1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff)

Discussion in 'Dedicated Source Components' started by purrin, Dec 5, 2013.
First
 
Back
387 388 389 390 391 392 393 394 395 396
398 399 400 401 402 403 404 405 406 407
Next
 
Last
  1. frenchbat

    Thought you had 10k in headphones. When it comes to speakers I'm not qualified sorry :) .
     
    As to the test, this is the straw that breaks the camel's back :
     
    What ? 77Ohms and they couldn't hear a difference on the HD800 ? These guys are either deaf or they don't know how to connect their set-up properly.
     
  2. jcx
    I looked at Tyll's InnerFidelity HD800 Z plot - say 350 mid audio, 650 at the ~ 100 Hz bass bump
     
    then do the divider math with 77 Ohms and I get 0.75 dB  - not immediately obvious that you can tell from Clark's ABX threshold plot for ~2 Octave and 100 Hz center (you have to visually interpolate adding to the guesswork)
     
    abx_crit.gif
     
    Clark, David L., "High-Resolution Subjective Testing Using a Double-Blind Comparator", Journal of the Audio Engineering Society, Vol. 30 No. 5, May 1982, pp. 330-338
    ABX]http://home.provide.net/~djcarlst/abx_crit.htm]ABX Amplitude vs. Frequency Matching Criteria
     
     
     
    certainly the bigger needed correction is for average level: 20*log10(350/(350+77)) = -1.7 dB
     
  3. evillamer
     
    Here's another pdf(see end of page 44 onwards) that covers the audio related implementation of zeroth-order hold, analog filters and anti-alias filters:
     
    http://www.analog.com/media/en/technical-documentation/dsp-book/dsp_book_Ch3.pdf
     
     
    I think we also have to note why analog is giving very basic information is because(quote from last page of pdf):
     
  4. evillamer
    Also it's quite intriguing on this quote from the pdf I just posted: (page 57)
     
     
    How did Schiit managed to do it with their closed form filter.. It's kind of a mystery.
     
  5. jcx
    the DAC's filter can't do anything about the mic-pe-filter-ADC chain
     
    studio practice is to capture at higher rez and they have more choices that you have no control over in the downsampling decimation/filter and wordlength reduction process - dither is also near universally used today to produce 16/44 release from the studio's internal format(s)
     
    Moffat went ballistic when several commented on dither and noise kinda making "bit perfect, dammit" claims rather curious
     
  6. prot

    Dont know about deaf but I'm pretty sure they know how to connect & test a few HW components. Dont have access to such a Realtek chip so I cannot say whether that dumb output impedance really affects the hd800. AFAIK, the hd800 works pretty well with OTL amps (e.g. bottlehead) which may have an even bigger output impedance.


    KeithEmo
    100% with you about the computer sounchips of yesteryear ... they were all quite cr*ptastic, you could hear all sorts of PC noises, even the mouse movement in some cases. However, the last gen chips & implementations seem to be a diff game ... more details in another headfi thread and from a PC forum ... and a sample lastgen implementation. There are many comments on headfi and elsewhere stating that those things sound quite a lot better than the usual MB chips of yore and it could very well be that (at least) some implementations are already transparent/inaudible.
    Times they are a changing :)

    P.S. people and biases not so much/fast apparently
     
  7. negura
     
    Out of interest what is your 10K speaker rig? Does it include a stand-alone DAC? Nothing wrong with questioning things, in general and even more so in audio. It's just unusual to see people with this level of investment having doubts about how computer chips vs good stand-alone DACs comparatively perform.
     
    Tony1110 likes this.
  8. prot

    May be quite unusual, I do not know ... personally, I just do not take anything for granted .. and even less when it comes to hifi :)

    Anyway, I'm not the one questioning DACs here, it's Tom's review. All I want to know is if anyone could find some serious issues with that review ... cause otherwise it is right and I gotta start questioning some of my audio investments ... at least some future ones. I did post some details about my setup in various threads (including a link not far above) but I'll just PM you cause there is no need to brag here.
     
  9. evillamer
    http://www.ti.com/lit/an/slaa523a/slaa523a.pdf
     
    page 11:
     
    It seems like PLL/VCO clocking adds phase noise.
     
    drez likes this.
  10. evillamer

     
    Looks like Rob Watts is claiming that his WTA Filter is better than filters that preserve the original data(aka Schiit Closed form filter)?
     
    And this:
    Quote:
     
    How does a DAC know what else to reproduce other than the data it is given(e.g. garbage in, garbage out)??? Unless Rob Watts have some kind of method/maths which compensates for analogue to digital converter's signal loss? This sounds like MQA type of solution.
     
    http://www.audiostream.com/content/mqa-ltd
     
     
  11. purrin
     
    Again, I wouldn't disagree with the review. The amping issue is a legitimate one. I know they were using built-in amps on the Xonar and DAC2, but that fact remains, they still are using different amps. Having owned the Xonar a while back, I know the headamp on that card sucks. The Xonar's LOs into a good headamp sounds much better. I would also assume the same for the DAC2: I'm sure the head-amp was more of an afterthought than the DAC circuitry. The OBJ2 amp: that amp is well known. So basically the comparison is this when you break things down:
     
    1) Motherboard DAC/amp out
    2) Xonar DAC | Xonar headout (with crappy headphone chip)
    3) BM DAC2 DAC | DAC2 headout (probably not as much attention paid to it compared to DAC section so likely a bottleneck)
    4) ODAC | Objective 2 headamp
     
    Of the tests, the only consistently identifiable setup was the motherboard out. I am not surprised. Basically you are comparing three mediocre / low-end setups and one really crappy one (which was easily identified.) It was a good test, but the conclusions are flawed. Also, it's obvious there was a huge confirmation bias thing going in (the photo of the Mcintosh tube amp which had nothing to do with the test is huge red flag of nwavuy syndrome) 
     
    Even they admitted they were amateurs. It would be like me comparing 5x7 prints from $3000 Nikons and $150 point-and-shoot cameras. Probably look all the same to me.
     
  12. Jones Bob
    Deleted
     
  13. frenchbat
     
    @JCX
    Well they actually did hear a difference since they supposedly corrected the volume matching later. Additionally they actually managed a 100% result on the Daft Punk track, but it didn't seem to satisfy them so they redid another test which failed, I'm still scratching my head on that. Besides that, well what Purrin said.
     
    The output impedance is one of the reasons why people use OTL amps with the HD6X0 and HD800, to get the bass bump.
     
  14. purrin
     
    Indeed, another sign of confirmation bias. I use the Daft Punk record a lot of for evaluating gear. The material on it is challenging for a lot of systems. Lots of bass, start-stop, low-end extension, effects, difficult waveforms, and even surprisingly good plankton.
     
    They get 100% result the first time, so they need to call "BS" - so they to try again (after their ears are fatigued) so they can do worse. A proper test would have been to conduct each test (not just he Daft Punk) several more times, perhaps on separate days.
     
    Confirmation bias goes both ways.
     
  15. Ableza
    there is a treatment for nwavguy syndrome: enjoy the music and stop obsessing about gear.  [​IMG]
     
    US Blues likes this.
First
 
Back
387 388 389 390 391 392 393 394 395 396
398 399 400 401 402 403 404 405 406 407
Next
 
Last

Share This Page