Frequency Splitter / Multiple Sends

I kind of missed this feature when i was trying to figure out how to do it on my own using 2 send devices on the same track.

What i was looking for was a way to send frequencies below one point to S01 and higher frequencies to S02. Well i figured it was doable by sending the same track to 2 seperate send tracks, and then filtering it with highpass on one, and lowpass on the other, but it would be really neat to have a special device where you can add multiple sends in different frequencies.

For instance sending …
40-200 hz to S01
200-3000 hz to S02
3000-5000 hz to S03
and anything above to S04

or whatever configuration and you want on as many send channels as you want.

Would be really neat when working with some old Mono Drumloops for example, then you could easily control stereo widht, panning, compression Reverb, or whatever on certain frequencies in that drumloop, or whatever sample or melody you want.

would also be cool if you could control the rate of bleed, hmmm… or what i mean is how sharp the edges between these freqency ranges would be.

Could be quite useful for mastering as well since there is no built in multi band compressor. Could aslo be used for some nice Panning automation on high frequencies in a Synth Pad or a lead synth for example.
There’s always the possibility to do some stuff yourself by using a duplicate of the track and filtering on your own but i do think there could be some things done more easilty with this kind of device.

You can ever do this isolation in a little more complicated way

channel 01 >> send to S1
S1 >> send to S2 (keep source) >> Eq band pass 0 to 200 hz >> effect (ex: sidechain)
S2 >> send to S3 (keep source) >> Eq band pass 200 to 3000 hz >> effect (ex: harmonic exciter)
S3 >> send to S4 (keep source) >> Eq band pass 3000 to 5000 hz >> effect (ex: stereo enhancer)
S4 >> Eq band pass 5000 to 20 000 hz >> (ex pingpong delay) ;)

you have to use a very straight EQ (db - inf) and a Q (+inf) which provides a good isolation of frequencies to prevent interferences. Try with a disco loop full of all frequencies it’s amazing how it works well !!!
But yes… if thats automatic it s better :)

i think its not 100% the wished method. because when you ‘keep the source’ the original audio will still sound underneath the isolated/removed frequencies.
but a good technique too.

Here’s a simplified version of what I usually do. Please forgive the ASCII :)

(On Track 1)  
Send Device > EQ 10 > Extra FX > Send Device  
Dest: Send Track 1 > Remove all > on only the > Dest: Send Track 2  
[x] Keep Source > high frequencies > low frequencies >  
  
  
(On Send Track 1)  
EQ 10 > Extra FX > Send Device  
Remove all > on only the > Dest: Send Track 2  
low frequencies > high frequencies >  
  
  
(On Send Track 2)  
Low + high frequencies get mixed back together here.  
Now we can apply any other 'master' FX like compression, etc.  
  

It doesn’t always work perfectly because unfortunately there is usually some slight overlapping of the frequency bands in the EQ. You could instead use a third party VST filter which has a very steep cutoff, such as Rubber Filter.

.

the original audio won t still because you cut frequencies after sending it.
the keep source is a way to send the same sound to all the send channels.
on the send 1 only the bass frequencies are send to master
on the send 2 only the mid frequencies are send to master etc etc

but as dblue said you need another powerfull filter to do it well

+1 for the original poster’s idea.

Nothing that can’t be done already as others have said but still, it would be very nice indeed.

Well, why waste filters if you’re always going to have them on? Just run the filter several time in the sample editor on your selected samples…that should make a steep cutoff point. Use the spec-an to make sure that the samples don’t overlap.

the correct way is to substract the filtered wave from the original wave, something that is currently not possible with renoise.

sry i misunderstood things. thanks for that tip again.

+1, would be an excellent time-saver

yes man !!!

okay, so here is a little workaround :

http://www.loozabeats.de/temp/filtersplit.zip

as I said, the only proper way to achieve a split at a frequency is to substract the filtered wave from the original wave.

Track 1 plays the loop and feeds it onto Sendtrack 1+2. In Sendtrack 1 it gets filtered and then inverted using the “Fleisch-Negator”-Plugin I put in the zip. Then the filtered and inverted wave is added to sendtrack 2 resulting in the original loop minus the filtered parts. The filtered part is inverted again and you have the full loop, split into two frequency bands in sendtrack 1+2, ready for whatevery you want to do with it. Hope this helps for starters.