Very moving indeed watching that clip on YouTube.
Schroedinger was really into Vedanta according to Wikipedia. But then again, he was a polyamorist - can a person still call himself a physicist when he only has the one wife?
But, they do do the job if you give them a hammering; that is, have them decode tracks that are full of high level, high frequency detail over an extended period. A good example, of "real" music for the job, is very raucous rock, where the drummer is giving the cymbal menagerie a good thrashing. It takes several hours for my system to truly come to life; the harder I drive it, the faster it wakes up.
Why this is so, I couldn't say precisely. One suspect is certainly the fact that capacitive mechanisms are intrinsic to the S-Ds, and capacitors hate cold starts ...
Typical would be trying some solo piano: a track put on first thing in the morning would be dead as a doornail, whole top octave lost in action. By mid-afternoon, the same track will demonstrate full, correct sparkle and vibrancy, sound like a completely different instrument.
Frank
Audiophiles seem OK with the concept of 'break-in' but I'm fairly sure that the market I'm interested in will consider it an inconvenience too far. So not an option. Just why audiophiles take it in their stride puzzles me somewhat - I guess its an aspect of the 'hair shirt' audiophile mindset.
You use Cirrus (Crystal) S-Ds so its perhaps a reasonable hypothesis for them. I've not played much with their parts - AD's and Ti's S-D chips aren't based on SC filters. I wonder - does the same effect apply to their parts too?
I was reading a thread over on AVS very recently where Charles Hansen mentioned his hypothesis about break-in. He seems to think its something to do with the dielectric materials in the PCB 'relaxing'. His products need into the 100s of hours apparently to reach their best. But the effect of 'relaxation' wears off so he says. No mention was made of the human side of 'break in' in his explanations which I find a little one-sided, personally. I know my ears 'break in' in the sense of become accustomed to a piece of equipment's sound.
This is an interesting aspect of design in general: what can be considered a reasonable "warm-up" period for any piece of consumer equipment or device, in any area of use, that will be considered acceptable by the user? You start a car in a frozen environment, it may take 10 minutes before you can actually get it moving satisfactorily down the road; perfectly normal for the people who live there, this would be considered an unacceptable joke for those living in warmer climes.
Of course, if you want to take the engineering to a further level there are always ways around such problems. If the converter is of good quality, but does require long periods of "conditioning", then add a special circuit to remain active so long as the power cord is plugged in, which maintains supplies to the converter, and constantly exercises it, 24/7.
There is a hell of lot that is part of the sound quality equation which ties up with material properties, second order effects. And, that all real components are not linear in low level behaviour because of manufacturing induced weaknesses. A simple one, is how well the leads on a capacitor are mated to the plates; are there any diodic behaviours there, for example?
I feel I'm fairly lucky with being able to, most times, critically listen to a system's sound. I have a batch of really "nasty" CDs that provoke very obvious, audible misbehaviour in a less than optimised setup, that tell me a great deal. If I can listen in a totally relaxed, comfortable frame of mind to such material at realistic sound levels then I know the system is in the "zone" ... . If I feel that my ears have to "break in" to the sound of a system, then I'm in the wrong zone!
Frank
Substitute 'boot up' for 'warm up' and its clear that the acceptable time is diminishing where computers are concerned. My current laptop is ready in around 3 seconds from pressing the 'on' switch, my previous one took considerably longer. But the machine I have now does want a hard reset every week or so. Perhaps the next generation will not require that?
Where the warm-up period is a function of the external environment is a separate issue. I call red herring here
I disagree that's higher level engineering, its a backwards step. Why not understand the mechanisms and design them out? Needing continuous power is not for the future methinks.
Still using leaded caps? I'm moving over to SMT as fast as possible myself.
Its a good point - how does the system deal with 'less than optimal' produced disks. I was listening to one last night and pondering this point. A CD I had enjoyed for many years on my less transparent systems of the past revealed more of its true nature. It wasn't unlistenable by any means but the recording defects I'd barely noticed before have become extremely obvious. To me this is a sign of progress in system transparency, so poor recordings are very useful markers of improvement.
Everything is relative: if the vehicle is an emergency services unit then this becomes crucial, engineering must be applied to yield a satisfactory behaviour.Where the warm-up period is a function of the external environment is a separate issue. I call red herring here
Obviously the "correct" solution is to understand the mechanisms. But, if for whatever reasons, there is a lacking there, oftentimes because it is all too hard, and there is not enough interest or money, then why just accept that? Engineer a "work around", and you have a better, more consumer friendly product -- your company is ahead!! And, in these days of energy harvesting chips and such like, there is no reason why a solution such as that can't be made to be green ...
