The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
DA/AD Degradation? Audio Interfaces
Old 8th March 2016
  #1
Lives for gear
 
JayTee4303's Avatar
DA/AD Degradation?

How high is the cost of using outboard? We just picked up some hardware, 1176, LA2A, M7, that does wonderful things to signals, and now I wonder about the cost of using it.

How many trips out of digital and back in do you normally make for OTB processing, before audible degradation?

What types of conversion degradation become apparant first, and are they even a concern, when compared to analog noise also incurred?

What converters are you using?

What steps, if any, do you take to minimize degradation on processing trips out of and back into the box?
Old 8th March 2016
  #2
Lives for gear
 
Strange Leaf's Avatar
 

I do everything ITB now, but I did a test years ago with a Alesis HD24 with the stock converters, and posted a blind test on another forum. I looped signal up to 10 times back and fourth through the converters, and no one could hear a difference between the original and 3 loops (3 AD + 3 DA). At 5 times, most people heard the difference, but the signal was not unusable in any way. At 10 times, it was starting to sound like a cassette, clearly inferior to the original, but not awful. Just muffled and dull.

I would say for 1-2 rounds through highend or high prosumer converters, I wouldn't worry about it at all.
Old 8th March 2016
  #3
Gear Addict
 

My tracks get 1-2 rounds all the time and I hear no difference.
I've tried up to 5 times but no big difference. nobody cares as long as it sounds good. Sending to hardware give you so much more sonically that AD DA degradation is negligent.
Old 8th March 2016
  #4
Quote:
Originally Posted by Strange Leaf View Post
I do everything ITB now, but I did a test years ago with a Alesis HD24 with the stock converters, and posted a blind test on another forum. I looped signal up to 10 times back and fourth through the converters, and no one could hear a difference between the original and 3 loops (3 AD + 3 DA). At 5 times, most people heard the difference, but the signal was not unusable in any way. At 10 times, it was starting to sound like a cassette, clearly inferior to the original, but not awful. Just muffled and dull.

I would say for 1-2 rounds through highend or high prosumer converters, I wouldn't worry about it at all.
hi . .I just read this + don't want to shoot myself in the foot as I am selling an HD24 right now, but ..

stock, they are nice converters, to my ears they sound much nicer than motu's musically but they are in the motu ballpark for noisefloor + I think this is the detectable element that adds up sonically.

there is also jitter noise but for roundtrip I think some of this cancels out so you will be mainly left with 24 bit quantization error from conversion but well below the noisefloor of the analog stage + therefore it should be very well dithered (at least equivalent to 4 bits @ 120db snr).

the noisefloor will accumulate over various DA-AD loops basically as a summation of random noise equation.

since this is a sum of squared standard deviations = variance , the rms measure (standard deviation) of noise will be growing according to sqrt(n) as we make n loops through the same da-ad.

in laymans terms rms noise will be growing at rate 1.41 (sqrt(2)) every time we double the number of loops. so to nearly double noise we make 4 loops , 1.41 *1.41, but to almost triple noise we need 8 loops , almost quadruple = 16 loops etc.

referencing a decibel scale it takes 100 loops to make a 10x = 20db increase in noise, but only 10 loops to make 3.162x = 10db increase in noise.

with a reasonable noisefloor roundtrip like a HD24 (something like 103db spec roundtrip) it will take 10 roundtrips for the noisefloor to become audible (equating this to > 16bit noisefloor = 96dbfs).

with a higher spec converter, perhaps 10 db better signal to noise ratio for a roundtrip, it would take 100x roundtrip for this same level of converter noise to become apparent.

Last edited by alexmaster; 8th March 2016 at 08:48 PM..
Old 8th March 2016
  #5
Quote:
Originally Posted by Audiofool View Post
My tracks get 1-2 rounds all the time and I hear no difference.
I've tried up to 5 times but no big difference. nobody cares as long as it sounds good. Sending to hardware give you so much more sonically that AD DA degradation is negligent.
call me weird but I actually often prefer some audio after a roundtrip DA-AD, the analog stage can introduce a light harmonic.

I used to like the sound of my lucid 88192 looped a few times, often better than the original.

