STAX SR-202/SRM-212 Measurements

Discussion in 'Audio Science' started by Serious, Dec 28, 2016.

  1. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Subjective impressions here: http://www.superbestaudiofriends.org/index.php?threads/stax-sr-202-srm-212-loaner-impressions.3612/
    This will go back to @sorrodje tomorrow and I wanted to get the measurements out before I I ship it away. Compare to Marvey's SR-207 measurements: http://www.superbestaudiofriends.or...170-sr-207-earphone-system-measurements.3595/
    My measurement rig is detailed here: http://www.superbestaudiofriends.or...nt-setups-post-your-rig.141/page-7#post-50142

    Frequency response (coupler):
    FR coupler.png
    As expected the coupler doesn't seal with the earpads, but that's not the point for this measurement. In terms of seal the in-ear measurements are much more accurate anyway. With this measurement we already get a hint at the 1.5kHz bump and treble ringing. BTW: Fantastic driver matching.

    CSD (35db range, right channel):
    SR-202 R CSD.png
    A resonance around 8kHz. Tiny ridges around 3, 3.4 and 4.2kHz. The left channel didn't look too different. Those tiny ridges don't look bad, but they always seem to be there with the STAX. There are actually a lot more when you drop the noise floor. This seems to be mainly what limits estats from getting as clean CSDs as something like the HD800. I think part of this is the driver itself and part of it could be reflections from the grille on the outside.

    No coupler measurement:
    This is with the mic just infront of the driver, about where the earpad ends. Maybe 1cm distance to the driver:
    SR-202 Open Air FR.png
    Fs L is about 196Hz, Fs R is about 199Hz. The difference is slight, but noticeable when blowing on the driver. The overall FR tendencies stay the same versus the coupler measurement.

    Ear canal opening mic measurements:
    SR-202 IE FR.png
    This is about as good a seal as I can get on my head. I wear glasses, but the glasses didn't seem to influence the result at all. I can only get flat to 20Hz with some pressure.
    The general FR now becomes too tilted upwards after 1kHz. As I said in the SR-207 thread I like to EQ everything above 1kHz down by about 4db. The 9kHz spot actually sounds worse than it looks, probably because of the huge dip just below it.

    Distortion at 95db, 500Hz (Left channel):
    SR202L 95db.png
    Note that I get a lot of 50Hz + multiples hum which shows up in the distortion results. This is also why I limited it to 30Hz, instead of 20Hz. There's a huge 25Hz 2nd order spike (you can see it rising already) because of the hum.
    Using the harmonic frequency as reference, which is why I get a lot of peaks after 2kHz. That's just the SNR varying with the FR. Distortion should stay similar above 2kHz. I generally get slightly more 2nd order distortion than the HD800. Distortion is low but not incredibly so.
     
    Last edited: Jan 8, 2017
  2. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Reserved for more experimental stuff.

    My EQ settings:
    SR202 EQ1.jpg

    Coupler-In-Ear measurement difference:
    SR202R coupler target.png
    This plot is basically an inverse of the effect the ear has on the FR compared to the coupler measurement. Because I think that flat on the in-ear measurements sounds flat, I like to use these as a way of looking at coupler FR targets for different headphones. Here you can also clearly see that I get a much better seal on the head than on the coupler. You can also see that at 500Hz the in-ear measurements are about 2db louder than the coupler. This is less than with other headphones (4.9db for the HD800).

    Level linearity:
    Level Linearity 96+.png
    I started at the highest volume it could take (which was only about 109db on my coupler) and then lowered the volume gradually until it was linear. This seemed to be around 96db. Here you can see the differences compared to 96db. For some reason it seems to become more efficient with higher excursions, but also loses some treble. I'm not sure how much of this is the SR-202 and how much of it the SRM-212.

    Subharmonics:
    200Hz 100db.png
    I tried to measure the first subharmonic. At about 100db at 200Hz, we get about 0.04%. Not bad, but dynamic drivers don't have this problem. The first subharmonic seems to be similar in level to the 3rd order distortion.

    High Bandwidth CSD (R channel):
    SR-202 R CSD100kHz.png
    This is a CSD that goes from 500Hz to 100kHz. My mic compensation is flat from about 40kHz and I only compensate for the DAC/ADC rolloff from there on. Seems to go to 100kHz without breaking a sweat, but there's a peak at about 33kHz.

    Impulse response:
    SR-202 R coupler IR.png
    Note: STAX phones are wired with inverted polarity. For easier comparison to other measurements I inverted the IR.

    Here are two more tests that I did:

    100 sample (192kHz SR) 10kHz sine:
    10khz.jpg
    Top is the SR-202, bottom one is the HD800. Both on the coupler. Leckerton UHA6smkII as amp for HD800. GO450 DAC.
    The SR-202/SRM-212 combination takes longer to reach the correct level and takes longer to decay.

    100 sample (192kHz SR) 15kHz sine:
    (Same thing, but 15kHz)
    15khz.jpg
    This is where it gets really fucky. Idk WTF is going on with the SR-202, but the results always looked similar. I think it's something to do with multiple arrivals from weird motions of the diaphragm that end up completely destroying the sine-waves.

    Tell me what other sort of measurements you want to see.
     
    Last edited: Dec 28, 2016
  3. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Frequency response with HE60 (yellow) and HD600 pads (blue) (in-ear measurement):
    he60 pads.jpg
    The HD600 pads give more 1kHz, which isn't a good thing with the stats, but is a good thing with the HD600.

