Motu UltraLite-mk4 technical measurements (DAC only)

Discussion in 'Source Measurements' started by atomicbob, Jun 5, 2020.

  1. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    Keep checking back as it may take multiple posting sessions to fill in all the sections listed. This report is a bit more complicated than most. Edits for clarity / grammar may occur over the next several days.

    Motu UltraLite-mk4 technical measurements (DAC only)

    Standard Prologue
    If you are unfamiliar with audio measurements please use a search engine with the query:
    "audio measurements" or "audio measurement handbook"
    Look for publications by Richard C. Cabot and also by Bob Metzler, both from Audio Precision. There are other useful publications as well. These will provide basic knowledge.
    Interpretation of the following measurements is beyond the scope of technical measurements posts.


    The data presented were collected as follows:
    1. PrismSound dScope III, picoscope 5243B, Cal Test CT2593-2 balanced probe
    2. Tecnec 75R spdif cable
    3. Balanced TRS to XLR cables Canare L-4E6S Neutrik XLR connectors
    4. Single Ended cables Mogami 2964 1 meter with Amphenol RCA connectors
    5. 100 Kohm load used for measurements
    6. dScope analyzer sample rate 48 KHz unless otherwise noted
    7. DAC 44.1 KHz sample rate, 24 bit depth unless otherwise noted
    9. Audioquest Forest and Schiit Pyst USB cables
    9. Vaunix Lab Brick USB hub
    10. Shielded 14AWG and 16AWG power cables

    Measurements are made in accordance with AES17:2015

    UltraLite-mk4
    Measurements for record commenced after warmup of 1 hour.
    Measurements were performed over a period of several days.

    Index
    Post 1 - measurement setup description, highlights
    Post 2 - spdif input - Bal TRS outputs
    Post 3 - USB input - Bal TRS outputs
    Post 4 - spdif input - SE output via TS on TRS outputs
    Post 5 - USB input - SE output via TS on TRS outputs
    Post 6 - Square wave response, Distortion 0 to -1 dBFS (Tech notes part 1)
    Post 7 - Distortion (Tech Notes part 2) and Bit Resolution
    Post 8 - Comparison with Matrix XSP
    Post 9 - reserved for additional data
    Post 10 - reserved for corrections

    Notable highlights:
    Employs es9016s 8 channel DAC
    Excellent performance, especially considering price and size
    Dynamic Range > 130 dB This surprised me as Motu specifies 120 dB
    Gain Linearity tight to -110 dBFS then less than ± 1 dB to -130 dBFS
    Channel Crosstalk is exceptionally low with separation > 145 dB Bal and > 128 SE
    Possibly highest performance / price ratio pro audio interface existing at the time of this post
    Has earned duty in my pro audio setup replacing RME Fireface UC for most lab and studio functions

    Measurement setup picture example:
    DSC_0460_small.jpg

    The UltraLite-mk4 targets the pro audio interface market. It is a lot more than a great performing Sabre DAC. Typical pro audio use will be balanced connections. There are several very important safety considerations for this interface.

    1. TRS outputs are active on both Tip and Ring. Never use a Tip-Sleeve phone plug for Single Ended connection, which shorts Ring to Sleeve and accordingly shorts one half of the active balanced output to ground. At minimum this will increase distortion. Worst case this might let the magic smoke out of components in the interface.

    2. Never hot patch outputs. Always stop playback of any signal before changing connections on the TRS outputs. Again, this is similar to 1 above with TRS outputs having temporary shorts as the plug is inserted or extracted. If the outputs are quiescent then risk of damage is minimized. Best to have system completely off but not always possible during a recording session.

    3. For mic input users: Never hot patch the mic inputs, turn off P48 power before connecting or disconnecting microphones and XLR cables to the mic inputs. A number of studio interns have learned the hard way why this is so important, and thus usually never worked beyond intern, often discharged immediately. At the minimum a loud pop will be sent through the system as XLR pin 2 and pin 3 make connection to the P48 at slightly different times. If the system is powering monitors or SR speakers this can cause major damage amplified to a VERy LOUD POP! Microphones also may sustain damage with differential voltage for that brief moment either pin 2 or pin 3 connects first to P48.

    This pro audio interface has a built-in patch bay and mixer with effects. There are also level trims to adjust. Here are example screen shots of internal settings to consider and were adjusted during measurements.

    Additional setup example screenshots

    Device page: clock source, mic P48 + trims, analog input trims, analog output trims, etc.
    UltraLite-mk4 trim -1 dBFS  main LR out.png

    Routing page: signal source to destination patch bay
    UltraLite-mk4 USB routing to main LR out.png

    Mixer: a reasonably featured mix console which was bypassed for measurements
    UltraLite-mk4 mixer.png

    UltraLite-mk4 nternal picture
    PCB_small_annotated.jpg

    ES9016S - DAC 8 channel
    CS5368 - ADC 8 channel
    CS3308 - Volume Control 8 channel
    PGA2505I - Programmable microphone preamp (Burr Brown)
     
    Last edited: Jun 5, 2020
  2. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    spdif input Bal TRS output measurements

    UltraLite-mk4 Dynamic Range spdif input Bal output
    20200523 UltraLite-mk4 dynamic range FFT spdif Bal.png
    Impressive for a pro audio interface, especially given the price point.

    Dynamic range, in simplified terms
    1. find maximum output voltage at 0 dBFS
    2. find residual noise with a -60 dBFS 1 KHz stimulus, removed from analysis by window notch
    3. Dynamic Range is the difference between maximum output and residual noise
    Specifics are in AES17 section 9.3 (measurebators do your homework please)
    Example here:
    https://www.maximintegrated.com/en/design/blog/spec-dynamic-range.html

    Why we have a -60 dBFS stimulus (included for certain measurement wankers that don't bother reading AES17 standards)
    Some clever codec designers include an output gate that shuts off when there is no signal present. This has the dual benefit of lowering output idle noise while also gaming the measurements. AES became wise requiring a -60 dBFS stimulus so any such gates are open during measurement and true Dynamic Range performance measured.

    UltraLite-mk4 A04 THD+N THD nth-HD FFT spdif input Bal output -1 dBFS output trim
    20200523 UltraLite-mk4 A04 THD+N THD nth-HD 4+HD+N 60Hz FFT -1 dBFS - spdif Bal.png

    UltraLite-mk4 Frequency Response y-axis highly zoomed FFT spdif input Bal output
    20200523 UltraLite-mk4 frequency response zoom FFT spdif Bal.png

    UltraLite-mk4 50 + 7000 Hz spdif input Bal output - Left Channel
    FFT_2_L1T13_1_A.png

    UltraLite-mk4 Gain Linearity spdif input Bal output - Left Channel
    G2_L1T3_1_A.png

    UltraLite-mk4 THD+N vs Frequency spdif input Bal output - Left Channel
    G2_L1T5_1_A.png

    UltraLite-mk4 Residual Noise Bal spdif input Bal output - Left Channel
    FFT_2_L1T16_1_A.png

    Complete spdif input Bal output analysis report pdf attached
     

    Attached Files:

    Last edited: Jun 5, 2020
  3. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    USB input Bal output measurements

    UltraLite-mk4 Dynamic Range USB input Bal output
    20200523-00 UltraLite-mk4 dynamic range FFT ASIO Bal.png

    UltraLite-mk4 A04 THD+N THD nth-HD FFT USB input Bal output -1 dBFS
    20200523-04 UltraLite-mk4 A04 THD+N THD nth-HD 4+HD+N 60Hz FFT -1dBFS - ASIO Bal.png

    UltraLite-mk4 Frequency Response y-axis highly zoomed FFT USB input Bal output
    20200523-05 UltraLite-mk4 frequency response zoom FFT ASIO Bal.png

    UltraLite-mk4 50 + 7000 Hz USB input Bal output - Left Channel
    FFT_2_L1T13_1_A.png

    UltraLite-mk4 Gain Linearity USB input Bal output - Left Channel
    G2_L1T3_1_A.png

    UltraLite-mk4 THD+N vs Frequency USB input Bal output - Left Channel
    G2_L1T5_1_A.png

    UltraLite-mk4 Residual Noise USB input Bal output - Left Channel
    FFT_2_L1T16_1_A.png

    Complete USB input Bal output analysis report pdf attached
     

    Attached Files:

    Last edited: Jun 5, 2020
  4. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    spdif input SE output via TS on TRS outputs

    A curious peak appears at 15 KHz on SE outputs only, though it is down 140 dB or more.

    UltraLite-mk4 Dynamic Range spdif input SE output
    20200604 UltraLite-mk4 dynamic range FFT spdif SE.png

    UltraLite-mk4 A04 THD+N THD nth-HD FFT spdif input SE output
    20200604 UltraLite-mk4 A04 THD+N THD nth-HD 4+HD+N 60Hz FFT -1dBFS - spdif SE.png

    UltraLite-mk4 Frequency Response y-axis highly zoomed FFT spdif input SE output
    20200604 UltraLite-mk4 frequency response zoom FFT spdif SE.png

    UltraLite-mk4 50 + 7000 Hz spdif input SE output - Left Channel
    FFT_2_L1T13_1_A.png

    UltraLite-mk4 Gain Linearity spdif input SE output - Left Channel
    G2_L1T3_1_A.png

    UltraLite-mk4 THD+N vs Frequency spdif input SE output - Left Channel
    G2_L1T5_1_A.png

    UltraLite-mk4 Residual Noise SE spdif input SE output - Left Channel
    FFT_2_L1T16_1_A.png

    Complete spdif input SE output analysis report pdf attached
     

    Attached Files:

    Last edited: Jun 5, 2020
  5. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    USB input SE output via TS on TRS outputs

    UltraLite-mk4 Dynamic Range USB input SE output
    20200604 UltraLite-mk4 dynamic range FFT WDM SE.png

    UltraLite-mk4 A04 THD+N THD nth-HD FFT USB input SE output
    20200604 UltraLite-mk4 A04 THD+N THD nth-HD 4+HD+N 60Hz FFT -1dBFS - WDM SE.png

    UltraLite-mk4 Frequency Response y-axis highly zoomed FFT USB input SE output
    20200604 UltraLite-mk4 frequency response zoom FFT WDM SE.png

    UltraLite-mk4 50 + 7000 Hz USB input SE output - Left Channel
    FFT_2_L1T13_1_A.png

    UltraLite-mk4 Gain Linearity USB input SE output - Left Channel
    G2_L1T3_1_A.png

    UltraLite-mk4 THD+N vs Frequency USB input SE output - Left Channel
    G2_L1T5_1_A.png

    UltraLite-mk4 Residual Noise USB input SE output - Left Channel
    FFT_2_L1T16_1_A.png

    Complete USB input SE output analysis report pdf attached
     

    Attached Files:

    Last edited: Jun 5, 2020
  6. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    Square wave response, Distortion 0 to -1 dBFS (Tech notes part 1)

    Square wave response

    Measurement conditions
    1. 20Hz square wave except as noted
    2. 44.1 KHz sample rate
    3. spdif input
    4. Bal output to Cal Test CT2593-2 balanced probe to oscilloscope

    DAC anti-alias filter ringing can clip on fast transition signals. Leading edges of percussive sounds often have transitions that are nearly that of a square wave which will stimulate filter ringing. All OS DACs have this trait. Most DAC manufacturers gain stage the DAC such that a 0 dBFS sine is just below the power supply limits and thus doesn't clip. When the filter rings, clipping occurs. Using digital attenuation can avoid this unwanted signal modification. How much attenuation is dependent on the filter design.

    UltraLite-mk4 100uS/div 0 dBFS
    20200531-20 UltraLite-mk4 20 Hz sqr 0 dBFS 20 Vpp 100uS div - spdif BAL - annotated.png
    Note the flat top circled above. Filter ringing on output hitting power supply limits.

    UltraLite-mk4 100uS/div -1 dBFS
    20200531-21 UltraLite-mk4 20 Hz sqr -1 dBFS 17_8 Vpp 100uS div - spdif BAL.PNG
    Filter ringing now below power supply limits and not clipped.

    Distortion 0 to -1 dBFS (Tech notes part 1)

    The UltraLite-mk4 exhibits an odd behavior above -1 dBFS. THD+N rises noticably and in the IMD measurement additional FFT components near 19+20 KHz appear. IMD measurement is unchanged which demonstrates again why it is so important to look at the FFT, not just the numerical results. This odd behavior occurs on both spdif and USB. spdif examples shown below.

    UltraLite-mk4 A04 THD+N THD nth-HD FFT 0 dBFS spdif
    20200531-01 UltraLite-mk4 THD+N FFT 0_0 dBFS - spdif Bal.png
    Scope trace appears normal though THD+N is approx 0.126%, FFT additional components are above -65 dBFS

    UltraLite-mk4 A04 THD+N THD nth-HD FFT -0.2 dBFS spdif
    20200531-02 UltraLite-mk4 THD+N FFT -0_2 dBFS - spdif Bal.png
    THD+N has dropped to approx 0.01%, FFT additional components are below -85 dBFS

    UltraLite-mk4 A04 THD+N THD nth-HD FFT -0.5 dBFS spdif
    20200531-03 UltraLite-mk4 THD+N FFT -0_5 dBFS - spdif Bal.png
    THD+N drops to 0.00047% and FFT indicates normal harmonic behavior with 3rd harmonic at -107 dBFS

    UltraLite-mk4 A04 THD+N THD nth-HD FFT -1 dBFS spdif
    20200531-04 UltraLite-mk4 THD+N FFT -1_0 dBFS - spdif Bal.png
    Essentially same performance as -0.5 dBFS generator setting above
    This is a convenient internal trim available on the UltraLite-mk4

    UltraLite-mk4 19+20 KHz IMD FFT spdif trim set to 0 dBFS
    FFT_2_L1T15_1_A_0dB_trim.png
    Observe all the additional FFT components near the stimulus frequencies

    UltraLite-mk4 19+20 KHz IMD FFT spdif trim set to -1 dBFS
    FFT_2_L1T15_1_A.png
    With output trim set to -1 dBFS most of the additional components vanish.
     
    Last edited: Jun 5, 2020
  7. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    Distortion (Tech Notes part 2) and Bit Resolution

    Effect of spdif transformer isolation on Gain Linearity

    Another peculiarity is an spdif input sensitivity but only when outputs are used SE, and then only on the Left channel. This does not occur with Bal output usage. dScope spdif BNC output is typically 600R shield to chassis ground.

    UltraLite-mk4 Gain Linearity spdif input SE output - Left Channel
    No spdif isolation transformer
    G2_L1T3_1_A_no_spdif_xfmr.png
    There is clearly some interference influencing the measurement. Several hours of experimentation were expended to determine root cause of interference.

    UltraLite-mk4 Gain Linearity spdif input SE output - Left Channel
    With spdif isolation transformer
    G2_L1T3_1_A.png

    UltraLite-mk4 Gain Linearity spdif input SE output - Right Channel
    No spdif isolation transformer
    G2_L1T3_1_B_no_spdif_xfmr.png

    UltraLite-mk4 Gain Linearity spdif input SE output - Right Channel
    With spdif isolation transformer
    G2_L1T3_1_B.png
    Right channel is not affected as severely as the Left channel

    Bit Resolution

    UltraLite-mk4 THD+N FFT -90 dBFS - spdif Bal
    20200531-06 UltraLite-mk4 THD+N FFT -90 dBFS - spdif Bal.png

    UltraLite-mk4 THD+N FFT -100 dBFS - spdif Bal
    20200531-07 UltraLite-mk4 THD+N FFT -100 dBFS - spdif Bal.png

    UltraLite-mk4 THD+N FFT -110 dBFS - spdif Bal
    20200531-08 UltraLite-mk4 THD+N FFT -110 dBFS - spdif Bal.png

    UltraLite-mk4 THD+N FFT -120 dBFS - spdif Bal - without spdif isolation transformer
    20200531-09 UltraLite-mk4 THD+N FFT -120 dBFS - spdif Bal.png

    -120 dBFS sine is ragged, but remember, this is -120 dBFS without the spdif isolation transformer
    See comparisons to Matrix XSP below
    Holo KTE Spring2, Matrix XSP, Soekris dac1541 measure better at -120 dBFS for reference.
     
    Last edited: Jun 5, 2020
  8. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    Comparison with Matrix XSP

    UltraLite-mk4 dynamic range spdif input Bal output
    20200523 UltraLite-mk4 dynamic range FFT spdif Bal.png

    Matrix XSP dynamic range AES input Bal output
    20200508 Matrix XSP dynamic range FFT AES Bal.png

    UltraLite-mk4 A04 THD+N THD nth-HD FFT spdif input Bal output
    20200523 UltraLite-mk4 A04 THD+N THD nth-HD 4+HD+N 60Hz FFT -1 dBFS - spdif Bal.png

    Matrix XSP A04 THD+N THD nth-HD FFT AES input Bal output
    20200508 Matrix XSP A04 THD+N THD nth-HD 4+HD+N 60Hz FFT 0dBFS - AES Bal.png

    UltraLite-mk4 Gain Linearity spdif input Bal output - Left Channel
    G2_L1T3_1_A.png

    Matrix XSP Gain Linearity AES input Bal output - Left Channel
    G2_L1T3_1_A.png

    UltraLite-mk4 Residual Noise spdif input Bal output - Left Channel
    FFT_2_L1T16_1_A.png

    Matrix XSP Residual Noise AES input Bal output - Left Channel
    FFT_2_L1T16_1_A.png

    UltraLite-mk4 sine 1 KHz -120 dBFS - spdif input Bal output - with spdif isolation transformer
    FFT_0_L1T18_1.png

    Matrix XSP sine 1 KHz -120 dBFS - AES input Bal output
    FFT_0_L1T18_1.png

    Matrix XSP has some measurable advantages but UltraLite-mk4 is very close.
     
    Last edited: Jun 5, 2020
  9. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    reserved 9/10
    additional data
     
  10. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    reserved 10/10
    corrections
     
  11. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Was the 0dbFS distortion issue specific to USB? Does it also occur with the SPDIF and optical inputs?
     
  12. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Anyway, I was going to say that I will experience this particular 0dbFS distortion behavior when using WDM or DirectSound. Windows messes with the bits. It's probably rounding error due to a mixup ot data types on the part of the developer(s). It's Microsoft so we shouldn't be surprised. There's so much garbage code on top of garbage with their operating systems.

    upload_2020-6-6_10-35-28.png
    upload_2020-6-6_10-37-30.png


    upload_2020-6-6_10-31-2.png
    upload_2020-6-6_10-30-10.png

    ASIO can be super iffy as well if it's a wrapper and not a native driver.
     
  13. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Now back to WDM / DirectSound. Even if we use the mixer and attenuate the input by -6db, the signal still exhibits those particular behaviors:

    upload_2020-6-6_11-15-42.png

    Output sent to WDM. -6db applied to input signal in the mixer.
    upload_2020-6-6_11-16-22.png

    Moral of the story: Don't use Windows WDM / DirectSound. Try to bypass Windows sound stack as much as possible.
     
  14. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    @atomicbob mentioned it was on both USB and SPDIF. FWIW I am not getting the power supply clipped squarewaves on my MOTU M4.
    I was very suprised to see more SNR than specified, so I measured my MOTU M4 as loopback. Somehow I also got slightly less noise than specified, though I'd want to recheck that. Specified as 120 dBA, I got around 125 dBA or 121 dB unweighted with 44.1 kHz sampling rate. About 116 dB unweighted with 192 kHz SR.
    I wonder how this 8-channel Cirrus ADC compares to the AKM velvet sound one used in the M4. Specified SNR is the same as the one in the Focusrite 2i2.
     
  15. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I think I found something for SPDIF. Try turning down the trim. I think there could be some math errors with both WDM and in the Motu. Need more time to investigate.

    upload_2020-6-6_11-30-45.png
     
  16. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,465
    Trophy Points:
    113
    Location:
    Winnipeg
    When I was shopping around, I was looking at the MOTU but the shop I was speaking with said MOTU was not worth it on Windows machines so I went with RME instead.
     
  17. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    All distortion presented in post #6 above used spdif input. This distortion anomaly is independent of the WDM issue which I have covered in this post .

    The distortion anomaly also appears using USB ASIO.
     
  18. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    Note that I determined setting UltraLite-mk4 output trim to -1 dBFS solves the issue for all digital input modes.
     
  19. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Oops. Some slightly bad math from Motu. Someone here should let them know so they can fix it in their next firmware update. Maybe it's an intersample over.
     
  20. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    Definitely seems to be between the digital patchbay and es9016.
    Here is a spdif loopback test and it is clean to -185 dBFS.

    Digital patchbay routing map
    20200606 UltraLite-mk4 spdif loopback routing.png

    Loopback test 0 dBFS
    20200606 UltraLite-mk4 spdif loopback 0dBFS FFT 200 dB scale.png
     

Share This Page