MOTU UltraLite-mk5 Review (Audio Interface)

[pkane]

Damn the Topping HS02 isolator doesn't work with Motu. I even put a USB hub in between which sometimes is needed with other USB DACs.

Cuemix doesn't see the DAC at all with Topping isolator in between. Damn

Did you try powering HS02 with a separate USB cord?

 

[Music1969]

Did you try powering HS02 with a separate USB cord?

I didn't know it take external power. Will try again

 

[mdsimon2]

I agree with @pkane, would have said 8 kHz is USB noise. I don't see that with a Mk5 loopback (Ch 1 out -> Mic 1, Ch 2 out -> Mic 2) when connected to a 2015 Macbook Pro on battery or plugged in or a mac mini 2012.

I see a little sliver of it with a MOTU M4 (USB powered) into the Mk5 mic inputs. It is easiest to see for me with a lot of averages (32) when observing silence from a DAC channel connected but not in use, with another M4 DAC channel playing something. Probably going to depend on the specific devices at play and how they are powered.

Michael

 

[Music1969]

I agree with @pkane, would have said 8 kHz is USB noise. I don't see that with a Mk5 loopback (Ch 1 out -> Mic 1, Ch 2 out -> Mic 2) when connected to a 2015 Macbook Pro on battery or plugged in or a mac mini 2012

my measurement is also Macbook on battery

So no "loop' for ground currents

 

[Davide]

I managed to create a script to perform real time BlackHole clock adjustment using the Ultralite Mk5 as reference.
It seems to work well and keep the drift constantly below 1 ppm.
Since, however low, the drift can still lead to buffer underun / overflow, I added a frame drift calculation so as to selectively compensate when it exceeds 1 (it seem to me a more robust solution than a control algorithm, that needs a lot of tuning and introduce instability risk).
To work properly I had to modify the BlackHole source code to use a 32 frame safety offset and directly receive the actual 24bit resolution coefficient from the script.
The script is extremely light, about 7MB of RAM and less then 0.1% CPU, and obviously much better than the ASRC performed by Core Audio with its Clock Drift compensation function.
Within the script I created the DAW opening command so I only click on it without having to remember to start it every time...
I could even let it started by the DAW, which is perhaps better ... I will evaluate.
Anyway, thanks to ChatGPT for its contribution, without which I would not even know where to start.

Having said that, I confirm the previously observed clock fluctuation, but in fact I can't know if it's that of the MK5 or the Mac because it is relative, and I have no other audio devices to test...

I hope it is the Mac, or that in any case the fluctuation is not present or is compensated by ASRC in DA-AD stadiums.

 

[Music1969]

but in fact I can't know if it's that of the MK5 or the Mac because it is relative, and I have no other audio devices to test...

Million dollar question.
Can you test on Windows or Linux to see if it is Mac?

How can i measure clock drift?

 

[Davide]

Million dollar question.
Can you test on Windows or Linux to see if it is Mac?

How can i measure clock drift?

You can use REW with 1kHz signal generator and real time analyser (RTA) with displayed distortion (it shows ppm in 0.0 resolution).
But it depends what devices you use. You need to set the output to one device and the input to another, and then route the audio between devices.

In theory I could do the same thing on Windows, to observe the fluctuation, but at the moment I don't have a configured PC...

 

[Music1969]

You can use REW with 1kHz signal generator and real time analyser (RTA) with displayed distortion (it shows ppm in 0.0 resolution).
But it depends what devices you use. You need to set the output to one device and the input to another, and then route the audio between devices.

In theory I could do the same thing on Windows, to observe the fluctuation, but at the moment I don't have a configured PC...

you often see greater than 1ppm ?

@mdsimon2 also ?

 

[Davide]

you often see greater than 1ppm ?

@mdsimon2 also ?

At 96kHz the relative drift between Mk5 and BlackHole (that follows Mac internal 24MHz clock) in my case is 32ppm, with a fluctuation of about +/- 0,5ppm. But the observed fluctuation is not periodic, very low freq components are present, and other tools are needed to accurately characterize it (like ultra low phase noise oscilloscope).

 

[Davide]

Anybody else see this noise with mic input? It is not there in rear TRS input.

