sidechain "multi-band compression" using NS effects?
hi all - a newbie mixing question here. I have a bass part with sounds across the whole frequency spectrum ( lows and highs) and I would like to sidechain-compress a bit for my kick drum to come through - but I've read that the best way to do this would be to only duck the lower frequencies (those which clash with the kick) rather than to duck the whole bass sound every time. That makes sense to me, as I can hear the higher frequencies in the bass part doing some unpleasant pumping effect when I use a sidechain compressor on the whole track.
My question is how to do this in NS2 without using external AUs (such as multi-band compressors) - because external AUs cannot be sidechained/automated ?
One idea could be to duplicate my bass sound into two tracks, one with a steep low pass filter and the other with a steep high pass filter, and only add the sidechain compression on the track with the low frequencies in it. But is there a simpler, more elegant way?
I have the same question for some wide-frequency spectrum pad synths - can I sidechain-compress only the lower frequencies of them, or do I have to sidechain the whole thing ( leading to pumping effect that I don't always want).
thanks in advance for your tech-knowledge :-)
Comments
@bobheads that One idea you mentioned with duplicate of that very same sound is what I would do. There is SC FILTER on ns2 compressor but I believe that is for incoming signal (of that side chain), for second part of your question that applies too as well since that is pretty much same scenario:/ would be nice to have native multi band compressor/ dynamic processor a very essential feature indeed;)
Luckily , that is if you using obsidian then you not going to runout of cpu;)
You can use send tracks to split the signal so you don’t need to duplicate the source. Just turn down the level fader on the source and set the sends to pre fader. Or use the source level as a wet/dry control (not volume matched, though).
Then if you want to get super efficient, save the channel layout into your default project so you only have to set it up once and it’s there for every new project.
This is actually something I started to try out yesterday, so it's good to know I am starting to think like a pro.... ;-)
I understand the first part but not sure exactly what you mean by 'volume matched' - you just mean that I have to manually balance the two tracks (source track and send/effected track) using their faders? If Yes, how would go about working out how loud to make each one/the total sound - maybe solo them both and watch the master meter?
Second Q (detail) - If I use only in-house NS effects (Stereo filter) so that I can use automation - then I have a question about setting the low-pass and hi-pass filters. They don't have 'brickwall/squarewave' options for the cutoff, so I have just set them as steep as possible around 100Hz, and they therefore overlap with each other a bit there - do you think that is okay or could cause a problem? ( it looks like this x instead of v, if you see what I mean!)
that's a top tip that hadn't occurred to me - great stuff!
Thanks so much both @Stiksi and @Cray23 for your thoughts and ideas
Yes, that’s it. I would maybe group the source and the sends together under a common parent track, then find a good wet/dry mix and use the group track’s fader to set the overall level.
This is actually a preference question because the steeper the cut, the more artifacts it will produce. This applies to all filters, not just the NS filter. NS’s filter goes all the way to 48 dB which is probably about as steep as any ”brickwall” filter will go but it’s good to judge the necessary cut with a critical ear. Less steep also means you get a little more leeway in the crossover point without affecting the overall EQ. Personally, I usually settle around 24dB, I use 48 if I want to cut off a heavy bass element or something like that. The cut dB number is per octave, so if you cut 24dB at 500Hz, it’s 48dB at 250Hz, 96dB at 125Hz and so on.
An X-shape is the normal kind of crossover point, it’s like a crossfade, which masks the split. If it’s a V-shape, you are introducing a notch (dip) in the EQ, which may or may not be what you are looking for. You can usually fiddle around a bit to find a good neutral crossover point or use it to dramatically change the sound if that sounds better.
The cut dB number is per octave, so if you cut 24dB at 500Hz, it’s 48dB at 250Hz, 96dB at 125Hz and so on.
@Stiksi Thanks so much for your responses.
In case anyone finds this topic interesting (and also for my own brain), here is a summary of what I ended up doing:
Bass track was a group of two bass synth sounds, already with an automated effect chain including filters following the song arrangement, and sidechain compression.
I removed the sidechain compression from the original Group track and then made two Send tracks which were fed by the original, the sends I set as 'Pre-fader' so I could put the fader on that original group track down to zero and still have the sounds sent to the Send tracks. Now I have two Send tracks, also grouped into one fader so I can mix the bass with one fader. The sound coming out of that fader/track follows the automation and filters already attached to the source track.
On one of the Send tracks, I hi-passed at 150Hz (tried 100Hz but it didn't work as well until I changed to 150Hz) and left it as is. On the other Send track I lo-passed under 150Hz and added a sidechain compressor activated by the Kick track. this means the lower part of the bass sounds always ducks when the kick hits. I adjusted the compressor to minimise any obvious pumping. I used a pretty steep filter but the sound works for me.
Lots of trial and error, but I think I got a good result in the end. I might even think about doing this on some of the higher-freq pad synths now! :-)