Go Back   Home > Forums > Blogs > abraxalito

Rate this Entry

Digital that sounds like analog

Posted 10th December 2012 at 04:18 AM by abraxalito
Updated 15th February 2013 at 06:11 AM by abraxalito

For those who missed Frank (fas42)'s link on a thread I started then here's where I'm continuing my minimal oversampling DAC developments for the time being : Digital that sounds like analog
Views 4023 Comments 108
Total Comments 108


  1. Old Comment
    Richard, can't reach you on the other forum, but just noted this latest comment:

    [QUOTE]When I was playing with my DAC earlier this week, I moved one wire and lost the soundstage depth almost totally. I had to scratch my head for a moment to work out how such a small change could make such a huge difference to the sound, so I retraced my steps one by one, listening as I went. Sure enough, the grounding position of one wire made that substantial difference[/QUOTE]I'm very familar with this nightmare scenario, in digital, and so sympathise. The dilemma has always been for me that the quality of digital sound is so fragile, and unfortunately the better it gets, so increases the sensitivity to the slightest irregularity.

    Of course this is because there is still a lack of understanding of these subtleties in the industry. A couple of things I've learnt: persistence in the face of extreme frustration in not finding a solution to an apparent problem always pays off; and no matter how good you think the sound has become, even if only for a transient period, it will always have the potential to be yet better again ...

    Posted 13th December 2012 at 09:22 AM by fas42 fas42 is offline
  2. Old Comment
    abraxalito's Avatar
    I didn't consider it a nightmare scenario this time, rather took it as an opportunity to learn something. At the time of searching around I was thinking of your comments about how fragile the awesome sound is so its apposite that you'd post up something. I tracked down just why the problem occurred - a shared grounding path. Otherwise called common ground impedance coupling. When you use I-out DACs its imperative that the ground return doesn't catch some other ground en-route back to its completion.

    The fragility of great sound is I think due to the unpredictability of the RF environment - I'm gradually learning what counter-measures to take against that. Just today noticed a ground loop (slight hashy buzz) and will experiment with various ground lifting inductors to see which one does away with it.
    Posted 13th December 2012 at 01:24 PM by abraxalito abraxalito is offline
  3. Old Comment
    This is going to be an amusing exercise ... sorry! ;)

    Noted your "progress", with regard to reactions to Reference Recordings. Sounds familar, ;), "audiophile" recordings become more and more boring as the replay sound is refined, the extreme sanitising and pre-digesting of these albums means that little emotional content subjectively is left, quite unsatisfying to listen to.

    For me, the real meat occurs when one uses the "worst" recordings one has to diagnose the remaining problems - the satisfaction you get when one of these finally blooms is immense ...

    Posted 15th December 2012 at 11:52 AM by fas42 fas42 is offline
  4. Old Comment
    abraxalito's Avatar
    Seeing as you're a restricted poster over there its fine that we have our own little chat gong on over here. Yes the RR recordings thing I noticed not too far down the road when other recordings kept getting better and RR stubbornly refused to increase in transparency. Made me speculate that when an HDCD recording is played back without the decoder then its actually compressed in that the softest sounds are boosted. This would explain why when the hash is gradually shifted nothing new is revealed, whereas unprocessed recordings benefit. I shall return to my reference sibilant recording right now and see where it takes me...

    Done - the sibilance still highly apparent but fewer lacerations of my eardrums on Alanis's voice. Not what I would call 'blooming'.
    Posted 15th December 2012 at 03:25 PM by abraxalito abraxalito is offline
    Updated 15th December 2012 at 03:31 PM by abraxalito
  5. Old Comment
    I presume you're speaking of Alanis Morissette ... I don't have any of her's, in fact never have heard her songs played on ambitious systems. Would this be Jagged Little Pill material?

    I can see why this would be good test data: recorded in mid 90's, studios coming out of the ultra deep and voluminous ambience phase, heading towards dreary, high raw compression. A distinctive, somewhat aggressive female voice, no softening at the edges -- yep, it all adds up ...

    I have no doubt that you can "tame" that sound; the female voice is excellent to refer to, because we "know" what it should sound like. When you listen, is this via speakers of headphones?

    That sibilance should evolve into intensity, sound with high emotional impact, gutsiness. Recordings from that general period were recorded such that they have very high impact, are quite overwhelming when reproduced correctly, but also meaning that they take no prisoners if there is significant added distortion.

