Page 1 of 17

Transient loss

Posted: Wed Feb 29, 2012 3:07 pm
by fredrikberg
I feel like I am loosing some transients when using Nebula. Especially noticable on snare and kick. I know about gainstaging so thats not the problem.
I feel it happens with all types of libraries from consoles, eq to tape

Is it because of volterra technique?

Can I tweak something to make it better?

Regards

Fredrik Berg

Re: Transient loss

Posted: Wed Feb 29, 2012 3:12 pm
by enriquesilveti
On preamp or tapes you can use Nebula non reverb version (but more CPU consumption) plus try to use TIMED engine on clean kernel (Nebula KERN page)

Re: Transient loss

Posted: Wed Feb 29, 2012 5:47 pm
by musicgreator
Isn't that characteristic part of the sampled hardware? If not, why doesn't nebula have the right settings to replicate the exact same behavior as the hardware by default? I didn't know, nebula reverb sounds different than nebula, i thought it's just tweaked for best performance with reverbs.

Re: Transient loss

Posted: Wed Feb 29, 2012 6:44 pm
by Tim Petherick
Ok ,


If you want super high quality this is what i do and if its wrong in theory someone please tell me.

Number 1 , use evf 17 in evf page.
2, use classc in kern page( split will be showing)

point the arrows so they are going towards freqd player lengths.

Then increase the timed kernals in order to complete the sample lengths


Its a cpu killer though

Re: Transient loss

Posted: Wed Feb 29, 2012 9:30 pm
by fredrikberg
Thank you both for tips.

I couldn´t try the first tips, cause Nebula non reverb version causes overload in the channel with like 100db.

I tried with Reverb version suggestions from Timp, and it really made a HUGE difference. Now all transients are there with beautiful coloration.

Thank you!!!

How did you come up with these tweaks. there is a lot of parameters to play with.

Regards

Fredrik Berg

Re: Transient loss

Posted: Wed Feb 29, 2012 9:56 pm
by Tim Petherick
Just by ear really .

But i think its more accurate in theory , Evf 17 is a debugged rms 17.

Timed player is a better algo than freqd . freqd is low on cpu though .

the only problem correct me if i'm wrong is that some samples cannot be completed because timed only goes up to 30 ms. but i'm guessing that would be for extreme lows of around 10hz and under , not sure.

If you take an low shelf eq for example.that starts at 2.2ms on timed . select the arrows to pointing to freqd player, you'll notice that the low frequency is lost because the sample is no longer long enough working in timed mode. So when you get to about 10 ms you can hear that most of that low end is back.

Maybe G can comfirm this because i'm really just guessing from trail and error.

Tim

Re: Transient loss

Posted: Wed Feb 29, 2012 10:10 pm
by Cupwise
timp- isn't that what split mode is for?
my understanding is that with split mode you get to use timed mode for first sliver of the sample, then freqd mode for the rest.
so you could have timed portion of clean kernel set to 10ms, then freqd set to 40, timed plays first 10ms, freqd plays last 40.
maybe you've already tested this and didn't like it as well as just using timed, but i thought this was what it was for...
?

stuff like this should have a little more discussion

Re: Transient loss

Posted: Wed Feb 29, 2012 10:19 pm
by Tim Petherick
Yes I was aware of the split mode's function .Well i guess in my tests i prefered the sound of non split mode, But I'll try again. The freqd player does not sound correct to me , but i want to be proven wrong because the cpu is so high using timed only. I find the everything sounds a little grainy without these tweaks. This includes editing the odd and even samples too .

Re: Transient loss

Posted: Wed Feb 29, 2012 10:29 pm
by fredrikberg
Do you guys think there will difference with these tweaks in Nebula vs Nebula reverb.

Would this also be the best setting for EQ, Compression and Tape.

I guess for VNXT Plate I could try Split mode.

Again, really BIG difference in sound.

Re: Transient loss

Posted: Wed Feb 29, 2012 10:29 pm
by giancarlo
yes you are correct. Imho freqd doesn't sound as much correct as timed but it's used just for the tail. Freqd is the classic fft algo used by best plugins around, so it's not so bad

Re: Transient loss

Posted: Wed Feb 29, 2012 10:38 pm
by fredrikberg
giancarlo wrote:yes you are correct. Imho freqd doesn't sound as much correct as timed but it's used just for the tail. Freqd is the classic fft algo used by best plugins around, so it's not so bad
Is that in split mode?. (when it reads Classd)

Re: Transient loss

Posted: Wed Feb 29, 2012 10:49 pm
by fredrikberg
Also, To get the best from reverbs should I use split mode with all 3 timed kernels set to 10ms.

...And is there a way to make these settings the default for all my programs or do I have manually edit and save them.

Re: Transient loss

Posted: Wed Feb 29, 2012 10:57 pm
by Tim Petherick
You'll need to use split mode and freqd player on harmonics for reverbs because the samples are so long.

Split mode -classc will show on screen

classc- split will show on screen

It's a bit confusing, the same for normal and economy quality.

Re: Transient loss

Posted: Wed Feb 29, 2012 11:34 pm
by fredrikberg
Wouldn´t it be great with a "Render" mode. Like with the Glue where you can set oversampling in rendering, but normal in realtime.

One could set "Split mode with 10ms kernels on timed and Evf17" as render mode.

I just did all these changes to an old mix and it turned out soo much better.

Re: Transient loss

Posted: Thu Mar 01, 2012 1:28 am
by babiuk
I´m reading this topic and I find it highly interesting because I can see I am not using Nebula benefits at 100%.
Anyway it´s a bit confusing to me. Could you help me to understand?
When I read this:
One could set "Split mode with 10ms kernels on timed and Evf17" as render mode.
Can I assume (CPU costs apart) that this is the best nebula setup for every preset?
And... is better (again CPU costs apart) using nebula reverb all the time instead nebula?