DDC warm up time technical measurements

Discussion in 'Portable and Other Gear Measurements' started by atomicbob, Mar 30, 2017.

  1. atomicbob

    atomicbob Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,186
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    On planet
    Clocks require a certain amount of time to reach thermal equilibrium. During this time jitter and resultant DAC residual noise is changing. Until stable we should reserve judgement on sound quality of DDC + DAC combinations.

    Here is the first of several DDC jitter + noise floor measurement installments.

    The DDC data presented were collected as follows:

    1. PrismSound dScope III
    2. Tecnec 75 ohm digital cable for SPDIF
    3. Lifatec optical cable for Toslink
    4. unbalanced output RCA from Modi MB
    5. 100 Kohm load used for DAC measurements
    6. 44 KHz sample rate, 24 bit depth
    7. USB input – Audioquest Forest
    8. Unbalance cable DH Labs Silver Sonic Air Matrix RCA

    Graph interpretation:
    1. Lower absolute level of extraneous spikes is better
    2. Lower overall level of extraneous spikes is better
    3. Fewer extraneous spikes is better

    In balancing the goodness factors above the three DDCs perfomrance rank as follows
    1. Singxer SU-1
    2. Breeze Audio DU-U8 (very close to SU-1) high performance / price ratio
    3. Gustard U12 - distant 3rd place

    Animated gifs depicting the jitter spectrum changing over time:

    Singxer SU-1 - requires approximately 90 to 120 minutes to reach thermal equilibrium

    https://gfycat.com/AdvancedLittleGrassspider

    Gustard U12 - requires approximately 60 to 90 minute to reach thermal equilibrium
    https://gfycat.com/NaiveParchedDodo

    Breeze Audio DU-U8
    *coming soon* - measurements from cold start yet to be completed

    Static graphs:


    SU-1 jitter spectrum at 0 mins:
    20170327 Modi MB inferred jitter - SPDIF - SU-1 - 00 min.png

    SU-1 jitter spectrum at 90 mins:
    20170327 Modi MB inferred jitter - SPDIF - SU-1 - 90 min.png


    U12 jitter spectrum at 0 min:
    20170327 Modi MB inferred jitter - SPDIF - U12 - 00 min.png

    U12 jitter spectrum at 90 min:
    20170327 Modi MB inferred jitter - SPDIF - U12 - 90 min.png

    DU-U8 jitter spectrum at 90 min:
    20170327 Modi MB inferred jitter - SPDIF - DU-U8.png
     
    Last edited: Mar 31, 2017
    ultrabike, Gemini, Dino and 10 others like this.

  2. atomicbob

    atomicbob Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,186
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    On planet
    reserved 2/5
     
  3. atomicbob

    atomicbob Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,186
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    On planet
    reserved 3/5
     
  4. atomicbob

    atomicbob Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,186
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    On planet
    reserved 4/5
     
  5. atomicbob

    atomicbob Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,186
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    On planet
    reserved 5/5
     
  6. Grahad2

    Grahad2 Friend

    Friend
    Joined:
    Jan 26, 2017
    Likes Received:
    626
    Dislikes Received:
    2
    Trophy Points:
    93
    Location:
    Singapore
    If you're using GIMP, there's a one click optimize button that works pretty well.
     
    atomicbob likes this.
  7. Azteca

    Azteca Friend

    Friend
    Joined:
    Sep 26, 2015
    Likes Received:
    1,008
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    Cincinnati, Ohio
    Might I suggest you upload to https://gfycat.com or imgur.com? Gfycat in particular is known for hosting large and high quality GIFs.

    Thanks, as always, for your work. Looking forward to it. So far it seems the changes are "different" rather than "better" from a strictly visual perspective. Perhaps that speaks to the quality of the products. But no matter what we want consistency so it is important to know when we reach a steady state.
     
    atomicbob likes this.
  8. atomicbob

    atomicbob Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,186
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    On planet
    Thank-you for the suggestions. First post has been cleaned and updated with gfycat links to animated DDC jitter spectrum warm-up. Looks like a jitter dance.
     
  9. Rockin_Zombie

    Rockin_Zombie Almost "Made"

    Joined:
    Jan 2, 2017
    Likes Received:
    121
    Dislikes Received:
    1
    Trophy Points:
    43
    Location:
    Toronto
    Does the SU-1 spectrum look slightly worse at 90 minutes when compared to zero? The two peaks between 8k and 10k specifically.
     
  10. atomicbob

    atomicbob Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,186
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    On planet
    At first glance it would appear that way. But look more closely at the group of spikes around the 11025 Hz sine. There are more of them and they are not static. Watch the gif animations and you will see those spikes begin in the cluster and then move out. Their amplitudes are difficult to assess accurately during initial warm up. A good example are the two spikes around 8 KHz and again at 14 KHz. Note the flat tops. The average of 16 long FFTs are smeared as those spikes shift frequency.
     
    Rockin_Zombie likes this.
  11. Grahad2

    Grahad2 Friend

    Friend
    Joined:
    Jan 26, 2017
    Likes Received:
    626
    Dislikes Received:
    2
    Trophy Points:
    93
    Location:
    Singapore
    Combining their 90 min spectra, seems like the DU-U8 actually has the lowest peaks:

    So it seems SU-1 ~= DU-U8 > U12

    [​IMG]
     
  12. atomicbob

    atomicbob Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,186
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    On planet
    That is one way to look at it. Here is another.

    Think of the spikes with respect to the 11025 Hz carrier, in particular those more than 300 Hz away.

    Breeze DU-U8:
    DU-U8 overall level.png

    DU-U8 has a cluster of spikes between 550 ~ 800 Hz and again at 2300 and 2800 Hz above the carrier.

    Singxer SU-1:
    SU-1 overall level.png

    SU-1 has strong but fewer spikes at 450 and 2600 Hz above the carrier.

    Now think about the ISO Contours of Equal Loudness (updated Fletcher-Munson) curves:
    iso226-2003 Equal Loudness Contours v2.png

    From this viewpoint the SU-1 appears slightly more clean than the DU-U8. Again, the two are very close. Pick your flavor of blackground.
     
    The Alchemist, EricR, batriq and 7 others like this.

Share This Page