Agree on that one ...
This has been very controversial, but I believe that it is possible to push a system quality to the point where those "defects" again fade into the background, subjectively. The pyschoacoustic factor, so to speak ...
Frank
I don't agree that there is such a thing as 'the "correct" solution' - this is the most effective solution. But that's only apparent over the longer term, in the short term then kludges come into play.
I accept it - its called 'laziness' and I'm by no means immune
But I also accept that laziness isn't a long term solution.Only in the short term. Then when electricity becomes intermittent because fossil fuels have become too expensive to extract, I'm going to pay for my laziness. All laziness has a consequence
Not particularly controversial with me, I think I'm hearing what you're saying. That the now increased emotional engagement with the music distracts us from its now more apparent technical flaws. Rather like appreciating one's own child's out of tune attempts at singing?
I believe, because I can see it happening now, that more and more "smarts" will be added to devices and aspects of our lives that effectively make them far less dependent on the grid. Think, electric cars with solar collection as part of the paint work and windows; and roofing materials that have energy harvesting as part of their intrinsic structure. This sort of thing is being vigorously pursued by many people ...Only in the short term. Then when electricity becomes intermittent because fossil fuels have become too expensive to extract, I'm going to pay for my laziness. All laziness has a consequence
Exactly. Though, a better analogy is listening to an excellent, real pianist, in a room where the air conditioning is making all sorts of obvious mechanical noises, adding sound clutter which is completely disconnected with the music making.
Frank
I'm not yet convinced that that analogy is a very good one, for the following reason. The noise of the AC and other things around is not correlated with the musical sounds. Correlation is a big thing with me - ISTM that our brains are really good at discarding uncorrelated sounds. We know for example the cocktail party effect - we can operate with a negative SNR (or what seems like one) provided the interference isn't correlated. Of course with the cocktail party effect there's a difference in direction too coming into play.
What I find annoying is correlated noise - the tweeter that always buzzes on the same part of the music, every time. There are people I've learned who really hate the fact that Glenn Gould vocalized along with his piano playing - another example of correlated (in his case though the correlation wasn't particularly strong!). But a noisy AC unit - not a problem really. Nor the sound of the sustaining pedal being lifted and re-pressed. If I hear that I know I'm getting better transparency, closer to headphones where that effect is often more obvious.
I agree entirely! I have found that many "bad" recordings have distortion which is not correlated, transcriptions from 78s are an excellent example -- cyclic clicks and other transgressions of the disk, the pops and crackles, are shifted into another acoustic space where they can be ignored.
You might say that there is a lot of distortion which is still correlated, say relatively primitive microphones which had the vocalist too close, obvious overloading on sound peaks. But my experience is that the better the system reproduction, the more the brain can separate the tonality, the spectrum of the raw music producing device, whether voice or instrument, from added distortion imparted by the recording and storage mechanisms. Somehow, when enough of the low level detail can make its way through the reproduction process without being significantly further degraded, a switch can flick over in the brain which allows it to seemingly effortlessly separate the music from these distortion strands. In spite of the fact that there is relatively strong correlation.
I've found that the "big trick" is to make it as easy as possible for the ear/brain to access all the content of the recording, with nothing added, which can only occur if the playback chain is scrupulously clean in its working ...
Frank
I don't call those kinds of recordings 'bad', rather 'historic'
To me a bad recording is one with harshness - sibilance - and that is very obviously correlated with the signal. Another would be one with reduced dynamics - aka noise modulation - very definitely a correlated thingy.Haven't really come across such. I do have a Chesky Records recording which for a long while I'd used as one of my tests. On loading it into Audacity I noticed its clipped at a couple of points. Hardly noticed it!
If the measure is apparent harshness then we are down to the heavy lifting area
! To put it bluntly, my goal is to eliminate the subjective impression of harshness in every case, not by adding honey on top, but by cleaning up the system distortion. Every time I thought I had a recording which was irredeemablely harsh, impossible to rescue, I was always ultimately proven wrong, it was really a more powerful magnifying glass, demonstrating that I still had work to do! Yes, just about everything seems to get clipped, at one or two points. Or, effectively so.
