Splitting Bass into Low/Mid/High

[FeedTheFaithless]

Hey,
I was wondering:
When you want to process the low, mid and high frequency content of a bass track seperately and copy it to three tracks (one for every frequency range), what to you use to limit the bandwidth of each?
Is it just done by using Logic's ChannelEQ (or similar) and setting steep lo- and hicut filters?

For example:
Low: hicut @ 100hz, 48db/oct
Mid: locut @100hz, hicut @ 800hz, both 48db/oct
High: locut @ 800hz, 48db/oct

Are these too steep? I'd imagine the steeper the better, but wonder if there are any negative side effects?

How do you do it?

Best regards,
FTF

[Sonika]

Yeah I just started toying with frequency splitting and I just cut off the other two freq groups in each track and split it into different tracks. Works fine for me, it's a really great method

[atticuh]

Are these too steep? I'd imagine the steeper the better, but wonder if there are any negative side effects?

By traditional standards, yes, that much resonance can definitely make your mixdown process more difficult by giving you huge PEAKS where there is already frequency crossover (which are usually problem areas already). At this point, you may be asking yourself WHY would a huge peak spring up where you're obviously making a ridiculous cut with your Alpaca VST. It has a lot to do with phasing and phase response.

To expand:

The only response to this that makes any sense is, ‘it depends’. Each case should be taken on an individual basis, and if you are going to roll off, do it only as gently/hard as necessary. Or do as much as you can without causing other negative effects. You don’t HAVE TO high pass everything with the sharpest possible filter. There’s always a cost. It depends Especially on a full tune, it should be clear from all of this that a low end high pass affects all the stuff above it. Seems crazy, but it is true.

[personal rant]People blindly piling up 58 high passes to make ‘da ulltim8 hipasz’ are very often causing themselves no end of problems. This stuff all applies whether you do it on low end, or snares, or whatever. The effect is generally more marked at the very low end due to longer wavelengths and higher amplitudes, but it is still there. Smearing the bottom end of a snare right up through the mids because you read you have to eliminate all frequencies below 200Hz = And you *may* be increasing peak levels anyway while mushing your sound up. So, why bother? Gentle rolloffs where necessary will usually do.[/rant]

The last thing to mention is, as radioactiv mentioned, using a linear phase eq. This obviously gets completely around this particular problem, though one should be very careful as it can introduce problems of its own with pre-echo etc. This really really pisses me off when you want a tight bottom end and suddenly things are speaking before they should. It’s a different kind of mush, but it is still mush. Not uuuuusually a problem with shallow HP shapes, but something to be aware of. No such thing as a free lunch etc.

[didi]

^^ This. basically.

Actually, Use a multiband compressor. Send your signal into three mixer tracks. Then stick a multiband compressor on each one, with high/mid/low gains set to -inf dbfs (except for the range you want on that channel).

This will give you 100% precise splitting, no phasing issues etc. etc.

Voila you have split a bass.

(btw if you use the search function you will come across an infinite number of better worded answers than mine)

[FeedTheFaithless]

Thanks for your replies guys!

Yes, the filter resonance was what I was afraid about. The idea with the multiband compressor seems really great, will definitely try that!

(btw if you use the search function you will come across an infinite number of better worded answers than mine)

I tried a quick search and couldn't find something exactly concerning this question but I have to admit I didn't invest too much time in digging through all the threads. Will promise to do so next time

Best regards,
FTF

[bri1]

Whats the purpose of splitting your bass track into low/mids/and highs? Would there even been Highs? Sorry if this is a stupid question, It clearly would give you more control over the sound. I'm just curious as to if there are specific reason for people to do this? What problems would this fix?

[Sonika]

Yeah freq splitting is for more in depth editing of the sound

[Acidhedz]

I usually use different synth for each part. If I am doing a more mid range sound I work on that first then add in a sine wave playing the same thing to give it a deeper sound, after that it's just a matter of adjusting the levels.

When I use a loop or sample I usually make a duplicate and just use the basic 3 band eq in Acid MS to cut out the bass freq on one, then cut the upper freq on the other, sometimes I use 3 copies and work with each freq band separately.

I use Acid Music Studio so I have F*** all when it comes to automation or fancy fx, I just screw around with sounds till they do more or less what I want.

[Ghost of Muttley]

If you don't have a multiband, or don't like to use one you can run the signal in parallel and invert the phase of one signal alongside the one you've split to get a more accurate cut too.

[bri1]

If you don't have a multiband, or don't like to use one you can run the signal in parallel and invert the phase of one signal alongside the one you've split to get a more accurate cut too.

I'm going to have to check how to set that up in ableton. Not sure if its possible with midi, does anyone know. Im going to check though the manual later this evening but if anyone knows that would make things much easier for me.

[Ghost of Muttley]

Yeah Ableton makes life easy with racks.

Make a rack on your bass track, two chains in it with an eq8 each (could prob do it with a filter too, personal preference).
One low cut somewhere between 100-200 hz, one high cut at the same freq.

Then on the chain with the high cut create another rack. In that rack you want to create a second chain that's just dry running parallel with the high cut. Put a utility plugin on the dry chain and invert the phase. So you end up layering two signals, one with all the freq's and the other with just lows. When you invert the phase of the dry signal it will cancel the lows and just leave the high's.

It's not as convoluted as it sounds.

I think I remember watching a tutorial on this awhile back, i'll see if i can find it.



You can do a similar thing in Ableton by putting a bunch of multibands in racks on one channel and soloing different frequency bands on each one

I think using racks as opposed to duplicating the channel using more cpu though, which is a bit odd and slightly annoying..

[Ghost of Muttley]

churrrr

Large up Mr Bill

[bri1]

badass, thanks guys Im going to try that out this evening hopefully!