+ Reply to Thread
Page 22 of 28 FirstFirst ... 122021222324 ... LastLast
Results 211 to 220 of 271

Thread: Digital Facts and realities?

  1. #211
    Join Date: Apr 2008

    Location: Cheshire, UK

    Posts: 2,829
    I'm Clive.

    Default

    Quote Originally Posted by NRG View Post
    Because the signal is bandwidth limited. It can break down if the signal is allowed to clip but that’s poor mastering, it shouldn’t be allowed to happen. The sampling is continuous not stepped or point in time so there are no gaps.
    So from 20Hz to 22.5kHz it's not actually sampling, it is in fact recording everything. Sampling suggests taking readings at set points.
    TT 1 Trans-Fi Salvation with magnetic bearing + Trans-Fi Terminator T3Pro + London Reference
    TT 2 Garrard 301 with NWA main bearing + Audiomods Series Six 10.5" + Ortofon 2M Mono SE
    Digital Lindemann Bridge + Gustard R26 with LB external clock
    Pre and Power Amp EWA M40P + M40A
    Bass Amp & DSP Behringer iNuke NU3000DSP x 2
    Speakers 1 Bastanis Sagarmatha Duo with twin baffleless 15" bass drivers per side
    Speakers 2 MarkaudioSota Viotti Tower

  2. #212
    Join Date: Feb 2008

    Location: Down South

    Posts: 2,413
    I'm Neal.

    Default

    My bad, poor choice of words, the samples are discrete points in time of a bandwidth limited continuous signal. If you wish to capture a signal which lays between your example of sample 4 and 5 IE: above 1/2 Nyquest frequency you need to sample at a higher rate...
    Listening in a Foo free Zone...

    Only a Sith deals in absolutes.

  3. #213
    Join Date: Aug 2008

    Location: Suffolk, UK

    Posts: 1,473
    I'm Paul.

    Default

    Quote Originally Posted by Clive View Post
    To change tack and go back to basics...

    First let me state that with a good DAC I'm very happy with 16/44.1.

    I often see it stated that red book sampling rate is perfectly good for reproducing a sine wave. That's dandy but we don't listen to pure sine waves. The waves we listen to have incredibly varied amplitude. I can freehand link up something like 12 dots to make a decent fist of a sine wave. If I tried to do the same with a wave of varying amplitude how many dots do I need to accurately show the direction changes? A lot more for sure. I'm not saying that 16/44.1 can't do this, clearly it can or it gets very close. It's that statement about accurately reproducing a sine wave which seems like the wrong test.

    Can 16/44.1 accurately reproduce a real-world jagged curve in all cases?
    I think what you are asking is something that I have wondered about in the past. This is what happens if a sharp transient initiates between sampling points? The transient will not be detected until the next sample point meaning that during playback it will be played slightly later than it actually occured in real life?

    Although it is easy to visualise this “change of direction” between the sample points This does not actually happen in real life with recorded sound, due to something called pre and post ringing. This is a gradual build up and fading away of a response to a transient or impulse that happens in electronic filters as well as microphones and our ear drums, which are essentially mechanical filters. Increasing the sample rate does improve the response to these transients of the electronic filters used in sampling and playback by reducing the pre/post ringing duration but it is argued that this improvement is not audiable and I think this is where the contention lies. Possibly with exceptionally well performing hifi equipment someone who knows their system and the music well and is skilled and experienced in listening can detect this, maybe. If they are able to decern what are essentially very small differences in timing then possibly they can hear a difference of a more accurate clock.

    I’m not sure that I have explained this very eloquently or completely accurately but maybe someone else can do a better job? I do openly admit that I am out of my depth somewhat.
    ~Paul~

  4. #214
    Join Date: Apr 2008

    Location: Cheshire, UK

    Posts: 2,829
    I'm Clive.

    Default

    That is indeed what I was getting but if I am correctly interpreting what Neal said, there are no gaps between samples. If so I don't get why sample graphs are drawn with gaps...or they should be blocks rather than lines but that would imply inaccuracy. Something doesn't stack up.
    TT 1 Trans-Fi Salvation with magnetic bearing + Trans-Fi Terminator T3Pro + London Reference
    TT 2 Garrard 301 with NWA main bearing + Audiomods Series Six 10.5" + Ortofon 2M Mono SE
    Digital Lindemann Bridge + Gustard R26 with LB external clock
    Pre and Power Amp EWA M40P + M40A
    Bass Amp & DSP Behringer iNuke NU3000DSP x 2
    Speakers 1 Bastanis Sagarmatha Duo with twin baffleless 15" bass drivers per side
    Speakers 2 MarkaudioSota Viotti Tower

  5. #215
    Join Date: Sep 2013

    Location: North Island New Zealand

    Posts: 1,757
    I'm Chris.

    Default

    Quote Originally Posted by Primalsea View Post
    I think what you are asking is something that I have wondered about in the past. This is what happens if a sharp transient initiates between sampling points? The transient will not be detected until the next sample point meaning that during playback it will be played slightly later than it actually occured in real life?

    Although it is easy to visualise this “change of direction” between the sample points This does not actually happen in real life with recorded sound, due to something called pre and post ringing. This is a gradual build up and fading away of a response to a transient or impulse that happens in electronic filters as well as microphones and our ear drums, which are essentially mechanical filters. Increasing the sample rate does improve the response to these transients of the electronic filters used in sampling and playback by reducing the pre/post ringing duration but it is argued that this improvement is not audiable and I think this is where the contention lies. Possibly with exceptionally well performing hifi equipment someone who knows their system and the music well and is skilled and experienced in listening can detect this, maybe. If they are able to decern what are essentially very small differences in timing then possibly they can hear a difference of a more accurate clock.

    I’m not sure that I have explained this very eloquently or completely accurately but maybe someone else can do a better job? I do openly admit that I am out of my depth somewhat.
    What actually happens is digital being sampled has 1's and 0's , so a error including jitter distortion interprets wrongly a 1 as a 0, ( naughty digital !!! ) and in the process changes the intended signal usually irreversibly,
    however two mechanisms attempt to correct this, error correction - and interpolation, which are both attempts at correction. The better solution is to try to avoid any mechanism for jitter to occur in the first place.

  6. #216
    Join Date: Sep 2013

    Location: North Island New Zealand

    Posts: 1,757
    I'm Chris.

    Default

    Which leads to an interesting point, instead of quickly making information up via interpolation, and sometimes still getting the data wrong, CD players were initially able to
    very slowly read all available data into a memory, there would then be a much better chance of data being read correctly when playing if the memory was referenced.

    The thing is, would we patiently wait for them to effectively buffer their content ? which might take a while.

  7. #217
    Join Date: Feb 2013

    Location: W Lothian

    Posts: 99,005
    I'm Grant.

    Default

    When Red book came out memory was silly expensive and pretty small scale. You couldn't save enough data to make it worthwhile.
    Also the system they use is pretty good if not absolutely perfect. (nothing is) and cheap to implement. CD was meant to take over from tapes anyway
    Regards,
    Grant .... ؠ ......Don't be such a big girl's blouse

    I've said it before and I'll say it again: democracy simply-doesn't-work
    .... ..... ...... ...... ................... ..... ..... ..... ..... .....
    FIIO K7 BT, M11 PLUS, BTR7, KA5 - OPPO BDP-103D - PANASONIC UB450 - PANASONIC 4K ULTRA HD TV - PIXEL 6 - AVANTREE LR BLUETOOTH - 2* X600 SOUNDCORE - HEADPHONES INCLUDE, FIIO, NURAPHONES', FOCAL, OPPO, BOSE, CAMBRIDGE, BOWER & WILKINS, DEVIALET, MARSHALL, SONY, MITCHELL & JOHNSTON - 2*ZBOOK'S- MERCURY BD ROM, ROON, QOBUZ, TIDAL, PLEX, CYBERLINK, JRIVER - MULTI HDD'S -

    Oh my god! There's nothing wrong with the bidet is there?

    “Nothing discloses real character like the use of power. It is easy for the weak to be gentle. Most people can bear adversity. But if you wish to know what a man really is, give him power. This is the supreme test. It is the glory of Lincoln that, having almost absolute power, he never abused it, except on the side of mercy".

