up. I wonder if there is anyone here that would be willing to whip up a player... >_>

Thanks for the SSF file.
by Lunar at 7:03 PM EDT on September 14, 2007: SSF format? LOVE. can't wait!

edit: i had no idea about SSF/DSF progress until today.

edited 8:05 PM EDT September 14, 2007
by Richter X at 7:24 PM EDT on September 14, 2007
by kingshriek at 1:59 AM EDT on September 15, 2007: ...sigh...
Okay one more.

Panzer Dragoon Zwei:
http://rapidshare.com/files/55833333/pdz_ssf.zip.html

As before, this rip is preliminary and may have to be redone when an accurate SSF player is released (as of right now, I don't know if sound errors are my fault or the players).

Also, I hacked together a (very preliminary) python script to dump ssf files using SSF (the emulator). The script requires pmdump (http://www.ntsecurity.nu/toolbox/pmdump/) and bin2psf (http://www.neillcorlett.com/psf/utilities.html) to run. See the script itself (comments) for how to run it.

ssfdump.py:
http://h1.ripway.com/kingshriek/ssfdump.zip

edited 2:27 AM EDT September 15, 2007
by marioman at 7:45 AM EDT on September 15, 2007: Thank you so much. If we can get a player going, I think that SSF ripping will really take off.
Your script will definitely be an invaluable asset. It is nice to see one of the so-called "dead" xSF formats getting some new life.
by Lunar at 11:03 AM EDT on September 15, 2007: Hm I dunno kingshriek, maybe ripping stuff would be a catalyst/the extra push needed for Neill to round it up and give it an official release. The only striking problems I've personally noticed with playback are some pitch issues here and there.

ADVBGM13.ssf - I've been addicted to this track from the Sakura Taisen set since hearing it.
by kingshriek at 12:37 PM EDT on September 15, 2007: Uploaded a updated version of the script, correcting a few areas where I shouldn't have been writing zeros to RAM (0x400-0x47F mainly). The driver seems to keep a history buffer of SCSP commands starting at 0x800 - this can speed up the process of selecting the right track bank/number.

ssfdump.py (ver 0.01):
http://h1.ripway.com/kingshriek/ssfdump.zip

@Lunar:
Not only pitch issues, but also many cases where certain sounds aren't playing at all. ADVBGM22.ssf in the Sakura Taisen set is a good example of this (among many others).
by Knurek at 12:40 PM EDT on September 15, 2007: Well, the Japanese sure like that kingshriek's script... Radiant Silvergun complete rip and a proof of concept of Yu-No available at 2CHN rip thread already...

Hopefully, we'll see a complete player soonish...
by marioman at 2:47 PM EDT on September 15, 2007: Good grief! I hope that they are as eager to write a player as they are to rip games.

Congratulations kingshriek, in about 11 hours you have successfully revived an undeveloped xSF format, AND have rounded up a group of SSF rippers in Japan. You now belong to a class of people like Steve Jobs who can come up with an improvement to something that already exists and have people drooling over it before the day is over. :P

I am fairly excited about this SSF development too. Can't wait to see how the next 12 hours goes...
by Lunar at 3:59 PM EDT on September 15, 2007: So what do peoples SSF wishlists look like? :D I'd be really happy to see Shining Wisdom, Gungriffon, Virtua Fighter 3 and Sonic 3D Blast.
by kingshriek at 4:03 PM EDT on September 15, 2007: I just hope that folks in Japan don't get too carried away with this. Without an accurate player, I'm unable to assess the accuracy/validity of the ssf dumper. I hope so too that this will stimulate SSF player development.

I have yet another short term update to ssfdump.py (fixed a few memory ranges, added option to avoid redundantly calling pmdump, put some effort into making the script more readable):

ssfdump.py (ver 0.02):
http://h1.ripway.com/kingshriek/ssfdump_v002.zip
by Knurek at 5:19 PM EDT on September 15, 2007: @Lunar: Aren't all those games using Redbook Audio for music? I'm pretty sure Sonic3D and VF2 (you meant VF2, right?) do that.

@kingshriek: I'd say it's too late for that... They've posted a patch for in_ssf already (not sure if it changes anything with emulation, or if it's just a translation of the plugins dialogs) and a rip for Black Matrix.

edited 5:24 PM EDT September 15, 2007
by Lunar at 5:48 PM EDT on September 15, 2007: 3D Blast sounds like redbook yeah, and VF3 is Dreamcast anyway. But I've got mp3s for both so that's fine.