    Frequency response for HE60 pads with HD600 front foam (left and right channel) (in-ear):
    SR-202 HE60 pads + front foam.png
    Pretty good. Still a tad bright, but it's manageable now. The thin foam you can buy at Michael's might be a bit better than the HD600 front foam here, but overall I really like the tonality like this.
     
    Last edited: Dec 29, 2016
  4. 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
    Looks like you confirmed the difficulty of getting an optimal seal (by this I mean flatline extension down to near 20Hz) even when worn on the head. The SR-207 / SR-202 pads seem less compliant and more rigid than the pads of the SR-407 / 507. I wonder if this is the cause. The SR-207 did not sound bright to my ears - many of the other STAX do. So maybe this could have been a minor change to the SR-207 from the SR-202.
     
  5. 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
    The 100 sample 10kHz sine waves seem to be of particular interest. Let me see if I can replicate. Might explain the STAX ethernal or plasticky quality and compression @OJneg bitches about.
     
  6. anetode

    anetode Friend

    Staff Member Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    479
    Trophy Points:
    93
    Location:
    How did I get here
    Home Page:
    Might make sense to replicate the 10khz test at different test tone and sampling frequencies. I'm still surprised by the rising distortion in the treble.
     
  7. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I tried to look for a measurement that could explain it since it bothered me subjectively. Idk, hard to say if the measurements do actually explain it since the timescales are so short. With lower frequency sine-tones the results were more similar, but the HD800 seemed to consistently decay faster at every frequency I tried. The slow rise-time is suprising considering that the SR-202 seems to go way higher than the HD800 from my measurements, but I guess it's not as simple in reality.

    I forgot, but I also wanted to include what a simple DAC/ADC loop looks like with those signals. Still, the HD800 probably follows the ideal better.
    Oh, forgot to mention. That's most likely just the harmonic frequency as a reference. Let me see what it looks like on the coupler. Could also be my microphone. I think the dips in the treble screw with the SNR and make the distortion look more than it is. It's probably consistent from 2kHz on up. Also the amp starts to clip or distort more at high frequencies first, so it might be that.
     
    Last edited: Dec 29, 2016
  8. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,467
    Trophy Points:
    113
    Location:
    Winnipeg
    On those 100sample graphs, my first thought was not to do with the driver but some sort of capacitive/whatever interaction between the headphone and amp. Possibly power supply related?
     
  9. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Another look at the same distortion plots, this time as absolute levels:
    Distortion 95db.png
    I think it's highly likely that what we're seeing here is the microphone distorting or something like that. Distortion should pretty much stay at a constant level between 0.05% and 0.1% from 500Hz on up. I also don't think all of the distortion can be explained by this. The HD800 was run at the same volume level at 500Hz and still measured with lower 2nd order distortion.

    Recorded the 100 sample sine waves through the same DAC/AMP setup they went through for the phones. I also realized that the 15kHz file was 100 samples at 192kHz sample rate and the 10kHz file was at a 44.1kHz sample rate.
    perfect.jpg
    This is what happens with no normalization. Top one is the 15kHz one, the bottom one is the 10kHz one. I forgot which dac filter I'm using with the GO450. I think it's the one it starts with when you don't press a button.
     
  10. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    RE:
    I didn't mention this earlier, because I didn't know what to make of it. As mentioned in my subjective impressions, the SRM212 hums, but the hum went away when I touched the case. Touching the case however also lowered the ultrasonic region (above 30kHz IIRC) by a few db. I mainly thought that this was an electrical issue (the results above are with me touching the case). Now I just had a different idea: What if the motion of the diaphragm from the 60Hz noise pushes the diaphragm out of its unstable area (between the two resting positions) and because of this increases the efficiency for very small excursions? I'd have to recheck my results for the level-linearity at the low end to see if they both show the same thing. This is just me thinking out loud and it's probably wrong, but I thought I should mention it. I don't think the effect should make such a big difference (a couple dbs), but there's something that irks me about how planars have to work with a diaphragm of fixed length. It has to strech and compress in real-time.
     
  11. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    RE: Distortion:
    Here's what the distortion measurements look like using the fundamental as reference (like how most other plots are). I think the 95db I measured at was closer to 92db because my microphone calibration was off by 3db, but they still give you a good idea.

    SR-202:
    SR202L 95db4.png

    HD800:
    HD800L 95db4.png

    4th order and up are below the noise floor with this measurement for both headphones.

    The SR-202 has much higher 2nd order distortion. Could be the SRM-212. The SR-202 should be at an advantage because of the large driver surface area and symmetrical drive.

    From the midrange on the HD800 seems to push the capabilities of my measurement rig. I prefer to look at the bass distortion with the harmonic frequency as a reference, but my measurements are generally too noisy to show a huge difference. The HD800 bass did sound cleaner to me. I'm pretty sure the rising distortion in the treble is just my measurement rig. I think the HD800 driver has some D2 at 3kHz when turned up really loud, which could potentially matter when using it as a tweeter for speakers, but I don't think it matters much with the headphone.
     
    Last edited: Jan 21, 2017
  12. mypasswordis

    mypasswordis New

    Joined:
    Jan 20, 2017
    Likes Received:
    5
    Trophy Points:
    3
    Location:
    US
    All Stax amplifiers as far as I know are single-ended, including the SRM212, so they have more even order distortion than push pull designs. Another thing to keep in mind is the output stage is biased at a very low current as the transistors aren't heatsinked at all, severely limiting slew rate. Also if you are using the stock switching unregulated power supply it can be injecting noise, and the internal stepup transformer most likely is causing the hum.
     

Share This Page