• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Measurements of Motu UltraLite mk4 (AKM DAC Version)

BE718

Major Contributor
forum experimenter
Patreon Donor
Joined
Mar 1, 2016
Messages
2,252
Likes
1,004
Location
Perth Western Australia
#1
As mentioned elsewhere I have just purchased a Motu UltraLite mk4. I already use a Motu 8A for driving my active 3 way DSP speaker system, however being only 8 channels I am unable to run the centre speaker from it. The EasyLite mk4 caught my eye as it has been upgraded over the mk3 to include the same ESS9016 dacs as the 8A and has the extra 2 channels I require. I have measured the 8A previously and its performance is excellent. The paper specs for the Ultralite claim they have improved the THD+Noise figure by 2dB to 112dB over the 8A.

More info about the product below. Casework is an excellent. Cost is around $595 US.

http://motu.com/products/proaudio/ultralite-mk4
http://motu.com/products/avb/8a

1.jpg


2.jpg


Measurements.

This just the DAC section, I will test the ADC later.

First off is 1kHz 0dB FS - Note that the noise floor is in reality lower (maybe 20dB) as the Measurement ADC noise floor is modulated by the high 1kHz signal level. I need a notch filter as Amirs AP does to remove the fundamental.

Balanced outputs, FS is 6.25v rms.

Disappointed straight away with multiple harmonics stretching right up the band. However I quickly discovered this disapears when you drop below 0dB. At -0.1 dB the spray of harmonics is gone. I wonder if this is a hardware issue in the Ultralite output amps or can be fixed in its internal DSP. Im going to contact Motu to see what they say. The 8A doesnt do this.

EL 0dB.png


-0.15dB - 1st and second harmonics are higher than I would like.

EL -0.15dB.png



-6dB

EL -6dB.png



-20dB

EL -20dB.png



-60dB - Odd spuria at about 14.8kHz

EL -60dB.png



-90.31dB Time Waveform - Fairly clean, seen better

EL -90.31dB TW.png


Noise Floor

EL noise floor.png



IMD SMPTE - Pretty Good

EL IMD smpte.png



IMD 19+20kHz - 1kHz IMD is down at -125dB, again pretty good

EL IMD 19+20kHz.png



12kHz tone Jitter - Nothing of note

EL 12kHz Jitter.png

EL 12kHz jitter Zoomed.png



Filter response and imaging

Filter and imaging.png


Overall a reasonable performance, I was hoping for a bit better. From memory it doesnt quite match the Motu 8A but I will post some comparison measurements later.

Also I wont be able to resist ripping the lid off so some electronic porn pics will be coming :)
 
Last edited:

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
4,049
Likes
2,235
#2
Nice write up, and nice test. Unit looks pretty good generally though I was thinking a bit short of the other Motu.

This would be a lot of trouble, but could you measure each Motu with the other Motu, and maybe do a loopback of each Motu.
 

bennetng

Active Member
Joined
Nov 15, 2017
Messages
213
Likes
61
#4
@BE718
Could you do some RMAA tests as well? If you are not using Windows I can attach the RMAA test signal as flac files, then you can attach the recorded files and I can generate the reports with my PC.
 

bennetng

Active Member
Joined
Nov 15, 2017
Messages
213
Likes
61
#7
RMAA allows people to perform any DA/AD tests with any audio devices. Like other software based suites RMAA knows nothing about I/O levels of the devices under test. The reviewer or tester needs to explain how the tests are performed to the readers.

One advantage of RMAA is the test parameters are almost fixed (at least in the free versions), users cannot tweak FFT size, averaging, windowing etc. That means the test results among different users can be directly compared if the same version of RMAA is being used.

The audio level of the recorded file should be same as the reference, RMAA will detect level difference and show it in the "Level change" section in the reports. Also, the playback and recording sample rate should be same as the reference file, that means a 96kHz test signal should be played and recorded at 96kHz.
rmaa645.PNG
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
13,990
Likes
6,337
Location
Seattle Area
#12
Question: did you put the ferrite core on the power cable or did it come that way? That is a good method (wrapping the wire multiple times) to get the filtering to go to lower frequencies.
 

Thomas savage

MQA all the Way
Moderator
The Watchman
Patreon Donor
Joined
Feb 24, 2016
Messages
5,928
Likes
1,869
Location
uk, taunton
#13
As mentioned elsewhere I have just purchased a Motu UltraLite mk4. I already use a Motu 8A for driving my active 3 way DSP speaker system, however being only 8 channels I am unable to run the centre speaker from it. The EasyLite mk4 caught my eye as it has been upgraded over the mk3 to include the same ESS9016 dacs as the 8A and has the extra 2 channels I require. I have measured the 8A previously and its performance is excellent. The paper specs for the Ultralite claim they have improved the THD+Noise figure by 2dB to 112dB over the 8A.

More info about the product below. Casework is an excellent. Cost is around $595 US.

http://motu.com/products/proaudio/ultralite-mk4
http://motu.com/products/avb/8a

View attachment 14093

View attachment 14094

Measurements.

This just the DAC section, I will test the ADC later.

First off is 1kHz 0dB FS - Note that the noise floor is in reality lower (maybe 20dB) as the Measurement ADC noise floor is modulated by the high 1kHz signal level. I need a notch filter as Amirs AP does to remove the fundamental.

Balanced outputs, FS is 6.25v rms.

Disappointed straight away with multiple harmonics stretching right up the band. However I quickly discovered this disapears when you drop below 0dB. At -0.1 dB the spray of harmonics is gone. I wonder if this is a hardware issue in the Ultralite output amps or can be fixed in its internal DSP. Im going to contact Motu to see what they say. The 8A doesnt do this.

View attachment 14079

-0.15dB - 1st and second harmonics are higher than I would like.

View attachment 14080


-6dB

View attachment 14081


-20dB

View attachment 14083


-60dB - Odd spuria at about 14.8kHz

View attachment 14084


-90.31dB Time Waveform - Fairly clean, seen better

View attachment 14085

Noise Floor

View attachment 14092


IMD SMPTE - Pretty Good

View attachment 14086


IMD 19+20kHz - 1kHz IMD is down at -125dB, again pretty good

View attachment 14088


12kHz tone Jitter - Nothing of note

View attachment 14089
View attachment 14090


Filter response and imaging

View attachment 14091

Overall a reasonable performance, I was hoping for a bit better. From memory it doesnt quite match the Motu 8A but I will post some comparison measurements later.

Also I wont be able to resist ripping the lid off so some electronic porn pics will be coming :)
Absolutely fantastic, your everything that makes this worthwhile for me Alan .

Thank you mate.
 

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
4,049
Likes
2,235
#14
RMAA allows people to perform any DA/AD tests with any audio devices. Like other software based suites RMAA knows nothing about I/O levels of the devices under test. The reviewer or tester needs to explain how the tests are performed to the readers.

One advantage of RMAA is the test parameters are almost fixed (at least in the free versions), users cannot tweak FFT size, averaging, windowing etc. That means the test results among different users can be directly compared if the same version of RMAA is being used.

The audio level of the recorded file should be same as the reference, RMAA will detect level difference and show it in the "Level change" section in the reports. Also, the playback and recording sample rate should be same as the reference file, that means a 96kHz test signal should be played and recorded at 96kHz.
View attachment 14096
I've not used RMAA for several years. Wasn't happy with it back then. I experimented with it week before last. It works better now than it did.

The results were broadly similar to results obtained manually. Differences between my results and RMAA were small. Small being 3 db or less for most parameters. So not bad for something quite convenient and quick. As noted easy to compare your results with those of some one else using RMAA.

For anyone running Linux, RMAA works under WINE in Linux. I didn't use it for running tests off the Linux machine. I did generate the analysis file. Run that file thru the DUT and record as a wav. Then let RMAA in Linux analyze the recorded wav file. This worked fine.
 

BE718

Major Contributor
forum experimenter
Patreon Donor
Joined
Mar 1, 2016
Messages
2,252
Likes
1,004
Location
Perth Western Australia
#15
Question: did you put the ferrite core on the power cable or did it come that way? That is a good method (wrapping the wire multiple times) to get the filtering to go to lower frequencies.
I did it, but it hasnt proved necessary, has no measurable or otherwise noticeable effect. Just curiosity on my behalf. I suppose it might reduce the radiated RF from the smps.
 
Last edited:

BE718

Major Contributor
forum experimenter
Patreon Donor
Joined
Mar 1, 2016
Messages
2,252
Likes
1,004
Location
Perth Western Australia
#17
As mentioned previously the QA401 analyser has an issue with noise floor modulation, ie it rises with increasing signal level. All A to Ds suffer to some degree along with their own distortion which adds to the distortion of the measurement. This which is why the AP that Amir uses has a notch filter to remove the high level fundamental signal leaving the best conditions to measure the distortion of the DUT. The QA401 seems to suffer more than it should do.

To show this I have measured the Ultralite at 1kHz FS output with the 8A. You can see the noise floor is a good 20dB lower and now allows us to see the distortion harmonics at a lower level. Just a note that the Ultralite appears to have a lower FS output voltage than the 8A, about 3dB, s I used the MOTU software to raise the input amp gain to bring the input close to FS.

1532238958492.png


Here is the MOTU 8A in loopback compared to the Ultralite in orange. 8A has slightly lower 1st and 2nd harmonics. The lower noise floor is predominantly due to the higher relative output level ( 9vrms v 6 v rms). What is clear is that the 8A has less noise like jitter. Difficult to see it being an audible issue at that low level however. FFT parameters are Kaiser 7 window and 131072 points.

Untitled.png
 
Last edited:

BE718

Major Contributor
forum experimenter
Patreon Donor
Joined
Mar 1, 2016
Messages
2,252
Likes
1,004
Location
Perth Western Australia
#19
Ultralite in loopback. Its A to D is not quite as low noise or distortion as the 8A, but this is expected and shown in the specs from Motu.

1kHz 0dB

1532243500475.png


-6dB
1532244907119.png


-20 dB
1532245418156.png


IMD SMPTE
1532243831937.png


IMD 19+20kHz
1532244062004.png


12kHz tone jitter -skirt of noise at low level
1532244071999.png
 
Last edited:

BE718

Major Contributor
forum experimenter
Patreon Donor
Joined
Mar 1, 2016
Messages
2,252
Likes
1,004
Location
Perth Western Australia
#20
Conclusions, well as mentioned the Ultralites A to D isnt as good as the 8A, however the D to A is a tough one. The performance is different, the Ultralite showing less spuria in some tests but worse jitter noise. Is any of this obvious in listening, I doubt it. Im going to have to listen and give it a try.
 

Similar threads

Top Bottom