    With that recording you mentioned, are all the backing sounds coming through 100% clean and only her voice a problem, or does everything sound as if it's almost at the point of being really uncomfortable to listen to?

    Posted 15th December 2012 at 09:18 PM by fas42 fas42 is offline
  6. Old Comment
    abraxalito's Avatar
    Its on that disc I already told you about months ago, 2nd track. I have another Alanis disc which sounds ok so I reckon its in the remastering they added the sibilance - its right across the board. I only listen on speakers - you can see which in another of my threads on WBF 'Frugal digital source'.

    I have found sibilance normally tracks individual mics - for example last night I was listening to a BBC made recording of the Xmas Eve service at King's College, Cambridge and they must have had various mics around the place - some sound worse than others. Its RF ingress into mic-pres that does it I reckon.
    Posted 16th December 2012 at 03:10 AM by abraxalito abraxalito is offline
  7. Old Comment
    Okay, post 2000 material is a real lottery, studio engineers have too [I]many[/I] tools to play with, they don't know when to stop fiddling with things. This makes it harder to reach into the highly manipulated acoustic environment and retrieve, subjectively, what is the essence of the performance. So, I will acknowledge that recordings of the last decade will be the severest test of all; as examples of this, Adele 21 takes every ounce of tweaking effort to recover a decent quality of replay, a best of Clapton compilation I recently acquired shows the disease in full flow, the ones recorded at the end of the 90's are so drearily flattened compared to the material of earlier years.

    Having said that, I don't believe that the sibilance issue can't be solved. Technically, it is [I]not[/I] a distortion of the recording process as far as [I]I'm[/I] concerned, rather the the combination of manipulations carried out make it that much harder for the sound to be unraveled, the slightest untoward distortion on the replay side sabotages the chances of an acceptable replay. When I first tried Adele 21, I thought this was going to be a mountain too high, but it finally came good, just precisely the right level of tweaking got it to clear up ...

    Postscript: Had a listen, via Youtube, to the offending album, and to me there's no obvious, via that mechanism, particular problem with its quality to me. What it[I] does[/I] have, is a level of reverb, or echo, in the mix which may cause problems in systems; I note that other versions on Youtube of that song, per the album and other acoustic renditions don't the same level of that effect added in, but that's purely an artistic decision by the producer. It's not a distortion, but an added "complication" to the sound, which should resolve naturally to the ear: the originating sound, followed by its echo.

    A good system should be able to make the various levels, and depths of echo in any recording clearly discernable and acoustically understandable; if however the sound comes across as being confused then that points to a possible problem.

    Posted 16th December 2012 at 10:01 PM by fas42 fas42 is offline
    Updated 17th December 2012 at 04:21 AM by fas42
  8. Old Comment
    Of interest, finally had a look at that ESS video you referred to back when, and I have to say "Hmmm ...". I'm afraid his explanations and conclusions aren't convincing, and his version of how noise modulation is the Darth Vader of S-D doesn't cut it for me -- the disparity in signal levels between what he shows as being measured and what I hear in poor digital is too great.

    I will certainly agree that interference, RF and other is at the core of the problems, and S-D may be particularly susceptible because its internals are chaotic with high frequency switching. But, that digital "veil" or haze can always be lifted, I've never found it to be something irreversibily embedded, no recording forever condemned by its influence ...

    Posted 17th December 2012 at 12:53 PM by fas42 fas42 is offline
  9. Old Comment
    abraxalito's Avatar
    Yet I've found several recordings with noise modulation irreversibly embedded - not least the Harry Christophers' 'The Sixteen' one I've just posted about in the other place After posting that I dropped a line to 'The Sixteen' on their website saying I'd noticed a technical flaw in their recording and what ADC were they using? If I get a reply on that, I'll be sure to post up what they say.
    Posted 17th December 2012 at 05:23 PM by abraxalito abraxalito is offline
  10. Old Comment
    The point I would make is that sibilance in of itself is natural; in many acoustic environments, listening to real voices, the prominence of that aspect of vocal sounds can be very strong. Particularly if one chooses at the time to focus on the 's" sounds, etc, you can make the process of listening to the live performance quite unpleasant for yourself. An accurate recording should precisely capture that element - then it can be a decision by the mastering engineer, or the end user if he has sophisticated signal processing capabilities to reduce the level of that characteristic.

    But the "cotton wool" is a problem ... I'm loath to say it's in the recording, because I have yet to find a classical piece where I can't listen deeply into the acoustic, assuming it's been recorded in some space. Combinations of recording technique and the performance acoustic on classical albums can easily highlight distortion problems in the replay system, a borrowed CD of a Scandinavian string quartet in a monstrously sized acoustic, recently recorded, was difficult to render correctly, but that's because it was shining a laser beam at the deficiencies of my system ...

    I might just repeat a point that I've made many times: "there's no such thing as a bad recording". I've used that mental approach to tweaking for many years now, and it has never failed me. It stops me letting myself get off the hook when I hear a problem in the sound, by saying it's nothing that I can do something about, it's got to be the other bloke's fault ...

    Posted 17th December 2012 at 10:51 PM by fas42 fas42 is offline
    Updated 17th December 2012 at 11:03 PM by fas42
  11. Old Comment
    Okay, just listened to the samples at Amazon of 'The Sixteen', and I can hear what you're talking about: this is in fact an example of post 2000 material being a lottery, and in this case it's because of the sloppiness of the sound engineers; not taking sufficient care to warm up and condition their equipment consistently between takes. I've heard some quite atrocious examples of this in recent classical recordings, nothing to do with the quality of the gear, all to do with the quality of the grey matter between the ears of the people doing the recording.

    The variability of tone in the piano is the giveaway, someone was doing their job poorly and didn't ensure there was consistency in procedures followed while doing each track ... just ask Bruce in the 'other place' what he thinks of the current batch of "sound engineers" ... :D

    Posted 18th December 2012 at 01:46 AM by fas42 fas42 is offline
  12. Old Comment
    abraxalito's Avatar
    The more times someone repeats something the more I'm convinced they're not totally sure of it themselves Rather like in one of my favourite movies 'Some Like it Hot' - 'I'm a girl, I'm a girl.....'

    As for Bruce in that other place - if he's slagging off the other guys he should have a read of Matthew 7 'take the log out of your own eye first' except let's recast it in terms of ears 'take the bean bag out of your own ear first before you attempt to remove the grain of sand from your brother's'

    As for the piano tone's variability, you noted its a period instrument (deliberately) did you?
    Posted 18th December 2012 at 03:38 AM by abraxalito abraxalito is offline
    Updated 18th December 2012 at 03:40 AM by abraxalito
  13. Old Comment
    Or, if a wife says to a husband each day that she loves him, that such an assertion should be treated with suspicion ... ;).

    Of course, if people are content to only have some of the music they own be fully satisfying, far be it for anyone to interfere with their freedom of choice. But, if one has to travel down a rough road, it may be helpful to some to be aware that there are at least a number of options in how to make the journey ...

    With regard to Bruce's comment, I mentioned some time ago in a post the bizarre variability of sound quality on a borrowed, very recently recorded CD of solo violin snippets. From one track to the next; it was quite extreme. And so followed Bruce's response.

    And I'm hearing a similar pattern on the MP3 snippets posted by Amazon, with regard to the piano sound. Are you saying they used different setups of piano, or pianos, between different tracks; do you hear the variation between the tracks?

    Posted 18th December 2012 at 05:14 AM by fas42 fas42 is offline
  14. Old Comment
    abraxalito's Avatar
    I hadn't noticed variation in piano sound between tracks as my attention was focussed on the noise modulation. But I will go back and listen again for that.

    I don't buy your argument about 'letting myself off the hook' for bad sound - I don't use that as an excuse for things I can fix in my own system. I choose the best recordings to show up defects in my system, not the worst ones. Though bad recordings can provoke nasty effects too - for example I was playing a Sting 'Brand New Day' CD the other day and wondered why it sounded so sibilant and was able to tame that largely by adding more ferrites to my filter. Still some was left but it was reduced considerably.
    Posted 18th December 2012 at 05:26 AM by abraxalito abraxalito is offline
  15. Old Comment
    [QUOTE]bad recordings can provoke nasty effects too [/QUOTE]
    That's the sort of thing I look for ... the highlighting that difficult recordings bring to the table. For me, it just makes it so much easier to identify that the system is still not there; with an "excellent" recording I find it easier to fool myself into thinking that I'm OK.
    I have found the pattern to normally work; if the "worst" recordings come together then all music in better shape, with fewer recording anomalies, also improves.

