Audio File Editing.......

[Blue Bear Sound]

Anyone know of any software that handles splitting audio files into chunks better than SF does?

Sound Forge's file splitting process (called "Auto-Region") is pretty kludgy.... parameters are set as a percentage rather than actual dB values, making setting specific sound levels as clip-points next to impossible.....

Anyone know if SF 7's implementation is better for that specific process or maybe there's another software package specifically suited to the task??

 

[SPINSTERWUN]

.....Anyone know if SF 7's implementation is better for that specific process or maybe there's another software package specifically suited to the task??

I believe that it is better.

Check you rPM in five (5) minutes.

spin

 

[Blue Bear Sound]

Back at ya, Spin......!

 

[mrface2112]

i dig cd wave

i used to use SF and i liked it.......until the day i found cdwave.

< http://www.etree.org/cdwave.html >

and

< http://www.cdwave.com/ >

let's say you've dumped a dat to hd and you've got 14 songs you need to split. you dump it as one big wav file, open the wav in cdwave, and simply click where you want your splits to be in the nice, pretty, easy as shit to use gui. it even automatically splits the tracks along sector boundaries and as a result i've never heard any sort of noise, pops, clicks, etc., from it. i've seeded a number of "jam band" live recordings which require flawless track-to-track/song segue continuity--all of them split with cd wave and nary a single problem.


oh yeah.....and a whopping $15 registration fee should you feel obligated to buy the thing.


cheers,
wade

 

[Blue Bear Sound]

Thanks Wade... but it isn't quite what I need...

If I have an audio file representing 2000 voice-over clips (with very clean, quiet space between each clip), I need an automated processed that allows me to set a dB threshold to trigger a new segment.... ie - split it up into clean chunks that can be exported as separate clips.

SF has the tool, but the parameter controls for it are essentially worthless - it would be equally as fast for me to go in and set the spots myself, since after running their tool, I have to check each clip-point it created anyways, because the controls are so imprecise.....

 

[RawDepth]

Bear,

I don't know if it is what you want or not but, check out this little utility and see.

http://ubhank.nl/hd24/

 

[Blue Bear Sound]

Hey - that could be pretty useful -- not for this particular application, but certainly in others... thanks for that link!

 

[Sonic Misfit]

Hey Blue Bear,

If you find a solution to this, please post the results. I use the strip silence feature in Protools, but this does not create individual files, it only specifies regions in the file.

Good luck on your hunting expedition.

 

[matty_boy]

There is a Cue Points function in Goldwave that will do just that. It will automatically detect silence to place cue points, and then split the file into separate files. The silence selection threshold is graduated in db. It's also available in a free demo.

http://www.goldwave.com

 

[Blue Bear Sound]

Sonic - I will... definitely!

MB - thanks - I'll look at that........!

 

[distortedrumble]

you could always lure an intern and give him the grunt work...may be slower than the computer but atleast you can kick him when you get pissed without destroying valuable hardware

 

[nwsoundman]

Nuendo 2.0

Use the detect silence and scissor tools. You can easily creat 2000 events in a few minutes. Look in the Help menu for details. I can answer questions but may take some time to reply. Very busy.

NWSM

 

[Blue Bear Sound]

NWSN....

Yes - I use the same function in Cubase SX2 - works very well, except how do I create distinct files of each segmented clip?

 

[nwsoundman]

Damn! I forgot you had 2000 clips. I will try to figure a solution that is easy when I get to the studio. What a pain it would be to save every event one at a time manualy.

NWSM

 

[cordura21]

hey Bruce, here are 2 to try:

http://tracksplit.sourceforge.net/

http://panic.et.tudelft.nl/~costar/gramofile/

Hope it helps.

Andrés

 

[Blue Bear Sound]

Damn! I forgot you had 2000 clips. I will try to figure a solution that is easy when I get to the studio. What a pain it would be to save every event one at a time manualy.

NWSM

Actually - 5000!

 

[Blue Bear Sound]

hey Bruce, here are 2 to try:

http://tracksplit.sourceforge.net/

http://panic.et.tudelft.nl/~costar/gramofile/

Hope it helps.

Andrés

Thanks Andre - those just might work....

OTOH... maybe not - I just looked --- they're both Linux-based....

 

[cordura21]

the second one has a windows version:

http://panic.et.tudelft.nl/~costar/gramofile/gramo_w_1.6.zip

Notes for this version: the .zip file will create the directory djgpp\gramofile-1.6\ and extract everything in there; it is advised to extract in C:\ to have the supplied .PIF file working correctly (it refers to the file C:\djgpp\gramofile-1.6\gramofile.exe).
You can start the pre-compiled program gramofile.exe immediately. If you wish to compile the program yourself, see the instructions in DOSPORT.txt
This version of the program does not support sound recording or playback, but track splitting and signal processing should be fully functional. For Windows systems, numerous .wav recorders/players are available, which you can use instead. But be aware that GramoFile only handles CD-quality .wav's, i.e. 44.1 kHz 16-bit stereo. Furthermore, it is a "console" program, meaning that it will run in a DOS window, and that clicking the "buttons" with the mouse will not work. Use the TAB key instead.
When running from DOS-only (i.e. not from Windows), you also need the CWSDPMI program (you probably have to put it in GramoFile's directory).
Bernd Leppla informed me that this version also runs well in an OS/2 DOS box.

Andrés

 

[nwsoundman]

Just checking

I had a chance to look into this some over the weekend. I didn't find a way to do it easy yet. Maybe looking at a different recording approach? What did you end up doing Bruce?


NWSM

 

[Blue Bear Sound]

Normally, for shorter VO sessions, I'd record right to the Masterlink and split the segments into tracks right there, creating instant separation.

But for a sound library of over 5000 words, sounds and phrases, that method simply wasn't practical. I recorded each day's session onto a track on the HD24.

So I had to manually slice 'n dice the individual full-session files into individual clips -- there was no other way but to hand-edit the session tracks, remove the unwanted takes, until I was left with a session file containing only the necessary clips, seperated by silence.

I then used Sound Forge's AUto-Region to split the clips up. Problem is, the parameter control on the AUtoRegion tool is pretty flaky and I couldn't get it to properly detect the long silence between clips. So again, once I set regions using the tool, I had to go thru each one and hand-check that the tool worked properly. Once the regions were set, I could use the "extract regions" tool to get each clip into its own file.

Cubase's "Detect Silence" tool would have been much better in terms of isolating each clip, and if only it had a way of separating each segment into its own file, then that would have been a huge timesaver.