NwAvGuy's Dream Cums Tru: Schiit Magni 3 Heresy

Discussion in 'Headphone Amplifier Measurements' started by purr1n, Nov 22, 2019.

  1. mscott58

    mscott58 Friend

    Friend
    Joined:
    Oct 13, 2015
    Likes Received:
    1,892
    Dislikes Received:
    8
    Trophy Points:
    113
    Location:
    Philly - Yo
    One of the problems is that when a sane/educated/experienced person like @purr1n tries to highlight issues such as the inconsistencies in their "science" the reactionary zealots translate it as SBAF as a whole not focusing on science at all. "Science" is a badge they appear to carry with pride without really knowing what it means - which reminds me of a certain quote from an awesome movie...

    [​IMG]
     
  2. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    11,185
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    On planet
    If I could work my will, the statement above would appear on the virtual equivalent of a billboard at all audiophile sites. As everyone should know by now, I loathe the reduction of complex measurement phenomena to simplistic numbers, a gross misrepresentation of actual performance. But the simple-minded fixate on numbers for comparisons and regurgitate bullshit masquerading as "The Trvth".

    I hadn't heard of the NoAmp plugin before I googled NoAmp to investigate a possible real device of which I might not be aware when @ultrabike humorously used the term. I'm afraid I haven't anything useful to offer on that topic.

    I am on record as loathing single numbers for use in attempting to categorize component performance. In many AES presentations Jonathan Novick (Avermetrics, formerly from Audio Precision) stated "Not all that matters can be measured. Not all that can be measured matters." I'm a firm believer in this philosophy given the present state of measurements. Still I measure in the journey towards reconciling measurements with perceived performance.
     
    Sherm, Psalmanazar, Joshvar and 7 others like this.
  3. elmoe

    elmoe Friend

    Friend
    Joined:
    Apr 23, 2019
    Likes Received:
    1,023
    Dislikes Received:
    104
    Trophy Points:
    93
    Location:
    Phoenix, AZ
    We should pit purrin vs Amir celebrity deathmatch style. The ring could be neutral ground like a thread on HF with Currawong as referee.

    Jude can be the waterboy.
     
    Sherm, pure5152, The Life and 2 others like this.
  4. redrich2000

    redrich2000 Almost "Made"

    Contributor
    Joined:
    Oct 28, 2017
    Likes Received:
    120
    Dislikes Received:
    0
    Trophy Points:
    43
    Location:
    Sydney, Australia
    This is what infuriates me about them. Their claim to science is pretty weak when it's one unqualified guy self-publishing with no peer review and non-transparent and arbitrary methodology and equipment.

    I want to start a website called Audio Temperature Review on which I measure the ambient temperature of audio equipment and argue that cooler is better. If any doubts me I'll just assume they hate science.
     
  5. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Friend MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,765
    Dislikes Received:
    16
    Trophy Points:
    113
    Location:
    Irvine CA
    Why hell, a grossly inefficient HE6 will put out a wopping 102 dB SPL with 4 Vrms.
     
  6. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    53,250
    Dislikes Received:
    89
    Trophy Points:
    113
    Location:
    Antarctica
    Rough guesses as to Vout for typical listening level peaks for a few headphones and IEMs. This illustrates why the concept of SINAD as a "one ring that rules them all" for headphone amplifiers makes no sense.
    upload_2019-11-26_21-17-40.png
     
    Psalmanazar, Don Miller and ultrabike like this.
  7. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Friend MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,765
    Dislikes Received:
    16
    Trophy Points:
    113
    Location:
    Irvine CA
    To me SINAD makes sense. But it's something that can be easily miss-used w/o proper context, and indeed useless w/o context at all.
     
  8. Thad E Ginathom

    Thad E Ginathom Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    6,734
    Dislikes Received:
    11
    Trophy Points:
    113
    Location:
    India
    NOOOooooooo!

    It doesn't matter how wrong Amir is, he will never shutup. Avoid. Do not attempt to engage, converse or interact in any way whatsoever.

    This is how SBAF avoided the Amircurse. Unlike Hydrogenaudio, where the late, lamented Arnie and co tried to pick him to pieces (Hydrogenaudio just can't resist a troll!), and it went on for dozens of pages, SBAF took the attitude, Oh, it's you: we're not interested. No conversation is the best conversation with Amir.

    Long may it remain that way.
     
    Sherm, Riotvan, tommytakis and 9 others like this.
  9. Lyander

    Lyander Too sensitive for SBAF

    Friend
    Joined:
    Feb 25, 2017
    Likes Received:
    4,901
    Dislikes Received:
    24
    Trophy Points:
    113
    Location:
    Pasig, Philippines
    Either your excellent Goldpoint or something similar in function, like a Schiit Sys? Please tell me it isn't cables :p

    Who is this person Thad is talking about? Why are we discussing them?
     
  10. Thad E Ginathom

    Thad E Ginathom Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    6,734
    Dislikes Received:
    11
    Trophy Points:
    113
    Location:
    India
    Don't ask. The curse is strong. I'm worried that I'm even reading about ASR.

