Creamware mixers alter the sound differently !!, tested

[musurgio]

Hi,
I tried the 2448 and the sound could not get phase deleted, there was a small but hearable difference.
1632 mixer way smaller so alters less the sound than the 2448
Using simple routings the signal could be phase deleted.
My method was:
I was sending from Paris via adat to pulsar (20bit adat) then back via adat toParis sound could get perfectly phase deleted.Of course I nudged the 15 samples latency which is for adat in and out on Pulsar
Using 1632 mixer at unity gain there was another 4 samples latency but the sound could not get phase deleted , a very small margin fo herable sound was evident.
This was worse on 2448 (eq and compressor bypassed using direct out).
I don't say that creamware mi=xers f... up the sound ,maybe they could make it better but anyway I wanted to try it.
So I am better using direct routings to effects and no mixer in beetween.
There I get perfect cancellation.
Regards,
Dimitrios

[hubird]

thanks for research
I recently for the first time encountered heavy phase distortion when I -in Cubase- added a distortion effect on the hihats audio track.
I was able to remove it by clicking the 'phase compensate' button or by sending the dist. effect via the same Cubase ASIO buss as the hat track used.
STM24/48.
You pay a high price for solving the phase problem by not using the full mixer

[djmicron]

it's normal,
each plug in introduces latency.
It's the same if you do the phase test with an hardware digital mixer or a digital effect processor.

[hubird]

the point was nor latency but phase cancellation...
edit: sorry, nor must be not

[djmicron]

the point was nor latency but phase cancellation...

of course,
phase cancellation is strictly related to latency.
I have done the test by connecting a wave source to the stm 2448 ch1, then the same wave source connected to adat destination 1/2, then adat source(from another scope system) 1/2 to stm 2448 ch2.

To obtain phase cancellation i must add a 0.479 msec of latency on the channel 1.

[hubird]

I understand that
Latency is normal and isn'ta problem normally.
The eventually resulting phase distortion however is a problem which should get solved, as you can hear it all too good.
The STM mixer is supposed to handle it by the phase compensate function, if it doesn't there is a problem!

[djmicron]

i was referring to the musurgio test and it's all about latency and the sound is not altered....

[musurgio]

OK,
To summarize again,
I have compensated for latency with sample by samplemovement.
Your ms calculations might be right but the samllest divsion is sample.
So I say again that without a mixer connecting just a device like compressor in series introduces 4 samples latency which I campensate and I get phase cancellation.
Now ith Pulsar mixers no matter what latency compensation you make sample by sample you are reaching in a situation where there is a small sound and no total phase cancellation.
That means that sfp mixers introduce an alteration in signal.
Hope that is more understandable.
Regards,
Dimitrios

[djmicron]

i understand what you say,
but i have done the same test and i have obtained phase cancellation with -109 db of peak, so it's not enough to think about sound artefaction....

[musurgio]

So one of us does it wrong, right ?
I will describe again my testing method.
I use Paris DAW that play an audio track.
This audio track is routed from Paris adat out to Pulsar scope.
There it connects to paris effect device in series and then to Pulsar adat out and back in to Paris adat in closing Paris insert.
Now on a adjust Paris channel I put the exact same audio track at exact the same position.
Now lets say I put a Pulsar effect like eq in Pulsar with at 0 db positions.
The track fromParis that was fed thru Pulsar effect has a latency of 15 samples (adat in/out) and a further 4 samples (all basic pulsar effects).
Now on the adjust Paris audio track I put voxengo sampledelay vst plugin and I delay it for 19 samples, switch the phase button and VOILA TOTAL phase cancellation !
Ok now in Pulsar instead of going direct thru effect device I connect 1632 mixer and and take the direct out and put it again to Pulsar adat out .
This way there is NO TOTAl cancellation only a samll sound and very low in volume ,but a sound, is hearable.
Using 2448 the sound is higher in volume.
Even delaying more or less in Voxengo does not cancel the sound to absolute silence.
THIS PROVES that sfp mixers DO NOT pass cthe sound unaltered.
There is a change.
Regards,
Dimitrios

[djmicron]

if you need to test the stm mixer, you can do a test inside scope without using all the external gear.
Try to use 2 stm 2448
The first mixer must have 2 channels connected to the same wave source
the second mixer must have one channel input connected to the direct out of one of the channels of the mixer 1
the other channel must be connected to the same wave source connected to the first mixer

Invert the phase of one of the 2 channels of the mixer 2

set the delay compensation on the channel of mixer 2 connected to the wave source

If the stm alter the sound you should obtain the same identical result as your routing wit hthe paris outboard, but mine is phase cancellation..

[musurgio]

Thanks I will try.
One thing that crosses my mind is the Paris adat is 20 bits only.
But under sfp mixer maybe there is an oversampling and then this gets truncated to 20 bits...
So maybve connecting a device is series does not alter the bit rate ...
Interesting.
Regards,
Dimitrios

[MCCY]

Djmicron is right that you can do the test inside Scope.
As I've done lots of similar phase testing, I would have come across problems with the mixers (using NO effects inside the mixers). But I found none.

I won't make any statement on right or wrong - even easy things can easily get complicated in a complex environment. Just wanted to let you know that my experiences were without any problems with soundaltering in Scope-Mixers.

Martin

[voidar]

I've done this test a lot within SFP using the De-vice tester tone-generator, and surely STM mixers are not necessarily phase-coherent.
This happens when the systems splits it's tasks onto multiple Sharcs and it's completely normal. Also, phase-coherentness varies - the "rules" change when you add or remove a device.

This means that you might not be able to totally phase-delete one signal put into the same mixer.
As some have stated, this happens in the analog hardware world too and my theory is that these "distortions" are to some degree what create the sonic characteristics of analog gear. I.e. "good sound".

If you need absolute phase-coherent mixing you will need to mix inside your DAW-software.

[MCCY]

"surely STM mixers are not necessarily phase-coherent"

Yes, they are not, but they seem (for me so far) not to alter the sound.

Martin

[Mr Arkadin]

At no point that i can see (i only glanced i'm afraid) have you stated you used the Phase Compensation button in the STM Mixers - did you?

[musurgio]

Arkadin,
If you ask me I am using phase switch inside Paris.
Regards,
Dimitrios

[MCCY]

So if you're not using the one on the mixers it could be, that L & R are slightly out of phase which makes phase correction impossible outside scope for both (correct one and the other will be out of phase) (if you sampledelay both stereochanels with same settings...).

[garyb]

the phase compensation on the stdm2448 only assures that the inputs are phase coherent.

[Mr Arkadin]

Arkadin,
If you ask me I am using phase switch inside Paris.
Regards,
Dimitrios

No, i was talking about the Phase Compensation in the STM - it's under Options - give it a whirl. Don't know if it'll make any difference using your method of testing (sorry, i didn't read all of it), but you might as well switch it on anyway.

garyb, not sure why that suggestion is so funny but there you go.