Syntrillium Backup Forums :: View topic

doctor_b · Location: USA

I've seen a number of neat delay and stereo-widening tricks mentioned in these posts, by Voodoo and others, where you make multiple copies of a track and delay some of the copies by a few milliseconds. I also like the trick where you record the same guitar part twice and pan one recording hard left and the other hard right. But as Graeme has emphasized, this may lead to tracks that are out-of-phase. I had done some experimenting myself, mainly with guitar tracks, and obtained great results and was feeling rather proud of myself until, late in the game, I listened to my mix in mono and heard things disintegrate from total or partial phase cancellation. Bummer!

My question is this: is there anyway to avoid these phase problems when using such tricks? I tried to move around my delayed guitar tracks ever so slightly, and match up the wave peaks, but this did not seem to solve the problem. Invariably, there must be other parts of the tracks where the wave peaks are not in sync.

Or is it simply hit and miss, where you can get away with it on some tracks but not others? If I wish to avoid serious problems with phase cancellation, I'm wondering if I should generally avoid such tricks, even though the effects were pretty darn cool in stereo -- and keep things raw and simple. Thanks in advance for your advice and suggestions. Great forum!

Syntrillium M.D. · Location: USA

zemlin · Location: USA

[ignorant] One thing I've been thinking about recently ... when starting on a mix would it be wise to set EQ, compression, other non-spatial effects, and levels with everything set to MONO? The theory is that when it sounds good as a MONO mix it would be a good starting point for a stereo mix. Then set panning, reverbs, and tweak levels.

Does this make any sense?
[/ignorant]
heavy padding has been secured in all sensitive areas, so feel free to flog ... it won't hurt

SteveG · Location: United Kingdom

Following on from what Synt said, I think that I can define the problem for you - but you will still have to experiment with your tracks somewhat, as he's pointed out. And as you will see, cymbals are going to be one of your larger problems...

If you have a time delay between two copies of the same signal, the equivalemt length of which is the same as any of the frequency content contained within them, then you will get cancellations and phasing problems. It's about that simple.

And what this means is that all of the really short delays are likely to cause you problems with HF signals - and cymbals have loads of HF spread over a wide frequency range - a bit like white noise. If you want to work out what the cancellation frequecy is, it's the reciprocal of the delay, so that for instance a 5kHz signal will cancel out, giving you phase problems at 1/5000 S, in other words 0.2mS. Yes, it's a pretty short delay! You can do experiments with CEP quite easily to determine the true extent of your problems, and as Synt says, you have to check things carefully on a mono mix, because this is inevitably where your problems are really going to show up. Technically, I could show you that there's actually a comb filtering element to this as well, but usually, if you know roughly what the harmonic content of your signal is, you can make a basic stab at seeing whether you are likely to get cancellation effects.

This also becomes something of an issue with mic placing - it's very easy to achieve the same effect with two mics at different, but fairly similar distances to a sound source. Here, you are interested in the signal's propagation as sound, so the sum is slightly different. The formula is Velocity of Sound = frequency x wavelength, and to save you working it all out, I'll tell you that in air, the wavelength of a 20kHz sound is 16.5mm - a little over half an inch. And it is the fact that these distances get to around the 'few inches' points that sometimes gives rise to some of the strange effects heard when recording drums with multiple mics.

SteveG · Location: United Kingdom

zemlin · Location: USA

I know CHECKINGa mix in mono is a good idea, but what I seem to hear folks saying is they mix in stereo and check it in mono. One of the things I've struggled with is EQing like sounding instruments so they have their own space. The one instance that comes to mind I had things panned and was relying on that somewhat to isolate the instruments (hammered dulcimer and tenor banjo). I was thinking the EQing might have been perhaps more challenging but in the end more productive if I had been working in mono and not panned things until I could hear each instrument in its own space. Working in mono to EQ would keep any stereo cues from clouding my senses - mono playback compatibility would be an additional benefit, but not really the goal of the technique.

Once things were panned and additional effects added, mono playback could be checked again to be sure nothing was screwed up in the process.

I have a couple of multi-instrument live recordings to mix down soon - I think I'll give it a whirl.

SteveG · Location: United Kingdom

If, as a technique, you find that working in mono and then panning afterwards works for you when setting up EQ, that's fine, because it means that you are almost guaranteed to get a mono-compatible mix out of it in the end.

The only thing I would say is that on more complex mixes, you can find yourself doubling up similar-sounding instruments and then panning them for clarity, which works fine in stereo, but the mix can get in a bit of a state when it's mono'ed up. If you did your original mix in mono, you might even be somewhat reluctant to put these instruments in at all, despite the fact that there may be an acceptable way to do it. I suppose the only thing that's worrying me slightly is that you might be tempted to leave things out of a mono mix, and then put them back into the stereo one with slightly less than wonderful results.

Now I know that most of these particular problems can be solved, but it's usually by repeated mono/stereo checking. Which is why the mono button on my system is external to the PC output, and I still mix fundamentally in stereo. I can't see anything wrong in principle with starting in mono, as long as you're aware that it can appear to shut off some options. I think that this is why Synt refers to mono mixing as a 'challenge' - and until ways of arriving at acceptable mono/stereo compatible mixes were arrived at, they were always treated separately... or were they?

I have a George Martin tribute 6-CD set which contains some faithful reproductions of early stereo mixes - where the vocals are all on one side, and the band/orchestra is all on the other! You don't need a vocal remover to extract one or other, just a balance control. This is typical of early attempts to get around the 'compatible mix' problem - and hilarious to listen to! But if you mono it up, it's fine!