How to match dissimilar vocal punch-in with original vocal?

Discussion in 'Tracking / Mixing / Editing' started by DogsoverLava, Aug 22, 2016.

  1. DogsoverLava

    DogsoverLava Active Member

    On one of the Stairway mixes this particular vocalist had to re-do a section of the vocal and what I was provided from her sounds environmentally and tonally very different from the original vocal.

    The good news is that this is in the heavily effects laden final vocal part so I can mask some of the difference but it's still noticeable. I thought it would be a simple task to essentially match it using EQ but I'm having difficulty. Is there a specific strategy or technique for this?

    Is there a common technique or goto technique for this?
     
  2. DonnyThompson

    DonnyThompson Distinguished Member

    Does your DAW have a frequency matching tool? Some FFT Eq's will analyze one section , allow you to save the analysis curve, and then apply it to another section:

     
  3. Boswell

    Boswell Moderator Distinguished Member

    Recording the same singer performing the same vocal part in two different studios with different engineers is going to result in more than just EQ differences. In fact, the EQ is likely to be one of the things that stays much the same, especially if you had asked for the track to be recorded raw with no added effects or dynamics.

    Differences in studio acoustic, model and positioning of microphone, and type of pre-amp are all going to contribute, and none of these is easily compensated for by twiddling with EQ.

    You can try the frequency-shaping tool that Donny suggested, but I think your only chance after that lies in the effect processing that you mentioned, and that may also mean some adjustment to the other tracks to get this one to fit in with them.

    What was the reason for needing the re-dubs? Perhaps you could start there. For example, if it were pitch correction, you could use the pitch envelope from the re-dubs to modify the pitch on the first track, keeping its original acoustic. Similarly with timing - taking the better timing from the re-dub and using it to slide the timing around on the original track could well result in a better fit than simply pasting in the newer version. Correcting wrong words is more difficult, but not impossible.
     
  4. Sean G

    Sean G Well-Known Member

    Personally, I'd re-track the vocal from start to finish.

    IMO there's no such thing as "fix it in the mix"
     
  5. DogsoverLava

    DogsoverLava Active Member

    Unfortunately Sean I wont be able to get any additional re-tracking with this vocalist. That FFT filter was pretty cool Donny - I'm using Reaper and it does have one. I've only used it in subtractive mode in the past as almost a noise gate on some acoustic tracks so it was interesting to play around with it and develop some more skills and have options for future use. In the end I found a decent balance in a couple EQ settings and through the use of an eventide plugin was able to kind of mess it up enough so that there's a less obvious difference... plus it comes right at the wettest part of the mix. The only thing I didn't try was squashing it with compression to take some of the life out of it (which I wonder may have worked as well?)

    The reason for the overdub was that she sang the wrong melody with two lines (right at the end). I almost wondered about the possibility of taking the pitch envelope itself and applying it as the correction but I don't seem to be very adept with the Reaper plugin for pitch. I may go back to that as a little learning project as I suspect the attempt will offer other learning opportunities. I can somewhat manually correct pitches by drawing the correction but I either don't have the skill & experience to be any more affective than that or the tool is perhaps limited in what it can do.

    I also had a couple of mispronunciations - the equivalent of wrong words but the overdubs were just too dissimilar and in a naked section of the track that I just went with the original take and wrote it off as part of the unique charm of the vocalist. Thanks guys - some great suggestions and a good point of learning for me.
     
  6. Sean G

    Sean G Well-Known Member

    SPL do a plug-in called De-verb...maybe you could use it to remove the reverb from both vocal tracks you are comping and then apply a common reverb once you have them comped together.

    This may help with getting them to sound like they are in the same / similar environment.

    Just a thought.

     
  7. pcrecord

    pcrecord Don't you want the best recording like I do ? Well-Known Member

    I'd try to put the takes on seperate tracks and try to make them sound the closest within the tracks ajustments and then send them to a buss and apply some common effects there. Some saturation, delay, slight chorus, reverb etc.. anything that can create a signature sound and make the tracks sound the nearly the same...
     
  8. DonnyThompson

    DonnyThompson Distinguished Member

    Just so I understand...

    Are you saying you printed the effects onto the vocal track?
     
  9. pcrecord

    pcrecord Don't you want the best recording like I do ? Well-Known Member

    Another thing that could lead to an unnatural vocal track is if the tracks have been edited to remove too much.. ex : no breathing or end of words cut too short.. etc..
     
  10. DogsoverLava

    DogsoverLava Active Member

    No -- just that they
    No -- just that they will be bathed in effects - the original tracks were raw. I did what was recommended and put each part in it's own track separate from the rest of the vocal then EQ'd each track individually to match as close as I could, then sent them to a common bus where I applied my FX etc for that section ... it sounds better for sure and blends better but it does also show the inherent issues of working with what I'll call compromised tracks. I'm just saying that I was lucky that the issue was in this section of the song. There was also one punch-in available to me in the main vocal (pre-solo) and there was just no way at all to even pretend to get them to sound even remotely close enough. The result is the Michelle PV7 file on the Stairway mix thread.
     
    pcrecord likes this.

Share This Page