Propellerhead Forum

Propellerhead Forum (
-   General Forum (read only) (
-   -   Reason has a hidden high pass filter (

wouterschakel 2005-04-26 00:21

Reason has a hidden high pass filter

This is quite a technical story, if you just want a HP Filter, just download the song (see link below).


I was trying to figure out how the stereo imager devided the sound in 2 seperate frequency bands. I figured it must be something like this:
- the original sound is put through a 12db Low Pass filter, beeing the low band.
- the low band signal is subtracted from the original, beeing the high band signal.
(Note, in this way 'low band' + 'high band' = 'original', this would not be the case when using a High Pass AND a Low Pass filter parallel)
Running sound through the stereo imager soloing the low band creates a signal very very very similar to a signal through the 12db LP filter (on the ECF-42 Filter). However, I couldn't test the same for the High Band because of the very old problem that Reason does not have a stand-alone HP Filter.

But then I realized...

Running a sound through the stereo imager, having both stereo knobs at no activity, turning the X-over knob like hell... does not have ANY effect. Meaning that 'low band' + 'high band' is indeed very accuratly the 'original' sound. Otherwise a slight notch or bandpass effect should happen at the X-over frequency.

Now I know, from creating my own stereo imager in NI Reaktor, that the 'HP + LP deviding method' does produce a signal that doesn't differ alot from the original, let's say about 90% the same.

In short:
- the low band of the stereo imager is 99% like the 12db LP Filter.
- '12db high band filter' + '12db low band filter' = 90% like the original.
- 'original' minus 'low band' is 'high band'

Take these three together leads to the conclusion:
"the 'high band' of the stereo imager looks alot like a 12db high pass filter!!"

Having this in mind and using the params on the MClass EQ for resonance can create a nice 12db HP Filter. Using 2 in a row creates a nice 24db HP Filter. I made a combinator patch that i put in a song:

It's .rps so you will have to copy it every step of the way. Be sure to copy all the numbers from the programmer, otherwise the resonance filter frequency isn't the same as the cut-off frequency or things like that.

One more note: due to limitations of the stereo imager, the cut-off frequency can only be set from 100Hz to 8000Hz. This will do 99% of all possible jobs but will constantly filter out all frequencies below 100Hz. Be sure to put the combinator on bypass when (at that moment not actively) using it on your master or on basses and stuff like that.



wouterschakel 2005-04-26 00:26

Re: Reason has a hidden high pass filter
Did I say max 8000Hz, it's 6000Hz :D

rpugwash 2005-04-26 00:30

hey great theory there...
do you think it would be possible to recreate this in version 2.5 using the VCF filter and perhaps some kind of spider audio config to just leave the high-pass?
just a thought... ive got three but for those of us on 2.5 it would be cool if this were possible i spose :-D

great stuff anyway
thanks for the knowledge buddy


rpugwash 2005-04-26 00:31

Re: Reason has a hidden high pass filter
Did I say max 8000Hz, it's 6000Hz :D

thats forgivable lol

its all relative... bit like incest ROFL!

rpug ;-)

wouterschakel 2005-04-26 00:45

Re: hey great theory there...
Well, it's impossible to 'subtract' one signal from another in Reason for as far as I know. For Reason 2.5 I guess the best method is this one:

But it has a bit of sound quality loss.

I used to use the EQ (in 512 mode) on the vocoder and 'draw' my HP filter shape in the band levels. Then with the shift button you can do some nice things. But you connot alter the resonance because you can only draw once and not automate the band levels.

I'm sorry, the further we go in Reason versions the closer we get. But we're just not there yet..



wouterschakel 2005-04-26 00:49

Re: Reason has a hidden high pass filter
"Relativity, a bit like incest"

You better not tell Einstein that!! :D

rpugwash 2005-04-26 01:29

Re: Reason has hidden hype-ass filters. seek & ye shall find
"Relativity, a bit like incest"
:You better not tell Einstein that!! :D

so einstein had an oedopus complex?

omg, thats how chinese whispers get started lol

i think i better not get on that one... i can barely handle sound engineering let alone quantum theory and psycho-analysis :-P

something tells me you'd cope tho hehe

cheers buddy!

Corsica_S 2005-04-26 02:10

Re: hey great theory there...
:But you connot alter the resonance because you can only
:draw once and not automate the band levels.

You can definately automate the band levels of the vocoder by setting up a sequencer track and recording the movements just like anything else. You can also set it up so your controller knobs move them too.
Corsica_S Refill A

wouterschakel 2005-04-26 17:43

Re: hey great theory there...
You're right, so you can indeed add resonance when you want it. But it's still lot's of work...

I must have been confused with the in/out puts of the band that didn't do something I once tried. Dunno it was though :).



Ogggy 2005-04-28 06:36

Re: Reason has a hidden high pass filter
Very interesting, thx.

I had wondered what the Q of the split was and found with quick tests the low was not so high and the high was...I want to say "very high" but I think I just found it was "different".


All times are GMT +2. The time now is 12:39.