I think the ssl alphalink sounds quite nice like this also.
Old 8th March 2016
  #6
Gear Addict
 

Quote:
Originally Posted by alexmaster View Post
call me weird but I actually often prefer some audio after a roundtrip DA-AD, the analog stage can introduce a light harmonic.

I used to like the sound of my lucid 88192 looped a few times, often better than the original.

I think the ssl alphalink sounds quite nice like this also.
I think so too. A roundtrip without hardware inserted sounds better.
So it means that two of us are weird.
Old 8th March 2016
  #7
Lives for gear
 
JayTee4303's Avatar
I appreciate the replies, guys. Alexmaster, especially useful formulas and theory.

A bit early for statistical conclusion, but I see a nexus around "1-2 passes inaudible" and " 5 passes noticable but useable."

From there' I believe we'll try to get up to 3db compression on one or two comps on the way in, if we think we'll need it. We are input and Daw processor adequate here, so we'll grab clean audio as well, in case we want original dynamics.

That gives me one DA/AD loop for additional processing, plus the FX send thru the Bricasti, for "free", so to speak, with a bit of "headroom" for additional trips out the box.

Thanks for the pointers!
Old 8th March 2016
  #8
Lives for gear
 
JayTee4303's Avatar
Dotting I's and crossing T's...I am assuming that trips outside the box are executed at capture SR and depth, no dither. Correct?
Old 8th March 2016
  #9
Lives for gear
 
DistortingJack's Avatar
 

Quote:
Originally Posted by JayTee4303 View Post
I appreciate the replies, guys. Alexmaster, especially useful formulas and theory.

A bit early for statistical conclusion, but I see a nexus around "1-2 passes inaudible" and " 5 passes noticable but useable."

From there' I believe we'll try to get up to 3db compression on one or two comps on the way in, if we think we'll need it. We are input and Daw processor adequate here, so we'll grab clean audio as well, in case we want original dynamics.

That gives me one DA/AD loop for additional processing, plus the FX send thru the Bricasti, for "free", so to speak, with a bit of "headroom" for additional trips out the box.

Thanks for the pointers!
I've never known why people don't compress on the way in, AND also track the clean sound. Best of both worlds. If you hear the compressed version and it sounds good, drop the other one. If you overdid it, compress straight away.

Done.
Old 8th March 2016
  #10
Quote:
Originally Posted by JayTee4303 View Post
I appreciate the replies, guys. Alexmaster, especially useful formulas and theory.

A bit early for statistical conclusion, but I see a nexus around "1-2 passes inaudible" and " 5 passes noticable but useable."

From there' I believe we'll try to get up to 3db compression on one or two comps on the way in, if we think we'll need it. We are input and Daw processor adequate here, so we'll grab clean audio as well, in case we want original dynamics.

That gives me one DA/AD loop for additional processing, plus the FX send thru the Bricasti, for "free", so to speak, with a bit of "headroom" for additional trips out the box.

Thanks for the pointers!
I've honestly made experiments with decent converters where maybe 8x DA-AD loops actually made some stuff sound better to my ears - but was some time ago, can't remember, at least 4x, try it yourself

converter quality will make all the difference.
Old 8th March 2016
  #11
Yup the cost in terms of digital degradation, will be that your mixes sound a lot less fractious and digital and begin to take on the sonic glue and cohesiveness of a mix done on expensive equipment.
Assuming you have A to D to A that is any quality at all and you are able to get good results phase wise without automatic delay compensation doing it all for you?.:-)X
Old 8th March 2016
  #12
Quote:
Originally Posted by Bassmec View Post
Yup the cost in terms of digital degradation, will be that your mixes sound a lot less fractious and digital and begin to take on the sonic glue and cohesiveness of a mix done on expensive equipment.
Assuming you have A to D to A that is any quality at all and you are able to get good results phase wise without automatic delay compensation doing it all for you?.:-)X
totally agree that flaky phase delay compensation can be a bit of a problem with the use of hardware insert. it can be a pain to match up some stuff if you want to do a lot of parallel processing busses with a different number of otb insert trips going on between each bus + the source.

possibly more of an argument for the whole use a console with decent analog routing for mults /patch etc thing, than are the extra conversion trips involved mixing itb with hardware inserts, but u can often work around the phase delay.

