Use of Dual Filters: examples or tutorials?
-
- KVRist
- Topic Starter
- 69 posts since 20 Nov, 2016
So far in my sound design journey, I've largely stuck to single-filter sounds and haven't gotten good results from messing around with adding a second (whether that second is in series or in parallel). Two of my main four synths (Omnisphere and Spire) have the ability to use dual filters, so I feel like I'm missing a lot of their potential. Does anyone happen to know of some well-explained examples of or tutorials for dual-filter sounds?
-
- KVRian
- 629 posts since 15 Jun, 2017
Not sure what you're after. So...a bit of a generic answer....
Generally filters are for shaping the spectrum of a signal. Filters (in subtractive synthesis) generally have 2 purposes:
- attenuate frequencies
- boost frequencies.
Sideline: Phase is often also affected. In the "extreme" case of an allpass filter "only" phase is effected.
That being said:
Filters in series are often used in 2 ways
- a Low pass + Low pass configuration. E.g. if you need a steeper slope (at the same cutoff). E.g 12 dB/oct + 12 dB/oct = 24 dB/oct. Or at different cutoffs the attenuate and/or boost at different places in the spectrum.
- a Low pass + High pass. E.g. where the cutoff frequency of the low pass is higher then that of the highpass. Generally to attenuate some of the (too) low frequencies that have passed the Low pass.
Filters in parallel are often used to create some liveliness by effectively mixing 2 independent and different "layers". E.g. at slightly different cutoffs/resonance and/or different modulation rate/depth.
This is nothing more then a general guideline...
When using filters in general, keep the spectral content (relative amplitudes of harmonics/partials) of a signal in mind. Because that's what you are generally trying to control.
Generally filters are for shaping the spectrum of a signal. Filters (in subtractive synthesis) generally have 2 purposes:
- attenuate frequencies
- boost frequencies.
Sideline: Phase is often also affected. In the "extreme" case of an allpass filter "only" phase is effected.
That being said:
Filters in series are often used in 2 ways
- a Low pass + Low pass configuration. E.g. if you need a steeper slope (at the same cutoff). E.g 12 dB/oct + 12 dB/oct = 24 dB/oct. Or at different cutoffs the attenuate and/or boost at different places in the spectrum.
- a Low pass + High pass. E.g. where the cutoff frequency of the low pass is higher then that of the highpass. Generally to attenuate some of the (too) low frequencies that have passed the Low pass.
Filters in parallel are often used to create some liveliness by effectively mixing 2 independent and different "layers". E.g. at slightly different cutoffs/resonance and/or different modulation rate/depth.
This is nothing more then a general guideline...
When using filters in general, keep the spectral content (relative amplitudes of harmonics/partials) of a signal in mind. Because that's what you are generally trying to control.
-
- KVRian
- 712 posts since 26 Jul, 2018 from Germany
typical setup for dubtechno chords. Two filters in series.....
1st filter: 12/24dB lowpass + envelope modulation
2nd filter: 12dB bandpass + lfo modulation
1st filter: 12/24dB lowpass + envelope modulation
2nd filter: 12dB bandpass + lfo modulation
-
- KVRian
- 629 posts since 15 Jun, 2017
Vangelis is very well known for using the mighty Yamaha CS80's 12 dB/oct Low pass + High pass filters to create many of his iconic sounds, tracks and soundtracks. Well...actually, the CS80 has to completely separate synth layers...including oscillators and amps. But hey, I'd say that's having both serial and parallel filter options....
-
- Patchmeister
- 1774 posts since 3 Nov, 2010
someone made a very nice sounding old-school house minor chord/stab with dual filters once in a soundpack (can't remember which)
I remade the patch in Sytrus so I remember it well.
It's a chord/stab that then carries on with the low-passed chord but not fully closed as far as the stab went.
Now you could make the "pluck" and simply open the filter up somewhat but it just didn't sound as good as when using 2 filters for whatever reason.
also I added a little attack to the continuing chord to make room for the initial "pluck" sound.
Another application is creating multiple resonance spikes or 2 band-passed sounds and feed it into a distortion to get some mad sounds out of it.
But for practical reasons 2 filters are rarely needed.
I remade the patch in Sytrus so I remember it well.
It's a chord/stab that then carries on with the low-passed chord but not fully closed as far as the stab went.
Now you could make the "pluck" and simply open the filter up somewhat but it just didn't sound as good as when using 2 filters for whatever reason.
also I added a little attack to the continuing chord to make room for the initial "pluck" sound.
