SPAT 2 announced

[Guy Rowland]

Yes, thanks from me too. Will be watching it closely.

[Jack Weaver]

Hey givemeenoughrope,

Did they happen to explain how an algorithmic plugin is going to be able to use impulse responses (as a convolution plugs does)?

.

[givemenoughrope]

Jack, I took this to mean a separate convolution reverb within SPAT. It now resembles a piece of software or rewire host as opposed to a plugin only.

[Jack Weaver]

Hmm...

Thanks a ton for the AES update, givemeenoughrope. I did watch a video on Gearslutz of a showing on the AES floor but of course the overwhelming ambient audio and the presenter with English as a second language all made it a bit challenging to get what was really happening.

Guess the real test will be the Piet Youtube review.

I'm hoping that I still use enough dry sample instruments to make SPAT worthwhile. Until I move into my next studio I won't be doing a lot of individual live instrument recording.

.

[givemenoughrope]

[Guy Rowland]

Thanks rope - interesting. It shows you can add any number of configurations of output, but it looks like it doesn't offer a stem workflow, which is sort of inexplicable. The most obvious uses I can think of for this - Post Production film and TV, Video Games, 360 audio, music mixing - all would benefit from stems. If it did and if the product was reliable, I'd be buying this right away I think.

Perhaps the architecture doesn't allow it. You want to control the same space, with the same tweaks, but keep some audio paths distinct. Perhaps that means multiple engines for each output and it would be CPU-prohibative.

The other thing that occurs is that the whizzsy 3D thing is all well and good, but representation seemed to be simply colours and shapes. You'd really want the name of each instrument / element surely?

[givemenoughrope]

I'm not sure that you can't route a stem workflow. That's what I thought the different room tabs to represent (could be wrong of course).

[Guy Rowland]

I'm not sure that you can't route a stem workflow. That's what I thought the different room tabs to represent (could be wrong of course).

You could be right - it would be most useful there to have identical rooms, just sent to different outputs. We wait and see - is there a release date / price or anything?

[givemenoughrope]

The routing matrix that you see in the video above at about :21 seems to point to that. I feel like a dope for not actually getting to the bottom of that while I was there (half the reason I lumped it downtown to AES). I've had SPAT for a little while now but I've only used it for trying to place things in their own room, like source cues or just samples that need to be pushed back into a space. This weekend I'm going to see how far I can get with trying to get 8dio strings to not sound wholly out of place with Sable (or whatever it's called now). I'm messed around a bit but I'd like to approximate mic positions as well. I know SPAT is ideally used on an insert and not on a send but maybe the C or even F could pass for the C mics in Sable and a quad out of the C mics could approximate the T and A mics. Not expecting to recreate Air L with a plugin but these are samples so what the hay.

[mr anxiety]

Regarding multiple stem output for film mixes and the use of Spat, this is what I currently do in ProTools. I have 6-8 instances of Spat, one for each of my stems for the orchestra (Strings-WWs-Brass, etc.) I send the corresponding sound sources (pre-fader) to these channels and buss them to my stem tracks for my music mix. In Spat, I place the instruments on the stage appropriately; the Strings up close, the WWs a bit back, Brass even farther back, etc. Yes, it is a lot of instances of Spat, but my MacPro 12 core is handling this so far. Sometimes I print a combined Orchestra tail stem that has a global hall setting sent from each Spat channel to taste. If I need to print verb tails for each stem, then I set up another set of channels to do this discretely. This starts hitting the CPU pretty hard if I don't use a DSP verb like ReVibe. I have no idea if Spat 2 will make this setup change for me. I hope this clarifies anyone still wondering how to use Spat for film music stem mixing.

[Guy Rowland]

(curious as to why you send pre-fader, mr anxiety? Doesn't that prevent you effectively mixing the audio volume before it enters SPAT?)

[coder]

https://www.gearslutz.com/board/12173542-post2.html is the video posted at Gearslutz of SPAT 2 being demonstrated at AES 2016.

[mr anxiety]

I have my instruments running into Spat pre-fader to allow all of the stage positioning to be handled within Spat. If I was not pre-fader, I would have direct plus the stage positioned version of the direct signal plus the ER room., so that would not work.

All my balancing happens via midi in my DAW (Cubase). I'm always running the DAW and Pro Tools,so any balance changes that need to happen are done in the DAW.

