Filter type(s) of the good ol' Soundblaster Live

Anything about hardware musical instruments.
Post Reply New Topic
RELATED
PRODUCTS

Post

Fellow Earthicans!

Is there any knowledge about the kind of the filter model used in the soundblaster live? I mean if it's more svf, diode, ladder (this one surely not ... :D )

I am converting some old projects from back in the days, trying to get away from .sf2 to mostly Kontakt. I recall the filter had a very nice, but disctinct sound where e.g. it let pass through some transient material even if the cutoff was close to 'closed'.

I am aware that there was the EMU10K1 on it, but that led to nothing, because (if I get this right) technically one could implement any filter type you like, of course limited to the dsp power of the chip.

I am also aware of sfz as well as many other soundfont players finding they're all not the real deal...

And yes! I really should just go by ear, take whatever sounds best...so if it's just out of curiousity I'd really like to know what kind of filter it might be... :hyper:

Thanks you!

Post

It wasn't really an analog-modeled filter in the way you might be thinking. It's described in documentation as simply a "resonant digital filter" of a "two-pole low-pass design," there's a description and diagram here:

https://web.archive.org/web/20140302033 ... /m2049.pdf

You can find the paper that doc references (which seems to be where the "special sauce" is) here:

https://ieeexplore.ieee.org/document/634131
Fugue State Audio - plugins, samples, etc.
Support the Union of Musicians and Allied Workers

Post

I kind of expected something like that, but it's cool to actually have the math of it available as a starting point...it will take some time to look into it I guess :hihi:

Thank you!
synthgeek wrote: Sun May 23, 2021 10:26 am It wasn't really an analog-modeled filter in the way you might be thinking. It's described in documentation as simply a "resonant digital filter" of a "two-pole low-pass design," there's a description and diagram here:

https://web.archive.org/web/20140302033 ... /m2049.pdf

You can find the paper that doc references (which seems to be where the "special sauce" is) here:

https://ieeexplore.ieee.org/document/634131

Post

You're welcome!
Fugue State Audio - plugins, samples, etc.
Support the Union of Musicians and Allied Workers

Post

You can measure it yourselves by running some white noise through it. I did it and measured it to be a 18 dB filter. However it is already some time ago and I could have made a mistake.

Post

Essent wrote: Tue Feb 03, 2026 11:45 am You can measure it yourselves by running some white noise through it. I did it and measured it to be a 18 dB filter. However it is already some time ago and I could have made a mistake.
Cool, I actually never thought about this. How does one do that? I assume you run the noise through a filter sweep with at least two different Q-values...and then?

Post

Orchidius wrote: Tue Feb 03, 2026 12:04 pm
Essent wrote: Tue Feb 03, 2026 11:45 am You can measure it yourselves by running some white noise through it. I did it and measured it to be a 18 dB filter. However it is already some time ago and I could have made a mistake.
Cool, I actually never thought about this. How does one do that? I assume you run the noise through a filter sweep with at least two different Q-values...and then?
  1. Just get or create a .sf2 file with a white noise sample.
  2. Disable the filter modulation envelope (set to filter cut off = 0).
  3. Set filter to say something between 1000 - 4000 Hz.
  4. Playback the sound (a long note) and record it.
  5. Analyse the recorded sound using a spectrum analyzer. My DAW is Logic, which has a good spectrum analyser (in the EQ)
Don't forget to analyse the Q as well.

I might be in the same boat as you. Got some legacy stuff (SoundBlaster Live / AWE64) with some crucial sounds / instrument, they really are a boaster to some of my (dance) tracks. So I definitely need them.

Post

synthgeek wrote: Sun May 23, 2021 10:26 am It wasn't really an analog-modeled filter in the way you might be thinking. It's described in documentation as simply a "resonant digital filter" of a "two-pole low-pass design," there's a description and diagram here:

https://web.archive.org/web/20140302033 ... /m2049.pdf

You can find the paper that doc references (which seems to be where the "special sauce" is) here:

https://ieeexplore.ieee.org/document/634131
If it is a 2 pole filter, according to that article, then it must be a 12 dB filter.

Post Reply

Return to “Hardware (Instruments and Effects)”