Nothing significantly obviously but just asking if others seen it

Front mic input - noise around the 8kHz area. It shows in all measurements, not just this multitone
View attachment 441857

Rear TRS input - no noise

View attachment 441859

@mdsimon2

Are you using the XLR input or the TRS input on the front panel?
Although the connector is the same they have two different impedances, and presumably two different circuits inside.
Front TRS is for instruments and has 2M impedance vs 2.8k of XLR and 20k of rear TRS. So front TRS is much more prone to EMI, like USB or other internal microprocessor, like the one for for the display.
If you are using XLR it's more strange...
But if it is internal interference you should see it even without connecting anything, just open the measure.
if it is external, well the connected cable is an excellent antenna if not properly shielded.
It must be said that on a balanced circuit it is however strange that interference is caught...

 

[Evgenioni]

In front, TS works as HI-Z, TRS switches to a linear, it just goes through the mic's pre. The rear TRS is connected directly to the DAC. Otherwise, they work identical.

 

[Music1969]

If you are using XLR it's more strange...

XLR on front

 

[Pancreas]

Last time I checked, axial fans had kind of a hard time sucking air through a hard surface...

A slight breeze going sideways across the interface would probably be more than sufficient. Maybe a pair of Arctic P8 Silent or Be Quiet Pure Wings 2 80mm with a fan pot, or two Pure Wings 2 PWM 80mm with a PWM fan controller thingy (+ 4-pin or 3-pin fan splitter cable as required). Make some provisions for vibration decoupling using thin rubber strips or whatnot. You could get a pair of Noctua NF-A8 ULN and just use the included 7V adapter, too... Noctuas are a bit ouch on the wallet but they do come with a nice package of accessories and have basic decoupling in the corners (while e.g. the Pure Wings come with a bare minimum package of 4 fan screws). I'm not sure I would trust the Arctics' fan bearings long-term when laid horizontally, but vertically I don't see a problem.

I pimped my office PC with Pure Wings 2 PWM 80 / 92 mm as case fans (using rubber nipples for decoupling) and a Noctua P9 PWM on the CPU. The Noctua can be slowed down even further and its tachometer signal seemed a bit more stable, but purely noise wise I have zero complaints about the Pure Wings. (The 120/140 mm versions have had some reports about some variability regarding bearing noises.)

Hmm, the Ultralite is a bit deeper than I thought at 17.5 cm, I might go up to 92 mm fans then. Arctic doesn't make any 92s, but Pure Wings 2 are available both as 3-pin and PWM (which is how I ended up with one to begin with), and Noctua's counterpart to the NF-A8 ULN would be the NF-A9 FLX.

the problem with those noctua is they are not usb, so now you have to go inside your case somehow which i dont wanna do. I have Corsair icue fans, is not gonna work with that

the board has fan headers but i rather not connect there, is more convenient usb

 

[redeyes]

Don't be insane. I've had mine on since i bought it 1.5 years ago. It doesn't get hot nor do you need to baby it.

 

[Pancreas]

Don't be insane. I've had mine on since i bought it 1.5 years ago. It doesn't get hot nor do you need to baby it.

considering they keep raising the price. I bought this for $595

Then it went up to $650, now is $700

There is literally zero change

Why is it ok to cool down PC components, even gaming consoles have fans, and all of those dont stay all the time

I keep this thing on all the time, maybe a USB fan connected to PC may help cool it down when is actually being used, one of those fans is $29, i dont see why not

Warranty is 2 years, so technically Im out of warranty now

 

[dasdoing]

Why is it ok to cool down PC components

because they can and will overheat while in heavy usage. For the Motu, this is extremely unlikely, unless you use it in a closed enclosure

 

[redeyes]

Why is it ok to cool down PC components, even gaming consoles have fans, and all of those dont stay all the time

Seriously? PCs put out hundreds of watts of heat while on. This tiny box takes 15w at most and likely way under that.

 

[Pancreas]

because they can and will overheat while in heavy usage. For the Motu, this is extremely unlikely, unless you use it in a closed enclosure

It still feels warm. Anything device can be can cool down, is always a good thing and the fan is $29. I don't see why not, or will you be replacing my Motu when it goes bad one day? lol

 

[Evgenioni]

It works in a rack 24/7 for 3 years, loaded almost completely. No problems

 

[phoenixsong]

It still feels warm. Anything device can be can cool down, is always a good thing and the fan is $29. I don't see why not, or will you be replacing my Motu when it goes bad one day? lol

"Anything device can be can cool down, is always a good thing" well that doesn't mean you have to liquid nitrogen-cryo treat everything