Creamware mixers alter the sound differently !!, tested

Post by **hubird** » Thu Nov 09, 2006 10:15 am

jeezzz...I was expecting that the anti phase button is applied...and it worked here

Music Manic · Post by **Music Manic** » Thu Nov 09, 2006 10:31 am

Certain channels are phase coherent.The chart was printed a few years ago by someone here after I did a phase test.

astroman · Post by **astroman** » Thu Nov 09, 2006 2:50 pm

you CANNOT test reliably for sample accurate operation - there is no instance in Scope that 'supervizes' the system in that regard.
You will get all and any result imaginable, projects (or their devices) may change over time.

It's independant from hardware, OS, and Scope software version.
My system currently cannot phase delete a mono signal that's fed to both left and right of a stereo channel... if I invert one, it's not even close to extinguish - delaying one of the lines doesn't help either, the displacement is simply kept.

It works on mono channels perfectly, the stereo gimmick is new...

I have one card more and the bloody fanless PSU blew up a couple of it's output caps, well effectively the non-blown caps look so bulky that they'd be soon to follow.
Maybe I have a hardware problem now, maybe a re-install will fix everything.
(this time it's definetely not my fault - the PSU had been in repair half a year ago and inspection after this damage showed that they hadn't fixed the screws between inner and outer cooler, reducing the effective metal by 50%, outch...

anyway, I had experienced (slighter) sample deviations before and verified them, and people with different hardware and OS versions have confirmed that it may happen (the usual 'spread processing over DSP boundary).

It's absolutely irrelevant for sound in 98% of all cases - and it can be easily verified if a couple of channels need extra precise treatment.
As mentioned, it's well known from the analog world - any capacitance shifts phase and those (almost omnipresent) parts have a hell of a tolerance (usually) ...

cheers, Tom

Post by **garyb** » Thu Nov 09, 2006 8:18 pm

Mr Arkadin wrote:[garyb, not sure why that suggestion is so funny but there you go.

using the button? not a funny suggestion at all. the sturm and drang? funny.

no hardware mixer is 100% phase coherent. 100% phase coherence is not an assurance of a good sound. inputs of the stm2448 are phase coherent on input, but not nessesarily on output(compressors, eqs and other inserts all will prohibit this). it's a little insane to stress the issue too hard.

a massive, obviously hearable phase issue is a big deal, this isn't.(all due respect to the participants of this thread, just mho)

Mr Arkadin · Post by **Mr Arkadin** » Thu Nov 09, 2006 8:31 pm

Oh, OK. i agree, analogue mixers do this and no-one complained and i would rather have a tiny bit of phase 'issues' (non-issues really) and use Scope's summing than have phase coherence in Cubase and have to put up with the summing engine in that. Storm in a teacup really.

Post by **garyb** » Thu Nov 09, 2006 10:20 pm

Stige · Post by **Stige** » Fri Nov 10, 2006 2:10 am

musurgio,

When comparing 2448, did you consider that the panning is not exactly at 0 by default? You have to double click each channels pan pots to reset them.
This could easily make bigger difference in phase when comparing to 1632.

tgstgs · Post by **tgstgs** » Fri Nov 10, 2006 4:32 am

phase tests a worthless!
algos testing algos
you trust them? me not!
you even dont know whats inside these algos do you ?

take a wav mono or stereo what bit or sapmlingrate ever;
play it with stsxK or vdat (INSIDE SCOPE!!) or are you knowing what asio wave driver round or sum or do?
route it over a mixer and record it again on stsxK or vdat;
now you have 2 files;
cut the begin that both samples start with the same sample and now compare sample by sample the values

this way you dont need phasecompensat buttons or whatever;

and plese dont do the comparison with wavlab (dont know paris!) or other wave editor course they round!!!
a normal editor to open a bytefile do the job
go down to the 4 bytes that translates in data on a texteditor forget the next 4 byte (they tell you how much samples inside the file) then the sampledata starts

phase coherent vibes from vienna

Immanuel · Post by **Immanuel** » Sat Nov 11, 2006 2:24 pm

I am tirred, and I have not read this thoroughly. Still I reply - because if you want to SEE 100% faze-cancelation between Scope mixer channels, you might need to reset faders. Drag the fader to another position and double-click it. For some reason they are not nescesarily 100% at the same level, when you open the mixer. This can explain, why you can not hit a level of "minus infinite" by moving samples.