2CAudio Precedence / spatialization

[Lawrence]

https://www.2caudio.com/products/precedence#_Overview

[Guy Rowland]

Interesting. This in particular caught my eye:

Like Breeze 2.0, Precedence is so efficient that it is easily possible to run over one thousand instances on the latest desktops and workstations such as Apple's new iMac Pro and Intel Skylake-X and Xeon based Windows machines. Running a separate instance for every performer in an entire 100-person orchestra to achieve new levels of realism for example, is now quite literally a breeze on modern CPUs. Breeze 2 and Precedence represent the first pro audio software of any kind to be optimized for Intel's new AVX-512 instruction set and performance is simply staggering on CPUs that support it. Performance on recent laptops and older machines is equally impressive.

Ultra low CPU usage is great, but there is one thing even better: zero CPU usage! Beginning with Precedence and Breeze 2.1, we have now introduced an intelligent Suspend-On-Silence feature in our products. If the plug-ins are not receiving any input signal, they will automatically disable processing, and enter an approximately zero CPU usage state! This is huge news for composers who use massive scoring templates and like to auto-load hundreds of instances. Maximum CPU usage will only be reached in such cases when all tracks in the project are active at the same time, which almost never happens in large sessions. Sparse arrangements can save orders of magnitude in CPU-resources, which in turn leads to more creative freedom and less annoyances such as heat and fan noise generated by powerful computers in the studio. Furthermore, if playback is stopped for a coffee break, lunch break, or overnight, processing is automatically suspended, potentially saving you significant money on your electric bill. In other words, we automatically turn the lights off for you, helping to save the planet in the process. We think that is pretty cool.

This all sounds wonderful. However, the demo audio didn't sound that great to me, but in part I didn't much care for the samples themselves. Would be interested in seeing / hearing a proper walkthrough video.

[Piet De Ridder]

New from 2CAudio: Precedence, a spatializer. I haven't tried the demo yet, and apparently, there's a few things we need to know first before making up our minds about it: below is what Andrew wrote on VI-C (I hope he doesn't mind me quoting him here):

the most important thing to realize regarding Distance is that for Orchestral / Scoring work, we expect/advocate the Precedence be followed by Breeze 2.1, directly on tracks. Enable P-Link in Breeze 2.1. Then set Breeze Mix/Balance to the same value as Precedence Distance. Now you will have the full expected depth/distance result.

Precedence provides 100% of the Left-Right positioning and about half of the distance positioning. Breeze 2.1 (or eventually B2 or Aether) supplies the other half.

Distance perception in enclosed spaces is very dependent on the "Direct to Reflected Energy Ratio" (i.e. Mix in standard music production terms) and also the "Initial Time Gap" (i.e. Pre-Delay in standard music production terms). Everyone knows intuitively "more verb means father away" to some extent", but there is wide misunderstanding about how to use Pre-Delay properly in a manner that agrees with other distance cues.

If you load Precedence and then Breeze on a track and set Precedence Distance and Breeze Mix to the same value the distance illusion will be completed perfectly. If you do this for 16 tracks in parallel and each has a different position in Precedence, you will have 16 instruments properly positioned with a unified space.

Without Precedence, using P-Link in Breeze will still help establish some distance cues and keep them agreeing with one another. It can be nice to work that way even without Precedence. It still gives *some* of the proper perceptual cues. Rather like simple Gain Panning gives *some* perceptual cues about Left-Right position. It works to some degree, but it can be much better when ALL perceptual cues agree with one another.

Breeze defines a space. Precedence can put the source sound anywhere you like within this space.

If you ignore "Direct to Reflected Energy Ratio" and "Initial Time Gap" cues which come from reverb, i.e. if you use ONLY Precedence, you will not get the full distance effect. But this is similar to saying if you mix your dry orchestral samples without reverb it won't sound like a real-world performance. We expect there will be reverb. And ideally that reverb will come from Breeze 2.1.

Without the distance cues related to reverb, the strongest ones are gain and high frequency loss. For music mixing you don't really want us reducing the gain of your tracks drastically based on distance (like it seems the example above does?). This would be physically accurate, and good for sound-FX and virtual reality, but for music mixing, it's best to let the track faders control mix levels IMHO. So we don't add any gain loss purely by distance.

We do add high frequency loss, but we do not do so in extreme manners bc this would also probably not be appropriate for music mixing. You can increase this effect by increasing the Freq Delta value.

We add other sophisticated distance cues in Precedence, but they are designed to work best in a model of an enclosed space: i.e. concert hall, or algo verb approximation thereof. The signal is pre-conditioned to expect the verb to supply the other cues. Breeze 2.1 will supply them completing the illusion. It's synergistic.

