Should amps greatly affect in-room response?

Discussion in 'Audio Science' started by moriya, May 12, 2020.

  1. moriya

    moriya Acquaintance

    Joined:
    Apr 20, 2020
    Likes Received:
    53
    Trophy Points:
    18
    Location:
    San Francisco
    So I was measuring the in-room response of my speakers (Paradigm 85f) with REW and a u-mik1 for shits and grins as part of testing a couple integrated amps to figure out which one I wanted to keep. I threw in my AVR as well for a baseline. I was expecting them to measure more-or-less the same, but was really surprised when I measured them.

    The amps were an Arcam SA20, my Denon x3500h, and a Schiit Ragnarok 2. The Denon and Schiit measured pretty much the same, although I was really surprised at exactly how close they were.
    [​IMG]

    Then on to the Arcam. It was wildly different:

    [​IMG]

    Here it is zoomed out a bit to 1/3 octave.

    [​IMG]

    Should amp affect measurements that much? It's funny because the Arcam's response squares with my understanding of the Paradigms (less energy between 2-4k), but thats not the Schiit and Denon's response at all. I'm weirded out in general that two amps would measure lock-step the same, and a third amp pretty vastly different (that room mode at 45hz is boosted by almost a full 10 dB)!
     
    Last edited: May 12, 2020
  2. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    This is EXACTLY the type of measurement I have been looking for! Just about all amps will measure flat with resistors and I have been curious what amp differences are with audio measurements.

    anyway, a lot of this might boil down to output impedance of the respective amps. Ragnarok and Denon are likely very low and would have similar interactions. Arcam seems a bit higher.
     
  3. rlow

    rlow A happy woofer

    Pyrate Contributor
    Joined:
    Jul 18, 2017
    Likes Received:
    7,735
    Trophy Points:
    113
    Location:
    Canada
    I’m not buying it. There has to be some setup difference between the SA20 and the other 2 amps. What is your chain for each amp? Does it differ at all in any way including cables? Did you buy the Arcam used and if so did you try resetting all the settings? You’re not plugged into the “phono” input are you? Need to employ Occam’s razor here - the differences are too big IMO especially in the bass. Also were you able to repeat this test multiple times going back to the other amps?
     
    Last edited: May 12, 2020
  4. Dr J

    Dr J Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    97
    Trophy Points:
    18
    My expectation that an equally capable amp would not change the room response. Equally capable = does not give up on low impedance speakers (powercube is decent), has low output impedance, measures the same in frequencey response and less than 0,1% distortion (or maybe even less than 1% distortion).

    Do not know you gear, but the difference is surprisingly large. Like rlow said, something in the setup is different. Maybe hard reset as the amp seems to be partially digital. Are you using the digital inputs and any processing?
     
  5. Dr J

    Dr J Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    97
    Trophy Points:
    18
    To cover all the bases: sure that the room or mic position did not change? I have had to re-run measurements after noticing that a cupboard or corridor door was open...
     
  6. moriya

    moriya Acquaintance

    Joined:
    Apr 20, 2020
    Likes Received:
    53
    Trophy Points:
    18
    Location:
    San Francisco
    I'll repeat it again tonight just to be sure, but I tried this several times and got the same response out of the Arcam. I was as surprised as y'all which is why I posted this - I was expecting maybe a few dB difference here and there at most, not 5-10. To answer your questions though:
    - both amps brand new
    - not plugged into phono stage
    - didn't touch the measurement mic, same location
    - same chain - macbook pro -> same generic USB cable -> modi 3 -> PYST cables -> amp -> bluejeans cables -> speakers. Not using the digital inputs on the Arcam to get apples-to-apples signal chain. For the Denon, I was using "pure direct" or whatever they call it, no processing/room correction, should be more-or-less as close as you can get to pure analog on that amp (even turns off the HDMI output and screen).
    - exact same room, no changes

    I'll have to run them again tonight, I didn't end up saving my other runs of the Arcam, but I did run the Schiit several times more later that day (I had moved the mic just a touch at this point) and got the same shape as a result (obviously small differences in volume). I just wanted to see if this was expected, and it looks like generally speaking the answer is what I would expect - "no".
     
    Last edited: May 12, 2020
  7. rlow

    rlow A happy woofer

    Pyrate Contributor
    Joined:
    Jul 18, 2017
    Likes Received:
    7,735
    Trophy Points:
    113
    Location:
    Canada
    You’re using the Paradigm 85f correct? And are you disconncting/reconnecting the PYST cables then on the amp end, and the same for the speaker cables? Make sure you’re not reversing the phase with the speaker cables as well, this could make a definite difference in bass response. In any decently powered/measuring amp, I wouldn’t expect this massive a difference...the one I would have expected to be slightly different (worse) would be the receiver, not the 2 integrateds...
     
  8. Biodegraded

    Biodegraded Friend

    Pyrate Contributor
    Joined:
    May 28, 2017
    Likes Received:
    7,975
    Trophy Points:
    113
    Location:
    Vancouver BC
    I think @Cspirou has it. Here's Stereophile's impedance profile of the Paradigm 95f (similar speaker to the 85f, which I couldn't find a measurement for; Fig. 1):

    https://www.stereophile.com/content/paradigm-prestige-95f-loudspeaker-measurements

    Stereophile measurements of the Rag 2 output impedance show it's very low across the board; into their simulated speaker load (Fig 4 here) there's almost no correspondence (Fig. 1):

    https://www.stereophile.com/content/schiit-audio-ragnarok-2-integrated-amplifier-measurements

    They don't have an SA20 measurement, but here are a couple of other Arcam amps for comparison (Fig 1 in both). Note the vertical scales are exaggerated compared to the Rag graph, but compare the numbers for the 3 amps given in the text:

    https://www.stereophile.com/content/arcam-fmj-p49-power-amplifier-measurements

    https://www.stereophile.com/content/arcam-diva-a85-integrated-amplifier-measurements

    Note also the amp measurements are in dBA, which will tend to minimize differences at the low end.

    Guessing that the 85f's low-end impedance peak is close to the room mode, thereby reinforcing it; that there's a dip rather than the 95f's peak around 2k; and that the SA20 has an impedance profile similar to the other Arcams whereas the Denon's is closer to the Rag.

    Edit: But hmm, the fluctuations do seem much bigger than they really should be given those impedance numbers...
     
    Last edited: May 12, 2020
  9. moriya

    moriya Acquaintance

    Joined:
    Apr 20, 2020
    Likes Received:
    53
    Trophy Points:
    18
    Location:
    San Francisco
    Yup, thats all correct! I'm going to be make extra sure I dont mess up the phase when I re-run the test later. Now that you mention it, the speaker terminals on the Schiit are +/- on one side, and -/+ on the other, so it's entirely possible I messed up moving from Schiit -> Arcam (which is +/- +/-), although I was pretty carefully watching that...
     
  10. moriya

    moriya Acquaintance

    Joined:
    Apr 20, 2020
    Likes Received:
    53
    Trophy Points:
    18
    Location:
    San Francisco
    Ok, just ran them again. The answer to my question is "no, it shouldn't". The answer to 'why do the graphs look that way then?' is "because you're dumb and didn't check your REW settings"

    Apparently I ran the Denon and Schiit on my left speaker only (because REW defaults to "left" for output) and the Arcam on L+R - my right speaker is unfortunately placed in a sub-optimal corner, which explains the bass hump. Here's both the Arcam and Schiit with L+R. There's differences, but they're 1-2db. Face, meet palm.


    [​IMG]
     
  11. Biodegraded

    Biodegraded Friend

    Pyrate Contributor
    Joined:
    May 28, 2017
    Likes Received:
    7,975
    Trophy Points:
    113
    Location:
    Vancouver BC
    Ok, that makes way more sense. The amps' impedance variations are way too small to account for it.
     
  12. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    Mystery solved!
     

Share This Page