Mr Anxiety

[Guy Rowland]

I'm afraid I still don't understand, Mr Anxiety (it's a Monday morning - the fog is perhaps slow to lift today). Do you mean that as well as being pre-fader, you don't route the audio direct to any busses within your DAW? (ie, it is sent only to SPAT?) That is how I'd expect SPAT to be routed, but the pre-fader part of it is still tripping me up. Effectively, it means the fader on your DAW does nothing at all - which now I read your reply again seems to be your exact situation, and you are constrained by only using midi to balance, which seems a long way from ideal to me. How I'd EXPECT it to work is post-fader, routed only to SPAT, so you can ride the level and as it goes up and down the room goes up and down with it. I can't see a reason why you don't do this - the critical part is having the instrument routed solely to SPAT, surely?

[mr anxiety]

The issue with having it post-fader the way you feel it should be is the way that the spatialization works, from my understanding, is you have two elements when introducing audio into Spat. You have the direct/Source signal and the room reflections based on where the direct/Source signal lives in the "space/stage". Unfortunately, you cannot have the signal/instrument "living" outside of the Spat environment and benefit from it's stage positioning/spatialization. If you have a "direct", unprocessed signal as well as the direct signal and room inside of Spat, you essentially have two direct signals and any latency offset from Spat will then come into play and you will have a phasing issues, etc. Now, you might think you could just use the room signal in Spat and keep your direct signal from the sampler input, but you lose the stage positioning aspect of Spat then. It serves a different task the; it becomes a ER supply but not a stage placement tool.

Does this enlighten the situation at all for you?

Now, if you want to adjust the level of your instrument in Spat, you can either do it before it enters Spat via your DAW, which is what most of the composer I know do it, or you could automate this in ProTools within Spat. You would automate the "Direct/ Source" fader in PT or your DAW. Maybe that is what you were thinking. The Spat plugin is a very complex tool and having your Direct/Source starting within Spat is very critical for it to do it's spatialization. The front to back placement effects the Source signal substantially and you would not have that if your direct signal was only coming from outside of Spat.

Now, this is not without some complexity if you have a mix that has Spat and non-Spat instruments together, in terms of delay compensation. That is why on my film scores I have a Spat instance for all of my stems, so everybody is on the same timeline/offset. It isn't difficult to mix both of these if need be, it just requires delay compensation.

Hope this was helpful. It's hard to explain this with words.........

[Guy Rowland]

I think we're definitely at cross purposes. At no point - not once - in this thread have I ever advocated having two signal paths. I've always understood it should just be one - into SPAT. My sole issue has been why that path is pre, rather than post-fader. In both cases the signal should go to SPAT and nowhere else - it doesn't get routed to a group in the signal chain as well, just SPAT. Only SPAT. Nowhere else. Just SPAT. That's it. When pre-fader, your physical fader (if you have a surface connected) does nothing and is effectively useless. When post-fader, it can affect the level of the instrument going into SPAT and nowhere else - which is just what you want, right?

[mr anxiety]

Hey Guy, Why are you making me feel like an adversary? I'm only trying to help in making you understand my workflow.

Thanks,

Mr A

[Guy Rowland]

Hey Guy, Why are you making me feel like an adversary? I'm only trying to help in making you understand my workflow.

Thanks,

Mr A

Apologies, not my intention of course. Just a bit of morning frustration that I can't seem to make my position understood. You have a method that works for you and that's all that matters - all the very best.

[mr anxiety]

I see. OK, maybe the issue that's confusing is that when I have my inputs post-fader in ProTools I am actually hearing the direct signal from two sources when the fader is up, due to the bussing of the input channel to the appropriate stem, which is left over in my template I used pre SPAT. To solve this I decided to just control the level from my DAW, as opposed to having another gain structure option to deal with. I get my samples from a variety of sources/farm PCs and they get summed into their appropriate stems into Spat.

This is probably the missing info you needed to understand my situation.

[Guy Rowland]

Ah gotya - that makes sense, thanks Mr A.

[givemenoughrope]

Mr. A, Piet or anyone,

Is there a method for simulating Close, Far, etc mics with SPAT? I'm trying with a piano recording (recorded with close mics) by using SPAT as a send effect routing LR to one bus and LsRs to another. I realize this is sort of a no-no but it sounds ok to me. But maybe there's a better way to do this to prevent phasing, etc.

[Hannes_F]

@givemenoughrope
The beauty of having Close, Main and Far (or even more) mic systems is that they are physically different microphones catching different aspects of the sound in the room. This means added content, added information embedded within the sound, even if they are mixed down to a stereo track. All this content originates from the sound source, so in analogy you are looking onto your sound source from more than one angle at the same time. Which makes the result more "3D" in a way.

Running the signal of one mic system through room simulators and then adding them up also adds information to the signal, but this is only information about the (simulated) room(s). The analogy would be to look onto your sound source always from the same angle but through different glass filters, prisms etc.

So yes, it could add an enjoyable contribution to the spatialisation aspect of the music, but not to the musical signal itself.

[givemenoughrope]

Thanks, Hannes. I think I'm following. (I thought I wasn't but maybe I was reading too into it.) Of course, it would be great to have mics 10 and 20 or whatever feet away and capture them. But the piano is basically in a large booth. Using SPAT like this has been the only way for me to actually get close to simulating a real room. but am I introducing too many other phase issues? I realize that I'm not really hearing different mic systems as you say but sounds ok to me. (I'll to post what I'm doing soon if I can.) I'm wondering if I can do the same wit trying to get 8dio strings (C mics i would guess) to sit with Sable.

Do you think I'm better off just using SPAT as an insert, putting the L and R source very close and then pushing the speakers further back into the room to simulate room mics? I actually tried that quickly and missed the sound of the close piano but maybe I wouldn't with strings.

Is this just going over my head? I'm not convinced that I totally understand what SPAT is doing from beginning to end.

[Hannes_F]

givemenoughrope, the most natural approach for using SPAT would be this: Run your piano signal fully through SPAT, no bypass paths. Usually that means that SPAT is used as an insert effect. And in case you have any knob to regulate the percentage of how much the signal is processed by a VST then that knob would have to be on 100%. Or, with still other words, you use SPAT 100% wet, no dry portion of the sound is allowed to come to the main output.

IO setup: The simplest method would be to have a stereo input and a stereo output. In your case I would pair the input into a stereo pair.

SPAT_Stereo pair.jpg (10.13 KiB) Viewed 15302 times

Go to the "Source" tab and check whether your combined (stereo) sound source is really in the middle. Set Azimuth to 0 if needed.

(Please note that while I am using the current version of SPAT for audio it does not look good on my screen because of an outdated graphics card, and that is why I use an older version for doing these screenshots. You should see two speakers in your version of SPAT instead of the one showed in this picture.)

Now if you use the default reverb of SPAT you perhaps will find that the result is too wet. With a conventional reverb you would fiddle with the wet/dry slider at this point. With SPAT you go into the Reverb tab and lower the "Reverberance" slider. I don't shy away from having this very low if needed, like 10 or 20.

Now if you want to have your piano further back you go to the "Source" tab and increase the distance.

If the loudness falloff is too much then you simply raise the output slider. Check whether the piano really has moved back (instead of just getting lower in volume) by activating/deactivating the whole plugin. You want a sound that is as loud as before but further back in the room.

There are more tweaking possibilities but hopefully this will walk you through the basics. You should have a natural and good sounding result at this point.

[Piet De Ridder]

There are several ways to balance the source and the room's response in SPAT.

Apart from the one Hannes suggests, there's also:

(1) Lower or raise the 'Room Presence'. Note however that, even when this parameter is set at zero, there's still a noticeable presence of room in SPAT's output. That is the sound of the room at its least present, but still not absent of course.
(2) Lower or raise the 'Source Presence'
(3) Increase or decrease the source's radial projection (with the APERTURE parameter). Narrow radiation triggers less of the room's response, wider radiation triggers more. Important parameter.
(4) Use the 'YAW'-parameter to rotate the source's projection away from the listener. One of my favourite SPAT-parameters and one that makes a huuuge and amazing difference. Especially in combination with APERTURE. (The more narrow the Aperture, the more effect changing the Yaw will have.)
(5) Balance (or edit) the Omni- and Axis-EQ's. 'Omni' deals with the omni-direction signal, whereas 'Axis' handles the signal that goes straigth from the source to the listener (and thus contains most of the dry signal). Lowering the 'Omni' makes for a dryer overal sound, lowering the Axis makes for a wetter sound.

All of the above are illustrated in the videos posted here.

And there's another set of parameters — which, to my embarrassement, I learned to work with only a few months ago, long after those videos were made — : the trio 'Drop Mode', 'Drop' and 'Radius'. These three are all connected and simulate the sound pressure-drop that comes with increased distance or, in other words; the (natural) phenomenon of sound becoming weaker with greater distance. Difficult to explain how it translates into practical use for us, so perhaps I ought to do another video to demonstrate it, because it's a really useful triplet of parameters.

Rope, if it would help: why not post an example of the piano which you're starting out with, as well as an example the sort of spatial sound you're hoping to end up with, and then we can maybe look for it together?

_