    Posted 18th December 2012 at 06:52 AM by fas42 fas42 is offline
  16. Old Comment
    abraxalito's Avatar
    Yep, its something you taught me that nasty recordings can set off effects in otherwise decent systems. However the whole premise of 'nasty recordings' existence goes against your philosophy doesn't it? You seem to be saying that there's no such thing as a nasty recording - that's a big point of disagreement for me. Sure nasty ones can be made to sound better, but they can't be fixed up.
    Posted 18th December 2012 at 07:21 AM by abraxalito abraxalito is offline
  17. Old Comment
    Here's where the big trick comes: the key words are "fixed up", which for me means that when I listen to the recording I'm not aware of the recording being defective. And what does that mean? For me, it's that the impact of the musical performance overrides my awareness of deficiencies in the playback. And this is where it most likely gets complicated, because it depends on how people listen to music in general. Some people seem to be very aware of acoustic niggles in any situation, live music included: say an excellent saxophone player is playing live in a room, and various resonances of objects in the room are excited by certain notes. If a listener can't get past this happening, and just enjoy the intensity of the music making in this situation, then he probably will never get a 'nasty' recording to work.

    OTOH, I don't have this problem. To me, the level of "betterness" that recordings can be lifted to is sufficient to overcome my sense that the reproduction is faulty. And to put in into the current context, take sibilance. The exercise, to me, would be to take that Alanis track and keep tweaking until I no longer feel repelled by that quality in the track; I would still acknowledge, intellectually, that sibilance was emphasised in the recording but that wouldn't get in the way of me enjoying the track. Of course, if one's hearing was such that people with strongly sibilant, natural speaking voices were irritating to listen to, then this wouldn't work. The aim is to feel at one's ease when listening to the music, and if this doesn't happen then it's not there, as far as I'm concerned.

    'Nasty' or difficult recordings certainly exist, I have a few real horrors in that regard: a "bootleg" Gene Pitney where the most egregious noise reduction process was applied, the person doing it must have been totally deaf at the time. I'll never get the noise reduction mess to disappear, but I can get it to the point where I can completely hear through that distortion, without discomfort, to the performance underneath. An analogy would be listening to a very hissy tape, or very noisy record on superb equipment; for me that would be enjoyable to do.

    Posted 18th December 2012 at 10:49 AM by fas42 fas42 is offline
    Updated 18th December 2012 at 10:54 AM by fas42
  18. Old Comment
    Just noted the CAD unit you mentioned in another thread; this review is making the right noises: the-ear-net.blogspot.co.uk/2012/11/cad-cable-controversy.html.

    As pointed out, when digital sound snaps together correctly it can be quite shocking to the initiate -- the good times, they are a-comin' ...

    Posted 18th December 2012 at 12:47 PM by fas42 fas42 is offline
  19. Old Comment
    abraxalito's Avatar
    OK I got your meaning now - you're talking about the technical details not coming in the way of the emotional connection. But my emotional connection is a little different - its a matter of choice. Thanks for the clarification anyway, very helpful I don't give discomfort feeling even when the recording's bad I just change to another one - exercising a choice.

    As for the review - the DAC's broken when a cable change makes 'night and day' difference - means its subject to CM noise. But I suppose its in his interests financially to sell a broken DAC then sell a cable to fix it up
    Posted 18th December 2012 at 02:13 PM by abraxalito abraxalito is offline
    Updated 18th December 2012 at 02:17 PM by abraxalito
  20. Old Comment
    Virtually all DAC's are broken by that measure -- over the years I've noted that they tend to fall into 2 camps: either all the low level detail is discarded or greyed out so that they are not obviously obnoxious; or they throw everything at you and you feel as if you're being assaulted by shrapnel.

    The "disappearing sound" variation to me is completely unacceptable; by luck or instinct I have always played with units in full frontal attack mode, and from experience these are the ones that will get me there. As you say, noise is the real enemy, and the hardest thing can be to eliminate that last source of noise - and when I do then I can experience a 'night and day' effect. Before, a very aggressive, everything and the kitchen sink, sound; after, big, rich, enveloping sound that you can walk through with your hearing, visiting each aspect and space within.

    Posted 18th December 2012 at 10:40 PM by fas42 fas42 is offline

New To Site? Need Help?
Copyright 1999-2017 diyAudio