Login

Does dsp buffer size affect sound of console mixbus like MFC

Tips & tricks, working results, technical support

Does dsp buffer size affect sound of console mixbus like MFC

Postby pleplo » Wed Dec 10, 2014 9:54 am

I know it's been discussed to death here. but since console and tape (specialy the console mixbus on the master channel and sub groups) is dynamic and by default use FredD mode, would the dsp buffer size (that changes the program rate) affect transient and the way the console dynamic output sound? I'm at 2048 buffer size. is that ok or I should go lower? it's all very subtle anyway what the MFC does but I hear the widening

EDIT:
OK this is a mix test with both version. one at buffer of 512 and the second at 2048. you can hear a difference. 512 is more controlled, and glued/compressed.

https://www.dropbox.com/s/7vw0cyj7da8ue ... 2.wav?dl=0

https://www.dropbox.com/s/qwnmum4rcm9el ... 8.wav?dl=0

Thank you very much :)
Patrick
Last edited by pleplo on Thu Dec 11, 2014 5:46 am, edited 1 time in total.
Plajia new album Piggie Park out NOW. (Pronounced plah-gee-ah) We make music, fock a genre. On Spotify and Bandcamp www.plajia.com
User avatar
pleplo
User Level XI
User Level XI
 
Posts: 181
Joined: Tue Mar 13, 2012 10:08 pm
Location: Montréal, Canada

Re: Does dsp buffer size affect sound of console mixbus like

Postby kels » Wed Dec 10, 2014 10:32 am

hi,
I don't have MFC so I cannot speak for that but on other console programs, I have perfect cancelation between 2 Nebula Reverb instances, up until @8192 and down until @1024.
Going down to 512 makes program rate to modify (divided by 2) and therefore affect the sound.

in all case, it is better to leave it default (8192) as it will allows more programs to pile up in your session.

There are however some compressors libs specificaly designed to run @512 with Nebula non reverb version.
kels
User Level XI
User Level XI
 
Posts: 166
Joined: Fri Oct 11, 2013 8:43 am

Re: Does dsp buffer size affect sound of console mixbus like

Postby pleplo » Wed Dec 10, 2014 10:39 am

Thanks a lot my friend . makes sens. that is why i use Nebula Setup and I have all separated my nebula dlls (mixbus, tape, eqs, comp, etc.) to separate the buffer. for my compressors xml (I use mainly Tim U76 and Compex) Tim said better have a 1024 buffer so its set to that and seem to work fine.

I guess for all the rest of I will use 4096. just to tweak my full mix and render. but that is a lot of latency. so when I change a nebula eq setting there is a annoying delay before you hear the change. no big deal for this session but the next songs I will be mixing I will lower the dsp buffers and do the same thing to start the mix. and as the mix go far with lots of nebula I will make buffer higher... I use mainly Nebula and busses since I also bounce all my tracks with nebula stuff on them (console, eq, tape, preamp, etc..)..

Thank you again for the responses :)
Plajia new album Piggie Park out NOW. (Pronounced plah-gee-ah) We make music, fock a genre. On Spotify and Bandcamp www.plajia.com
User avatar
pleplo
User Level XI
User Level XI
 
Posts: 181
Joined: Tue Mar 13, 2012 10:08 pm
Location: Montréal, Canada

Re: Does dsp buffer size affect sound of console mixbus like

Postby pleplo » Wed Dec 10, 2014 10:43 am

anyway I will do a test where I will have mostly all nebula instance set to around 4096 than bounce a mix. than change all instance to 512 buffer and bounce and do A/B test of the mix to see if I hear something different.

Will keep you posted
Plajia new album Piggie Park out NOW. (Pronounced plah-gee-ah) We make music, fock a genre. On Spotify and Bandcamp www.plajia.com
User avatar
pleplo
User Level XI
User Level XI
 
Posts: 181
Joined: Tue Mar 13, 2012 10:08 pm
Location: Montréal, Canada

Re: Does dsp buffer size affect sound of console mixbus like

Postby nonstandardryder » Thu Dec 11, 2014 12:47 am

pleplo wrote:anyway I will do a test where I will have mostly all nebula instance set to around 4096 than bounce a mix. than change all instance to 512 buffer and bounce and do A/B test of the mix to see if I hear something different.

Will keep you posted


Would love to hear this test. I am also very curious.

Fingers crossed for a simple "Extreme Quality" setting that changes everything needed under the hood for the best offline or freeze renders. I don't care if it takes hours to render. heh I just want the best sound.
nonstandardryder
Member
Member
 
Posts: 201
Joined: Sat Apr 24, 2010 5:23 am

Re: Does dsp buffer size affect sound of console mixbus like

Postby pleplo » Thu Dec 11, 2014 1:06 am

I will post audio demos of the test i've done

It will mainly affect the consoles busses (mixbus, guitar groups, synth group, drum group, etc..), the MFC Panner (I use it on all tracks that I pan) mostly hard left and right. and on the reverbs (I dont think it will make much difference on the reverbs).

I also use the new Tim Petherick Gyrator eq its amazing. it uses Timed Mode so DSP size doest affect program rate I think. but I still leave it to default 512 buffer. so it can get heavy on cpu quick with this dynamic eq.

Note that I have MFC console Mic Pre et Line In on every channel but those are alrady all bounced and rendered. so this I leave at 512 buffer since cpu no problem. same with tape if I use them. but I often use them on busses too so that might change the sound of the mix.

