Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Mark McFarlane]

I recently added a Motu 896mk3 (8*10 interface) to my 01V96v2 rig to use as extra preamps and as a measurement interface.

I was thinking, as I started to rerack, maybe I should run a single ADAT cable to the MOTU and use it as 8 XLR outputs, bypassing the 01V96 mains and omnis. 

So this begs the question, besides aurally comparing the 01V96v2 main and omni outs to the external converters, what can I do to electronically compare them? 

I don't have an external function generator,...the 01V96 has a limited one internally,... I was trying to think of some approach with something like Smaart using transfer functions but using an external CD player as the reference I would be capturing (and comparing) the AD as well as the 01V96 or MOTU DA.  I also have a dual-trace scope, I could overlay the traces but again no way to really know which is more accurate.

[James A. Griffin]

Are you just trying to compare the output levels or are you trying to measure something more vague like "audio quality?"

[John Roberts {JR}]

There are multiple confounding issues that make this difficult. Unlike the old days when we were looking for fractions of a percent distortion, modern digital paths are so good that the resolution required for signal sources and test benches to explore limits are difficult to easily assemble.

Further even performing simple null testing, just to identify if there is a difference (not which one is more accurate) will give false null products if there is any delay between the stems.

You can try a null test and ignore the HF null results (from short delays) but this is far from a definitive test.

Do you have a specific concern, or wondering if expensive boxes are actually worthwhile. You can probably guess where I come down on that. 

For some reason people find it hard to believe that it is possible to put multiple high performance paths inside a common chassis. If there are more than two of them, they MUST be somehow inferior.   

JR

 

[Mark McFarlane]

Thanks James and John for taking the time to respond.

I'm primarily interested in sound quality, but there are also some cabling and feature issues I'll discuss below..

I have three options to output to mains and monitors:

1) 01V96v2's built-in 2 analog XLR and 4 Omni outs (6 out - most reliable solution, no external gear, no extra optical cables, but only 3 monitor sends and 1 sub send)
2) Send ADAT to Motu 896mk3 (8 outs, a little cleaner wiring at the console, just one cable to my DSP and EQs. Unfortunately using one of the 2 independent Motu ADAT ins means I can only connect one of my two Audient preamps to the Motu for direct recording to computer.  I wanted to use the Motu for 24-track recording as a backup for my HD24 - actually the HD24 would probably be the backup since I could avoid transferiring fies from the HD24 which is always a PITA.)
3) Send ADAT to an ADA8000 (same benefits as Motu, but not in my output rack currently and I don't have an extra 1U unless I dump a 2U drawer which is darn handy ))

Sending ADAT to the ADA8000 for 8 outs would be the best solution, assuming sound quality and reliability are adequate.

I've read of people complaining about the DA section in Yamaha consoles so I did some ear tests myself one day and thought the XLR outs on the 01V96 sounded 'slightly' better than the ADA8000 but it was close enough it could be completely psychological.  I have some better speakers now (RCF TT22A) so I need to repeat the listening tests.

I have no problem electronically comparing the DAs but I can't seem to determine which one is more accurate because the AD is in all the measurements if I compare the source before the AD...  I understand the conversion chips themselves are probably all fine, but I'm concerned that the analog drivers behind those chips may be of different quality.

Probably I am tilting at windmills again,.... it happens.   But learning is fun.

Thanks again for your valued time.

[James A. Griffin]

Sending ADAT to the ADA8000 for 8 outs would be the best solution, assuming sound quality and reliability are adequate.

Probably I am tilting at windmills again,.... it happens.   But learning is fun.

It sounds like Option 3 fits your workflow best and you're looking for permission to do it.  Go for it.

I used an ADA8000 with my 01v96v2 for years to send Aux's to monitors and never had quality or reliability problems with it.    I don't know how to measure or quantify that, but my satisfaction in that area was very high.       This is one box Behringer got right.

[Mark McFarlane]

I went ahead and made some measurements (my wife is on vacation and I'm bored), which I'll share as 3 separate posts for everyone's amusement.

Experimental setup is as follows:

Audio source is a latop, routed through a peavey USB-P interface set in mono. Steeley Dan Aja album as the test music.  Left channel out of the USB-P serves as the Smaart reference trace.  Right channel of mono USB-P is sent to AD1 (ch 1 preamp) on a Yamaha 01V96.  Smaart is being fed by a Motu 896mk3 as the preamp.

My goal is test 3 different DA devices, so I patched the 01V96's Stereo L to 2 different ADAT outs, one driving a Behringer ADA8000, the other a Motu 896mk3.  The graph below shows transfer functions comapring the Peavey USB-P reference signal to the 01V96v2 XLR out, the ADA8000 out, and the Motu out, at both 96k and 48K.

Ignoring the waviness of the traces at HF at 48k (I'll discuss the likely source of the waves in a later post), my basic conclusion was that the response of all three DA's were essentially identical in this test. 

I'll stick my neck out and interpret the smooth phase swing at high frequencies as simplying not being able to get an accurate enough delay to time align the reference and measured signal.  Whack me in the head if I'm off base here and these HF phase swings are something else, or at least meaningful.  The two 'ripples' in each trace in the high frequencies is another story, and I'll comment on them in the next post.

My basic conclusion, backed by listening tests, is that all three DA's performed essentially the same.  The listening tests were done through two RCF TT22A's.  At first I thought I could hear a difference between different DA converters, but later after changing which speaker was used for which converter I decided that the difference between the two TT22A's was larger than the difference between converters.  I was actually a little dissapointed that the TT22A's weren't matched better.  Each sounds great, but one seems slightly brighter than the other.

I tried very hard to get repeatable measurements but in a few cases when I went back an hour or two later and measured something again the phase trace changed a little bit (e..g. it swung up or down 10-30 dgerees more than the previous test).  I'm not sure what that was all about. I was fairly careful, and sober.  In most cases after performing 3 repeat measurements over several hours for the same setup I got the same result each time, but at least twice I got slighlty different swings at the HF when repeating a measurement. Has anyone seen this?

The first graph below shows the Peavey USB-P as the reference compared against the 01V96v2, ADA8000, and 896mk3, all at 48K.

The second graph shows the same comparisons at 96k.  The 01V96 and 896mk3 traces are 'on top of each other'.

[Mark McFarlane]

The two 'ripples' in the HF region of my first 48k tests annoyed me.  One fatal design flaw in my test was that the reference trace didn't go through the same AD process (it was straight from the latop into Smaart through the Motu interface.  The reference never sees the 01V96v2 preamps, so I was not really comparing the effect of only the DA devices, but all DA devices had the exact same preamp/AD path.

For fun I reran the tests using the 01V96 XLR out as the reference, and made comparisons against the ADA8000, the 896mk3, and an Omni out on the 01V96. 

The HF ripple is gone, implying that the preamp section (or AD) of the 01V96 was causing this problem.

[Mark McFarlane]

To confirm that the 2 HF ripples at 48K sample rate were caused by the preamp/AD of the 01V96, I connected the USB-P reference trace to ch1 of an Audient ASP008 preamp with built-in AD and ADAT out, and went ADAT out into ch1 of the 01V96.

Voila, the ripples are gone.

My interpretation is that ch1 of my 01V96 is introducing these phase ripples, either in the preamp or in the AD section.  I didn't test other channels.

In the chart below, the Motu and ADA8000 curves are on top of each other.

[ProSoundWeb Community]