Mixing Engine exceels on T2

[GYang]

It's not joke, but after carefully preparing 22 stereo tracks project with acoustic guitars, piano, live drums (all together with lot of harmonics) I tried both 32 bit and 64 bit mixing engine.
I understand that many user DO NOT HEAR difference due to average (or bad) monitors, but difference is impressive with good amps, converters etc.
Also, no meaning to mix synths and sources that are initially 'simple sounding'.
When the signal is downgraded to CD quality lot of magic is indeed lost (on MP3 completely).
I found that when listening the mixdown of some 12-14 various instruments you can easily locate and isolate each in 64-bit mix. It is same sound, same picture, but all details are there. Like difference between hi-fi system and high-end. You just feel like you have cleaned your ears.
CPU power is significantly higher in complex project, so only the most powerfull PCs in good configuration actually allow so many tracks mixed (incl. effects done on Powercores or UADs to release CPU from that task).
At the end of the day, pure quality of mixing (I don't talk about system as whole) in 64-bit is better than in any other DAW I know presently, incl. Pro Tools HD or Nuendo as references I used alot.
So, T2 for me reached the target in musicality of complex mixes done in 64-bit. I doubt that many users will benefit very much on that, but I am completly sold on this.

Happy mixing to all.

yang

[Barbarossa]

Would you mind to upload a phasetest-file verifying your observations?

[M'Snah]

I need a good 128-bit host.

[Mr. Tingle]

Would you mind to upload a phasetest-file verifying your observations?

How would a phase test prove there is more detail in the hi end?

[GYang]

Would you mind to upload a phasetest-file verifying your observations?

Sorry Barbarossa, but I am not scientist or chartered tester qualified for properly done comparison tests in ideal environment, measuring equipment etc. I think Mackie would be able to do it and show what we are talking about.
Similar endless discussions on other forums are related to differences between sound of various cables, converters or valves, but here we have something easily perceptible by anybody who has good converters (Apogee or better) and monitors (as rule in 2500$ + price range) wishing to mix many channels of sonically demanding acoustic sources.
What I actually suggest is that subjective perception of soundstage, details, articulation and separation of instruments within complex or very complex mixes is overall better when mixed in 64 bits and difference is not subtle.
I am happy with the result and for the time being do not have more requirements on quality of mixing engine for DAW, as for further improvements I would need to invest quite significantly in monitoring chain.

yang

[rpc9943]

gyang u rule keep mixing keep excelling in these things, you know your ears and we know them too

RonC

[Barbarossa]

Would you mind to upload a phasetest-file verifying your observations?

How would a phase test prove there is more detail in the hi end?

No, it would show, if differences are in the audible range.
I'd say mixing 20 24-bit-tracks with an resolution of 32bit-float should produce absolutely no difference.
Heavy "dsp-processing" is necessary to hear the limits of the 32bit-float format.

I would really like to hear the two files...

[Frippertronix]

I like it. Doesn't seem to eat too much CPU, so why not use it?

Gotta justify the upgrade somehow.

[djsubject]

IMO (witch isint worth much) it gives you space

you can use it to get more out of your compressor when pushing sounds & its good for fully bloated 14 instrument tracks

i hear a difference!

i think the "its like you just cleaned out your eres" quote said it all

Subz

[GYang]

No, it would show, if differences are in the audible range.

I would really like to hear the two files...

Of course that some waveforms or similar could show the difference, but again you cannot read the sound from the display.
In one well argumented research it was proven by respected expert that converter clocked by external master clock (no matter of how good it is) should always sound worse than same converter clocked by well implemented internal crystal.
It was apparent to all reading the report that some kind of 'smearing' is present in graphical displays related to outer clocking. In same time, absolute majority of studio engineers incl. some of 'golden ears' supported quite opposite opinion and continue to use particular world clocks to actually improve sound of converters.
That is just to show relativity of measurable facts related to subtle differences in digital audio.
Again, I believe it is quite simple test everyone can do. Minimal requirement are converters and monitors of upper range. Record let say 4 stereo tracks of acoustic guitars, 4 stereo tracks of vocals, 4 tracks of percussion/cymbals etc + couple of other as bass, piano etc. Use good reverbs and compressors, as well as, try to position each instrument/vocal to its place. Than mix both ways.
You'll hear the difference, if not your chain is not worth this difference, but in the mix it will be present you can hear it on better systems later.
Make A/B blind test with other people, 90% will tell you correctly what is better.
CD will slightly degrade this effect and MP3 is not worth discussion on this issue at all (I use 24/96 Masterlink format to keep the quality for mastering purposes).
If MP3 is targeted format for your music even 24 bit recording/mixing is more than enough.

yang

[Barbarossa]

The human ear is not useful for proving technical facts and absolutely everyone can make the difference between the 32bit- and 64bit-float calculations hearable by a file comparison.
Why do you rely on your subjective ears and make technical conclusions from that, although you could easily make the prove with an objective test?
Convince me with facts, that your 64bit mix of 20 tracks sounds better than the 32bit one...

[.jon]

Why, Barbarossa... why?

If it sounds better, it sounds better.

[Barbarossa]

If you believe it sounds better on one monitoring-system in one room, it sounds better?
Not really.
But if it sounds different, then it can be measured.
And in this case the difference of the two mixes could be calculated very easily.

I don't understand why people are fearing such evidences.
Be happy if the difference is that tiny, that you can save the CPU-power on something else.
Or in the case the difference is really in the audible range, then be happy that you can get a better sound quality with the click of one button.
So where's the problem?
It's not a matter of believe, it's a matter of facts.

[Rock]

GYang you might want to try using "Tape it" on your masters and then dithering down using an external audio editor to see if you can keep some of that shine in you mix once it is done. Let us know.

[Barbarossa]

No dithering. The differences he "hears" when monitoring, come from the hard-rounded 24bit master-out signal. So simply rendering the output to 24bit should deliver the most authentic results.