Metroid Prime 2 Echoes Demo DSP by valiant at 7:08 PM EDT on July 24, 2009
Hey there, hcs!

I've just recently started properly ripping the first two Metroid Prime games since all the previous releases were not very satisfying. The first title is already done and will be released soon, but I've got a problem with the second one. Since I always include all beta and trailer tracks in my rips, I got my hands on the Metroid Prime 2 Echoes Bonus Disc and, playing the demo, found some compositions different from the final releases. The THPs and RSFs play just fine with vgmstream (using the latest revision), though most DSPs won't play.

Three of the DSP files are working anyway, they just sound a little off (played back at half the original tempo and with the pitch just as low):

dark.dsp
first-ruins.dsp
ridley.dsp

Those tracks were originally from Metroid Prime, so they could be some intermediate test tracks for the new DSP format. I uploaded the files with corrected pitch and tempo and included their Metroid Prime DSP counterparts. It seems that the demo files are the exact same as the left channel DSPs from the first game once they are corrected.

The other sound files are still up on the old mirror.

I'd appreciate if you could look into it and fix the track playback for one of the future releases.
by Mouser X at 7:36 PM EDT on July 24, 2009
If I recall correctly, HCS is aware that the MP2 Demo tracks don't play properly in in_cube. Since (IIRC) their support was "ported" over from in_cube into VGMstream, it's expected that they don't work in VGMstream. The problem is, since it's a demo, HCS doesn't really care about this DSP variation, since not only is it only one game, but that game is a demo, with very few files. As such, HCS was not interested in looking at this "hybrid" (from the first to the second game) format, back then. Perhaps now, his opinion has changed. But, if it's still not working properly, I woudn't expect that it has.

I just thought it'd be useful to point out why MP2 Demo files don't currently play properly, and that, based on that information from the past, it's not likely that it'll be fixed soon. But as I said, that was awile ago. Perhaps HCS has changed his mind since then. Mouser X over and out.
by bxaimc at 12:46 AM EDT on July 25, 2009
hmm, GENH perhaps?
by valiant at 9:35 AM EDT on July 25, 2009
Didn't know it existed.


Tried it with both a final and demo version track but of course didn't work since I have no idea which values are correct.



I assume the interleave value 36608 (0x8f00 for RS03 DSP) is right, but the channel coefficients are missing. I know they start at 0x20 and 0x40 in the DSP header respectively, but how do you get the decimal value for the Generic Header Creator from those?

Also, if I save the final file with the GENH, the old header is still included. Shouldn't "Header Skip" cut it?
by snakemeat at 2:21 PM EDT on July 25, 2009
Also, if I save the final file with the GENH, the old header is still included. Shouldn't "Header Skip" cut it?

Sorry can't help with the rest, but GENH will not alter a file in any way that will cause data loss. The reason for this is that it will allow you to get the original file back if the format is later supported in its native form (without requiring a re-rip).

Header Skip tells vgmstream (or any other plugin that offers GENH support) to skip the number of entered bytes at the start of the file. Basically, to "skip the header".
by bxaimc at 3:20 PM EDT on July 25, 2009
Remember, GC DSP needs coef values for GENH. I'll take a look at when I get home later today.
by hcs at 7:14 PM EDT on July 25, 2009
0x20 = 32
0x40 = 64
by valiant at 8:11 PM EDT on July 25, 2009
Umm, thanks. That was anticlimactic. I thought the actual 32 bytes at these offsets were the channel coefficients, not the offsets themselves... Duh...

The final version DSPs now work with the GENH, albeit with quite a few clicks here and there. The demo file is still horribly garbled. The RS03 identifier (first 4 bytes) is missing there, but I took that into consideration, so that's not it. Could it be that the interleave value is different and that that's the only problem?


Go to Page 0

Search this thread

Show all threads

Reply to this thread:

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

HCS Forum Index
Halley's Comet Software
forum source