Last edited by alexmaster; 9th March 2016 at 01:50 AM..
Old 9th March 2016
  #13
Lives for gear
 
AlexK's Avatar
 

Quote:
Originally Posted by DistortingJack View Post
I've never known why people don't compress on the way in, AND also track the clean sound. Best of both worlds. If you hear the compressed version and it sounds good, drop the other one. If you overdid it, compress straight away.

Done.
Personally I prefer it when clients just send me one version of something. Having compressed and uncompressed vocals just slows me down.

Very rarely receiving badly compressed vocals nowadays thankfully.

Drums on the other hand... Why is everything so hypercompressed nowadays. Sounds like toss and gives me no options in mixing...
Old 9th March 2016
  #14
Lives for gear
 
Strange Leaf's Avatar
 

Quote:
Originally Posted by alexmaster View Post
hi . .I just read this + don't want to shoot myself in the foot as I am selling an HD24 right now, but ..

stock, they are nice converters, to my ears they sound much nicer than motu's musically but they are in the motu ballpark for noisefloor + I think this is the detectable element that adds up sonically.

there is also jitter noise but for roundtrip I think some of this cancels out so you will be mainly left with 24 bit quantization error from conversion but well below the noisefloor of the analog stage + therefore it should be very well dithered (at least equivalent to 4 bits @ 120db snr).

the noisefloor will accumulate over various DA-AD loops basically as a summation of random noise equation.

since this is a sum of squared standard deviations = variance , the rms measure (standard deviation) of noise will be growing according to sqrt(n) as we make n loops through the same da-ad.

in laymans terms rms noise will be growing at rate 1.41 (sqrt(2)) every time we double the number of loops. so to nearly double noise we make 4 loops , 1.41 *1.41, but to almost triple noise we need 8 loops , almost quadruple = 16 loops etc.

referencing a decibel scale it takes 100 loops to make a 10x = 20db increase in noise, but only 10 loops to make 3.162x = 10db increase in noise.

with a reasonable noisefloor roundtrip like a HD24 (something like 103db spec roundtrip) it will take 10 roundtrips for the noisefloor to become audible (equating this to > 16bit noisefloor = 96dbfs).

with a higher spec converter, perhaps 10 db better signal to noise ratio for a roundtrip, it would take 100x roundtrip for this same level of converter noise to become apparent.
I don't know enough theory of converters to agree or disagree with that. But it seems to me you are saying that it's impossible to hear audio roundtrips less that 10x loops. You can keep that calculation if you prefer it over real world tests.

But I just posted the result of a blind test, which was relevant to the thread. The audio I used was a live recording with a big audience, so any noise was well masked by the audio.

And as I said, after 5 rounds through the DA-AD's, most people could hear the quality loss. In this blind test, some of the files was copies of others, to confuse the listeners further, and the results was consistent. 3 was hard to hear, 5 pretty easy, and 10 impossible to miss.

You have a HD24 and if you doubt me, you can test this for yourself.
Old 9th March 2016
  #15
Quote:
Originally Posted by Strange Leaf View Post
I don't know enough theory of converters to agree or disagree with that. But it seems to me you are saying that it's impossible to hear audio roundtrips less that 10x loops. You can keep that calculation if you prefer it over real world tests.

But I just posted the result of a blind test, which was relevant to the thread. The audio I used was a live recording with a big audience, so any noise was well masked by the audio.

And as I said, after 5 rounds through the DA-AD's, most people could hear the quality loss. In this blind test, some of the files was copies of others, to confuse the listeners further, and the results was consistent. 3 was hard to hear, 5 pretty easy, and 10 impossible to miss.

