Subject:More on SF Pro 10 Spectum Analysis tool

Posted by: winnie

Date:8/31/2010 5:17:43 PM

I am facing a weird problem with the Spectrum Analysis tool in SF Pro 10. When I analyze stereo files with this tool the spectrum seems to mirror at 1/4 the sampling rate, rather than 1/2 of it, i.e., if the SR is 44.1k one expects the magnitude spectrum to be mirror symmetric around 22.05k but in my case it mirrors at 11.025kHz.

This is not the case with mono files. Has anyone else encountered this issue?
Thanks

Subject:RE: More on SF Pro 10 Spectum Analysis tool

Reply by: ForumAdmin

Date:9/13/2010 11:05:46 AM

Yes. We are looking into this.

J.

Subject:RE: More on SF Pro 10 Spectum Analysis tool

Reply by: GPD

Date:9/14/2010 8:13:41 AM

Yes.... something is broken in Spectrum Analysis in 10b.

Try this:

1) New File 96K, 32bitFP.
2) Insert Synthesis: White Noise 16sec.
3) Resample using Izotope SRC to 44.1, and set "Cutoff scaling" to 0.50.

4) Perform Spectrum Analysis. (normal display)

We SHOULD see something like a brick wall filter at 11025hz with no energy above this.

Instead we see a very minor dip, and high energy the whole way to Nyquist...

Additionally if we switch to "sonogram" view, the sonogram takes FOREVER to draw. MUCH Slower than previous version. (using a 3.2Ghz Nehalem Xeon)

Subject:RE: More on SF Pro 10 Spectum Analysis tool

Reply by: GPD

Date:9/14/2010 8:18:05 AM

...confirmed that the errant result does NOT occur when using a mono file. The result is correct on mono files...