    (Or are you doing that clever thing where you just look past someone and deny their very existence? Powerful stuff, that.)

    Back to the amplifier...
     
    Lyander likes this.
  11. wadec22

    wadec22 Acquaintance

    Contributor
    Joined:
    Sep 28, 2018
    Likes Received:
    58
    Dislikes Received:
    0
    Trophy Points:
    18
    Location:
    USA
    I don't think he needs to go away. I think he needs;
    • a very specific test/check/procedure list that is performed with each unit
    • a reference post with each test not only scientifically explained, but given real-world context and why it should or should not be audible (tossing out tests that are irrelevant)
    • better balance his testing between what he is sent from users and free chi-fi product. the latter has become too much a focus, simply b/c those companies are exploiting/leveraging his page views to advertise their new products
     
    Sherm, obsiCO, ultrabike and 3 others like this.
  12. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    53,250
    Dislikes Received:
    89
    Trophy Points:
    113
    Location:
    Antarctica
    You guys want to know why Chi-Fi products measure so well? They use cookie-cutter designs based on chips. It's not hard to do when the chip company themselves provide evaluation boards. There's some Kung-Fu required in board layout, but it isn't anything like layout for computers or network communications gear that runs in the GHz region.

    The real world aspect and approach to proper visualization and presentation is a big one. I will address what I mean in a bit.

    I would add another one for ASR: blind tests in regards to measurements. Have someone else perform a consistent standard suite of relevant measurements. ASR then provides commentary and color without knowing what the product is.
     
  13. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    53,250
    Dislikes Received:
    89
    Trophy Points:
    113
    Location:
    Antarctica
    There, fixed it with real-world context.
    upload_2019-11-27_11-12-1.png

    Maybe too aggressive, but this should illustrate the point. A fuzzy zone between 60-80db would be more appropriate. I don't have Photoshop on this PC.
     
    ultrabike likes this.
  14. Serious

    Serious Inquisitive Frequency Response Plot

    Friend BWC MZR Executed
    Joined:
    Sep 28, 2015
    Likes Received:
    1,709
    Dislikes Received:
    0
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Come on Marv, 30dbA environmental noise might be normal in noisy parts of NYC, but I don't think too many people live in such noisy areas. I doubt your area is that noisy. The kind of things I can hear when I don't have my music running at night are...
    • My breathing and heartbeat
    • Occasionally a car or a train
    • In the winter: The oil heater in the basement
    • Maybe coil whine from charging my phone (charger is behind a pillow)
    • The power supply in my LCD and CRT monitors when they're turned on
    • My plasma TV when it's on
    • The noise the transformer in my LED lamp makes when it's on
    • etc
    I don't have a mic that has such low self noise that I can say how quiet it is here, but I doubt it's louder than 10dbA. I've done hearing threshold tests in my room before and I also think that the bit-depth tests and 30kHz sine-tone tests I wrote about some time ago wouldn't have been possible in a noisy room.
    I'm not just saying this because I think my room is super quiet, I just don't really buy that 30dbA environmental is a best case scenario like you've said before.
     
    Psalmanazar likes this.
  15. wadec22

    wadec22 Acquaintance

    Contributor
    Joined:
    Sep 28, 2018
    Likes Received:
    58
    Dislikes Received:
    0
    Trophy Points:
    18
    Location:
    USA
    Exactly. Your post here helped me start to get it; https://www.superbestaudiofriends.o...ically-incorrect-audio-blog.7261/#post-249242

    Then another forum member over there actually had a not so scientific example that helped me better understand as well. He recorded pink noise at some variable levels in separate recordings. I think it was like -90db, -80, -70.... up to -40 (I get it there are probably flaws in his recording setup but bare with me).

    Setting volume on my amplifier to my typical listening levels, unplug the fridge in my office, set my pc fans to "silent" and plug in my most sensitive headphones I was barely able to detect the -70db recording. If I cranked the amp to unbearable levels, I could pick up the -90.

    Not super scientific, but it reiterates your point. -80db or lower is going to be good the vast majority of scenarios.
     
  16. Lyander

    Lyander Too sensitive for SBAF

    Friend
    Joined:
    Feb 25, 2017
    Likes Received:
    4,901
    Dislikes Received:
    24
    Trophy Points:
    113
    Location:
    Pasig, Philippines
    @Serious I was tempted to dislike your post but would hate to break your streak.

    Just going by relative population densities and how that correlates with environmental noise, I'd say your logic there could maaaaaaaaybe use some tweaking ;)

    Your setting, if it genuinely is that quiet, is a wee bit of a massive fucking outlier. I've been out in the boondocks (this word, interestingly, stems from the Filipino word "bundok", which translates to "mountain") and you'd be surprised how much noise nature itself makes. Of course I have no idea how quiet things get when it's snowing because I imagine that loose and packed snow both absorb a fair bit of noise, but again, I don't quite buy that 30dBA is quite so outrageous. Hell dude, busy parts of NYC I imagine would be worse than what counts for a busy street here, and that's 60dBA+ easy, maybe 40 indoors.

    Edit:

    Was curious and did a quick search. Study for S. Korean residential spaces, looks about right: https://www.sciencedirect.com/science/article/pii/S0003682X17300403

    "Jeon et al. [12] measured noise levels while the apartment was empty and the windows were closed. Lai et al. [13] measured the noise levels for 15 min in 32 residential apartments and the average noise levels for 15 min were found to be 67.1 dBA with a variation from 52 to 77.9 dBA. Noise levels for one hour were also measured in urban residential buildings under a natural ventilation condition [14]. Similarly, Pujol et al. [7] measured the noise levels in bedrooms and the main rooms to analyse children’s exposure to environmental noise at home. They found the averages of noise levels for day, evening, and night in 44 dwellings were 51.3, 53.6, and 36.9 dBA, respectively."
     
    Last edited: Nov 27, 2019
  17. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    53,250
    Dislikes Received:
    89
    Trophy Points:
    113
    Location:
    Antarctica
    10db is the threshold of hearing for most people. 20db is about a super insulated mix bay. Most small treated edit bays are about 30db SPL ambient.

    My house is in a suburban area of Greater El Lay County. It's not a rape dungeon. In the quietest room, I don't get anywhere near the worst insulated bay at any of the production sites at work.
     
  18. julian67

    julian67 Friend

    Friend
    Joined:
    Oct 5, 2015
    Likes Received:
    245
    Dislikes Received:
    22
    Trophy Points:
    43
    Location:
    England (Proper England - Vilayet).
    They often measure horribly. Cases in point from ASR:
    https://www.audiosciencereview.com/...-lp-2020a-stereo-audio-amplifier-reivew.9806/
    https://www.audiosciencereview.com/...-measurements-of-pifi-dac-pi-sound-card.8614/

    I just picked those, out of many, as they are both extremely cheap items that have seen a lot of overly positive reviews and comments in different places. They do apparently sound how they measure i.e. not great or even good.

    I know ASR gets a lot of heat here but step back a little from the personal frictions and so on and his measurements are actually very useful for identifying some really horrible products that get/got a pass elsewhere. I wouldn't claim that you can tell everything important about an audio product by its numbers (you do have to use it and listen too!), nor even that you can always discover or describe audible deficiencies with measurements, but lousy measurements can certainly rule out the junk and also often diagnose design and manufacturing foobars.

    I've owned fotm things which measured great but had audible problems (because who tests a DAP and measures the icky noise *between* tracks or the hideous pops on power up/down!?), and fotm things with hype behind them but no numbers which I thought were crap, but I don't recall owning anything which measured like shit but was a pleasure to own or use.
     
  19. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    53,250
    Dislikes Received:
    89
    Trophy Points:
    113
    Location:
    Antarctica
    First of all, did you perform blind tests on the linked products? The amp is borderline, but the DAC isn't (seeing from measurements). The only reason they measured like shit is because someone there in their commentary TOLD YOU they measured like shit. Personally, I doubt you've even gone to the step of determining your own standards for what constitutes a bad measurement.

    Valhalla 2 is a product that measured like shit on ASR. So did Yggdrasil. And I've done numerous triple blind tests on the Schiit DACs to 0.000 stuff like X-Sabre Pro and 0.00 stuff like Convert-2 so I know ai can tell the difference.
     
    ultrabike likes this.
  20. wadec22

    wadec22 Acquaintance

    Contributor
    Joined:
    Sep 28, 2018
    Likes Received:
    58
    Dislikes Received:
    0
    Trophy Points:
    18
    Location:
    USA
    I agree that he's finding some useful information through testing and if you are ready with dual screens and some coffee, you can make some useful comparisons between products.

    My frustration is more about the lack of strict testing guidelines, adherence and context.

    Some examples off the top of my head;
    He was overly critical of the Monoprice 788 desktop measurements because the internal amplifier did not test the way he wanted it to - so he completely wrote off the DAC. He also failed to properly communicate how most of it's short comings he was measuring were inaudible.

    Also the Topping D50s - he loved the thing and gave it much praise. He also did not initially test it for THD+N vs FR - a test he usually runs. When asked to test it later, he did and posted the results several pages back where it was requested. The D50s showed clear problems on that graph (indicating to me that it would cause fatigue with high frequencies). When asked to put the results on the front review page, he did so but was dismissive of the results and would not discuss them. It was a product he wanted to do well.

    Same thing with one of the Topping DX7 units - one of the variants, I can't recall. He initially didn't test it's output impedance (or did but did not post it). Later on he did and it tested with a high output impedance, something he is normally extremely critical about. Seems odd when again, he loves the Topping product.

    I believe his personal bias (again, maybe even subconsciously) is weighing into what he is doing with certain products. His hardware and time invested could be much more valuable if he reset and tackled things a bit differently.
     
    Joshvar, crenca, BenjaminBore and 4 others like this.

Share This Page