• Register
  • Help
Page 1 of 2 12 LastLast
Results 1 to 10 of 13

Topic: What does it mean to "collapse the stereo image"?

  1. #1

    What does it mean to "collapse the stereo image"?

    Is this just a fancy phrase for converting stereo to mono? And is there a proper way of doing it?

    I sometime just load a stereo Gig into a gigastudio slot, trigger it from a MIDI track panned 100% left, and take the mono result out of a single mono output into my soundcard mixer, so that I can work with lots of mono rather than stereo tracks and control panning in the mix.

    Is this the wrong way to do it? Truth to tell, I\'m not really happy with the sound I get this way and I\'m investigating various possible reasons why. Is there a better way to get from stereo samples to a mono signal in the mix?


  2. #2

    Re: What does it mean to "collapse the stereo image"?

    When you pan two channels hard l or r, you actually mute the quiet channel, thus not only loosing possible stereo but also sounds that\'s in there. Perhaps, you could assign the two channels to seperate mono tracks and pan both. Then you don\'t loose any info. Sometimes, depending on the source, it sounds awful though, mixing the left/right together. Depending on how the micing was done, the result will be from perfect mono safe to ugly chorus effects.

  3. #3

    Re: What does it mean to "collapse the stereo image"?

    Collapsing the stereo image actually means partial mixing of the two channels (i.e. final_left = original_left*k + original_right*(1-k), final_right = original_right*k + original_left*(1-k) to yield a new stereo image that is more focused toward the center. The amount of mix i.e. coeff, can vary from 1.0 (preserving the original image) to 0.5 (yields two identical channels, corresponding to a mono signal), to 0.0 (channel inversion).

    This can be easily carried out in CoolEdit:
    Effects->Amplitude->Channel Mixer.

  4. #4

    Re: What does it mean to "collapse the stereo image"?

    Does that mean that in order to do it with gigastudio sounds you have to export the samples individually, collapse them in an editor such as cool edit, then re-import them?

    I was under the impression it could be done \"on the fly\" just by the way things are routed. Sounds like a lot of work.

    Is it something people do just to take the individuality of ambiance away from particular sample libraries, so that they can be made to sit in the same space better?

  5. #5

    Re: What does it mean to "collapse the stereo image"?

    If you only need to collapse the stereo field in Giga, there\'s a much simpler way. Click on DSP station and you\'ll get a series of vertically oriented sliders. Above each of them, there are two smaller, horizontally oriented sliders, marked as r & l. By simply moving both cursors towards the center, you\'ll get collapsing of the stereo field.

  6. #6
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Dallas, Texas

    Re: What does it mean to "collapse the stereo image"?

    Often, however, the microphone techniques used in a given setting are not mono-compatible. The mono-technique can actually be the better choice, both at those times, and in any number of cases.

  7. #7

    Re: What does it mean to "collapse the stereo image"?

    Thanks everyone.

    As an experiment, I took some of the samples I was unhappy with and A/B\'d them through two different routings. One was just panning them left in Giga and sending them out of it mono (as I\'ve done so far in my tracks), the other was sending them out of Giga stereo to my mixtreme soundcard, and using the Soundscape mixing application to mix this stereo input to a mono output.

    It made no difference that I could hear - I certainly didn\'t notice anything \"left out\" in the first routing that came through in the second. I\'ll try Giorgio\'s method of collapsing them and keeping them stereo and see how that sounds.

    But I suspect the problem may be more in the way I\'m mixing them.

  8. #8
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Dallas, Texas

    Re: What does it mean to "collapse the stereo image"?

    We may all be saying the same things, but just one point I\'d reiterate is that even with stereo panning, there are plenty of cases in instrument sample libraries where you don\'t want to pan the stereo channels. If the library is recorded with an a/b type stereo (spread mic), then generally, panning that stereo track\'s individual channels together will start introducing comb-filtering artifacts because a/b stereo is a stereo of phase as well as directionality, vs. something like mid/side or X/Y. The latter are \"stereo of direction\" but of common phase, and do not create the same kinds of comb-filter artifacts.

    Or you may be able to actually live with the comb filtering. But two channels are not necessarily better than one when you\'re building an image of the complexity of an orchestra.

    Say you have a clarinet sound, which is recorded in spread stereo and is naturally panned to the left side by the recording process when both channels are played unaltered. That means the left side is more direct, the right side is more ambient. In essence, that gives you two good mono choices, one which will be drier and closer, one which will be wetter and more distant--in addition to your stereo \"left leaning\" image. But panning that \"left leaning\" image towards itself could have radical tonal effects, and create a hollowing or shifting of the ambient content, which would be avoided by simply picking one side or the other to use in mono. Subsequently, your mono choice would be \"re-stereoized\" by feeding it into the reverb processor you\'re using to create the entire image.

    I always encourage library producers to consider using mid-side. GigaStudio will decode mid-side in realtime if you program the library for it, and in my book, that offers the end user really excellent flexibility. You could essentially rig the mod-wheel or another controller to determine your stereo width, and just sit there in realtime dialing it in per instrument, in context, as the mix plays.

    The only downside is that some samplers don\'t decode mid/side natively, so you\'d have to batch out different files for other samplers at a fixed width. And Giga\'s current mid/side implementation is a bit wonky--you have to program it in dual-mono and use a dimension split to handle the decoding. I have asked them to simplify it and to achieve the split on a single-dimension program, but they obviously already had a lot on their plate, and I don\'t know if it happened. Another possibility is decoding the mid/side split via GigaPulse. I\'ll see if I can get that info.

  9. #9

    Re: What does it mean to "collapse the stereo image"?


    I\'m missing something. Would it be a problem decoding the mid/side audio, pre the sampler? Why decode in Giga? When decoded, the samples are fine to be passed on to any platform - mid/side capable or not. No?

  10. #10

    Re: What does it mean to "collapse the stereo image"?

    Originally posted by KingIdiot:

    If you have an ambient sample (say trumpets), that are recorded slightly to the right of the mics, there will be more direct signal in the right channel. If I take the MIDI panner and pan it 100% left, I will lose the right channel completely.

    Anotehr example (and more obvious) is to have a stereo mix of two guitars playing. Acoustic on the left channel and Electric on the right. Using the MIDI pan control, if I were to pan 100% left, the Electric guitar woul \"disappear\".

    this is what will happen with basic Power Panning of Stereo signals. It turns down the channel are panning \"away\" from

    If I go into the DSP mixer (or use a mixer with stereo panning controls), I can take the right channel\'s pan controller, and move it to the left. This forces the right channel to come from teh left, rather than not sounding at all.

    Stereo Panning allows you to shape the stereo widthe AND position the signal in the overall stereo mix.
    <font size=\"2\" face=\"Verdana, Arial\">King, do you mean that you pan the right channel to the left AND the left channel to the left, and output it as mono? Or do you mean what was referred to before - that you pan both channels to the middle and output it as stereo, but with a narrower image? (Or do you mean something else entirely?!)

    This stuff is more complicated than I thought.

Go Back to forum

Tags for this Thread


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts