The other day I was playing around with my bowed cymbals samples and found that they didn’t timestretch particularly well with Elastic Audio or otherwise. I like Elastic Audio for many purposes but I keep running into it’s limitations. Stretching bowed cymbals to 200% is one of them, I later discovered that most other methods aren’t much better either. I’ve created some examples so you can come to your own conclusion about which is the ‘best’ way. As you’ll hear, some of these have the painfully obvious audio equivalent of THIS IMAGE.
Download RAR file with original file and 17 stretched files. 16 bit, 44.1kHz (same as original recording)
- original
- TCE Digidesign
- TCE Timeshift Default
- TCE Timeshift Stereo Mix
- Audiosuite Time Shift default
- Audiosuite Time Shift follow transient
- Audiosuite TimeCompressionExpansion Default
- Audiosuite TimeCompressionExpansion accuracy 5
- Audiosuite TimeCompressionExpansion accuracy -5
- Elastic Audio Monophonic No Event markers
- Elastic Audio Monophonic
- Elastic Audio Polyphonic default no event markers
- Elastic Audio Polyphonic default
- Elastic Audio Rhythmic default
- Elastic Audio Rhythmic default no event markers
- Elastic Audio X-Form default
- Elastic Audio X-Form Formant
- Elastic Audio X-Form No Event Markers
In my opinion and for this specific situation there is one method of stretching that is clearly better, the least “stretch marks” it’s also the most time consuming.
Let me know what you think about the test. What would you like for round 2?
Also Interesting:




It’s so hard for a digital process that doesn’t perform exactly as advertised to to be cool.
Yeah, I’ve found the same. If I’m using EA in a project, I generally set it to polyphonic while I’m getting the timing right, and then switch to XForm to render it, and go walk the dog while it’s doing its thing. It’s the only one that gives tolerable results to my ears, on practically any instrument. I’ve tried on clarinet, trumpet, violin, percussion, upright bass, and oud, and keep coming back to this conclusion…
this is why i still use sony’s acid 6 or 7, for time stretch. much better for than digi’s half baked algorithm….
@mediawest – Let’s hear it.
Nothing beats the wonders of Pro Tools.