1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

Phase! somebody said Phase?

Discussion in 'Recording' started by Serge locat, Sep 19, 2002.

  1. Serge locat

    Serge locat Guest

    Phase problems/anomalies are pratically a thing impossible to correct on a ProTool daw(and maybe on most daws).

    Here I don't mean just switching out of phase or tweaking with sample shifts BUT SUBSAMPLES SHIFT, expressed in samples or, the better, in degrees rotation.

    Someone has clues where to find TDMplug-insusing DSP)?
     
  2. A couple of possible solutions:

    Check out digidesign.com and look up SpectraFoo from Metric Halo and Surround Scope from Digidesign. Both have scopes on them.
    Phase correcting at subsample levels seems like a tricky deal. You could perhaps try rerecording the original track to a higher sample rate (44.1 kHz rerecorded on 48 kHz) and slosh the 48k track by a sample in either direction. Using 48k would allow you to bump it at finer than one sample of 44.1k. You could feed the reference signal and the 48k signal to a scope and take a peek at phase. Just a wild thought, never done it. Doc
     
  3. Serge locat

    Serge locat Guest

    Hi DOC!

    Well! I don't wanna "see" the phase problem, I can hear it anyway! I want to correct it!

    It I play and shuffle with sample rate, I destroy the fidelity!

    There should have a plug-in who does some sort of Oversampling (like 180 times higher to the session's samplerate) then I could shift by 1 degree(at least) and between this, just revert phase, then add some other degrees(coming hotter!) just to be able to slide 360 degrees total!

    Not so complicated! Just DSP power!!! :D :D :D
     
  4. mrbeecroft

    mrbeecroft Guest

    Can smebody please explain Sub-Sample phase shifting to me?

    As I understand, one sample is the most basic time related unit for digital audio. The only way I can imagine you could alter the audio with finer detail, is to is to increase the resolution and store(process) it using the new, higher resolution format.

    I suppose Jitter on the other hand could be desctibed as Sub-Sample Phase Anormalities occuring during the transfer of digital audio, but that relates to the PCM signal itself, not the audio that it represents.

    Am I missing something?

    Tom
     
  5. Serge locat

    Serge locat Guest

    I'm just trying to imagine a way to shift in the "sub samples" maths.

    I'm not an ingineer, i'm a musician and mixer.

    I've just imagined to take for example(like they do in converters) an oversampling process. Let's say 128 times oversampling.(You could go higher than that). Then while in the higher and now new high sample resolutions(128X 44K for ex.) you could say shift theese samples by one sample increement, then store temporarely this information, then RE-SAMPLE this back to 44K (or any original session's sample rate) an then Bingo! you have done sub-sample shift. :p
     
  6. Serge locat

    Serge locat Guest

    Excuse my poor english(I'm french) :w:
     
  7. Serge,
    As regards your earlier reply:
    "I play and shuffle with sample rate, I destroy the fidelity!"
    The solution I suggested would definitely require going analog. I agree that a phase correction plug-in would be terrific. Do you have a specific situation that is driving your question? Doc.
     
  8. Serge locat

    Serge locat Guest

    As I said in a previous post
    this should not destroy anything. And I'm not talking to re-sample from analog domain, just re-sample it in a higher sample rate just to be able to slide the audio file in a smaller increment than the original samplerate. Then re-sample bask it to the original SR.

    I just submit a way to do it. I'm not guru of digital and software programmation but I give a possible way to achieve it.

    IMHO
     
  9. Serge locat

    Serge locat Guest

    And in this process, there is no jitter induced because all the information (samples) is done in digital domain. This is just calculations, no conversions from analog to digital.

    I took the example of an A/D converter because there is a possibility to go much higher than 192Khz SR.

    But "this" very high SR is only for shift purposes.
     
  10. I suspect that what you do to correct the sgnal at one spot on the wave will cause a corresponding mistake at some other part of the wave, just considering the basic physics of sound waves. Of course you may be able to get it to sound better than it originally did... Best of luck! Doc.
     
  11. Serge locat

    Serge locat Guest

    The purpose of that, is to bing the phaseshift done by different mikes position from the source to zero, so no comb filtering or smearing occurs. If the target is to remove phase anomalies then the Phase Corrector Plug-in would do the job. If the phaseshift is needed(like stereo micking) to do what we want, great! :)
     

Share This Page