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

Objectivists board room

Discussion in 'Sound Science' started by joe bloggs, May 28, 2015.
280 281 282 283 284 285 286 287 288 289
291 292 293 294 295 296 297 298 299 300
  1. Allanmarcus
    Whoever said DACs "are supposed to be audibly transparent"? Did you poll every DAC vendor to determine this? Man, what an unscientific statement, and very likely untrue. DACs are "voiced" by their maker to offer the sound the maker wants them to have. The goal, most of the time, is for the DAC manufacturer to say their DAC sounds the best using all the foolish audiophile terminology they can fit int he space allotted to them. I doubt any DAC makers have live singers and musicians in their regular testing process to validate the sound from their equipment sounds like live. It's possible they do this occasionally as a marketing stunt, but I'll bet most DAC makers rely on measurements and listening to determine if the DAC is producing the results they desire.

    As for my testing, no one said one was inferior to another. I said I preferred one over the other. Sony is known to voice their stuff on the warmer side (ever list to they Z1 or their DAP - very warm), and the UDA-1 was a little warmer than the Schiit. Also, it sounded a little closed in compared to the Schiit, but that is very subjective. On a purely objective observation, the FR didn't sound the same, and I preferred the Bimby.

    Ug. I get to type this again. The small volume variation was in favor of the DAC that the folks didn't think was better. The general consensus is that the louder the better. Unfortunately, my google foo is failing me. I read about the testing here on HF. I think it was at atomicbobs house. I'm sorry I cannot find more info. If I do I will post.

    As for my testing...

    The Sony and the Schiit (spec wise) have the same output. Listening matching for levels using test tones, to the best of our ability said they were matched. Both were fed by identical USB cables (AudioQuest Diamond Platinum Obsidian! - JK, just regular USB cables, probably from mono price) and identical BJ RCA cables I got for a good price used. I set the Mac to output to both simultaneously, then I used a home made RCA switch box between the DACs and the amp. My son would turn away from the computer and I would flick the switch (with no music playing) to randomize the start. That became position 1. The other DAC was position 2. I would start the music, and he would listen. When he wanted a switch, he would raise 1 or two fingers.

    We also did a version of this test were he would listen to one, then I would pause the music, flick the switch a few time, and he would listen and say if it was the same DAC.

    While it possible the outputs of the DAC were not perfect, it was clear that the sony was warmer; the FR wasn't the same. Not worse or better; just different.

    Perfect testing? no. It was a close as we could get.

    I'm not sure what beating this all to death is accomplishing. Are we trying disprove that there is any variation in DACs, other then output level? That has been proved over and over again by measurements (you know, science). Are we trying to say that any variation in DACs is inaudible? There would be like saying prove god does't exists. You can make that statement with authority as there is no viable scientific way to prove it, although it's easily disproved with blind testing.

    I'm not even gonna proofread anymore. I'm just posting.
  2. orderingrabbits
    A DAC is a digital to analog converter. An ideal DAC converts digital to analog perfectly. Coloration is deviation from the pursuit of this ideal form. Yes, they are supposed to be audibly transparent, as that is a much lower standard of "perfect to our ears" rather than "perfect to our highly sensitive measurement tools".
  3. Allanmarcus
    Sorry, this is defined where? Also, everyone in the world agrees to this definition?

    Please define "perfectly".

    To me "perfectly" might mean no variation from the original content. It also might mean sounds as close to analog as possible. I apologize if "perfectly" was defined in other postings and I missed it.

    From what I understand of AD and DA conversions, something is always lost in the translation due to how the signal is spliced into little bits (pun intended) and then reconstructed. From what I understand, the higher the bit rate and the higher the sample rate, the less interpolation needs to happen. Also, there are various ways to reconstruct the signal. There are filters, oversampling, interpolating, pulse density conversion techniques, feedback loops, noise shaping, and other aspects to DACs that are implemented differently, thus the resultant sound may not be the same on each implementation.

    Personally, I don't care if the signal is reproduced as closely to the same signal that was used to created the file. If the file was created in an all digital path, then how can you even determine if the resultant analog signal is the "same" as the input that went into the digital representation of the sounds? Unless you also had an analog recording of the original content to compare to, there is no way to validate the results.