You have a HD24 and if you doubt me, you can test this for yourself.
sry strangeleaf, but I think you didn't understand my argument.

what I am saying is that you quite likely can hear the degredation of the HD24 after 10x roundtrip, its not impossible that you could hear it after 2x roundtrip.

but, that's because the HD24 has but a reasonable noise spec ~ 103db snr roundtrip.

most high end converters nowadays are > 110 db snr roundtrip, maybe even ~ 120db.

as I was showing in the math this converter noise will actually will accumulate as a multiple that is a sqrt of the number of roundtrips. given that decibels are themselves logarithmic units we find that it takes a great deal of roundtrips to generate a certain number of decibels of noise accumulation.

effectively a 10x roundtrip for a high end spec converter would be equivalent to each 1x roundtrip for the HD24. it would take 100x roundtrips high spec converter to equal the 10x roundtrip hd24.

a better spec number might be noise + THD (includes distortion aswell as noise) but the relative gap between mid and high end converters is often similar.

you need to set up a similar test with a high end converter but you might find results conclude 100x roundtrip instead of 10x.

if you take the lower of any converters AD and DA spec , then subtract 3db you have a rough estimate of roundtrip.
the difference between 123db (see ssl alphalink quoted dynamic range spec http://www2.solidstatelogic.com/stud...ions#&panel1-1) and 103db (hd24 spec) would yield 1000x.

the noise+THD numbers are maybe more relevant though and perhaps these differ less extremely, and will also include harmonic distortion (that could be considered desirable), or intermodulation or jitter effects (both less desirable).

you have to understand the fact that the relationship between noise level and the number of roundtrips is in orders of magnitude.
btw random noise does add up but it is based upon this relationship.

nb. it's worth considering that some of the distortion effects, particularly the harmonics, are based to some extent on the signal and therefore not fully random (includes inharmonic distortion that is cyclical) . as such they could build up in a more linear fashion, ie ~3x roundtrip >> +10db, 10x roundtrip >> +20db.
(amplitude ratio accumulating linearly with roundtrip (not as a square root process as per random noise), but decibels still increasing as a log of the power ratio)
I think that linear buildup of any cyclical distortion will actually limit the sense of sonic integrity with far fewer roundtrips than the noise.
(it's worth noting that a decent converter might exhibit cyclical distortion some 20db lower than a less decent converter, as such a high end unit could roundtrip 10x before it becomes comparable to even 1x roundtrip of a lower end unit. given the numbers in this example it would take 30x roundtrip for the high end unit to produce distortion at the level of 3x roundtrip for the low end unit)

Last edited by alexmaster; 10th March 2016 at 01:04 PM..
Old 9th March 2016
  #16
Lives for gear
 
DistortingJack's Avatar
 

Quote:
Originally Posted by alexmaster View Post
sry strangeleaf, but I think you didn't understand my argument.

what I am saying is that you quite likely can hear the degredation of the HD24 after 10x roundtrip, its not impossible that you could hear it after 2x roundtrip.

but, that's because the HD24 has but a reasonable noise spec ~ 103db snr roundtrip.
The noise level of A/D D/A converters is just hiss.
It's not going to degrade your audio any more than any other type of hiss will.

The things that will actually degrade audio will be:
  • treble attenuation, phase shift and aliasing due to the lo-pass filter
  • jitter transient smearing
  • harmonic distortion from the analogue components

I wouldn't go over 2 trips unless I really had to.
Old 9th March 2016
  #17
Quote:
Originally Posted by DistortingJack View Post
The noise level of A/D D/A converters is just hiss.
It's not going to degrade your audio any more than any other type of hiss will.

The things that will actually degrade audio will be:
  • treble attenuation, phase shift and aliasing due to the lo-pass filter
  • jitter transient smearing
  • harmonic distortion from the analogue components

I wouldn't go over 2 trips unless I really had to.
actually I just had a think about the whole accumulation of noise thing with reference to THD.

the THD component should actually stack in a linear fashion, since it is based on the original signal in a less random way.

harmonic distortion will be present at exactly the same frequencies.
I think this is why I sometimes prefer audio after a few DA-AD loops.
the very slight harmonic distortion can add up with roundtrips.
this possibly good distortion could increase by roughly 10db over x3 roundtrips, +20db over 10x roundtrip etc.

jitter distortion is possibly more random, reflecting the random timing differences, perhaps with some less random cyclical component.
bad (cyclical) jitter would appear as frequency modulation, inharmonic peaks around the signal + could also accumulate in a more linear fashion.

intermodulation of the analog stage components can also create a low level distortion similar in appearance to jitter, inharmonic peaks related to the signal.
again this could accumulate linearly.