Clipping like this is very clean, that is, effectively unnoticeable. What is noticeable, is when the power amp compresses because its voltage rails are starting to get very ragged, this is what a lot of people call clipping ...
Frank
If the measure is apparent harshness then we are down to the heavy lifting area
Apparent harshness does seem to be somewhat acceptable to audiophiles of the digital persuasion. They often report 'no trace of digital harshness or glare' as if that's something magnificent to admire - yet the analog die-hards have been there for years.
Claiming that absence of harshness is the goal to me is a sign of considerable dysfunction in the marketplace. Can you imagine if the goal of famous cooks was 'complete absence of strychnine' in their cooking? Or if a restaurant reviewer managed to eat a dish and his key focus in writing about it was 'totally free of MSG'?
Can't say as I find that a realistic goal at present. Cleaning up system distortion for sure, 100% with you, but how can I hope to clean up distortion that's already there on the recording? ISTM only by adding extra honey over the top.
Then I have the very disk for you ....
Agree, you can't touch the recording distortion, but I have found the nature of that distortion is of a type that "intermodulates", shall we say, with that contributed by the playback in a way that is particularly noxious, particularly grating to the ear. Good example, massed brass instruments in the background of a relatively poor end recording, add a good dollop of system misbehaviour, and it can be torture at decent volumes. Yet I've found that by steadily whittling away at the problems at my end, I keep lifting the apparent quality of that brass accompaniment, until it is of an acceptable musical standard. At least, for me ...
And that would be ...Then I have the very disk for you ....
? Frank
Based on my own experience you'll not fix up the sound of brass before you abandon S-D converters wholesale and switch over to multibit. But I haven't heard all S-D converters yet so this is in part conjecture.
And that would be ...
Amazon.com: The Very Best of MTV Unplugged: Various Artists: Music
Yes, you're talking of the usual failing of these converters to handle the upper harmonics correctly. Again, this can be largely overcome by thoroughly conditioning them before serious listening. Plus, try the power down and switch back on after 30 secs reset. Then you can get the intense bite and sheen of the true brass sound, minus the unpleasant distortion overtones.
Because the dynamics have been squashed, creating a very intense, in your face sound? I have a collection from the Jools Holland show stable that may have the same "issues" you're possibly referring to ...
Frank
Last edited:
I already said no thanks to the conditioning suggestion - the customers won't buy it. Nor will they buy resets. Besides, multibit DACs are a cheaper overall solution for me, perhaps because they're old designs and currently deeply unfashionable. No S-D DAC gets close in terms of cost effectiveness.
No - the first reviewer puts it very well, let me quote:
The recording and mastering of this album is so bad that I get an headache if I listen to it with an headphone.
It has pure sibilance, on every track. Compression would be heaven compared to this.
I was curious what searching diyAudio would bring up about sibilance ... there have been debates about whether excessive such characteristics were FR aberrations, or distortion. Well, I come down strongly on the side of distortion, and in particular issues with the playback chain. Yes, particular recordings will provoke that misbehaviour more, perhaps much more, than others, but ultimately they are doing you a favour: pointing out where there are weaknesses in the setup, the idea I mentioned earlier.
Put it this way: my system when humming puts out very intense, dynamic high frequency content, but the word "sibilance" never pops into my mind ...
FRank
If you ever record singers in a studio you'll find that avoiding sibilance can be a major issue and sometimes it is impossible to fix by mic choice or placement.
The last resort in those cases is either a de-esser (a frequency specific compressor) or digital editing.
Or leaving it in because sometimes the cure is worse than the illness.
Basically what I am saying is that true sibilance is a problem that arises at the mic before any electronics touch the signal although some electronics and mic capsules exacerbate the problem, usually through tiny frequency response anomalies as every mic capsule essentially superimposes its own eq curve onto the incoming signal.
The last resort in those cases is either a de-esser (a frequency specific compressor) or digital editing.
Or leaving it in because sometimes the cure is worse than the illness.
Basically what I am saying is that true sibilance is a problem that arises at the mic before any electronics touch the signal although some electronics and mic capsules exacerbate the problem, usually through tiny frequency response anomalies as every mic capsule essentially superimposes its own eq curve onto the incoming signal.
So let me see if I have this clearly. You're saying that playback chains can and do have weaknesses which provoke sibilance, but recording chains never do?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.