is this a bug in layer filters? - Fxpansion.com

Forum

FXpansion Forum

is this a bug in layer filters?

Support for Geist

Moderators: Drew_BFDTeam, clare_fx, SKoT_FX, Steve_FX, Mully_FX, mayur_FX, Angus_FX, Andreas_FX, Rory_FX, Rhi_FX, Paul_fx, Moderators

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

is this a bug in layer filters?

Postby M-Goldie » Tue Nov 23, 2010 4:25 pm

simple example - i've got a snare drum pattern playing, i select the snare pad and go to pad/layer. i turn on the layer filter and by default its low pass 2 LPF2. the cut-off is at max (22500hz) but the filter is still taking the high frequency out of the snare sample? not a little bit either, its taking loads of top end out of it.

if the cut-off is at max then the LPF should not be affecting the sample at all?

everything is at default - resonance is 0%, drive is at 30% (shouldn't default be 0%??), Vel-Cutoff is 0%.

am i missing something or this a bug?

also if you do actually turn the drive to 0% then with a LPF the sample is virtually inaudible even with the cut-off at max (22500)

surely this is not correct...

Drew_BFDTeam
Posts: 3882
Joined: Fri Jul 21, 2006 5:32 pm
Location: London, UK

Postby Drew_BFDTeam » Tue Nov 23, 2010 4:42 pm

Well the drive is measured in percentage, but actually I think anything below 30% is a negative drive (ie -10db) ... I'll have to double check that.

Will have to check whether the tool-tip is wrong, or the filter itself.

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

Postby M-Goldie » Tue Nov 23, 2010 5:00 pm

drive defaults to 30% so i presumed that was 'no drive' but would be good to know.

but there certainly seems to be an error in the LPF layer filters.

by the way i have not set any filter cut-off or drive paramenters in the pattern graphs.

HPF seems to be unaffected by the problem but its difficult to tell with the other filter types. would be good if you could investigate and report back :)

cheers.

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

Postby M-Goldie » Wed Nov 24, 2010 11:16 pm

FXPansion devs - have you been able to check this yet?

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

Postby M-Goldie » Tue Dec 07, 2010 12:27 am

bump :)

Drew_BFDTeam
Posts: 3882
Joined: Fri Jul 21, 2006 5:32 pm
Location: London, UK

Postby Drew_BFDTeam » Tue Dec 07, 2010 10:41 am

What the dev's have told me is since the filter is modelled on an analog filter, it will always affect the audio - even when fully open.

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

Postby M-Goldie » Tue Dec 07, 2010 11:25 am

hello drew, thanks for the reply.

however i disagree and with all due respect would ask your devs to please check it again. even if it is analog modelled it is still incorrect and means geist is not functioning correctly.

for example, on a snare hit, i am using LP4 and have it all the way open (20khz no resonance), but i lose LOADS of the top end.

comparing it to soundtoys 4 pole filter (which also have a reputation for very accurate modelling) i roughly estimate that your 'open' filter is cutting off at around 7000khz. i also have mooger fooger analog lo pass filter and this is not the behaviour. if the filter is all the way open then the sound is basically unaffected - if there are subtle 'analog' effects then it is certainly nowhere near the highly audible cut-off that geist filters are doing! :)

i believe this is incorrect behaviour of the filter? it means that you can't filter sweep it properly because you can't get back to the sound of the filter being open, you can just get back to a half open sound.

would you mind checking again please and seeing if they give this a test with actual samples.

cheers :)

M

Drew_BFDTeam
Posts: 3882
Joined: Fri Jul 21, 2006 5:32 pm
Location: London, UK

Postby Drew_BFDTeam » Tue Dec 07, 2010 11:34 am

No problem; will take another look. In the meantime, can you upload an export just to see if we're experiencing the same thing. It doesn't sound anywhere near 7khz to me, so I just want to be sure we are hearing the same thing.

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

Postby M-Goldie » Tue Dec 07, 2010 11:44 am

yes no problem - i'll do it today.

cheers
M

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

Postby M-Goldie » Tue Dec 07, 2010 1:23 pm

hi drew.

can't seem to upload MP3s or wavs here i'll send it to support by email.

cheers
M

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

Postby M-Goldie » Tue Dec 07, 2010 1:31 pm

drew i have a problem - i can't contact support because they don't have 'geist' listed under what product i am contacting support about?!

i tried putting 'guru' but as i'm not registered for that i can't proceed.

so i'm stuck.

i have an MP3 to send you.

h e l p

M

Drew_BFDTeam
Posts: 3882
Joined: Fri Jul 21, 2006 5:32 pm
Location: London, UK

Postby Drew_BFDTeam » Tue Dec 07, 2010 1:45 pm

Hey,

Do you have a Dropbox account? Try that, or use Rapidshare, then just PM me a link to the file.

M-Goldie
Posts: 35
Joined: Wed Nov 17, 2010 9:33 am

Postby M-Goldie » Tue Dec 07, 2010 2:12 pm

ok will do :)

ENohoKuri
Posts: 9
Joined: Tue Nov 23, 2010 7:34 am

Postby ENohoKuri » Wed Dec 08, 2010 11:46 am

I'm also getting this behaviour. I don't mind coz Geist has enough outs to use 3rd party fx if the internal ones don't do what I want. Pretty much the story of VST for me. But, yeah, those filters have a noticeable cut to these old ears. Even when they are wide open.

stoutlyric
Posts: 62
Joined: Sat Mar 19, 2005 5:48 pm

Postby stoutlyric » Sun Mar 13, 2011 11:06 pm

I noticed this as well when importing a guru 1.5 kit into geist and trying to get it to sound the same.

It defiitely kills the top end when conventional wisdom has it that it should not mess with the top end when the cutoff is all the way up.

I also ntoiced the filter stops working in one pattern then in another pattern in the same engine it works. Definitely a bug, I need to test more, but i was using timestretch a lot and noticed my cpu going up, this might mess with the filter. The pattern where the filter was working was just playing the one sample and not the timestretched samples.


Return to “Geist”

Who is online

Users browsing this forum: No registered users and 4 guests