test A/D D/A - without personal bias & without ears!

[space2012]

Hi,

this thread/post is the followup of myrtlebacker post:
http://www.gearslutz.com/board/3684233-post11.html
taked the file:
http://www.gearslutz.com/board/attac...s-dac-test.wav

...and made some loop tests:

Roland MMP-2 ===> Roland MMP-2
"MMP-2 can only work AD or DA, two are needed for the loop test."
MMP-2 all bypass, no EQ, no compressor, no Mic/pre emulation,
just AD or DA.
with RME Hdsp9632 pci, buffer size was 1024, with AD+DA total sample latency was 1127 measured with free Voxengo Latency Delay Plugin.

In/Out +/-dB was near as perfect as i could. calibrated with free MDA Testone generator 1Khz Sine @ 0dB, no clipping in the converters. the Roland Outs are a bit hott than the Ins. so i had to reduce gain in the master digital fader in Sonar arround -1dB. = Signal was Hot.
I should had used the free SSL X-ISM to calibrate In/Out dB levels.

tested loops:
A) both Rolands with the RME SteadyClock(TM) from the s/pdif signal chained from the Hdsp9632.
vs.
B) both Rolands with Drawmer M-Clock DMS-1, s/pdif signal chained.
vs.
C) first MMP2 with RME SteadyClock(TM)==analog==> second MMP-2 with M-Clock.
vs.
D) +4dBu vs -16dBu with M-Clock = two tests with same clock.
vs.
E) original .wav file.
------------
DDS turned off at all times in the Hdsp9632 settings, for better ext.wordclock thru performance.
nice 99.997% OFC Rca--to--Rca cables for s/pdif.
6x Meter XLR--to-->XLR Proelgroup.com DieHard cables full OFC, 24k gold plated contacts.
all machines exept the PC were plugged AC 120v to Furman AR-15 series II,
old PC. p4 1.6ghz 400fsb, compaq 4700
------------
i like Roland MMP-2 becouse they have transparent mid-range.
maybe the highs are not like Mytek or hyped like behringer ddx3216, but the mid range its accurate and totally transparent, and i like the dynamic range, emu1820m has way better specs measured with audiorightmark RMAA, but to me the Roland MMP-2 sounds better.
some serial numbers have a +-2dB roll off in the freq. range., but anyway i like it.

also used Sonar 7 32bits with 64-Bit engine active. becouse sounds totally transparent in the mid range. the 32 @ 32 does not.

taked screen shots of the zoomed waveforms,
the original vs. the DA-->AD loops waveforms with diferent wordclocks, all test were @ +4dBu, and one at -16dBu.

the sine waveforms looks 99.99% flawless in all tests,
only screen shots of the Saw at start, and Saw at the end.

the test i didnt had the WordClock Module for the Hdsp9632, i had to use the s/pdif as wordclock, and made a serial chain.
Drawmer M-Clock--s/pdif-->MMP2.(A)-->hdsp9632-->MMP2.(B)==analog==>MMP2.(A)-->Hdsp9632

the wordclock signal was not paralell, but serial.

the serial signal: "thru" s/pdif wordclock/audio signals seems to add a bit of jitter in each step.
jitter & drift is less with parallel distributed wordclock.

anyway, here are the results:::
take note that all softwares render visually diferent the waveforms.

[Reptil]

hey man thanks for doing the tests.
something is going on with steady clock..