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

Testing audiophile claims and myths

Discussion in 'Sound Science' started by prog rock man, May 3, 2010.
823 824 825 826 827 828 829 830 831 832
834 835 836 837 838 839 840 841 842 843
  1. Don Hills
    No, the amplifier has to be able to supply high voltage at high frequencies, not high power. The impedance of the coils rises with frequency, requiring more voltage to be able to drive enough current into the coils. A constant current drive high power amplifier is a tricky beast, it's much easier to use a conventional amp capable of high voltage swings. It's the same problem when driving tape record heads, but at lower powers of course. There, it's usually practical to have constant current drive.
  2. analogsurviver
    Believe me - NOBODY is better aware of this than myself...

    The defects of analog records are nowhere as plainly audible as any steady state signal - particularly pure sine wave. I kHz is "nice" - compared to 3.15 kHz steady state tone used for wow & flutter measurements ... 5 minutes of it in one go - imagine "joy" of discovering that the test record drops in few places from the nominal 3150 Hz to below 3k ... - another useless freesbie masquerading as a test reference tool.. If > 10 turntables show the SAME deviations at the SAME spot/time on record, and all > 10 turntables measure OK on another test record, that is the only possible conclusion.

    It is a miracle that analog record does sound as good as it does - part of the answer is that most of the real music is nothing but an endless succession of transients - where analog record, if recorded and reproduced to quality level, exceeds the RBCD - at least twice.

    Sustained tone nightmare par excellence for analog - Horn Concerto with Organ Contionuo. Horn produces the purest close to sine wave waveform with very little, if any, harmonics - and can be played "steady state". Organ is SLOW, SLOW, and again - SLOW; no transient worth worying about. Any distortion, noise, speed deviations, tape sensitivity modulation, ticks and pops of records, you name it - WILL stand out, as a sore thumb, in such a ( hopeffully hypothetical only ) case ...

    The Judgement Hour for RBCD ? Jazz Big Band ... look no further, although there ARE even more pronounced and audible examples of wrongdoings of RBCD.

    I can - and do - record in any digital from MP3 192kbps to DSD128 . An do have analog test record digital recordings for approx - at very least, would have to check and count them one day .. - 50 cartridges, with approx 1 hour of pure technical signals per cartridge/arm/ turntable, etc )

    That is LOTS of "bzzz/fiyueEWEWEee" - besides the nominally recorded signals....

    Still, with real music, using - State of The Art equipment, if required - analog does smoke RBCD on MOST of the music.

    That's WHY I only attempted doing CD-R transfer of analog records twice - or , MAYBE - three times - it was simply NOT ENOUGH.
    Last edited: Feb 4, 2019
  3. analogsurviver
    I have stopped studying the details about record cutting amplifier requirements back in late 80s - when it has been decided, that the only cutting lathe ever to be known in then capital of Republic of Slovenia within Yugoslavia, now the capital of Slovenia, will be never installed and has been, still in wraps - sold.

    It was a Neumann VMS-80, with all the bells & whistles of the later models.

    If it comes to the realization of my dreams ( some really good DSD recording - mastered as a half speed analog record ), I might want to refresh my knowledge regarding amplifier requirements. I know the resistance of coils is around 5 ohms, but have no data regarding the inductance of a cutting head... That sounds similar to me as a normal dynamic speaker coil.

    Be it as it may - at 1/2 speed mastering, absolutely demanded for the > 20 kHz content of the DSD master, amp voltage/power is the last thing to worry about. But I will go and check cutter head electrical parameters anyway in a few days.
  4. KeithEmo
    We seem to have a sort of disconnect here......

    I don't disagree at all about all the ridiculous complications involved in mastering to vinyl. (And I would be very interested to see, for example, THD measurements of the entire process.... measured from the output of the microphone preamp in the studio to the output of the listener's phono preamp.... with a clean sine wave and no processing.... compared to those for RBCD. I'm quite sure there are good reasons you never see THD and IMD specs for cutting lathes, phono cartridges, or the mechanical accuracy of the shape of the wiggles in the vinyl.)

    I was specifically addressing the claim that "the 44.1k sample rate was chosen for CDs because the audio masters HAD to be able to be recorded on U-Matic tape systems - which only supported 44.1k". I was simply pointing out that there was no specific need to ever store the data on U-Matic tape.

    Excluding all the details about the equipment involved. Before digital media came along you took the ANALOG MASTER TAPE to the place where they used the cutting lathe to produce the lacquer master. After the switch to CDs, you cold have just as easily taken the same ANALOG MASTER TAPE to the facility where the CD masters were produced. There was no specific necessity to convert from analog master to digital master, on U-Matic tape, AT THE STUDIO before transporting it.... and I'm not aware of any reason why it would have had to be transferred to U-Matic tape after the conversion either (I'm guessing that the CD mastering studio could have used one of those 550 pound hard disc drives if they wanted to. Also, unlike with vinyl production, there was no tweaking for the engineer to do at the point of conversion. Other than, possibly, making minor adjustments to the ADC, there is no reason for the mastering engineer to be involved in the physical process of converting from analog to digital.

    Incidentally, unlike analog data, digital data isn't so fussy about being stored in one continuous segment on a single piece of media. Once the digital audio data for the CD was created, and assembled on the "CD cutting equipment itself", you could have easily stored one CD on four of those 200 mB digital data tapes. There's no issue whatsoever with allowing a digital data backup set extend across multiple tapes. It is routinely done and there is no loss of quality or reliability. (And since, with digital, there is no generational loss of quality, once the CD data stream has been created, there is no issue whatsoever with making multiple master copies, and multiple backup copies, and changing between formats when convenient.)

    When the CD standard was first written - there were no CD disc recorders, and no CD mastering facilities.... so, at that point, they could just as easily have been built to a standard of using a 48k sample rate on 5" discs for 45 minutes.... or 96k on a 12" disc for an hour. I was simply pointing out that there were no significant technological reasons that made other choices impractical. The reasons were more at the level of: "Most studios already have U-Matic machines, and know how to use them, so they'll be happier if they don't have to purchase and learn new hardware, or have to ship their precious master tapes to a CD mastering facility after they leave the vinyl mastering facility". The reality is that "audibly perfect audio quality with a reasonable safety margin" was not the primary consideration when the format was chosen... it was just one of many factors considered.

    They did NOT conduct extensive listening tests, in a huge and impartial scientific study, at a wide variety of different sample rates and bit depths, to determine which one sounded better... or if higher sample rates were audibly better than 44.1k. And they most certainly did not test whether recording it onto a CD would audibly degrade content from a high-quality master, recorded and converted on the best equipment available in 2018, and played back on the best quality playback equipment available in 2018 - for obvious reasons. What they did was to develop a standard that met or exceeded all of what they considered to be the practical requirements, then conducted some listening tests to confirm that it was adequate for the market, and produced no obvious audible problems when tested on the equipment available at the time. (Please note that there is nothing terrible, or even unusual, about doing this. You are NOT driving the safest car, or the fastest car, or the most efficient car, that could be built using current technology either. However, we all know that there will be "better" new models next year. However, when it comes to audio, there seem to be a lot of people determined to believe that 'the game is over, we now have the best possible, and there is simply no point in looking for, or hoping for, anything better". I am quite convinced that RBCD was audibly transparent, when tested with the available audio content, available audio equipment, and listening acumen available at the time.... but the time was the 1970's.)

  5. analogsurviver
    Err.... there WERE good measurements for phono gear.... THD and IMD included - published in Italian and German audio press, but measurements themselves being the labor of love by
    Instituto Alta Fedelta, Roma, Italia.
    The pdf files of these objective reviews are like tide - they come and go, but I did find at least the front page and content of the September 1977 issue of Suono :
    For those not speaking Italian, "fonorivelatore" means phono cartridge. I do have this issue in my archive, it has been instrumental for me buying Supex cart a year or so later. I will try to dig up the reviews proper - there used to be similar coverage as for american magazines, which are still online.
  6. bigshot
    I don't think AS owns any LP records or turntables. They're just fetish objects to him. Sperg fodder.
  7. analogsurviver
    Oh dear.... - and to think - somewhere along the way - I wanted to furnish you with a VERY rare phono playback device that would render your warped LPs to play like perfect copies ....

    With your attitude, you blew that chance long ago.
  8. 71 dB
    What would make CD not transparent? This: If there is a transparent format (say 24 bit / 96 kHz). You downsample it to 16 bit / 44.1 kHz and you calculate the difference signal. If you can hear the difference signal using normal listening level there is a possibility you can hear the difference when the music is playing. But you have masking so in practise you are much less likely to hear the difference. If you can hear the difference then CD is not transparent. If you can't hear it CD is transparent. I don't know about other people, but I can't hear the difference at normal listening levels so CD is transparent for my ears. If you are a bat then you are a bat. I am a human being with human hearing.

    If you are a bat and need hi-res then USE the ****ing hi-res! IT IS AVAILABLE NOWADAYS FOR CHRIST! If not then it's capitalism not RBCD to blame.
  9. james444 Contributor
    As an IEM afficionado who doesn't give a damn about hi-res audio files or DAPs, I can tell you why I still welcome the hi-res craze when it comes to these tiny transducers.

    IME, manufacturer's efforts to extend bandwidth to 40kHz in a halfway decent manner have (as a side effect) led to improved treble performance below 20kHz. So, while I don't really care about what hi-res certified IEMs might bring to the table for bats, I generally appreciate what they can do for me in the audible range.
  10. bigshot
    How does that work? The resolution of high data rate audio below 20kHz is no different from 16/44.1
  11. james444 Contributor
    I'm talking about reproduction quality of IEM transducers, not about high data rate audio. In general, improved performance in the treble range where it matters (6 - 16kHz), higher quality drivers, less harshness, better linearity.

    One could argue that IEM technology would have evolved anyway, but I think the hi-res specification for headphones has helped push the envelope.

    Imagine tyres with a top speed rating of 130 mph vs. 75 mph. You probably wouldn't be surprised if the former performed better at 75 mph than the latter.
  12. bigshot
    Ah. I've never heard the term hires applied to transducers before. Yes, I can see how a wider frequency extension might balance things better below.
  13. KeithEmo
    That makes perfect sense..... and, at the same time, you will have confirmed that whatever sample rate converter you used produces no audible side effects.
    However, I would be sure to perform the experiment with at least a few hundred different files, or even better a few thousand, just to make sure there isn't a difference that is extremely audible, but only on a few of them.
    And, just to be accurate, you cannot directly subtract samples at different sample rates.... you would have to use some sort of interpolation to enable you to do so.
    Of course, you could convert them to analog, and then do an analog difference compare.... but then you have any errors that might be introduced by the DACs....

    It's also worth noting that the opposite could occur.
    There could be quantitatively significant differences which might turn out not to be audible.... for example, a high level of THD at 15 kHz would probably not be audible, since the first distortion product would be at 30 kHz.

    Of course, I would suggest trying the experiment again every few years, just to confirm that, as mastering technology improves, the situation hasn't changed.
    And, of course, your ability to hear differences will be limited by your playback equipment and your ears.

  14. bigshot
    Let's just do it with every file in the world before we decide. Because it's possible that in some alternate universe "yes" could be "no" and "up" could be "down".

    Get back to me when you're done and I'll tell you to do it all again because something may have changed.
    Last edited: Feb 4, 2019
    gargani likes this.
  15. bigshot
    But it isn't better. There is no audible difference.
823 824 825 826 827 828 829 830 831 832
834 835 836 837 838 839 840 841 842 843

Share This Page