precise stereo correlation analyzer plugin-does it exist??

[ericzang]

http://www.soundradix.com/products/pi/

I know, but its for mix only...and useless for repairing badly recorded stereo samples

I have used Pi to improve phase correlation issues in a stereo file. First I separate the stereo track into 2, (left & right), and apply Pi. Might this apply to your needs?

[kvaca]

http://www.soundradix.com/products/pi/

I know, but its for mix only...and useless for repairing badly recorded stereo samples

I have used Pi to improve phase correlation issues in a stereo file. First I separate the stereo track into 2, (left & right), and apply Pi. Might this apply to your needs?

so how did you liked the results? Pi seems like a magic box to me...or have you figuered out how it actually works? Im still unable to find any reasonable info on their site...apart from words like "dynamic phase rotation" which can be thousands of things /starting with leslie box etc/

in any case - I prefer full manual control over processing...so no mystery magic box needed, also no dynamic rotation needed - my samples are completely static... all I need is some sort of static spectral delay with precision down to 1 Hz, where I can input only worst colliding frequencies found on correlometer analyzer...the rest of the sound must be unaffected after processing, so probably linear phase design is the way to go...I can only dream that somebody will invent it in the future

[Bhan]

Hmmm- the way I think about it - There is no such thing as In phase and in stereo from your ears perspective. Phase differential is what primarily determines Left to right panning information.

[kbaccki]

Hmmm- the way I think about it - There is no such thing as In phase and in stereo from your ears perspective. Phase differential is what primarily determines Left to right panning information.

I've been struggling with this a bit as well. The spectra of left and right change over time, so at any given time X a given frequency will either be mono compatible or not, but not necessarily at X+1. I was looking at the correlation of a stereo piano sample (a single key)... so your primary harmonics are OVERALL static for a given note, but in general the spectra is dynamic over the full sample, which is also part of the stereoness of what the ear perceives, right? I guess if you lock in on specific frequencies (eg, known harmonic series) and just treat those for the full duration, then maybe that's the goal.

EDIT: oh, I see that kvaca just mentioned above that his samples are static...

[ericzang]

For static situations (multiple stationary mics and stationary sound source, drumset typical example), have you tried MAutoalign (Melda) or Autoalign (Radix)? I use the melda one and it has LP and HP detection filters, so you can have it base its adjustment on a certain range of frequencies (though of course all of the audio is still adjusted as a whole).

With Pi, it has a full range mode and a low freq mode (below 800hz). I had a situation where I only wanted it to act upon less than 100hz, so I divided my tracks with a linear phase crossover and put Pi only on the now separated <100hz tracks. I suppose this would work for any frequency range. But sounds like you're wanting more precision than a general range.

I don't know how Pi specifically works, and the controls are limited to grouping, weighting (which tracks get changed less or not at all), and frequency range as I mentioned.

I mostly use it for bass issues, but I have used it on a stereo recording. Here's an example of such for comparison from just a simple living room recording of a friend playing the Persian santur (hammered dulcimer). I didn't place my mics in an optimal phase situation. just a rough AB arrangement. The zip file linked here is with a couple humble mini omni mics. No other processing on the tracks. Dry, MAutoAlign, and Pi.

10 second samples here:
http://ericzang.com/tmp/Santur_Dry_MAutoAlign_Pi.zip

Full video here: (sorry didn't know it would put the whole video in the message)

---- Hmm, listening to the video now I think I notice some odd shifting on some of the sustained notes. Maybe this is Pi messing with things. On the video the larger mics are in the mix, though much lower in level, and Pi is applied to all 4 of them.

-------- I went back to the project and checked and I was able to improve the shifting by grouping the stereo pairs to their own groups first. I hadn't been using Pi for very long at the time I did that recording.

https://www.youtube.com/watch?v=yXMJ03IrZYQ

[kvaca]

Hmmm- the way I think about it - There is no such thing as In phase and in stereo from your ears perspective. Phase differential is what primarily determines Left to right panning information.

I've been struggling with this a bit as well. The spectra of left and right change over time, so at any given time X a given frequency will either be mono compatible or not, but not necessarily at X+1. I was looking at the correlation of a stereo piano sample (a single key)... so your primary harmonics are OVERALL static for a given note, but in general the spectra is dynamic over the full sample, which is also part of the stereoness of what the ear perceives, right? I guess if you lock in on specific frequencies (eg, known harmonic series) and just treat those for the full duration, then maybe that's the goal.

EDIT: oh, I see that kvaca just mentioned above that his samples are static...

yes, these are pipe organ samples, completely static...and recorded with wide AB mic configuration in very close distance to avoid church reverb as much as possible, this is where phase issues are impossible to avoid...