so the accumulation of distortion would likely be far greater than noise, but again it points to the quality of the converter as to how good/bad the clocking and analog stage actually are.

if you look at graphs of low level converter distortion shown in this article https://www.soundonsound.com/sos/jun...sterclocks.htm
you can see that the noise + distortion floor varies quite a bit between converters (here they are showing effects with bad clocking but also without, the raw distortiuon for each converter).

it has actually given me an idea of how to measure such converter distortion without the need for expensive dscope or suchlike, by repeating loopback of test tone you might be able to see more clearly how much any non-random, cyclical distortion effects build up above the noise.

personally I try to record and work at 96khz to avoid any treble attenuation or ripple from any lowpass anti-aliasing filter.

Last edited by alexmaster; 10th March 2016 at 12:51 PM..
Old 10th March 2016
  #18
Gear Nut
 

There is a low pass filter (anti-aliasing filter) in front of your A/D converters, it is mandatory.
That's probably the thing that makes your sound dull after a few loops.

This filter has a cut frequency which moves according to the sample rate used.
Higher the sample rate is, higher the cut frequency is.
That explains that in certain gear (cheap) you can hear a difference between 44khz and 96khz etc.. There is more air with higher sample rates due to higher cut frequency of this low pass filter.
There is another thread talking about samples rates, i should reply there too.

DSD take care of this issue (and more). Due to a high sample rate there is no need of an anti-aliasing filter (low pass filter).
Old 10th March 2016
  #19
Lives for gear
 
DistortingJack's Avatar
 

Quote:
Originally Posted by _ben View Post
There is a low pass filter (anti-aliasing filter) in front of your A/D converters, it is mandatory.
That's probably the thing that makes your sound dull after a few loops.

This filter has a cut frequency which moves according to the sample rate used.
Higher the sample rate is, higher the cut frequency is.
That explains that in certain gear (cheap) you can hear a difference between 44khz and 96khz etc.. There is more air with higher sample rates due to higher cut frequency of this low pass filter.
There is another thread talking about samples rates, i should reply there too.

DSD take care of this issue (and more). Due to a high sample rate there is no need of an anti-aliasing filter (low pass filter).
DSD still has a filter.


Filters are inaudible if the sampling rate goes above 60 kHz.
Well-designed filters are pretty much inaudible for 44.1 kHz too.
Old 10th March 2016
  #20
Quote:
Originally Posted by _ben View Post
There is a low pass filter (anti-aliasing filter) in front of your A/D converters, it is mandatory.
That's probably the thing that makes your sound dull after a few loops.

This filter has a cut frequency which moves according to the sample rate used.
Higher the sample rate is, higher the cut frequency is.
That explains that in certain gear (cheap) you can hear a difference between 44khz and 96khz etc.. There is more air with higher sample rates due to higher cut frequency of this low pass filter.
There is another thread talking about samples rates, i should reply there too.

DSD take care of this issue (and more). Due to a high sample rate there is no need of an anti-aliasing filter (low pass filter).
general excessive layering of noise and distortion (without eq) can also create a dulling, muddying, cloudy (word of choice) effect.

excessive harmonic distortion on bass heavy material can also do this.
Old 10th March 2016
  #21
Gear Nut
 

Quote:
Originally Posted by DistortingJack View Post
DSD still has a filter.

Filters are inaudible if the sampling rate goes above 60 kHz.
Well-designed filters are pretty much inaudible for 44.1 kHz too.
Nothing compare to a brick wall.
BitPerfect: On DSD vs PCM … again

I'm not agree, some filters are audible, especially with cheap audio system, above 60khz.

Yes, some good filters are inaudible at 44khz and more but after 10 loops?
Old 10th March 2016
  #22
Lives for gear
 
DistortingJack's Avatar
 

Quote:
Originally Posted by _ben View Post
Nothing compare to a brick wall.
BitPerfect: On DSD vs PCM … again

I'm not agree, some filters are audible, especially with cheap audio system, above 60khz.

Yes, some good filters are inaudible at 44khz and more but after 10 loops?
The filter on PCM converters is nothing like a "brickwall" at 88.2 kHz sampling rate or above.
If you have an audible filter at that rate, then you have a truly terrible converter. Of course, here it's a "personal" matter, and I can't prove you can't hear it, but it really is rather easy to design such a filter. Much more than a 44.1 kHz filter, for sure.

