biquad~ 'clear' message bug report?

Dec 1, 2006 at 2:04pm

biquad~ 'clear' message bug report?

Dear Cycling Gurus

Here is a potential bug report, is someone can confirm the behaviour.

Summary:

The clear message does not work without restarting the DSP chain.

Steps to Reproduce:

1. open biquad~ help
2. boost the b1 factor to blow the filter (poc!)
3. bring the b1 factor to 0.
4. press clear.

Expected Results:

The dsp chain should resume with the filter reset.

Actual Results:

The filter is still blown! To resume, stop and start the dsp..

System:

Powerbook 15″ 1.67MHz MacOS 10.4.8 Max v.4.6.2 with jitter 1.6.2
installed

pa

#28995
Dec 7, 2006 at 9:28am

Pierre Alexandre Tremblay wrote:
> Dear Cycling Gurus
>
> Here is a potential bug report, is someone can confirm the behaviour.

I’ll add some experiences I made with blown patches: even switching
on/off audio couldn’t bring some of them back, I usually have to close
the patch and reload.
I never had the time to track it down that’s why I never reported
anything. But recovering blown dsp seems to have a problem.
My ears tell me there is a high dc (possibly maximum instead of NAN)
resting somwhere in the dsp chain…

Stefan


Stefan Tiedje————x——-
–_____———–|————–
–(_|_ —-|—–|—–()——-
– _|_)—-|—–()————–
———-()——–www.ccmix.com

#89680
Sep 29, 2007 at 4:49pm

Salut!

I’m having trouble with biquad~. It seems that it introduces, sometimes, a high DC when I change the cut-off frequency through a filtergraph.

If you know this bug and you know how to work around, it could be usefull for me because, actually, this object is no use for me in these conditions… I need to clear the biquad~ to remove the DC which is very anoying in a live performance or when the biquad~ is in one poly~ among others…

Merci!

#89681
Sep 29, 2007 at 6:18pm

I solved my problem with biquad~: sending a cut-off of 22050 is not a good idea even it is theorically correct so I limit the cut-off frequency to 22000 and everything looks fine.

Still, I have problems of DC in another patch with cascade~.

#89682
Sep 30, 2007 at 10:37am

Hey PA,

> 1. open biquad~ help
> 2. boost the b1 factor to blow the filter (poc!)
> 3. bring the b1 factor to 0.

Bizarre, but I can’t actually get the filter to blow on my system… I can’t figure out why not – anyway – no matter how high I push b1 it won’t blow….

I suspect maybe something quite different is happening under intel when the feedback goes above the max float value – the values I’ve got coming out are pretty big – just not max – and I can’t set up a situation which blows the filter – returning the b1 factor to 0 always results in correct output.

System:

MBP 17″ 2.4Hz Intel Core 2 Duo – OS 10.4.10 – Max 4.6.3

Alex

#89683
Oct 1, 2007 at 7:21am

wow, this is a very, very old post… like 5 or 6 months ago!

Did you just receive it?

>> 1. open biquad~ help
>> 2. boost the b1 factor to blow the filter (poc!)
>> 3. bring the b1 factor to 0.
>
> Bizarre, but I can’t actually get the filter to blow on my
> system… I can’t figure out why not – anyway – no matter how high
> I push b1 it won’t blow….

Funny, I have not tried this under my new Intel-based machine, and
you are right, it does not blow. It freaks out, then when returned
to normal the clear button is not even needed…

Any explanation from the cycling crew?

pa

#89684

You must be logged in to reply to this topic.