I'm fairly sure Shining Wisdom is sequenced though. It certainly sounds it from the youtube videos I saw (the OST CD has arranged tracks so it's not a good guide.) As for Gungriffon... no idea whatsoever. I can't even find a rip of the arranged CD for it, let alone a line-out rip or even some videos. Very elusive indeed, so here's hoping it is sequenced.

edited 7:08 PM EDT September 15, 2007
by marioman at 6:06 PM EDT on September 15, 2007: @Knurek: in_ssf? Is that a Winamp plugin? If so, can you please give a link?

@Lunar: I am very sure that Sonic 3D Blast is redbook. Nights into Dreams would be a good candidate for a SSF rip though. I know that it's sequenced.

edited 6:06 PM EDT September 15, 2007
by anewuser at 6:21 PM EDT on September 15, 2007: http://www.neillcorlett.com/etc/htdemo/ dunno about the 2chn patch.
by marioman at 6:46 PM EDT on September 15, 2007: For some reason I did not know that Highly Theoretical also played SSFs. Oh well. Thanks for the link.

Since DSFs and SSFs are so similar that they can both be played by the same player, does this mean that there could also be a ripping script written for DSFs?

Also, what on earth is 2CHN? I hear about it everywhere but I don't know exactly what it is.
by unknownfile at 9:31 PM EDT on September 15, 2007: stupid japanese bbs
by Mouser X at 1:03 AM EDT on September 16, 2007: marioman: Unfortunatly, while SSFs and DSFs may use similar hardware (based on the fact that they're played using the same plugin. I also recall reading that somewhere from Neill), their ripping method is quite different (as I recall). DSFs are ripped similar to USFs and PSFs, and such, and require knowledge of ARM<number> assembly to rip. I *REALLY* want a full rip of Skies of Arcadia... Which is actually possible, since one of the samples available is from that game. However, it'd still require a person to be good with hex editing, since you'd have to find all the sequences, and alter the memory of the current sample (which has the driver in it) to work with the sequences. According to Neill, he thinks the current sample file may actually contain all of the audio from SoA already in it (sample banks. Possibly sequences? He wasn't clear on that, but sample banks for the entire game are almost certainly in there). Apparently, the sample file is much larger than it should be...

Anyway, these SSF developments are pretty awesome. I *really* hope a new (and fixed, at least for Saturn audio) version of HT comes out (or, a new player entirely, if Neill is unable to get it working, or doesn't have the time). Mouser X over and out.
by kingshriek at 5:09 AM EDT on September 16, 2007: Looking at the SoA DSF, the sound driver looks very similar to the standard SDDRVS.TK driver found in many Saturn games. The Dreamcast driver uses very similar sound commands.

I haven't been able to find anything else in the SoA DSF except for a sound effect bank. To access, run psf2exe on the dsf, open up the resulting file in a hex editor and jump to 0x160114. You should see a 01 00 00 04 00 command there (this initializes track 0 on sound bank 4). The sound effect bank is bank 1, so change the command to 01 00 00 01 <track#>, where <track#> is user specified. Try 01 00 00 01 12 for a short music track.

Also, I've updated ssfdump.py yet again. This time I added a history buffer output, dumping relevant SCSP commands to a text file. This saves a lot of time in selecting the correct script parameters.

ssfdump.py (ver. 0.03):
http://h1.ripway.com/kingshriek/ssfdump_v003.zip

edited 5:13 AM EDT September 16, 2007
by marioman at 3:31 PM EDT on September 16, 2007: Does anyone know where that patched version of in_ssf is available at? While I am on the subject, what kind of improvements are made in the patched version?
by Knurek at 5:48 PM EDT on September 16, 2007: http://pc11.2ch.net/test/read.cgi/dtm/1174838154/750

The archive contains the patcher program.
Once again, I'm not sure if it changes anything or if it just translates the plugin's dialogs to Japanese.
by marioman at 8:39 PM EDT on September 16, 2007: OK thanks.

Previous Page | Next Page
Go to Page 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Search this thread

Show all threads

Reply to this thread:

HCS Forum Index
Halley's Comet Software
forum source

User Name		Tags: bold: [b]bold[/b] italics: [i]italics[/i] emphasis: [em]emphasis[/em] underline: [u]underline[/u] small: [small]small[/small] Link: [url=http://www.google.com]Link[/url] [img=https://www.hcs64.com/images/mm1.png]
Password
Subject
Message