Running filters in parallel versus series
-
- KVRist
- 257 posts since 23 Oct, 2005
Thanks to my mad mspaint skills, you can see that filters in series dumps the output of the first filter right into the next. This is can be used to get all sorts of bizarre filters. (Note that LP and HP are just examples. It could be two LPF's, a band pass and a notch, a formant and a shelf, etc) Some ways you can use this is to put a HPF first w/ a LPF second. Set the HPF cutoff lower than the LPF's and you have a rather tweakable band pass filter. You can also use a LPF first followed by a BPF tuned to the 2 or third harmonic for a twist on the plain LPF. Becareful not to kill the signal at the first filter tho. (For example, a HPF set to it's highest cutoff followed by a LPF w/ a low cutoff.)
Running in parallel is a little simpler, just imagine that you have two independent filters, each processing the same signal. This is fun for making notch filters that sweep up, down, and get smaller and smaller til the bandwidth goes negative (ie the filtering effect disappears.) You might be left w/ some phase oddness tho, but that depends on how aal the programmer was that day.
Hope that helps, anybody else care to come up w/ a more concise explanation?
-
- Skunk Mod
- 21249 posts since 10 Jun, 2004 from Pony Pasture
Parallel: You get the sound of the input as heard through two filters, mixed together. Typically fewer frequencies will be subtracted, as those that one filter removes might pass the other.
Serial: Filters the sound twice in succession, the first removing some frequencies and the second possibly removing others from what comes out of the first.
Given the same settings on LP and HPF in underface's diagram, serial processing is likely to have a more drastic effect on the sound.
Add in resonance on one or both filters and things become a little more complicated, but that's the basics.
An optical analogy: White light is passed through two pieces of colored glass, one fairly intense red and one deep, vivid green. This corresponds closely to the subtractive filtering of sound. In series you get very little light out, as most of the predominantly red light coming out of the first filter will be blocked by the green glass. If you shine white light through both independently, which is to say in parallel, then mix that light together, you get yellow light -- the combination of red and green light. Dunno if that helps but there it is.
Serial: Filters the sound twice in succession, the first removing some frequencies and the second possibly removing others from what comes out of the first.
Given the same settings on LP and HPF in underface's diagram, serial processing is likely to have a more drastic effect on the sound.
Add in resonance on one or both filters and things become a little more complicated, but that's the basics.
An optical analogy: White light is passed through two pieces of colored glass, one fairly intense red and one deep, vivid green. This corresponds closely to the subtractive filtering of sound. In series you get very little light out, as most of the predominantly red light coming out of the first filter will be blocked by the green glass. If you shine white light through both independently, which is to say in parallel, then mix that light together, you get yellow light -- the combination of red and green light. Dunno if that helps but there it is.
-
blackboyrockinit blackboyrockinit https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=182229
- Banned
- 433 posts since 5 Jun, 2008
from the the information/theories given here it seems as if parallel will give you more comb filtering.