Previous Page | Next Page
- by nensondubois at 11:14 PM EST on January 19, 2008
- Don't want to sound like an idiot but I don't know where to download the new version. It's not on the previous page.
- by JILost at 11:25 PM EST on January 19, 2008
- This might be a really dumb question, but how do I get the Ace Combat 4 files to play? Do I have to rename them to .vag or something?
- by hcs at 12:08 AM EST on January 20, 2008
- in_cube
I don't know how to play the Ace Combat 4 files, either. I'll get 'em and see (probably just need renaming).
- by Mouser X at 12:10 AM EST on January 20, 2008
- Nensondubois, did you consider looking on the in_cube webpage? I certainly wouldn't have thought it was that hard to find. It's certainly right where I expected to see it.
As for Ace Combat 4, do they have a file extension? Is there a readme or info file? I don't have those myself, so I can't say for certain how to get them to play. I would have figured that they'd just work. Sorry I can't help more. Mouser X over and out.
- by JILost at 1:06 AM EST on January 20, 2008
- They're in /bonus, under PS2ADPCM. The file extension is .PS2 but all the filenames contain "VAGBGM", so I figured they were "VAG (with dual channels)" as stated in the hist.txt file. Am I missing something? Are they some weird alternate format, encrypted, or something like that?
- by snakemeat at 3:10 AM EST on January 20, 2008
- They are interleaved files, and as such, have varying increments depending upon the game. You'll need to use FastElbJa's ADPCM Player to hear them. Grab it here. It requires .net 2.0. Load the files, and enter the interleave in the bottom right corner (you can type it in, you don't have to use the arrows). The Ace Combat 4 set needs a value of 800, as the folder name indicates.
- by hcs at 6:37 AM EST on January 20, 2008
- JILost, you are quite right about the ADX volume bug, I'll see what's up there.
Ah, I had increased the volume on ADXs a while back to 0x4000 but I didn't update the configuration routine (it was still at 0x2000). I'll have a new version up (with some other fixes) in an hour or so.
edited 6:48 AM EST January 20, 2008
- in_cube 0.28 by hcs at 7:04 AM EST on January 20, 2008
- in_cube 0.28 now up.
- by Gohdan at 8:46 AM EST on January 20, 2008
- I found a bug on in_cube 0.28: I tested this version with the Super Paper Mario streams. At the point where the loop ends the music sounds defectively, but only on that point! For example: evt_star1_44k_lp.brstm at 0:27, or evt_zigen_appear1_44k_lp.brstm at 1:11.
- by hcs at 9:51 AM EST on January 20, 2008
- Good catch. This is because of a phenomenon I call "short chunk".
Streamed audio lesson:
brstm (and indeed a lot of streamed formats) are "interleaved". This means that, in the file, first we have a little chunk of data for the left channel, then a little chunk of data for the right channel, then a second chunk data for the left channel, and so on, alternating chunks between the two channels. This is done so that the file can be read sequentially, which is important on storage media where seeking is costly, such as optical or magnetic disks. The chunks of data are typically all the same size and this size is called the interleave for the file. Largish chunks are used so that the decoder has to only occasionally switch between them (switching wastes time), but the larger they get the more memory is needed to store them. Because of this an equilibrium must be reached, and you get widely varying interleave values depending on what the developers thought was appropriate.
One side effect of interleaving is that if the size of your data is not divisible by the interleave you need to waste a bit of space at the end of the file. When the left channel ends you may need to fill in junk (usually silence) to pad out the last chunk to the interleave size so that the right channel can appear at the correct position. Often the interleave is small enough relative to the size of the data involved that no one worries about this. Certain drivers, however, avoid wasting space by making the last chunks in the file smaller.
This is the "short chunk" phenomenon. It means that I can't blindly use the same interleave value for the whole file, I have to check if this chunk set is going to go past the end of the file, and if so I need to recalculate the interleave to be half of the remaining file. I don't always do this in in_cube, only for certain types, because some types keep the same interleave but end the file when the right chunk ends, which is a different situation but looks the same.
When I added the check for Super Paper Mario filenames that need to be treated specially, I added a new type. I neglected to add that type to the list of types that need short chunk processing.
So I'll get that into a new version, shortly.
in_cube 0.29 now up.
edited 10:06 AM EST January 20, 2008
Previous Page | Next Page
Go to Page 0 1 2 3 4 5 6 7 8 9 10
Search this thread
Show all threads
Reply to this thread:
HCS Forum Index
Halley's Comet Software
forum source