WAV loop points in SF vs Cool Edit -Sound Forge - PC

Subject:WAV loop points in SF vs Cool Edit Posted by: tsphillips Date:9/8/2001 2:11:17 PM Could anyone here explain (or direct me to a reference on) the technical differences between SF and Cool Edit (by Syntrillium) regarding the encoding of loop points in WAV files? I have found that loop points written by Cool Edit will not properly transfer to an ESI-2000/4000 sampler using SMDI, while SF loop points will. (This is using the Esi-Win utility for SMDI transfers.) I had thought that the encoding of loop points was standard according to the Sampler Chunk (smpl). Tom
Subject:RE: WAV loop points in SF vs Cool Edit Reply by: tsphillips Date:9/10/2001 10:15:17 AM I have forwarded this question to Syntrillium. If they reply I will post their answer here.
Subject:RE: WAV loop points in SF vs Cool Edit Reply by: tsphillips Date:9/18/2001 9:43:50 PM Syntrillium support replied, but does not know what the difference is. I looked at the sample chunks in wav files saved with SF and Cool Edit, and it seems Cool Edit is inserting extraneous data in the loop section. I don't know what sort of extra information they may be inserting, and their support didn't seem to know either. :-( Tom
Subject:RE: WAV loop points in SF vs Cool Edit Reply by: ljackson Date:9/18/2001 11:30:02 PM I've got a little tool called RIFFView that lets me do a bit of probing into WAV files with RIFF chunks, including smpl chunks. I did a quick test by creating identical loops in SF5 and CE2K. Here's what RIFFView said about them. SF5 loop - RIFF 'WAVE' (wave file) (format description) PCM format 2 channel 44100 frames per sec 176400 bytes per sec 4 bytes per frame 16 bits per sample (waveform data - 1261508 bytes) (sample parameters) Sample period 22675ns (44101.4Hz) MIDI root note 60 Loop #0 mode 0 (forward) start 0 end 315376.00 CE2K loop: RIFF 'WAVE' (wave file) (format description) PCM format 2 channel 44100 frames per sec 176400 bytes per sec 4 bytes per frame 16 bits per sample (waveform data - 1261508 bytes) (cue points) #1: sample 0 (playlist position 0) LIST 'adtl' (associated text) (region description) #1: 315377 samples as rgn (text label) #1: "Loop 1" (text comment) #1: "Sampler loop" (sample parameters) Sample period 22675ns (44101.4Hz) MIDI root note 65 Loop #1 mode 0 (forward) start 0 end 315376.00 For starters, it looks like CE2K is mis-identifying the loop. It's calling it Loop #1, where SF5 calls it Loop #0. This might throw off some brain-dead hardware and software. If you want to get the loop files I created, contact me. Please feel free to forward this info to the proper people at SF and Syntrillium.

Subject:WAV loop points in SF vs Cool Edit

Posted by: tsphillips

Date:9/8/2001 2:11:17 PM

Could anyone here explain (or direct me to a reference on) the technical differences between SF and Cool Edit (by Syntrillium) regarding the encoding of loop points in WAV files?

I have found that loop points written by Cool Edit will not properly transfer to an ESI-2000/4000 sampler using SMDI, while SF loop points will. (This is using the Esi-Win utility for SMDI transfers.) I had thought that the encoding of loop points was standard according to the Sampler Chunk (smpl).

Tom

Subject:RE: WAV loop points in SF vs Cool Edit

Reply by: tsphillips

Date:9/10/2001 10:15:17 AM

I have forwarded this question to Syntrillium. If they reply I will post their answer here.

Subject:RE: WAV loop points in SF vs Cool Edit

Reply by: tsphillips

Date:9/18/2001 9:43:50 PM

Syntrillium support replied, but does not know what the difference is. I looked at the sample chunks in wav files saved with SF and Cool Edit, and it seems Cool Edit is inserting extraneous data in the loop section. I don't know what sort of extra information they may be inserting, and their support didn't seem to know either. :-(

Tom

Subject:RE: WAV loop points in SF vs Cool Edit

Reply by: ljackson

Date:9/18/2001 11:30:02 PM

I've got a little tool called RIFFView that lets me do a bit of probing into WAV files with RIFF chunks, including smpl chunks. I did a quick test by creating identical loops in SF5 and CE2K. Here's what RIFFView said about them.

SF5 loop -

RIFF 'WAVE' (wave file)
(format description)
PCM format
2 channel
44100 frames per sec
176400 bytes per sec
4 bytes per frame
16 bits per sample
(waveform data - 1261508 bytes)
(sample parameters)
Sample period 22675ns (44101.4Hz)
MIDI root note 60
Loop #0
mode 0 (forward)
start 0
end 315376.00

CE2K loop:

RIFF 'WAVE' (wave file)
(format description)
PCM format
2 channel
44100 frames per sec
176400 bytes per sec
4 bytes per frame
16 bits per sample
(waveform data - 1261508 bytes)
(cue points)
#1: sample 0 (playlist position 0)
LIST 'adtl' (associated text)
(region description)
#1: 315377 samples as rgn
(text label)
#1: "Loop 1"
(text comment)
#1: "Sampler loop"
(sample parameters)
Sample period 22675ns (44101.4Hz)
MIDI root note 65
Loop #1
mode 0 (forward)
start 0
end 315376.00

For starters, it looks like CE2K is mis-identifying the loop. It's calling it Loop #1, where SF5 calls it Loop #0. This might throw off some brain-dead hardware and software.

If you want to get the loop files I created, contact me. Please feel free to forward this info to the proper people at SF and Syntrillium.