We are exploring now establishing communication between Precedence and Breeze instances, so that changing position in Precedence automatically updates Breeze as well. This will be the eureka moment! We hope to have that ready in the next 30-60 days. Until then for orchestral/scoring work:

Add Precedence.
Add Breeze 2.1
Enable P-Link in Breeze 2.1
Set Distance in Precedence
Set Mix/Balance in Breeze 2.1 to match Distance in Precedence

_

[Hannes_F]

To me it sounds very good, however I am on Laptop speakers here right now.

I remember discussing such a plugin for orchestral use with Andrew Souter at Galbanum / 2caudio as early as 2009! This is what I wrote then:

" We orchestral composers do not need much right-left panning because that can be done with a combination of power panning and delays. What we need is to move the sound source backwards in the room, hopefully without making it much wetter.

A good positioning algorithm would definetely incorporate the contribution of the floor (underestimated IMO), air absorption and absorption by bodies (musicians and audience). All that happens before the ER kick in."

Can't wait to try it out when I'm back in my studio.

[Piet De Ridder]

Have meanwhile spent some time with the demo version and my conclusion so far is: totally useless plugin. It doesn't even qualify as a fun toy, in my opinion. It started off promisingly though, because I really liked the 'stereo air' that Precedence creates around a dry source (The Trumpet, in my tests), but sadly, that stereo signal turns out to have zero mono-compatibility (meaning: a very weak and fragile stereo-signal) which is, for me, in itself already reason enough to immediately stop taking this software seriously.

I'm also not too impressed by the difference between a value of '0' for distance and one of '100'. To my ears, Precedence makes that sound like a difference of a few centimeters at most (where you'd expect at least several meters, no?).
See, when I test SPAT in Belgium, I can make it sound as if the source is located somewhere in South-Africa or thereabouts. With Precedence however, the best I can do, is make it sound as if the source is positioned one or two feet behind my computer screen.

Also a problem is the fact that just about every parameter change is accompanied by severe 'calculation crackling'. Which means that you can forget about automating any of Precedence's parameters. (And subtle automation of certain positioning parameters can really add surprising amounts of life to a mix. Which is partly why I love SPAT so much.) Again, in SPAT, I can drag sources to the left and to the right, push them all the way to the back or bring them up close, make them rotate around their axis ... and you'll never hear SPAT cough up even a single crackle or click. (The only time SPAT produces such crackling is if you change certain reverb parameters like its length, or the size of the room.)

Not a problem as such, but puzzling nonetheless: why does Precedence include a settings-randomizer? I mean, who in his/her/its right mind is going to leave something so delicate and precise as spatialization, to the roll of the virtual dice?

I'm aware that testing Precedence without it being linked to Breeze is apparently an incomplete and unfair test, but even so: even without Breeze, one should at least hear some hint of the power and musical usefulness of Precedence, no? I'm afraid that all I heard is that Precedence has no power or musical usefulness at all.
Odd that 2CAudio, who have quite a reputation when it comes to reverberation and such, should release something so underwhelming.

_

[Guy Rowland]

Thank you Piet for once again being so diligent at the coal face. One less product for me to worry about.

[SirKen]

I'm also not too impressed by the difference between a value of '0' for distance and one of '100'. To my ears, Precedence makes that sound like a difference of a few centimeters at most (where you'd expect at least several meters, no?).

This was my experience as well. I did not feel any major changes based on the distance from the listener.

Also a problem is the fact that just about every parameter change is accompanied by severe 'calculation crackling'. Which means that you can forget about automating any of Precedence's parameters.

This also makes fine-tuning a big hassle.

I'm aware that testing Precedence without it being linked to Breeze is apparently an incomplete and unfair test, but even so: even without Breeze, one should at least hear some hint of the power and musical usefulness of Precedence, no? I'm afraid that all I heard is that Precedence has no power or musical usefulness at all.

Breeze makes a difference. The changes in distance became more audible when you add it into the equation. However, users have to manually enter the distance value from Precedence into Breeze wet/dry ratio field. They promised that this will be automated in an upcoming Breeze update before the end of the year.

Odd that 2CAudio, who have quite a reputation when it comes to reverberation and such, should release something so underwhelming.

The implementation with the other reverbs (Aether and B2 will be updated soon) seems to be at the very early stages at this point. I also don't see how Precedence could work well with any other reverb outside of 2cAudio products. Which makes me wonder if this should have been directly incorporated into the Breeze etc directly.

I think I will stick with the positioning in EAReverb 2 for now.

