Na gearslutzu imaš sample, kar v nekaj threadih. Liquid doda nek svoj specifičen prizvok, dokaj v redu (baje) "emulira" barvo samplane komponente, ne pa tudi njegovega delovanja.
Evo nekaj postov enega tipa, ki ga je imel, imel je tudi original sintefexovo
makino. Je še najbolj nepristranski (kolikor se je meni zdelo) in ne hvale in ne graja preveč, pač pove kako in kaj.
Da ne bo kdo rekel, da nisem bil priden!!!
http://www.kvraudio....h...id&start=75
Well, according to some of the tests I ran with a for computer made dynamic convolution (this can unfortunately not be distributed due to patents) I could run one instance of 1024 samples long impulse response with 128 layers and it took 90% of my Athlon 3200+ CPU. The Sintefex FX8000 has 6 channels and runs 2048 long IR's. Now, the new toy is supposed to run 32 channels? Hmmm.. makes one wonder a bit.
True, DSP power has gone up since 1998 when the original unit was introduced but has it cost effectively really gone up that much? Could be, who knows.
Cheers!
bManic
___________________________________________________
Well, I can tell you right away that the Liquid mix EQs will be very accurate (both the EQ itself AND the "vibe" of the original unit) but the compressors are NOT accurate at all. The "vibe" or "sound" of the unit will be very accurate but the compression action is way off. It's not even nearly the same. It's just a fairly basic (but very high precision) compressor algorithm that is "pasted on" to the dynamic side of the effect. It's not a bad compressor, not at all, but it does mean that every emulation will sound about the same, except for the basic vibe/character of the compressor. It does NOT emulate each models personal compression behaviour.
Cheers!
bManic
----
Snooky: "Without me balancing the insanity around here, it will all go down the drain...trust me."
Whyterabbyt: "If you were any more full of yourself, you'd be a Klein bottle."
_______________________________________________________
I've demoed the Liquid channel too and as far as I could tell the dynamic side of the thing is identical to the sintefex. They actually use a FX 8000 to capture all their "models" (mentioned in their manual) then they just port it over to the liquid channel.
For instance, the 1176 model in the sintefex doesn't compres at all as the UAD-1 version does but the overall "sound" of the unit is IMHO closer to the real thing (though I've never tried a real revision E or D blackface so this makes my statement rather vague. I have experience with early 1178 and revision H 1176). I have captured some compressors myself and it is impossible to get the compression action itself identical to the original unit. You can get pretty close but not nearly as close as the UAD plugins are. However, on the other hand you can get the "tone" of the box so close that I have not been able to find it in a blind ABX test (I tried this on my crappy DBX 376 tube channel thingy). This also applies to EQ units, they can be captured pretty much identical to the original which makes the Liquid Mix very good value.
So, to sum it up, liquid mix is: Good for EQ/tone, mediocre for Compression emulation (but the compressor itself is better than average).
Cheers!
bManic
----
Snooky: "Without me balancing the insanity around here, it will all go down the drain...trust me."
Whyterabbyt: "If you were any more full of yourself, you'd be a Klein bottle."
_____________________________________________________
______________________________________________________
http://www.gearslutz...ighlight=bManic
I'll copy/paste what I've answered in other forums:
-----------------------------------------------------------------
I've demoed the Liquid channel too and as far as I could tell the dynamic side of the thing is identical to the sintefex. They actually use a FX 8000 to capture all their "models" (mentioned in their manual) then they just port it over to the liquid channel.
For instance, the 1176 model in the sintefex doesn't compres at all as the UAD-1 version does but the overall "sound" of the unit is IMHO closer to the real thing (though I've never tried a real revision E or D blackface so this makes my statement rather vague. I have experience with early 1178 and revision H 1176). I have captured some compressors myself and it is impossible to get the compression action itself identical to the original unit. You can get pretty close but not nearly as close as the UAD plugins are. However, on the other hand you can get the "tone" of the box so close that I have not been able to find it in a blind ABX test (I tried this on my crappy DBX 376 tube channel thingy). This also applies to EQ units, they can be captured pretty much identical to the original which makes the Liquid Mix very good value.
So, to sum it up, liquid mix is: Good for EQ/tone, mediocre for Compression emulation (but the compressor itself is better than average).
-----------------------------------------------------------------------------
.. again, this tool will probably be quite good bang-for-the-buck but do NOT expect it to do any miracles. The compression section of it (the compression action itself) is VERY simple and does not allow you to do stuff that the original units do. It's simply a transfer function that responds to the generic compressor code. It has no program dependent release nor attack or any other oddities. It has only a simple Peak and RMS choise (with +/- input cycle check, that is, does it react to the positive incoming wave or negative). It's not a bad compressor at all, actually it sounds rather good on complex material but it does not in any way, shape or form replace the original units. It does a far worse emulation of the compression action than many plugins even.
Hope this clears some things up. I'm still hoping that sintefex (who are responsible for all the code) creates a more complex, user tweakable compressor that allows us to truly replicate the original units.
Cheers!
bManic
___________________________________________________________
http://www.chrismiln...o...ic&start=15
Bottom line: Liquid channel emulates the compressors quite badly (read: not even close) but the TONE of the emulations are spot on. The EQs only require the tone so they are very much like the original units (which makes the focusrite unit tremendous value).