Often people hear an audible difference between sampling rates due to intermodulation distortion from ultrasonic frequencies doubling down to audible ones.

This is caused by analogue equipment (power amps, speakers, etc) that isn't designed to handle frequencies above 20 kHz.
In that case, a sampling rate like 48 kHz is actually ideal, removing the crud at the top with enough wiggle room to make its phase shifting and ringing inaudible.
Old 10th March 2016
  #23
Gear Nut
 

I remember having had a 1616M PCMIA and the difference between 44, 96 and 192 was clearly audible. More air each time with an upper frequency rate.
Old 10th March 2016
  #24
Lives for gear
 
JayTee4303's Avatar
The primary Daw in the control room has a Motu PCIe 424 card, with HD-192, 2 2408 mk3s, and a 2408 mk2 attached. S/N is 120db on the 192, 108db on the mk3s, and 105 db on the mk2.


All the outboard is hung on the 192, with the Bricasti serving a master reverb buss on AES, so there shouldn't be any DA/AD degradation there.

I'm currently running the 1176 chained into the LA2A, (all thru XLR patchbays so I can reverse or seperate them if need be) and line level preamped mic signals use their own 192 inputs.

I assume that this signal routing constitutes one DA/AD loop, even though we're not all the way into and out of Daw software. I believe that since I see signals inthe DSP software, the DSP zero-latency monitoring still uses the AD and DA converters.

If I have all this straight, my...least affected...signal, would be from a pre, directly into the comps, and only then into an HD192 input, for the "on the way in" processing pass?

Am I also correct in assuming the master reverb pass thru the Bricasti on AES incurs no DA/AD penalty?

We generally record at 48k, as most of our audio is associated with video, but the 192 goes to 192 and I'll look into potential benefits of avoiding Nyquist filtering at higher SRs.

I'm not ignoring anyone's recommendation that a few DA/AD loops can improve mix cohesion, but I do want to know where I stand in DA/AD loop count, and where ny least affected signal lies.
Old 10th March 2016
  #25
Quote:
Originally Posted by JayTee4303 View Post
The primary Daw in the control room has a Motu PCIe 424 card, with HD-192, 2 2408 mk3s, and a 2408 mk2 attached. S/N is 120db on the 192, 108db on the mk3s, and 105 db on the mk2.


All the outboard is hung on the 192, with the Bricasti serving a master reverb buss on AES, so there shouldn't be any DA/AD degradation there.

I'm currently running the 1176 chained into the LA2A, (all thru XLR patchbays so I can reverse or seperate them if need be) and line level preamped mic signals use their own 192 inputs.

I assume that this signal routing constitutes one DA/AD loop, even though we're not all the way into and out of Daw software. I believe that since I see signals inthe DSP software, the DSP zero-latency monitoring still uses the AD and DA converters.

If I have all this straight, my...least affected...signal, would be from a pre, directly into the comps, and only then into an HD192 input, for the "on the way in" processing pass?

Am I also correct in assuming the master reverb pass thru the Bricasti on AES incurs no DA/AD penalty?

We generally record at 48k, as most of our audio is associated with video, but the 192 goes to 192 and I'll look into potential benefits of avoiding Nyquist filtering at higher SRs.

I'm not ignoring anyone's recommendation that a few DA/AD loops can improve mix cohesion, but I do want to know where I stand in DA/AD loop count, and where ny least affected signal lies.
I think what actually has the potential to cause the greatest amount of sonic degredation in that setup is the choice of central master clock.

motu systems are pretty bad at distributing clock signal to my recollection
Old 10th March 2016
  #26
Gear Guru
 
drBill's Avatar
Digital is all about capturing the clarity and definition of the source....Analog is about about capturing a musical impersonation of the source. Most people prefer the latter. 9/10ths of the plug ins out there are about reaching for the latter.

A couple of trips through DA/AD? LOL I don't even think about it. If you want the clarity of digital, don;t ever leave once you're in. if you want the musical aspect of analog, go in and out as many times as it suits your production esthetic.

