KVR Audio

Jeff McClintock · Post by **Jeff McClintock** » Thu May 24, 2012 12:01 am

The math does my head in. But I did build an iterative solver for a game physics engine (n-body problem), so I could possibly figure out a filter based on the same principles.

I would like to see someone release 'cookbook' zero-delay filter source code though, like Vinnie Falco made for traditional filter types..

https://github.com/vinniefalco/DSPFilters

..may someone could enhance Vinnie's??

camsr · Post by **camsr** » Thu May 24, 2012 12:16 am

I haven't been able to wrap my head around the math, but Vadim's paper made the layout of events exceptionally clear. Is it wrong to say the signal is filtered twice in a first order integrator?

JJBiener · Post by **JJBiener** » Thu May 24, 2012 12:17 am

Jeff McClintock wrote:The math does my head in. But I did build an iterative solver for a game physics engine (n-body problem), so I could possibly figure out a filter based on the same principles.

I would like to see someone release 'cookbook' zero-delay filter source code though, like Vinnie Falco made for traditional filter types..

https://github.com/vinniefalco/DSPFilters

..may someone could enhance Vinnie's??

Hi Jeff, I think you are selling yourself short. I've seen your work. I am confident that if you put your mind to it, you could come up with a workable solution.

aciddose · Post by **aciddose** » Thu May 24, 2012 12:37 am

camsr wrote:I haven't been able to wrap my head around the math, but Vadim's paper made the layout of events exceptionally clear. Is it wrong to say the signal is filtered twice in a first order integrator?

yes, it is wrong. although how far you'd have to stretch your imagination to make it right would depend on what sort of integrator you're dealing with.

if you define a filter as a function, so any x = ..., you could say it's right. it depends upon how you define a filter though. if we're talking about the difference between input/output, no it wouldn't be correct to say that.

Caco · Post by **Caco** » Thu May 24, 2012 7:23 am

Yes, I have been using brute force iterations so far that have been smoking my CPU but I am looking forward to getting some free time to dig through mystran's code and Vadim's book to try the more optimized approach

mystran · Post by **mystran** » Thu May 24, 2012 8:03 am

If you can't do it yet, don't panic.

I'm working on a solution that will hopefully evolve into something idiot-proof.

Ichad.c · Post by **Ichad.c** » Thu May 24, 2012 10:10 am

mystran wrote:If you can't do it yet, don't panic.

I'm working on a solution that will hopefully evolve into something idiot-proof.

Well, the votes are once again 50/50 - so you'll be helping a lot of people.
If you need a guinea pig from idiotland to test - I'll always be available.

Music Engineer · Post by **Music Engineer** » Thu May 24, 2012 11:25 am

hmm..well, just implemented a (linear) TPT/ZDF-SVF from the info in Vadim's book. does that already qualify as "build"? because, actually it was nothing more than a mere "translation into code".

george · Post by **george** » Thu May 24, 2012 11:26 am

We did

Music Engineer · Post by **Music Engineer** » Thu May 24, 2012 11:27 am

mystran wrote:If you can't do it yet, don't panic.

I'm working on a solution that will hopefully evolve into something idiot-proof.

hey, i'm looking forward to that.

hehe - it seems, finally i did now jump on this (already long running) train.

antto · Post by **antto** » Thu May 24, 2012 11:32 am

i think, there is enough brain potential in this forum, that when you guys work together - you can make a revolution in music dsp (and maybe not just in music)

i'm happy to be around such guys

obiwanjacobi · Post by **obiwanjacobi** » Thu May 24, 2012 11:50 am

I am already struggling with a log mapping function I need....
Man, Math really sucks big time!

So, anything DSP is out of my league. I did copy/paste a simple delay once and sort of understood that, but that's about it...

A: NO

mellotronaut · Post by **mellotronaut** » Thu May 24, 2012 11:56 am

well, my girlfriend uses her zero delay filter all the time to get rid of unwanted noises, but i ... no, i can't

davidguda · Post by **davidguda** » Fri May 25, 2012 12:04 pm

I voted no, for at the time being I haven't figured it out yet, on the other hand I haven't even tried, I've just read through the threads here on KVR. I'm sure I would eventually get it after reading them through a couple of more times and trying and failing a couple of times.
I "just" need time that I can put hard focused effort in to it.

stratum · Post by **stratum** » Thu May 31, 2012 6:54 am

I wonder the practical motivations for building a zero delay filter. I had thought IIR's are good enough? Why anyone should bother about 1-3 samples of delay? (As you can see this topic is new to me:) )

update: nevermind:) google has found the answer at http://www.xils-lab.com/pages/Zero-Dela ... lters.html

KVR Audio

Poll - can you build a zero-delay filter?

Can you build a zero-delay filter?