Slate Virtual Console Plugin is now available

[_dada_]

Just tried the most recent "internal" (1.2.2).

Just when I thought, I'll send a PM ...

One being the already known (and earlier fixed) "open plugin/GUI results in audio dropout" bug.

Luckily not happening here ...
But I was already running pretty smooth, with 1.1.2.
( at 44.1, without realtime oversampling and grouping, with projects running at 30% to 40% ASIO meter )
I guess projects with higher CPU load may be more revealing.

bye, Jan

[Compyfox]

Could this be, that this problem really resolves with different sampling frequencies as well? I'm running 48kHz stock.

Oh and I don't go by any ASIO meters, I go by the CPU load meter in Windows.

[Mercado_Negro]

I don't know but there is a difference in sound/meters ballistics/headroom at higher sample rates, that's for sure. I work at 96kHz and VCC is rock-solid here with 1.2.2 (not with 1.1.2, the official versions didn't work for me so far). Btw, since I work at 96kHz I don't use oversampling while mixing, only for rendering and just 2x.

[Compyfox]

No oversampling while doing tests on mx end. While mixing, no OS either.

Did any of you reconstruct that "slowdown on GUI load" bug?
Try this: close all VCC GUI's, wait a couple of seconds, let your song loop. Then open one random instance.

On my end (i7), this results in an audio "dropout" until the plugin is loaded.


Now here's the fun part... leave one instance of VCC open... click around a bit, do other stuff. Now open another instance of VCC. Two things can happen: 1) actually nothing - no dropouts since there is one VCC instance running already, 2) sporadic happening audio dropout again.


I really don't know of this is due to VCC writing the settings file, or VCC acessin the iLok2. I didn't encounter this with Devil-Loc (the freeware), neither did I with PODfarm 1.


Mercado:
Did you forward that bug to the developers? I really want to see the VU/headroom fixed. This is THE ESSENCE of the plugin. Without it, I can't be sure if all mixes are off. And I don't want to overdrive all channels just to "hear" then this friggin thing bites - then it's too much already.

[Mercado_Negro]

Slowdown on GUI load:

No, I can't reproduce audio dropouts nor glitches on my end when I open one VCC plug-in. REAPER's UI freezes for 1 or 2 seconds but that's it.

Core i5 750 @ 2.67GHz
4.0GB Double-Channel DDR3 @ 666MHz
Motherboard Intel DP55WB (J1PR)
GFX card GeForce 9400 GT 1GB DDR2
REAPER 64 bit
Windows 7 64 bit

With 32 bit FP audio files and 96kHz for both audio files and project

VU/headroom problem:

Yes. As far as I know, they're working on it

[_dada_]

Compy, does changing the soundcard latency make a difference ?
( My RayDat is set to 1024 )

AFAIK, VCC is still the only plugin, using the strongest protection class
of the new Pace Eden ( iLok2 ).
That's why other iLok plugins will not show such issues ... I guess.

bye, Jan

[Compyfox]

Haven't tried higher latency values, I'm usually a low latency guy myself. Do you think that would make a difference?

In theory, I don't think this would work. The "dropout" occours while loading VCC (e-button in Cubase), but fortunately not while working regulary. On top of it, the tests I did were with barely any tools in parallel, last test even without wave files (only tone generators).

So yeah. I'm curious what the developers have to say to this. At the moment, it's really no fun using VCC on the long run.

[_dada_]

I'm pretty sure, higher latency can help.
Obviously, something needs a lot of resources, when opening a VCC GUI.
A higher latency gives more 'headroom' to the audio processing,
keeping the audio stream running, even if the system is bogged down for
a few msec.

At least, it may help to tell Slate, Your running on low latency ...

bye, Jan

[Compyfox]

Guess I should go update my ticket then.

[Slate]

Hi guys.. I've been getting a lot of emails today asking whether the VCC interacts with the actual DAW workstation's summing math. I want to discuss a few things about that.

An analog console adds dynamic nonlinear characteristics to an otherwise perfect audio signal. To review, these nonlinear characteristics are often component saturation, noise, harmonic distortions, phase distortions, crosstalk, and frequency manipulation (dynamic).

A DAW mixer sums by a very simple math equation and adds NONE OF THIS.

So if we take ANALOG CONSOLE SOUND and minus DAW PERFECT SOUND, the product = DYNAMIC ANALOG ARTIFACTS.

So what the VCC does, is recreate those dynamic analog artifacts and allow you to plop it into your DAW.

There is no need to interfere with the actual math equation of the summing of the DAW.. what would that do? Add artifacts that we are already adding within the Channel and Mixbuss plugins??

