Page 1 of 1

Harbal As A Channel Strip

Posted: Mon May 05, 2008 1:05 pm
by Carmichael
Let's get some opinions on how best to eq a track with harbal. There is brief reference to it here:

http://www.har-bal.com/ipw-web/bulletin ... php?t=1000
http://www.har-bal.com/ipw-web/bulletin ... .php?t=904

But, this sounds like a bigger topic. Here's my 2 cents:

Currently the way I mix is to analyze every track with har-bal. I don't bounce anything extra, I just direct harbal to the proper file in the audio files folder, then I set my track eq to simulate the harbal filter. What's the best filter for a track could be a complex subject, but at the least, it helps to see what you're dealing with.

For me, the best track eq is to treat it like a master. I like a fairly flat response for most instruments throughout their range. So if a vocal comes in at 250 hz, I like to see it relatively flat up to ~ 10 khz (but dont get carried away from 4khz to 9 khz). This forum has some excellent posts about how best to approach this, and I think it takes hundreds of hours of experimentation to get good with harbal.

Most people would disagree with this approach (flat response for each instrument). For instance, this is the opposite of the "give each instrument it's space" advice that is usually given - to notch holes in instruments so other intruments / vocals can peak through. I find that concept is useful at times, especially to protect the lead vocal, but doesn't work for me on a songwide basis.

Anyway, just my 2 cents (and I'm probably wrong). But, I don't know how I would live without harbal. What's your approach?

Posted: Mon May 05, 2008 7:59 pm
by HarBal
As long as by flattening out the spectrum of each track in the mix, you mean in a loose sense then I don't see a problem. The only problem you might have is that if you take it too far (literally flattening out) then you'll lose the character of the instrument.

Given that you are using HB for analysis and reconstructing the frequency response with your channel strip EQ then you are probably somewhat restricted (by virtue of what the channel EQ is capable of) from extreme smoothing, which you could say is a good thing. In any case I have experimented with that approach with good success before so I hear you!

The issue of instruments needing a space of there own spectrally can be valid in some cases but so too can your approach. The issue overlooked in the spectrum allocation view is temporal sharing. Two or more instruments can happily share the same space if they are sharing time (ie. you go first, now its my turn) or if they are playing harmony. When instruments don't temporally complete then the spectrum allocation doesn't matter.

Cheers,


Paavo.

Posted: Tue May 06, 2008 2:12 pm
by Carmichael
Thanks Paavo.

Yes, flattening the curve in a loose sense.

I mostly use IntuitCurve to approach an idealized reference spectrum as described by uncajesse here (my curve is similar to his):
http://www.har-bal.com/ipw-web/bulletin ... .php?t=763

Posted: Sun May 18, 2008 11:31 pm
by Hitmaker
Hi ,

I'd concur with your thoughts that having each instrument present a 'flattened' spectrum does seem to cause less intra-instrument masking .... I'm presuming this is because the masking is reduced within that sound ... so it's also less able to mask others ....
If 'bracket' equing ( top'n'tailin' ) is required , Har-Bal makes it particularly easy to work out where .... and put your shelf on at whatever slope you like ...

Cheers ,

Evan .