UAD vs Liquid Mix: theory

[supersonic]

There is an old article in UA webzine

http://www.uaudio.com/webzine/2004/july/index2.html

explaining why dynamic convolution fails to reproduce every possible state of a compressor.

This is way too scientific for me. Anyone with engineering background to put it in the real world terms?

What characteristic of sound does suffer from dynamic convolution compression? Is it worse attacks and transients? Or is it distiortion? Whatever..

Do these nonlinearities occur in the everyday use of a compressor or only in specific situations? And finally, can this be HEARD or at least MEASURED?

 

[Eric Dahlberg]

Dynamic convolution can do a good job of capturing the sound of something but not its behaviour. UA's case of late has been that the sound of a compressor is tied to the way it behaves to incoming signal. That is true in large part, though I'm sure they recognize that many, if not MOST, of the highly regarded compressors out their have a distinct & desirable sound of their own even when signal isn't passing threshold.

Anyway, the Liquid Mix ought to do a good job with your standard VCA-type compressors (SSL, DBX) while the UAD-1 ought to be better for the less linear Fairchild's, 1176's, LA2A's, & 33609's.

 

[neil wilkes]

The easiest way to explain this is to think of convolution as a \"snapshot\" of a process, which is why it is so good for reverb.

As the article says - it is great for a linear process such as reverb where the decay is a known factor and can be accurately predicted.
Non linear systems cannot be accurately modelled in this manner, but a close approximation of the character of the unit can be achieved.

 

[bmanic]

I've been trying to tell you guys this thing over and over and over again, now listen:

Sintefex does NOT use anything special for the COMPRESSION in itself. It is just a normal compressor that uses an 'index' as it's data (plotting the transfer curve).

The DYNAMIC CONVOLUTION itself is there just for the 'TONE' of the vintage unit. It has NOTHING to do with how the compression behaves. None, zero, zipp, nada.

So, Sintefex units -> Dynamic Convolution for the \"tone\"/\"color\" of the unit + a generic compressor that follows the sampled units transfer curve. = Simple, not that special, but rather good sounding because of the good basic compressor algorithm.

How close to the original sampled compressor? I'd say about 20%.. a bit depending on the type (the fairchild seems to behave a bit more like the original because of it's \"continuous\" knee).

Why am I preaching this stuff? Because I don't like Sintefex marketing. It's utter bullshit and I also don't like their silly patent. However, I am a quite happy FX 8000 owner and can highly recommend the unit for it's variety and flexibility of tone that it offers. I also hope that Sintefex would upgrade the rather basic compressor and make it more complex, this would enable us to more closely re-create the compression of known units.

Cheers!
bManic

 

[neil wilkes]

I've been trying to tell you guys this thing over and over and over again, now listen:

Sintefex does NOT use anything special for the COMPRESSION in itself. It is just a normal compressor that uses an 'index' as it's data (plotting the transfer curve).

The DYNAMIC CONVOLUTION itself is there just for the 'TONE' of the vintage unit. It has NOTHING to do with how the compression behaves. None, zero, zipp, nada.

So, Sintefex units -> Dynamic Convolution for the "tone"/"color" of the unit + a generic compressor that follows the sampled units transfer curve. = Simple, not that special, but rather good sounding because of the good basic compressor algorithm.

How close to the original sampled compressor? I'd say about 20%.. a bit depending on the type (the fairchild seems to behave a bit more like the original because of it's "continuous" knee).

Why am I preaching this stuff? Because I don't like Sintefex marketing. It's utter bullshit and I also don't like their silly patent. However, I am a quite happy FX 8000 owner and can highly recommend the unit for it's variety and flexibility of tone that it offers. I also hope that Sintefex would upgrade the rather basic compressor and make it more complex, this would enable us to more closely re-create the compression of known units.

Cheers!
bManic

Have you got a Liquid Mix/Liquid Channel unit?
Do you know how it sounds?

 

[bmanic]

I have not heard the Liquid Mix yet. I have plenty of experience of the Liquid Channel and I own a Sintefex FX8000 box (basically the same as the liquid mix and channel though I suspect it uses longer IRs than the liquid mix).

As for the compressors and how they work: It's 100% identical to the Sintefex (AFAIK). Same look-up table kind of thing for the transfer curve and dynamic convolution for the \"sound\".

For more info on the Sintefex FX8000 (the box with the technology they use to actually capture the Liquid Mix and Liquid Channel 'samples') you can get here.

Cheers!
bManic

 

[Eric Dahlberg]

It's too bad you're not local, I would've sent you a Liquid Mix to compare otherwise.