[ZeeCount]

Here's a mockup I've been working on-and-off on for a few years from Jurassic Park. All the brass is sample modeling, and the woodwind VSL. Both placed using precedence on each instrument into seperate instances of B2 for the woodwind and brass. Percussion and harp is spitfire, and strings are Adventure Strings and Soaring Strings, using precedence to create a violin 2 section on the other side of the stage.

https://www.dropbox.com/s/mq7v6oc4kdx9q ... o.mp3?dl=0

[Piet De Ridder]

A bit of a problematic sound you got there, Count, if you don’t mind me saying. Though I don’t think Precedence is to blame for most of it. I can’t really describe what, to my ears, is wrong here, but if I were to try, I would reach for words like filtered, doctored, thin, glassy, exaggeratedly EQ’ed and over-produced to describe some of it. I also hear a collection of spaces (all similar but different), rather then one single space in which everything is happening.

And the brass, I’m afraid, doesn’t really show SampleModelling in the best possible light: those trumpets, for example, produce what I find a quite aggressively annoying sound — and I fear this is as much a programming as a production short-coming —, the horns seem unable to articulate even the most basic phrases (except the one at the very end), and the brass, as a section, seems also in constant danger of phasing.

How exactly Precedence is involved in all of this, is difficult to say, but to my ears it did not bring solutions to this difficult mix but rather more problems.

_

[Guy Rowland]

...though I think you should get a few points for even attempting, Zee. If you've done that midi arrangement by ear, its a really good job.

That said, I too found myself wishing away all the Sample Modelling - like Piet I really didn't like the trumpets especially as they lost that effortless sheen of the original entirely. If all the brass were replaced by, say, CineBrass, I suspect the whole thing would sit much more comfortably - not perhaps perfectly, but much improved. But I realise that would entirely defeat the purpose of this exercise, which is to make different sources blend and you can''t get much more different than Sample Modelling and any ambient library.

So SPAT is still the undisputed king of spatialisation. I've gotten slightly lost as to the recent update - is it now all done? How is resource use?

What I wish for in a spatialisation tool is something of the form and structure of VE Pro. 1 dedicated app, separate from the DAW, and then tiny send plugins on each channel which appear on the virtual soundstage in the app. The really good bit would be the ability to then route the outputs by stem - group together all woodwinds, brass or even more detailed if desired down to trumpets, trombones etc and those return to the DAW groups. I guess logically the app would require a copy of each engine for each stem - so if you routed everything to one output the CPU use would be very low, five stems would be 5x that etc. Incidentally, I think this would be very good for audio post use too.

I don't think this is quite how the new version of SPAT works but not perhaps a million miles off?

[ZeeCount]

I appreciate the honest feedback and criticism as always. Undoubtably cinebrass would get closer, since it's very similar players in exactly the same place as the original recording, but my aim was to see how close I can get sample modeling to matching it.

[tack]

I still feel inclined to give a shout-out to EAReverb2 for spatialization. Cheaper than Precedence+Breeze, and I think it does a much better job from what I've heard of the demos so far. Yeah, it's no SPAT, but to my ears it gives reasonable results at a significantly lower cost.

[Linos]

Guy, the setup looks so complex that I can't be sure. But to me it looks like that should be possible in Spat Revolution. If you set up one room per stem within the Spat app, you can then route the output of each room to a separate bus in your DAW. But don't take my word for it. An email to Flux should clarify. Apart from the outreagous price, the need for virtual routing through audio drivers is a huge downside on my setup. I have virtual drivers to route all audio through DRC, and it was a huge pain to set up and having it run reliably. Having that same hassle in my DAW as well is not a tempting outlook at all.
I don't understand how Flux can make a tutorial video explaining that you can ctrl-click in Spat Revolution, but no video about how to set up virtual midi cables and how route Spat Revolution within your DAW.

About Precedence not having too much of an effect: on Vi-C Andrew wrote that Precedence should do approximately 50% to add the depth, the other 50% being provided by B2. He wrote that Precedence doesn't attenuate the signal with distance because he thought that should be taken care of during mixing. Also, he wrote that the high frequency dampening was applied subtly by Precedence, because this, too, he thought should be taken care of during mixing.
That's an odd choice in my opinion. It would have made more sense to have an option to control these two things withing Precedence in my opinion, instead of having to deal with it seperately.

[rJames]

Thanks for the reviews. I respect you guys and this saves me time from trying it out. (with my terrible hearing).

[Tanuj Tiku]

The only thing I liked was the stereo spatialization. But I have also checked demos from UAD OceanWay and it does something similar but better - at least to my ears.

The choice of demos are not great. And it already sounds like a 'halo' of reverb hanging in there. It doesn't sound like it is putting sounds in a desired space.

I am reacting purely based on the demos. Not that impressive as per their usual standards. At least, not to my ears. May be I will give it another listen at a later time.

SPAT is light years ahead. However, I am very disappointed that they seem to have discontinued the older SPAT version and gone for a new multi-channel plugin that is priced at 1,400 EUR which is OK for its intended use but I very much like the old version which has exactly the features music production requires and sensibly priced.

I hope they are not going to abandon updates and continued support on the old version. That would be a disaster!