[kvaca]

For static situations (multiple stationary mics and stationary sound source, drumset typical example), have you tried MAutoalign (Melda) or Autoalign (Radix)? I use the melda one and it has LP and HP detection filters, so you can have it base its adjustment on a certain range of frequencies (though of course all of the audio is still adjusted as a whole).

I know, I have Melda Total so I have used Autoalign, unfortunately its useless for organ samples from obvious reason /to correctly detect the point where pipe organ sample exactly starts is hard task even for very experienced people...completly imposssible for simple plugin/

With Pi, it has a full range mode and a low freq mode (below 800hz).

so as I have expected its completely useless for my task where I want to correct only some specific harmonic overtones from many and with the best possible spectral precision while preserving the rest of the sample unaltered as much as possible

[KTlin]

Very interesting thread. Like the OP, I'm also looking for a precise correlation analyzer. Nugen Visualizer is great, but at $250 is way over my budget.

As for the out-of-phase correction, the "Prevent Antiphase" option in the Imager module inside Ozone7 seems to resolve some phase issues, but it might not have any effect on completely static material like the organ samples.

[MogwaiBoy]

As for the out-of-phase correction, the "Prevent Antiphase" option in the Imager module inside Ozone7 seems to resolve some phase issues,

Dude, I did not even know that was there! Thanks for the heads up. Just went looking for it and found it in the Options menu under the Imager tab.

The Ozone 7 manual says:

Prevent Antiphase
When checked, the Imager will automatically prevent any settings from being applied that would result in phase cancellation of the stereo signal, when summed to mono.

[KTlin]

Dude, I did not even know that was there! Thanks for the heads up. Just went looking for it and found it in the Options menu under the Imager tab. I'm going to go and read up about it..

Yes, it's strange they've put it in the Options menu. It should definitely be in the Imager's main screen.

[Aloysius]

For those of us who didn't upgrade, it's there in Ozone5 too.

http://help.izotope.com/docs/izotope-ozone5-help.pdf

Pg 131

Phase Meter
Prevent antiphase: Prevents mono compatibility problems when widening the stereo image.

[KTlin]

The Ozone 7 manual says:

Prevent Antiphase
When checked, the Imager will automatically prevent any settings from being applied that would result in phase cancellation of the stereo signal, when summed to mono.

Hmm... I don't know about that, but I tried it on some material with out-of-phase issues (correlator moving to -1 most of the time), and without me changing any of the settings in the Imager, just with the "Prevent antipahse" option checked, it corrected the phase problems (with the corellator closer to 0).

[MogwaiBoy]

The Ozone 7 manual says:

Prevent Antiphase
When checked, the Imager will automatically prevent any settings from being applied that would result in phase cancellation of the stereo signal, when summed to mono.

Hmm... I don't know about that, but I tried it on some material with out-of-phase issues (correlator moving to -1 most of the time), and without me changing any of the settings in the Imager, just with the "Prevent antipahse" option checked, it corrected the phase problems (with the corellator closer to 0).

By the description in the manual it sounds like it only takes care of phase issues generated by itself, not on the existing stereo waveform input.

But if what you're saying is true, then... damn that's cool!! I will do some of my own tests.

Might be really handy to throw an Imager module (provided you have Advanced) somewhere late in your mastering chain - without any stereo imaging effect, just simply the Prevent Antiphase activated. Set and forget style.

Seems almost too good to be true?

[KTlin]

By the description in the manual it sounds like it only takes care of phase issues generated by itself, not on the existing stereo waveform input.

That description seems rather inaccurate to me. I mean, the "Prevent Antiphase" option certainly does resolve some out-of-phase issues on the existing material, even when all the other settings in Ozone are left untouched. It's easy to test it: just apply the Imager to some sounds that contain out of phase frequencies, then turn on and off the Prevent Antiphase option and watch the correlation meter; it should move closer to 0 when the PA option is enabled.

Seems almost too good to be true?

Well, don't expect some black magic to happen, though. As I said earlier, it seems to work only on certain materials that don't have huge out-of-phase issues. If, for example, you play a continuous pure tone that has one of the L/R channels inverted, the PA option won't have any effect.

[woggle]

audition has a range of phase correction and viewing tools - very sophisticated.

https://helpx.adobe.com/audition/how-to ... is-cc.html

izotopes' RX might let you change phase in very tight bands for any amount of time - not sure on the visualisation side of that, but you can rotate left and right phase independently and that should be able to be applied to a selection of the time-frequency spectrum - I have not tested this