We modeled the entire path of the console's channel traveling into the mixbuss.. and applied those nonlinear models to the plugins. So when you use the VCC on your channels and Mixbuss, it is adding all of the nonlinear artifacts that we were able to find in the real desk. Interfering with the simple addition plan of the DAW's summing is pointless and wouldn't gain us any benefit at all.

Let's say for example that we WERE able to mess with the DAW's summing... Ok so then we would put in an algorithm PRE MASTER fader.. that would alter the phase, harmonic profile, and adjust the headroom as each channel was being added to the summing network... so in a sense we're recreating the summing amplifier and center section artifacts PRE master fader... Then it would go to the master fader.

But wait... our mixbuss plugin placed on the first insert DOES alter the phase, harmonic profile, and adjust the specific kind and curve of saturation (depending on desk), adding the exact same artifacts as each channel is added to the mixbuss!

My point, there is no difference. Regardless of whether or not you put the artifacts right before the master fader or in the master fader insert, you JUST NEED TO ADD THE DYNAMIC ARTIFACTS to make it sound like a recreation of the analog signal path.

Hope this explains it well, and hope you guys are making great music.

[Compyfox]

I think this was directed at the Sonimus SatSon crowd. And basically all these type of "summing" plugins do the same after all. Thanks for the clarification.


On the other hand, what about the updates, Steven?

I still have an open VCC ticket (bug reports, I'm probably not the only one), most notably the shifted VU, and the drift mode is not available either. Which is why I still stick to beta 2.5 (most stable, most acourate IMO, with some setbacks but still).

I know you're probably stuffed with studio work and the SSD v4 drums. But can you maybe give us a slight hint when things might move forward again?

[Slate]

Hi Compyfox, the discussion that is happening lately has spurred my inbox to be bombed with questions of whether or not the VCC "takes over" the DAW summing so I hope the above post clears it up! Also, for anyone who wants to discuss things with me personally, please use slate@stevenslate.com rather than PM.

I think we're gonna have an update very soon that fixes remaining reported bugs that you might be experiencing. We're also making great progress with 64bit and WinRTAS... this alone will take care of a lot of things too.

The drift will come a bit later!

Thanks,
Steven

[aciddose]

"takes over the mixing", ...

i mean the fact that people actually believe mixing / summing / adding aren't all the basic operation known as: addition! a + b for example. there can be only one answer to this regardless of how it's implemented. if the answer isn't a + b, it's incorrect.

of course in an analog console the operation is still exactly a + b, but the difference is what you put into a and b, and how the result is modified after being added.

the most important factor in my opinion as i've said before is the imperfection of where you place the controls. the level, eq, gains, bus panning, send levels and so on all will have small (5%? more?) variations that are not random - they're where you turned the knob. it isn't humanly possible to perfectly center a knob, even knobs with a center detent tend to have small offsets.

any cross talk, distortion, noise, filtering or otherwise is secondary to that but ultimately it's still a matter of:

[input] -> [do stuff like input gain, eq, etc] -> [add to bus] -> [do stuff to bus]

"add to bus" is always going to be a simple "+", nothing more.

[Compyfox]

I still hope the VU bug (shift) will be adressed (along with the feature request to add +/- 24dB for the gain/trim knob). 10dB off in terms of the meter sure is a bit much. And I still think having an accourate VU linked to the headroom setup (-18dB RMS = 0 VU with 300ms integration time should resemble the VU being at 0 if I send in a -18dB sine signal) is essential while doing postproduction. At the moment I NEED a third party plugin to see the correct level, and I don't know if the headroom is shifted as well. Especially not if I work subtle (aka not overdriving more than +1,5dB to +2dB on the VU).

Which is the reason why I still use Beta 2.5!


I still wait for the manual update regarding the VU reponse times as well (Fabrice mentioned he wanted to implement that). And it would be nice to add the corresponding "pan rule" for each console as recommendation. I read somewhere that the SSL uses 4,5dB as panning rule. Don't know off hand about Neve, API or Trident.




One final question for you, Steven.

You mentioned the nonlinear artifacts that your plugins create to (dare I say) "analogify" the digital mixing environment. Crosstalk does count to this as well, no?

Your initial presentation mentioned this as essential feature. Can I take it it (crosstalk) is still there or did something suddenly change and "bam - gone"?

[LeVzi]

OK I understand the theory, but in practice, how does this actually sound ? No meters, no visual aids, just straight up on or off A/B , listening to both.

Does Analogue summing actually make "That" much of a difference ?