    “You see these dictators on their pedestals, surrounded by the bayonets of their soldiers and the truncheons of their police ... yet in their hearts there is unspoken fear. They are afraid of words and thoughts: words spoken abroad, thoughts stirring at home -- all the more powerful because forbidden -- terrify them. A little mouse of thought appears in the room, and even the mightiest potentates are thrown into panic.”

    "You don't have free will. You have the appearance of free will.”

    “There's a war out there, old friend. A world war. And it's not about who's got the most bullets. It's about who controls the information. What we see and hear, how we work, what we think... it's all about the information!”


    ***SMILE, BE HAPPY***

  8. #218
    Join Date: Feb 2008

    Location: Down South

    Posts: 2,413
    I'm Neal.

    Default

    Quote Originally Posted by Clive View Post
    That is indeed what I was getting but if I am correctly interpreting what Neal said, there are no gaps between samples. If so I don't get why sample graphs are drawn with gaps...or they should be blocks rather than lines but that would imply inaccuracy. Something doesn't stack up.
    Basically for convieneince. The correct way is a lollypop graph, the stair step graphs that we've been fed with for decades are technically incorrect and confusing, there are no gaps in a sampled, correctly bandwidth limited signal. This is still a great video from Xiph.org FF to 6mins, explains it all....and at 20:55 https://www.xiph.org/video/vid2.shtml you'll need chrome or firfox to view.
    Last edited by NRG; 27-03-2019 at 08:14.
    Listening in a Foo free Zone...

    Only a Sith deals in absolutes.

  9. #219
    Join Date: Apr 2008

    Location: Cheshire, UK

    Posts: 2,829
    I'm Clive.

    Default

    Quote Originally Posted by NRG View Post
    Basically for convieneince. The correct way is a lollypop graph, the stair step graphs that we've been fed with for decades are technically incorrect and confusing, there are no gaps in a sampled, correctly bandwidth limited signal. This is still a great video from Xiph.org FF to 6mins, explains it all....and at 20:55 https://www.xiph.org/video/vid2.shtml you'll need chrome or firfox to view.
    The video helps greatly, thanks Neal. I suppose I'll just have to take the scientific view that the space between two lollipop heads can ever show any variance from a perfect curve. Whether it can or not I expect in reality is immaterial or 99.9% likely to be so. Other factors such as jitter are of far greater importance.

    I still can't help having a nagging doubt in the back of my mind. The digital conversion we're told is perfect but is it really 100.00% perfect? Maybe it's 99.99% perfect but is a digital error of say 0.01% better or worse than an analogue error of 0.1% or even 1%. The analogue error maybe euphonic or habituated so even though it's an order of magnitude (or more) that any remotely possible digital error the aural effect may differ, some people may prefer one or the other. It's all a bit academic, even I hold my hands up and admit my digital playback is superb and every "bit" as good an my analogue. Vinyl has its issues with noise and dirt and computer audio has it's issues too as it's not as stable as I'd hope (there's often some connectivity issue or update).
    TT 1 Trans-Fi Salvation with magnetic bearing + Trans-Fi Terminator T3Pro + London Reference
    TT 2 Garrard 301 with NWA main bearing + Audiomods Series Six 10.5" + Ortofon 2M Mono SE
    Digital Lindemann Bridge + Gustard R26 with LB external clock
    Pre and Power Amp EWA M40P + M40A
    Bass Amp & DSP Behringer iNuke NU3000DSP x 2
    Speakers 1 Bastanis Sagarmatha Duo with twin baffleless 15" bass drivers per side
    Speakers 2 MarkaudioSota Viotti Tower

  10. #220
    Join Date: Nov 2008

    Location: Valley of the Hazels

    Posts: 9,139
    I'm AMusicFanNotAnAudiophile.

    Default

    Quote Originally Posted by Clive View Post
    computer audio has it's issues too as it's not as stable as I'd hope (there's often some connectivity issue or update).
    If you have stability issues then you need to explore other ways of doing things.
    My Squeezebox setup is very, very stable, but there again, I'm not trying to reinvent the wheel all the time, and I'm not drawn in by the latest in visual rendering.
    My server is local, so doesn't have to go to the www. whenever I want to play local material (which is not the case for some player distros), and all of the connections for critical equipment are wired.
    Chris



    Common sense isn't anymore!

+ Reply to Thread
Page 22 of 28 FirstFirst ... 122021222324 ... LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •