Feature Request !!..Best Idea Ever...

[Jake68]

OK Guys I have had one of my momentous ideas.

In Cubendo you can create VST connections for outputs, these can be virtual or real.

Imagine if UAD created a digital model of the summing from a world class console Neve or SSL. They have summing character which is often nicer to listen to in my experience.

Why dont UAD create a powered plugin VST Effect array, that allows you to insert these effects across multiple VST Connections. The First Insert across the master mix output pair re-routes the audio to a stacking virtual Neve summing buss. You add another plugin to a VST Connection and it creates another stereo channel strip in your virtual Neve and Sums the tracks back down to the output of the original Connection so that you can render the virtual summing bus. The most connections or even groups (?) you make that you add the routing plugins to, the larger your virtual console becomes. This could also allow for an improvement in digital summing resolution as well I guess.

What dya think. Hell of a project for UAD, but a damn good way of selling shed loads of DSP cards, especially the 4 in 1 cards that they gonna HAVE to bring out sooner or later. Personally I think this would be THE big seller!

An automateable virtual classic console with all the characteristics of classic summing buss modeled in a high internal precision DSP plugin. Its a winner!

 

[Paul Woodlock]

Full marks for thinking 'outside the box' for an 'in the box' solution

personally I think something like this will eventually be implemented in the DAW program itself, where the whole signal path can emulate the Famous Desk of your choice.

Imagine being able to select between Neve, SSL, API, etc ( and of course clean digital ) within Nuendo?

That's where Iv'e always seen it heading.

However it's an innovative idea Jake.


Paul

 

[Empty Planet]

Not to rain on your parade, but the problem with this kind of approach is that once the sound has hit your outboard pre and gone through the converters, it's already been wanged. In other words, unless you're tracking through fairly high-level outboard gear, the track has already been sonically comprimised in a way that will forever prevent one from reaping the benefits of an ITB Neve or SSL emulation.

I didn't really get this till I got my first high end pre, but after getting familiar with it and beginning to get a grip on what it was doing for me, I realized that my low- and mid-level pres had been subtly constricting the frequency range of all my tracks, that they had been introducing unwanted artifacts (resulting mostly from the use of mid-level or cheap parts), and the level of self-noise from those units had been coloring every sound I recorded in what was, more often than not, an unpleasant way. Unfortunately, one can't hear the difference until one auditions really nice gear in the environment in which one works -- I was the exact same way.

All I'm trying to impart here is that the frequency range, lack of noise, and quality design is a big part of the sound on those tracks by your favorite artist. Part of what makes high end consoles so fab is that you have amazing frequency response, incredibly low noise floor, and an expensive and well-thought-out signal chain taking care of your audio at every point in the path. This adds up, it really does.

If a person is using only mid- to low-level front end gear, the very best such emulations could do is impart a distant flavor of high quality gear, but without crucial elements of the actual quality which, alas, makes those units so damned expensive.

:?

Cheers.

 

[Jake68]

Well thats not exactly true.

I have mixed several tracks down from PT (recorded in Nuendo) into a Neve console.

Analogue summing of this quality still has a quality that it imprints on any sound.

Obviously it would be better to record the entire shouting match with that quality level, but even tracks that arent recorded that way are often mixed that way. And it does bring something unique to the show. Something different and imperfect that sounds desirable, that digital mixing just doesnt have.

Not all material recorded on first class consoles even benefit from high quality signals. But any DAW could benefit from an analogue mixing model.

Also please consider that within your example it wouldnt be possible for anyone but high end users to benefit from any improvement in effect technology like UAD's other products. But they do.

Try this. Take a Pultec and insert it across a buss, make sure its a no boost or atten. initial setting and turn it off. Play your mix, half way through turn it on. See what happens.

IMHO anything sounds better summed through an analogue console and surely it should be possible to model those characteristics.

 

[Empty Planet]

Well gosh, if I had the opportunity to mix down my ITB tracks through a Neve, I'd definitely do so too!

But I'm not talking about the fact that it's ITB. I don't think the medium is important. I'm talking about what it went through on it's WAY to the box. That's where the sound is so often getting compromised.

Sounds created in the box? Absolutely. I'm with you. A summing boc might impart some character that might fit the song.

But if we're talking about sounds you actually recorded, it's important to realize what the expensive gear actually does to the sound, that it's not just about \"character,\" and to recognize that you can't put back what has been taken away by a mediocre recording process.

Cheers.

8)

 

[Empty Planet]

BTW, I realize my posts sound like a downer from a certain perspective. I don't mean it that way, I'm actually just trying to give you some information that will help in the larger picture. I was spending a lot of money, over time, on software solutions. If I'd saved up and directed those resources to a decent front end sooner, I'd have gone farther, faster, at least in the direction of getting that \"record\" or \"pro\" sound. That's my situation, recording things like guitar and vox. If an artist's work begins and ends all in the box, it's a different situation entirely. At any rate, best to you in your efforts.

8)

 

[Jake68]

No mate I fully understand your point. Any mix can be improved by first class pre's. That still doesnt stop hit records being made on Mackie 8 busses! And it doesnt negate the point that DAW summing is unsympathetic.

But in my view, if anything requires improvement in DAW its the summing.

Taking your point there may be no requirement to generate any superior DSP products for anyone.

If you aleady had an SSL why would you want superior DSP, you could just get a MADI solution and stream your channels into you VCA or flying fader first class console and compress with your hourds of Urei's etc and your classic plate and acoustic chambers maybe a 480 or something.

But the point is that if we can improve our ITB mixes with UAD's then we can improve the mix further with more sympathetic summing. And I think that word is key \"sympathetic\"!

Anyone know how to find out what UA think of this idea, I'd be really interested!?!

Likewise, dont think of me as being argumentative, but I am convinced that if an LA2A is a good product for a DAW then so is a Neve summing buss!

 

[Empty Planet]

No mate I fully understand your point.

No, mate, clearly you do not. To wit....

Taking your point there may be no requirement to generate any superior DSP products for anyone.

God only knows how you got that from anything I said.

That still doesnt stop hit records being made on Mackie 8 busses!

Not the point I was making. A lovely, defensive, entrenched response to a point of view that you've decided to envision as opposing.

Ugh. Why do I bother?

But in my view, if anything requires improvement in DAW its the summing.

Which you propose to solve...by DAW summing.

Gotcha.

I wouldn't wait up too late for UA to call.

:roll:

 

[Paul Woodlock]

.....But in my view, if anything requires improvement in DAW its the summing.
....!

The problem is Jake, is you never hear JUST the summing of a Neve or SSl, etc, you hear all the rest of it too.

Summing in ANY hardware desk is just a bunch of resistors connected in a star config to one virtual ground earth point. The bus itself is just a piece of wire.

So what makes a Mackie 8 bus different to a Neve aint the sum bus it's the other stuff.

So really you don't need the convoluted routing required by your suggestion. You need a Neve channel strip emulation for the channels, and Neve output section for the DAW output sections. 2 plugs basically.

That's they I see it anyway.


Paul

 

[Jake68]

???

LOOOK...

We ALL understand that audio is better recorded through high quality front end.

I DO understand you point.


You have even said youself that you would like to be able to dump your ITB mixes through a Neve summing buss. An analogue buss has character and that character can be modelled.

It doesnt matter WHAT you recorded and how you recorded it, if anything analogue summing is more forgiving.

We all KNOW that its best to have a complete signal path. But as I said I have just been experiencing the benefits of mixing down on a world class console. Material that was inexpensivley recorded. When it comes to models we have the EQ's, We have the compression. What we dont have is the summing.

Perhaps its my point you are missing?
Even if you have something that is poorly recorded, the character of an analogue summing bus may deliver a desireable charactistic. After all, why sell shed loads of UAD's to kids who work out of their bedrooms with Focusrite Platinums?

 

[Jake68]

Even if the EQ's are set to zero, or lifted out of the signal path completely.
There are op amps, and other circuitry that give it \"character\".
There are analogue summing boxes available as a completely seperate product. If what you say is the total \"sum\" of the situation what is the viability of these products?

Analogue summing is not just a matter of combinations of signals.
There are op amps in each channel AND in the centre mix section of the console. Each one of these components aside from the EQ's add to the character of the console.

But this is the beauty of the idea. Not only can you model the behaviour of these op amps but you can change the bit resolution and even the sample rate of the summing engine itself.

\"Gotcha\"...not quite. Currently our DAW's sum at the host sample rate and with 32 bit floating point. What I propose isnt merely DAW summing. But summing that models the op amps used in classic consoles for their frequency response etc, but one that may also up sample and increase internal precision. No reason to merely use the hosts depth or rate.
UAD currently make world classic EQ emulations, one of which up samples as I remember. We have the EQ's, what we currently realy on is unsympathetic 32bit Floating point precision.

The idea is that flexible. Three is no need for me to defensive in any fashion, I just havent heard anything yet, with respect I say again, I am not looking for any confrontation, that negates the idea of a model for analogue summing.

Paul I could understand you point if these classics where passive.

The idea is so flexible, why stop with the summing.
Say UAD sell the Summing Buss as one plugin for $200 say, then they sell console EQ's as another add on?

The scope of the UAD itself is not to influence the input in anyway, but to merely make a better or even just different job of effecting the sounds recorded than those offered by other DSP designs. They do this by \"modelling\". Anything can be modelled. Ampegs, Marshalls, Microphones (!), Even with a datum its possible to model mic amps.
Why on earth would anyone want to model tape saturation? But they do? Because sometimes, it sounds good.

Of course we should all have a Neve sidecar, but we dont. Sometimes we have VLZ's or Behringers. All of these guys are usings UAD compression, if so why not use UAD Summing. The Models and internal precision could even be switchable!

 

[TomW]

Hi Jake,

In fact I suggested this very thing like 3 years ago...no-one took any notice.

My idea was a little different though and covered both angles. I can see Pauls point of view and yours.

I suggested a whole plug-in mixer with modelled summing. It would obviously require a whole re-write of the VST engine but anyhooo. You know how you can create mixers in Soundscape or SCOPE? Well imagine if there was just a blank FX slot in 'Cubendo / PT' whatever, that you could load a DSP builder into. There would need to be a basic framework for I/O streams etc.

Dial up some API models for drums, SSL4ks for geetars, fx and some Neve for vox...whatever. Along with the already excellent stuff like the LA2A, 1176, EQP etc. Wouldn't it be nice (theoretically) to have a SCOPE like patch system so processors can be chained in all manner of configs, complex parallel processes etc before being returned to a mixer I/O stream. The DAW just edits and streams to and from the DSP mixer.

I all honesty I'd love to see PT go in this direction as its one of the only things thats capable (apart from Pyramix or Soundscape). There is just not enough DSP on a UAD-1, maybe not even in a UAD-4. In fact apart from the modelled summing this sounds exactly like Soundscape!

I really have no problem with DAW summing its proven to perform very well, if its colouration you seek, go OTB. Over the last three years I've come to realise that I just don't care about having the above as there are many options now, too many that get in the way of the music.

Fuck it lets go analogue.....I dont care if it sounds better - it feels better!

To add to Pauls comments on analogue summing. There is a lot more going on in a summing network than just the 'bunch of resistors' connected to the sum node. Its make or break time after that, the way the combined signal is dealt with and amplified, especially in a virtual earth mixer. Frankly the way SSL or Amek get the performance out of their large consoles is scary, way scary.

Couple this to the yards and yards of supporting circuitry (eqs etc) and as Paul said the sound you heard on the Neve desk is hardly just summing. But you surely cant say that a Mackie sum-amp is going to perform as well (let alone sound as good) as a superb high-end consoles' mix buss.

I really think this is going to be a tricky thing to model, possibly impossible when you add complex harmonic distortion and saturation from things like output transformers....and if its done this DSP system will not be cheap as it will need to be POWERFUL. If you can afford a HD system worth thousands then you could also afford one of the various analogue mixing solutions that are going to be the real deal.

It actually seems like a complete waste of R&D time because I'd rather the guys got that Ethernet based UAD-4 out so I can run more plugs...the only nice aspect is the complete freedom to construct whatever, only limited by your imagination and available DSP in digi land!

What was the Neve console BTW? If you have access to one then why worry about a fake modelled solution? Mix on the real thing mate.

I know I would.

Anything can be modelled. Ampegs, Marshalls, Microphones

Yeah but how many times have you ever heard anyone say that the modelled mics, marshalls etc sound anything like the real thing? Sometimes its close maybe even 90% but if you want the sound of that Neve you're going into that last 10%.

I also dont see the point of adjusting internal precision - surely you want the most precise model - whatever it takes...or would you want to take a 98% model and tweak to be 75% accurate??

You seem adamant that you loved the analogue summing thing...why not just do that very thing? Seems simple to me, plenty of products to choose from, you wouldn't be alone in that choice.

To be honest it seems crazy to have a request for a DSP modelled mixer as it would take forever, cost a lot of R&D and possibly still not get it 100% right - there are available analogue solutions NOW!!! why wait, spend less time thinking, save some money, go analogue and create!

Of course to be totally accurate you would have to model the dust/corrosion on the faders, the time of day, how old the caps were, how long the desk had been powered up, and what DA converters were feeding the console....................digital has a long way to go in terms of emulation thats for sure!!!!!

-Tom

 

[Jake68]

Yeah but how many times have you ever heard anyone say that the modelled mics, marshalls etc sound anything like the real thing? Sometimes its close maybe even 90% but if you want the sound of that Neve you're going into that last 10%.

Surely that what we have been accomplishing here for ages 90%?
OR does the UAD 1176 sound 100% like the original with any program material?
Why does the character we require only live in the last 10%? I am not sure I fully understand that point.
The point is that Neve Summing has a forgiving "nice sounding" character and any model that moves digital summing in that direction is viable. More viable the further it goes.

I also dont see the point of adjusting internal precision - surely you want the most precise model - whatever it takes...or would you want to take a 98% model and tweak to be 75% accurate??

Adjusting internal precision "UP". I am not sure of the viability of using UAD DSP's for increased bit depth. But the Pultec upsamples for greater model accuracy, so could this "idea".

The limitations and benefits to this idea are the same as the ones in use in the current range of UAD plugins!..No one fully expects a perfect representation but getting damn close and attracting all the pleasent and in this case sympathetic audio characteristics is surely desireable. Especially as DAW 32 bit Float summing can be so harsh.
Surely if we are using pultecs and 1176's and LA2A etc, we want to take advantage of the qualities of the summing in a classic console as well? If anything this proposal takes the UAD user even closer to the goal and the point of buying the product in the first place.

 

[mightymike]

I'm with you jake!

The only thing that would make it difficult is the DA conversion and emulating that! On a real ssl you would have a signal converted from say pro tools to the ssl which will make a tiny difference.... how the desk receives the signal is important, no?

But the only other thing UA would need to do is make a plugin that represent a ssl channel with no EQ or compression on but would have to emulate of how the faders and attentuentors work and then say another plugin which you put on the main ouput which receives all the signals and \"sum\" for the ssl experience!


Personally this would be easier for the DAW software maker rather than the a plugin maker!!

mm

 

[Jake68]

I dont think the DA is relevant.

After all we are providing a model from the input back. As if we were summing a PT mix down on a Neve which I see taking place very often.

I think the difference between ADDA of resonable class is surely neglagable.

After all if this was part of the equation what would be the purpose of the other models we use? DO you model in input source on an LA2A or do you just model the box itself. We want to present a model for what it would do with ANY sound, rather than what it would do with a sound that was PT ADDA presented or RME ADDA presented. Any modelling process surely just takes a datum signal?

Others may know more, but that bit doesnt seem relevant to me.

Even if this is just 80% of the forgiving nature of a Neve Summing bus then it has to be better for extreme frequencies than the blunt 32bit Float summing we have at the moment.

 

[TheEastGateMS]

from my experience with the Dangerous 2 Bus and the Roll Music Folcrom...

100% ITB Summing vs Analog Summing (16 channels) = Analog Summing sounded much better.

100% ITB vs. Digital Summing (passed through the same preamp used in the Analog Summing test) = Digital Summing (passed through pre) sounded much better.

Analog Summing (16 channels) vs. Digital Summing (passed through the same preamp used in the Analog Summing test) = audio nearly made perfectly null when phase reversed and lined up to be sample accurate.

from these tests... i highly agree with Paul's statements.


then again, maybe i would hear more of a difference now that i am using lynx conversion as opposed to the HD192 i had at the time.

 

[Jake68]

I am not so clear on what you are saying here Eastgate?

As you saying that When presented with exactly the same material recorded with the same pre analogue summing nulls completely with digital summing?

Whar do you mean by pre, do you mean mic pre?

 

[diggo]

Even if this is just 80% of the forgiving nature of a Neve Summing bus then it has to be better for extreme frequencies than the blunt 32bit Float summing we have at the moment.

This is not correct. The registers extend to 80 bit wordlength when summing within bog standard Windows. Digital summing is simple and trivial to do well. I have no idea what you mean by "blunt 32 bit", except that (at some stage) the wordlength will be reduced to 24 bit or less. This would also apply to any so-called "summing emulator"....

At issue are other factors which affect the eventual output, such as dither, driver settings and more.

The only effect you obtain from any form of analog summing which has a "sound" is distortion (coloration), which may be euphonic and therefore (in some cases) desirable. But this is not accuracy, it is an effect. It would be relatively easy to model such an effect, but the designer would not need to recreate a virtual summing bus to do this. The effect occurs within the bus: all the designer has to do is model the bus (if the designer discounts the cumulative coloration within each analog channel - note that this coloration is per channel and would therefore be separately modelled as a plugin channel insert ). The user then inserts the resulting plugin over the master channel in the DAW. It would be a very subtle effect in most cases.

I think this kind of plugin would have a very limited use and do not expect to see developers bother with such endeavours. Personally, I prefer accuracy. If I want coloration, there are already plenty of coloration options in the plugin world....or I can send my output through an appropriate analog device.

Sean

 

[twinbeatzrmp]

No mate I fully understand your point. Any mix can be improved by first class pre's. That still doesnt stop hit records being made on Mackie 8 busses! And it doesnt negate the point that DAW summing is unsympathetic.

But in my view, if anything requires improvement in DAW its the summing.

Taking your point there may be no requirement to generate any superior DSP products for anyone.

If you aleady had an SSL why would you want superior DSP, you could just get a MADI solution and stream your channels into you VCA or flying fader first class console and compress with your hourds of Urei's etc and your classic plate and acoustic chambers maybe a 480 or something.

But the point is that if we can improve our ITB mixes with UAD's then we can improve the mix further with more sympathetic summing. And I think that word is key "sympathetic"!

Anyone know how to find out what UA think of this idea, I'd be really interested!?!

Likewise, dont think of me as being argumentative, but I am convinced that if an LA2A is a good product for a DAW then so is a Neve summing buss!

u are absolutely right a bout the summing because ive tried exactly what uve tried with a number of plugs on the master bus and the best seems to be the pultec because it acts as if its the analog piece warming ur tracks even if no knobs r touched. personally i suggested a dither plug to throw at the end of the master bus that runs off the uad card. because almost everyone hear uses a different DAW and they all process differently during mixdown even tho we are all using uad plugs. take Pro Tools for example, this program even has its own built in dither that im sure is ran off its own internal dsp processors. theres an option in the settings to set this parameter, adding this feature i think would improve mixes because instead of depending on ur native daws' dithers we'd all have one universal dither which would improve mixes across the board and obviously better number crunching when it comes down to the mixdown process

 

[iode131]

Eastgate. I don't quite understand your tests. What do you mean by \"digital summing with same pre..blah...blah...\"?

Do you mean you sent all your tracks to an external mic pre then back to DAW for summing or that you sent your mix (already digitally summed) through a mic pre?

Anyway I second Jake. Basically he says there's no reason not to try to upgrade the summing features in DAWs. To make them better. One way that seems to work is modelling so why not try it with summing?

On the other hand the whole 90%-10%, \"the-original-beats-the-fake\" debate is kindda silly. Don't you think? I mean we all know that -by essence- a copy will never be alike the original. That's almost philosophy here but it stays true. What is sure -and I think that's the cool thing about modelling- is that you often have to study the masters who made it all before to get inspiration. For instance, I take UAD's 1176LN for what it is: a great compression plugin made by people who love the old faithfull analog 1176 and got inspired by it.

Cheers all

Max