    All I care about is, "do I like it?" Yes, that is subjective. But that is not the point of my original assertion that some DACs sound different, and that those with trained hearing and/or sensitive hearing can distinguish. The minor variation in measurable output, when reproduced in certain combinations of preamps, amps, and speakers, may be distinguished and preferred, and possible describe (using a very subjective set of terms). My other assertion is that if, in a blind test a difference is heard, and the test was statistically significant, there is a high probability there is a difference, even if not measurable using measuring equipment available to the testers.

    And yes, volume matching quite probably accounts for most identified variation, but to the trained and sensitive ears, other variation (as discussed above) may also be present. A trained and sensitive ear is more likely to identify variations that sensitive measuring equipment would also identify.

    How about we switch the topic. I'll start: Isn't MQA great? Way better than anything we have now!

  4. sonitus mirus
    If some DACs sound different, the differences are measurable. If the differences in measurements are magnitudes beyond what our hearing can detect, it is audibly transparent. Your ideas fall in line with the industry marketing blurbs we see all too frequently. Not surprisingly, you also brought up MQA and suggested it was great. Hope that was sarcasm.
  5. RRod
  6. orderingrabbits
    At a basic level, "perfect" decoding would be something like input digital number (sample) of 0.2 -> output 0.2 volts, followed by a new sample and a new voltage that happens in an infinitesimal amount of time. Of course, this is not possible in the real world.

    Yes. The answer is "may". You are essentially debating where the floor of audibly transparent is.

    That's literally saying you don't care about fidelity. The circle of confusion is not an argument for an electronic device which has very well defined ideal parameters. Removing audibility and audio concerns, the job of a DAC is to convert digital representations of analog signals into an accurate analog counterpart. It doesn't really matter the stuff upstream of the file is.

    I'm only taking issue with your statement of there is no "perfect" for DACs and DACs shouldn't be audibly transparent. That's like saying there isn't a perfect resistor, or perfect capacitor so we shouldn't try to make these parts within a certain spec. Of course not, but we have mathematical representations to which we can strive. DACs should certainly be audibly transparent. I'd say the same about amps. Transducers actually interface with the human body directly so there's no way to make a transducer that is neutral for everyone. Of course, this is under the premise that you care about fidelity. Sometimes, I DON'T care about fidelity. That's why I own a tube amp, because it's cool. The psychological cool factor > inaudible or negligibly audible higher distortion than SS amps.

    LMAO nice
    Allanmarcus likes this.
  7. Allanmarcus
    Nope! I think MQA is the best thing ever! It will revolutionize music! I've switched everything I own over to it!


    Yes, it was sarcasm. If MQA were to sound better to me, I would care. I don't have an MQA enabled DAC, but I did try some software MQA and I could not tell the difference. But I can't hear the difference between 44/16 and 96/24. Lucky me.

    You sure we have figured out how to measure everything? You have all the equipment necessary to measure everything? You take no stock in blind testing? You one of those people that believes every scientific discovery worth making has already been made? I've met physicists at LANL that believe that. They were humbled by the Higgs boson discovery. I'm in the camp that we don't know everything, and we never will.

    And in no way do I believe the marking bulls*t most audio vender spew. Very few of them will submit to blind testing. If they did, their house of cards would collapse. I've sat in Nordost and Audio Quest demos and watched them lead the audience, and the ooos and aaaahs from the audience just made me laugh. They would switch something, tell the audience what they were doing, and say "isn't that better". Their whole market is based on expectation bias. I despise that.

    If you look at my equipment list you don't see any fancy cables, power conditioners, LPS, or expensive vibration control equipment. I make my own headphone cables with reasonably priced mogami or canare quad cable and AliExpress connectors, and nuerik plugs from Markertek. I played a little with silver wire, mostly because I got some decent wire for a very good cost. If I think I can hear a difference, I know I need to prove it to myself before I can attribute it to the cables. I do subscribe to the school of thought that suggests that plugging all the cables in the audio rack into the same power strip can reduce ground hum, because they all share the same ground, but that's it for power. I have some HVAC vibration control that I got for a whooping $0.25 each at supply house, and they do reduce very audible vibration sounds when I walk around near the table my Crack is on, and they add extra ventilation spacing too other equipment, which I have no proof is necessary, but for a couple of bucks, I figured could hurt.

    I do believe is "trust your ears", but only in the context of blind testing. I've never seen that on a brochure.
  8. sonitus mirus
    If some DACs sound different, it is measurable. No need to bring in quantum mechanics or theories about dark energy. The principles behind digital audio are not some huge mystery yet to be discovered. You are enjoying digital music precisely because the theories have been put to use in practical applications.

    Show us any ABX test where anyone identified a difference between DACs where no measurements commonly used to engineer the device show a difference large enough to be considered audible. I have never seen anything conclusive or repeatable. Seems outrageous to me.
    Last edited: Mar 25, 2018
  9. Allanmarcus
    @sonitus minus
    Thanks for the post. Seriously.

    I would like to understand how various aspects of how people describe sound are related to measurements. I understand frequency response and how it relates to "warmer" and "brighter". What measurements relate to

    Sound stage?
    Instrument separation?

    I'm not trying to put out a trap here. I've always wondered about this, and the folks here seem to be very knowledgeable.

    By the way, my original assertion was simply that DACs are not all the same and that a person with trained and/or sensitive hears might be able to distinguish a difference. The conversation then went sideways into measurements. If it's well accepted that everything audible variation can be measured, that I will accept that. It doesn't change my original assertion that not all DACs are the same, and that a person with trained and/or sensitive hearing may be able to distinguish. This doesn't say that others can't hear the difference. just that some may be to hear the difference, and it might be more likely for a trained ear and/or a sensitive ear to better hear measurable differences.
  10. bigshot
    It is the function of a DAC to convert digital audio to analogue audio. That's its only job. A good DAC accurately does that to beyond the range of human hearing. The amplifier has tone controls that are intended to color the sound. Everything else in the chain leading up to that should be neutral.

    If a DAC sounds different, it is either 1) defective or 2) deliberately made to sound different (which I describe as defective by design). A DAC should be audibly transparent. If it isn't, I would recommend returning it for a refund.

    Measuring sound has been done for nearly 100 years. (Google Bell Labs if you are interested in this subject.) We may not be able to measure everything, but we certainly can measure more accurately and completely than ears can hear.

    Soundstage is covered under channel separation. Instrument separation falls under distortion levels (and frequency response when it comes to masking). However, in practice, both of those things are more a matter of the quality of the mix than they are the quality of the playback equipment. Room acoustics can drastically affect both of those things too.

    The primary aspects of recorded sound quality in DACs are frequency response, distortion, dynamic range, and timing errors. None of these should be affecting the sound within audible levels, even on the cheapest CD player or portable DAC.
    Last edited: Mar 25, 2018
    Allanmarcus likes this.
  11. old tech
    Perhaps you should define what you mean by 'sounding close to analog as possible".

    I presume you mean as close to the original analog input signal, rather than what is possible with an analog recording or playback equipment?

    If you mean the former, any decent DAC will reproduce the original analog signal closer than is possible with any analog equipment - ie magnetic analog tape or a vinyl record. If you mean the latter, well where is your evidence for that?
  12. Allanmarcus
    Sorry. I meant as close to the actual analog sound (e.g., the voice and instruments).
    old tech likes this.
  13. bigshot
    That is mostly a function of the miking and mixing done in the studio, not playback.
  14. Allanmarcus
    thanks for the explanation. Is this the same as crosstalk? For example, the Schiit Mj2 specs says "Crosstalk: >-75dB, 20 Hz-20KHz". The Vali spec says "Crosstalk: -68dB, 20Hz-20KHz". Does that translate directly into "perceived sound stage"?

  15. bigshot
    Yes, crosstalk is channel separation. Soundstage isn't the same thing as channel separation, but it can be affected by it. Soundstage is the interaction of the mix, speaker dispersion and room acoustics. That can be adversely affected by crosstalk. It can also be adversely affected by incorrect speaker placement in a room, bad room acoustics, and a lousy mix. Both of those specs you cite represent inaudible levels of crosstalk. When it comes to that spec, they should sound identical.

    Take a look at the three links in my sig file. There's a lot of great information in there and it can help you understand what the terminology and numbers in specs mean.
280 281 282 283 284 285 286 287 288 289
291 292 293 294 295 296 297 298 299 300

Share This Page