You want to talk sonic "degradation"? Talk about bouncing tracks on tape. Didn't seem to really hurt the music that much in those days. In fact, that's the sound so many are seeking....
Old 11th March 2016
  #27
Lives for gear
 
JayTee4303's Avatar
Quote:
Originally Posted by alexmaster View Post
I think what actually has the potential to cause the greatest amount of sonic degredation in that setup is the choice of central master clock.

motu systems are pretty bad at distributing clock signal to my recollection
Complex system here...4 rooms, 8 PCs, 12 interfaces, and at least 5 additional clocked devices, most, but not all, currently clocked via ADAT over Toslink, thru a Z-Sys Optipatch Plus, which reclocks the signal.

Have not tested every possible configuration, probably won't but...

A 2 khz test tone records cleanly at every primary and supporting Daw, from every source tested to date, across the facility. Those include internal clocking on five Daws, and Slave clocking using the Motu rig described above, a Profire 2626, and an M-Audio C-600 ( video edit PC audio interface), as Master Clock.

To date, there are two...potential issues...we've discovered so far. With the Motu as Master, audio from an ADA-8000, Toslinked thru the backbone to the 2626, ( clock by proxy?), records clean, but audio coming back thru, headed for an artist's headphone mixer, has occasional pops and clicks.

Artist's monitor audio in another room, has similar artifacts, routed one way, but not others. There's a good chance, this is also related to that ADA-8000, but I haven't chased it down that far, yet.When first noticed, I just re-routed to get thru the session. That ADA-8000...is third line gear...or less. It currently takes signal from a DX-7, a stereo feed from a Casio VZ10m, that's about to get upgraded to a tier two input on the 2626, when a long overdue guitar FX switcher arrives and frees up the ins, and a stereo feed from our 3rd line full time guitar processor, a Pod Pro we haven't used in a couple years. It also heats the room significantly.

Other than those issues, I'm satisfied with the Motus as masters and slaves, in the configs we use frequently. Of course, that may mean I'm happy with Z-Systems cl8ck, not the Motu.

I like the Motu Cuemix spec-a, Ozone has some useful tools, and even Span tells me good stuff, once this thread plays out, and we move from discussion to analysis, the scopes should tell any tales my ears might not assimilate.

Back on topic...we need to end up at 48k for video production. Would that not imply dithering, if tracking and processing at higher SRs in the early stages?

If so, what does your gut, and any formulae you care to contribute, say about the tradeoffs implied?

Last edited by JayTee4303; 11th March 2016 at 07:52 AM..
Old 11th March 2016
  #28
Lives for gear
 
JayTee4303's Avatar
Quote:
Originally Posted by drBill View Post
Digital is all about capturing the clarity and definition of the source....Analog is about about capturing a musical impersonation of the source. Most people prefer the latter. 9/10ths of the plug ins out there are about reaching for the latter.

A couple of trips through DA/AD? LOL I don't even think about it. If you want the clarity of digital, don;t ever leave once you're in. if you want the musical aspect of analog, go in and out as many times as it suits your production esthetic.

You want to talk sonic "degradation"? Talk about bouncing tracks on tape. Didn't seem to really hurt the music that much in those days. In fact, that's the sound so many are seeking....
Back in the days when Cakewalk and DOS were it, when Wi dows were but a ripe, juicy (red!) gleam in Bill's eye, and sequencer sync to tape ran a cool quarter mill...my sequencer and analog tracks aggregated onto cassette tape, one bounce at a time.

The key to getting specific final results was a product of realistic expectations, and innovative planning. There were workarounds that avoided extra tape generations, some "legal", some not. With sequencer memory volatile, and (cassette!!) data storage cumbersome and unreliable, you had to accurately assess tradeoffs in advance. A part of that included knowing precisely what generational losses you'd incur from this tape on this deck, to that track, on that tape, in that deck.

Part of that body of information could only come from hands on experience, but I always found it more efficient to discuss the issues w other engineers first, then experiment with a pre-selected range of options, hopefully narrowed to the most productive spectrum, than to conduct exhaustive, Edison style shotgun empirical testing of every conceivable possibility.

This thread, falls under a similar effort, a general description of which might be, "pre-discussion".

;-)
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump