9 Replies Latest reply: Jun 21, 2014 4:13 AM by SteveG(AudioMasters) RSS

    parallel compression

    therealdobro Community Member

      I'm getting what I think is really bad comb filtering using two different compressors in Audition CC, so I'm wondering if I'm doing something wrong.  Here's my procedure:

       

      1 Izotope tube-modelled compressor on a mono bus.  Threshold -60, Gain +5, Ratio 5:1, Attack 5 ms, Release 50 ms.  Bus Volume at 0 dB.

       

      2 Vocal's being sent to that bus pre-fader.  Send level's at 0 dB.

       

       

      I'm getting about 10 dB of gain reduction with this routing.  If I mute the compressor, the gain jumps up 10 dB.  I'm thinking it must be comb filtering, but what do I know?

       

      I've tried the same setup with a second compressor (Fabfilter compressor) I get about 5 dB of gain reduction.  But if I'm doing this right, and if the plugins are compensating for latency, then I should be getting no gain reduction at all, right?  (In fact, I'm thinking that I should be getting a modest gain increase because of the extra channel (the compressor bus) being added to the dry signal.  No?)

        • 1. Re: parallel compression
          Bob Howes MVP

          Compressors reduce gain--that's how they work and why there's an output gain or "make up gain" control on them.  Although it's easy to think of them "bringing up the quiet bits what they in fact do is reduce the level on all parts of the signal above the threshold you've set by the amount you've set.  You then add the output gain to whatever amount is needed to give you the signal at whatever level you want but with reduced dynamic range.

           

          Comb filtering generally shows up, not as reduced gain overall but rather as a very strange frequency response with frequencies alternately boosted and cut.  It's a pretty distinctive sound when you hear it.

          • 2. Re: parallel compression
            therealdobro Community Member

            Okay, so it's not comb filtering.  But what is it, then?  Why would adding send to a compressor drop the level of the dry track?

            • 3. Re: parallel compression
              ryclark Community Member

              Perhaps you mean "phase inversion" rather than comb filtering?

              • 4. Re: parallel compression
                SteveG(AudioMasters) Community Member

                If you're using them in parallel and there's even the slightest processing delay difference (although there shouldn't be) then comb filtering is a possibility, because you'll have the equivalent of a 'flange', only fixed in time.

                • 5. Re: parallel compression
                  therealdobro Community Member

                  Is it this: the Izotope compressor doesn't have any make up gain setting.  Is that it?

                  • 6. Re: parallel compression
                    therealdobro Community Member

                    Stupid.  Of course it has make up gain.  That's not the issue.  Okay, I really don't get it.  Why would sending the dry track to a buss with the Izotope compressor on it  drop the level of the track so significantly? 

                    • 7. Re: parallel compression
                      ryclark Community Member

                      Because the Izotope signal is in anti phase with the direct one, thus causing partial cancellation of the audio signal. Because the Izotope signal is compressed the level is not equal to the direct one all the time, hence not complete cancellation.

                      • 8. Re: parallel compression
                        therealdobro Community Member

                        "Because the Izotope signal is in anti phase with the direct one, thus causing partial cancellation of the audio signal."

                         

                        Hm?  Anti phase?  What's that?  And if it's in 'anti phase', why is there not complete cancellation?  And why does the Izotope compressor do this anyway?

                         

                        "Because the Izotope signal is compressed the level is not equal to the direct one all the time, hence not complete cancellation."

                         

                        Hm?

                        • 9. Re: parallel compression
                          SteveG(AudioMasters) Community Member

                          therealdobro wrote:

                           

                          "Because the Izotope signal is in anti phase with the direct one, thus causing partial cancellation of the audio signal."

                           

                          Hm?  Anti phase?  What's that?  And if it's in 'anti phase', why is there not complete cancellation?  And why does the Izotope compressor do this anyway?

                          He means that it's signal inverted - actually it hasn't got anything to do with phase at all. No I don't know why, and I haven't tried it - although it seems somewhat unlikely, I must say.

                          "Because the Izotope signal is compressed the level is not equal to the direct one all the time, hence not complete cancellation."

                           

                          Hm?

                          That's easier to answer. Compression inevitably means that the signals at different amplitudes are going to be at different levels when it comes out of the compressor. If they weren't it wouldn't be compressed... so cancellation would only be complete when the inverted signal was at exactly the same level as the original non-inverted signal. Where there was a level change so that the signals weren't at the same level, then cancellation couldn't be complete.

                           

                          Mind you, I'm not convinced (without hearing it) that this has anything to do with what's going on at all...