Lastly I did not get into that "changing all programs to TimedD" mode. I read a lot and decided not to go this way. stuff is sounding good already. so I leave all third party libs to their default. exept compressor of course all the latest good ones are well set and timedd.

cheers
Plajia new album Piggie Park out NOW. (Pronounced plah-gee-ah) We make music, fock a genre. On Spotify and Bandcamp www.plajia.com
User avatar
pleplo
User Level XI
User Level XI
 
Posts: 181
Joined: Tue Mar 13, 2012 10:08 pm
Location: Montréal, Canada

Re: Does dsp buffer size affect sound of console mixbus like

Postby pleplo » Thu Dec 11, 2014 3:25 am

Ok guys just arrived at my studio. with fresh ears. did some test with good headphone to hear the 2 same mixes with different DSP buffer size in nebula xml mainly for consoles busses.

First of. in the null test they dont cancel but this means nothing as I remember doing a test with the same mix settings bounced twise and they did not cancel. might have to do with the realtime mixdown and how nebula react dynamically so its never quite the same.

but in the blind tests I just did with different parts of a complex mix. almost all of the time I chose the same file that I found was more glued and full. to my surprise it was the mix I had the buffer to higher. like 2048 or 4096 I think. the other one as set to 512 on nebula..

I can post the 2 audio file if you guys are interested to make the test.

cheers
Plajia new album Piggie Park out NOW. (Pronounced plah-gee-ah) We make music, fock a genre. On Spotify and Bandcamp www.plajia.com
User avatar
pleplo
User Level XI
User Level XI
 
Posts: 181
Joined: Tue Mar 13, 2012 10:08 pm
Location: Montréal, Canada

Re: Does dsp buffer size affect sound of console mixbus like

Postby nonstandardryder » Thu Dec 11, 2014 5:01 am

pleplo wrote:Ok guys just arrived at my studio. with fresh ears. did some test with good headphone to hear the 2 same mixes with different DSP buffer size in nebula xml mainly for consoles busses.

First of. in the null test they dont cancel but this means nothing as I remember doing a test with the same mix settings bounced twise and they did not cancel. might have to do with the realtime mixdown and how nebula react dynamically so its never quite the same.

but in the blind tests I just did with different parts of a complex mix. almost all of the time I chose the same file that I found was more glued and full. to my surprise it was the mix I had the buffer to higher. like 2048 or 4096 I think. the other one as set to 512 on nebula..

I can post the 2 audio file if you guys are interested to make the test.

cheers


Very interesting, Yes please post the files.
nonstandardryder
Member
Member
 
Posts: 201
Joined: Sat Apr 24, 2010 5:23 am

Re: Does dsp buffer size affect sound of console mixbus like

Postby Tim Petherick » Thu Dec 11, 2014 5:06 am

pleplo wrote:I will post audio demos of the test i've done

It will mainly affect the consoles busses (mixbus, guitar groups, synth group, drum group, etc..), the MFC Panner (I use it on all tracks that I pan) mostly hard left and right. and on the reverbs (I dont think it will make much difference on the reverbs).

I also use the new Tim Petherick Gyrator eq its amazing. it uses Timed Mode so DSP size doest affect program rate I think. but I still leave it to default 512 buffer. so it can get heavy on cpu quick with this dynamic eq.

Note that I have MFC console Mic Pre et Line In on every channel but those are alrady all bounced and rendered. so this I leave at 512 buffer since cpu no problem. same with tape if I use them. but I often use them on busses too so that might change the sound of the mix.

Lastly I did not get into that "changing all programs to TimedD" mode. I read a lot and decided not to go this way. stuff is sounding good already. so I leave all third party libs to their default. exept compressor of course all the latest good ones are well set and timedd.

cheers


Hi Pleplo,

Just to let you know I'm using Spith this uses both timed and Freqd. It's fine tuned that way, you shouldn't need to go pure timed for it so keep it the same.


Cheers,

Tim
User avatar
Tim Petherick
Expert
Expert
 
Posts: 1351
Joined: Sat Apr 17, 2010 4:07 pm
Location: Bath , Uk

Re: Does dsp buffer size affect sound of console mixbus like

Postby pleplo » Thu Dec 11, 2014 5:16 am

Tim Petherick wrote:
pleplo wrote:I will post audio demos of the test i've done

It will mainly affect the consoles busses (mixbus, guitar groups, synth group, drum group, etc..), the MFC Panner (I use it on all tracks that I pan) mostly hard left and right. and on the reverbs (I dont think it will make much difference on the reverbs).

I also use the new Tim Petherick Gyrator eq its amazing. it uses Timed Mode so DSP size doest affect program rate I think. but I still leave it to default 512 buffer. so it can get heavy on cpu quick with this dynamic eq.

Note that I have MFC console Mic Pre et Line In on every channel but those are alrady all bounced and rendered. so this I leave at 512 buffer since cpu no problem. same with tape if I use them. but I often use them on busses too so that might change the sound of the mix.

Lastly I did not get into that "changing all programs to TimedD" mode. I read a lot and decided not to go this way. stuff is sounding good already. so I leave all third party libs to their default. exept compressor of course all the latest good ones are well set and timedd.

cheers


Hi Pleplo,

Just to let you know I'm using Spith this uses both timed and Freqd. It's fine tuned that way, you shouldn't need to go pure timed for it so keep it the same.


Cheers,

Tim


Thanks Tim! but does having different DSP buffer instead of the default 512 can change the sound of the Gyrator since its dynamic. I see the prog rate change to very high value on high buffer size. Thanks for letting me know. With the consoles test (im doing some more tests) I think it makes a small difference in the sound.

Cheers :)
Plajia new album Piggie Park out NOW. (Pronounced plah-gee-ah) We make music, fock a genre. On Spotify and Bandcamp www.plajia.com
User avatar
pleplo
User Level XI
User Level XI
 
Posts: 181
Joined: Tue Mar 13, 2012 10:08 pm
Location: Montréal, Canada

Next

Return to Working with Nebula

Who is online

Users browsing this forum: No registered users and 7 guests