Another application is creating multiple resonance spikes or 2 band-passed sounds and feed it into a distortion to get some mad sounds out of it.
But for practical reasons 2 filters are rarely needed.
CHOOSX Remakes on my Youtube Channel
-
- KVRian
- 629 posts since 15 Jun, 2017
Except maybe for the Low Pass + High Pass combination that is used in many patches to attenuate some of the lower partials/frequencies. Like in strings and brasses.
Also...apart from the intended high pass filtering: a High Pass may do more then just filter out frequencies below the cut off. It may also change the phase of parts of the spectrum above its cut off frequency (in the frequencies that are supposed to have "passed"), thus effecting higher bands/partials as well (due to phase cancellation effects).
-
- KVRist
- 282 posts since 24 Aug, 2017
You’re definitely missing out, it’s like you only had an EQ with one band.
Check out a demo video of something like soundtoys filterfreak, you can do all sorts of overlapping and curve shaping with two modulated filters.
Check out a demo video of something like soundtoys filterfreak, you can do all sorts of overlapping and curve shaping with two modulated filters.
-
- KVRian
- 629 posts since 15 Jun, 2017
Most synths come with only 1 or 2 filters per "voice" (in series or parallel). Generally also limited in types. LP, HP, BP and slopes 24, 12, 6 dB/Oct being the most common. I guess that using those "in-synth" options is what the thread starter is after.
I guess we all realize that external filtering exists. And in many forms. EQ's being the most common. But there are all sorts of external filters. Static and dynamic. Some with their own modulators and/or envelopes and/or note and/or envelope follower(s). And there are many ways to use them. Ranging from subtle spectrum shaping to outrageous effects.
-
- Patchmeister
- 1774 posts since 3 Nov, 2010
I agree. but if you browse through your own presets, how many in % use more than 1 filter?
I definitely have some but I'd say 2 - 5% max
I definitely have some but I'd say 2 - 5% max
CHOOSX Remakes on my Youtube Channel
-
- KVRian
- 712 posts since 26 Jul, 2018 from Germany
-
- Patchmeister
- 1774 posts since 3 Nov, 2010
-
- KVRian
- 629 posts since 15 Jun, 2017
Actually, if the synth offers a LP + HP, I tend to use an added HP a lot. Either at a fixed cut off frequency (sort of EQ-ish, effecting the Lower notes only) and/or following notes as a shaping tool.
The "classic" oscillators (Saw, Square, Triangle) have really strong/pronounced first partials. These amplitudes are not always desirable (and are in need of some attenuation).
Not saying there's a right or wrong here. But some form of HP can often improve a "preset".
-
- Patchmeister
- 1774 posts since 3 Nov, 2010
I tend to do that after the synth in the DAW. I leave most of my presets free of things like reverb/delay etc because if I end up using them I always go and change them depending on what I'm using it for. So they sound "bland" but they're for me anyway
CHOOSX Remakes on my Youtube Channel
-
- KVRian
- 712 posts since 26 Jul, 2018 from Germany
Another "trick"....
At times of mighty wavetable synths like Serum it's not really necessary anymore. But if you are restricted to basic waveforms for example, it's possible to use one of the filters to shape the harmonic content of a oscillator.
OSC 1 -> Filter 1 (Bandpass -> Keytrack = ON) -> Filter 2 (Lowpass)
OSC 2 -> Filter 2 (Lowpass)
At times of mighty wavetable synths like Serum it's not really necessary anymore. But if you are restricted to basic waveforms for example, it's possible to use one of the filters to shape the harmonic content of a oscillator.
OSC 1 -> Filter 1 (Bandpass -> Keytrack = ON) -> Filter 2 (Lowpass)
OSC 2 -> Filter 2 (Lowpass)
-
- KVRian
- 629 posts since 15 Jun, 2017
One more usefull thing...
Many synth only offer either serial and/or parallel filters. Not all synths offer control over the filter input and/or output levels. A synth like Fathom offers a wet/dry mix per filter. A synth like Surge let's you set the filter mix/balance. This way you can have a pretty steep Low Pass filter at a low cutoff and still pass/mix higher frequencies (from the dry signal) too. Very usefull for strings and pads.
Many synth only offer either serial and/or parallel filters. Not all synths offer control over the filter input and/or output levels. A synth like Fathom offers a wet/dry mix per filter. A synth like Surge let's you set the filter mix/balance. This way you can have a pretty steep Low Pass filter at a low cutoff and still pass/mix higher frequencies (from the dry signal) too. Very usefull for strings and pads.
