KVR Audio

Den* · Post by Den* » Mon Jun 15, 2015 12:52 pm

Andrew
We need down sampling for the next versions of Aether or else please.
I need it for some closer old hardware simulations.
Thanks....

Andrew Souter · Post by **Andrew Souter** » Mon Jun 15, 2015 12:54 pm

Den* wrote:
V0RT3X wrote:
den* wrote: In the nature there is noting above 7K in natural reverbs.
Really where can i read more about this?
Up to 8Khz, and no more.
http://reverb2014.dereverberation.com/w ... 899165.pdf

Den, this is a paper on DSP enhancements for mobile phones:

"EXTRACTION OF ROBUST FEATURES BY COMBINING NOISE REDUCTION AND FDLP FOR THE RECOGNITION OF NOISY SPEECH SIGNALS IN HANDS-FREE MODE"

I guess extreme fidelity for pro audio in film, TV, games, music etc is NOT their major concern. Their concentration, based on a 10sec skim of the paper, is to improve speech inteligiblity for mobile devices.

You will notice they use a sample rate of 16K. (likely so save battery life in mobile phones -- ie. it is reasonable comprise -- it is not the perfect choice for fidelity.). Thus of course they do not have any energy above 8K!!

You're ears are AWESOME! Your presets are AWESOME! We love your work! If you feel less high freqs are needed in your presets and you design them that way, wonderful! Perfectly great!

...but let's not get into pseudo-technical arguments about why we should or should not use OS. Suffice to say we know what we are doing and we feel there are reasons to use such things in some cases...

I don't want to be rude or make you feel unappreciated. You are DEEPLY appreciated by us, and all our customers. Maybe just say something like "My opinion lately is that... xyz" instead of absolute statements that sound like they are officially coming from the company as sometimes our opinions do not align perfectly... that's OK. It's a free world and they don;t have to align always. Debate is fun -- we learn things this way. Just please be careful not to say things that sound like they are "official 2Caudio policy" or whatever... Merci.

Andrew Souter · Post by **Andrew Souter** » Mon Jun 15, 2015 12:57 pm

Den* wrote:Andrew
We need down sampling for the next versions of Aether or else please.
I need it for some closer old hardware simulations.
Thanks....

Sure, such things may be possible and interesting as "Creative choices". It's a cool idea and maybe useful for SOME cases. Thanks.

Andrew Souter · Post by **Andrew Souter** » Mon Jun 15, 2015 1:04 pm

A little technical info to clarify:

1) Feedback based algo verbs use delay lines.

2) These delay lines are modulated usually for various reasons.

3) Modulating delay lines requires "interpolation"

4) Most forms of "interpolation" create the side effect of High Freq Loss. Interpolation is effectively a filter.
(All-Pass interpolation does NOT theoretically create HF loss, but it creates a TON of noise quite easily. There exist techniques to minimize this, but they are not perfect and some are patented.)

6) IMHO while liberal amounts of "controllable" HF loss is generally advisable to emulate real spaces, "uncontrollable" HF loss is NOT desirable, and emulating real spaces is only a small subset of all the creative things you might like to do with a reverb/spatial-processor... i.e. the Damping Filter curve should be the only thing effecting the time-varying spectrum of the reverb decay... i.e. "what you see, should be what you get"...

7) Most digital filters exhibit freq wrapping and other artifacts at freqs close to Nyquist. This is generally not desirable.

For these reasons and others, we feel oversampling can be beneficial as it helps solve all of these things (and more.) Perfection is simply not possible without Oversampling IMHO. "Good enough" may be possible, but we get more excited about "perfection" than "good enough." That's just us.

V0RT3X · Post by **V0RT3X** » Mon Jun 15, 2015 1:13 pm

Galbanum wrote:Den, this is a paper on DSP enhancements for mobile phones:

Yah i scoured the paper and kind of figured this wasn't for pro audio stuff

Den* · Post by Den* » Mon Jun 15, 2015 1:36 pm

Yes. I see now I was wrong about 7Khz natural reverb response.
Sorry.

Den* · Post by Den* » Mon Jul 20, 2015 5:52 pm

In the case that you missed that M7 preset here is the link:

http://www.kvraudio.com/forum/viewtopic ... 5#p6176905

Echoes in the Attic · Post by **Echoes in the Attic** » Fri Jul 24, 2015 2:52 am

What's all this talk of oversampling... nano quality mode is the best!!! I'm afraid to oversample it!

Echoes in the Attic · Post by **Echoes in the Attic** » Fri Jul 24, 2015 3:08 am

Hey by the way, was it a B2 preset responsible for the sound of the piano in the track below? It's sort of set back and dark and I'm curious how much is due to B2 processing. And do you have the preset?

https://soundcloud.com/2caudio/mix-dedr ... audio-b2-1

Andrew Souter · Post by **Andrew Souter** » Sat Jul 25, 2015 4:53 pm

Echoes in the Attic wrote:Hey by the way, was it a B2 preset responsible for the sound of the piano in the track below? It's sort of set back and dark and I'm curious how much is due to B2 processing. And do you have the preset?

https://soundcloud.com/2caudio/mix-dedr ... audio-b2-1

I'm actually not sure exactly who made that example so unfortunately I can't comment of the preset(s)/setting(s) used. I think one of our beta testers made this example. It was a while ago now.

Andrew Souter · Post by **Andrew Souter** » Sat Jul 25, 2015 5:15 pm

Echoes in the Attic wrote:What's all this talk of oversampling... nano quality mode is the best!!! I'm afraid to oversample it!

Well density/diffusion mode (nano, lo, hi, xtrm) settings don't have too much to do with oversampling. They are sort of two orthogonal/independent settings. It's ok to try oversampling with Nano mode.

Density/diffusion mode will control how quickly the reverb tail reaches maximum density, as well as how flat the spectrum is, as well as how much CPU usage the plug uses.

Nano is cool with large sizes of you want sparse/grainy type of presets. These can be very nice on medium to slow attack synths and can give a very nice retro feel.

Using one of the AP or Air filters in nano mode can give quite interesting results also.

Overs sampling won't undo any of this. It will simply allow more precise control over the frequency decay profile.

-a

Echoes in the Attic · Post by **Echoes in the Attic** » Sat Jul 25, 2015 5:41 pm

Galbanum wrote:
Echoes in the Attic wrote:Hey by the way, was it a B2 preset responsible for the sound of the piano in the track below? It's sort of set back and dark and I'm curious how much is due to B2 processing. And do you have the preset?

https://soundcloud.com/2caudio/mix-dedr ... audio-b2-1
I'm actually not sure exactly who made that example so unfortunately I can't comment of the preset(s)/setting(s) used. I think one of our beta testers made this example. It was a while ago now.

Oh I thought that was one of your tracks.

thanks for the other info.

Cinebient · Post by **Cinebient** » Sun Jul 26, 2015 2:51 pm

I let out a bit my melancholy here....
drenched in some B2 reverbs of course!

https://soundcloud.com/cinebient/remember-me

lnikj · Post by **lnikj** » Sun Jul 26, 2015 5:33 pm

That's beautiful!

Shades of John Tavener in that cello.

Den* · Post by Den* » Wed Jul 29, 2015 8:51 am

One new "Den Large Hall SD" for B2, one engine preset for sharing.
+1 more
Recommended

2CAudio B2: Full Body. Maximum Attitude.