Cheers!
bManic
___________________________________________________________
UAD forum
I have a sintefex FX8000, 6 channel version. It does mimic the captured hardware's "sound and character" very well. I'd say up to about 90% or so but the compression action itself is not very accurate. Here I would say the UAD-1 emulations kick the shit out of the sintefex.
It just so happens that I got an email from Mike Eden today (one of the founders of sintefex) who said that they are refining the compression algorithm.
As soon as you stop buying the sintefex for it's "emulation" capability and just buy it as a box full of tone choises (think of it like the super-elite-mega-Colortone Pro) you will be much happier with the unit and actually come to love the features and sound!
My advice is: If you can get the box cheap, then it's well worth it but I wouldn't pay the full asking price for the top of the line range (fx8000, 8 channel version used to be over 10k$).
Cheers!
bManic
______________________________________________________________
All of the top linear phase EQ's, in their highest quality mode, have imho truly no "grain" at all. These are iZotope Ozone 3 (set the 'digital' eq section to 16k and 8k buffers), PLParEQ (set to quality level 7) and Algorithmix Red and Orange.
Earlier I stated that I could run Algorithmix at full quality without my CPU dying but unfortunately I had forgot to "reload" the plugin after insertion which is the only way to change quality setting (I think this is true for PLParEQ too). At the Xtra quality, the algorithmix plugins absolutely kill my CPU. Wavelab reports over 400% cpu usage. Pretty heavy for an eq, eh? More importantly though, they sound absolutely superb at this level. I'd say these are the best digital equalisers available (all 3 of the above).
iZotope Ozone being the cheapest of these (well, PLParEQ3 was free for a limited time!) so I suggest checking that out. You can also email algorithmix to get a functional demo of a limited time.
For analogue modelling I still think that the technique that TriToneDigital is using (and Sintefex, even adding the dynamic part) is superior to anything else. Can't wait to hear AngelTone and the latest Hydratone 1.3 beta sounds superb too!
Cheers!
bManic
http://gearslutz.com...hlight=Sintefex
23rd January 2006, 02:32 AM
Hmm.. I have a sintefex FX8000, 6ch version and I have had some very good and some not so good experiences with it. First of all, when you guys say that the compression models are spot on it makes me go GGGGNNNNNnnnnn! Sure, the "tone" of the processor is pretty much as the original but the compression action itself is not so accurate. I have only compared the SSL Bus compressor, the alesis crappy compressor (yes, they have that unit replicated too.. that 3462 whatever model, that 1u rack horror compressor) and then of course my own replicated compressors (Tube Tech LCA 2B, Joe Meek SC2, Dbx 376) and all of these are pretty far from the originals. The "tone" or "character" is there but the compression action is not identical, not even close. Every replicated or sampled compressor gets a certain Sintefex signature sound, that is, it's own compressors sound and it is a digital sound indeed (a very GOOD digital sound I might add but it's still there).
I've emailed the people at Sintefex several times and asked if they might consider further developing the compression attack/release behaviour detection algorithm so that we would get more realistic compression action to go with the already superb dynamic convolution. Like it looks now the sintefex unit sweeps the target with a sine wave signal and gets the transfer curve but this of course does not take into consideration any program dependent behaviour. I'm sure that this could be done with a more intelligent capture signal and better compression algorithm in general. As it is now we have basic Peak (+ and - meaning it reacts to either positive or negative cycle of the incoming wave information) and RMS (also + and -).
I would love to see the huge amount of DSP that sits in this box to be used to it's fullest. I'd also like to be able to tap into the vast amount of sample data that is available for the liquid channel. It does feel a little like Sintefex has been bought off by focusrite and that they are not really taking care of their own brand, probably due to lack of time. I also wonder how many units Sintefex has sold. AFAIK I have the only Sintefex FX8000 unit in Finland and I don't know about too many of them in Scandinavia either. If any of you guys reading this have one, let me know!! Would love to share some sampled devices!
All in all I've got some very good results from the unit during mixing. It interfaces perfectly with the RME HDSP trough ADAT which gives me 6 separate mono channels or 3 stereo with extremely flexible M/S routing.
Cheers!
bManic
Oh, you all probably know this but here is a sintefex fan site with some extra downloadable samples of various units. Many thanks go to Salvator from TriToneDigital who has created a bunch of the better ones. I wish he still had his FX8000 to supply me with a lot of goodies!
http://sintefex.pitonelab.ru/
Cheers!
bManic
I've had a love/hate relationship with the sintefex. The love came first, then the regret and because of the extremely good price we paid it nearly made us sell it again (for profit) but now lately I've began to love it more again. Once you stop treating it as a wanna-be emulation of vintage units and treat it like an independent awesome effects/tone box you can't help but smile. This thing sounds fabulous!
However, I still hope Mike and Michael will continue develop this platform and extend it's pallette. I think that the compression transfer curve sampling can be made more intelligent to get a more authentic sound and I also hope that they give us even more options for the basic compressor itself (like program dependancy of various kinds, don't care if they are based on some real units or not as long as they sound good! ).
Rob, I'll email Mike about the test tone as he can provide it in digital form. This way you can run the process with your superior (most likely) DA and AD instead of me using the crappy RME HDSP 9632 converters. You can clearly hear the difference by taking extreme care of the signal all the way. Like in any recording/mixing situation!
Cheers!
bManic
