Track preload affects sonics - HELP!!!

[BrokeLinuxPhile]

23 minutes ago, sandyk said:

For real life calls to sound like crap every time it rained, it should have only been possible due to partial short circuits or insulation losses to earth in either the customer's riser or main pair cables, or the cables from the premises to the external pit.

(Unless of course, it was in a rural area with aerial cables)

 Unfortunately, this is more of a problem these days, especially with ADSL,FTN ( Fibre to the Node) etc.  with aging local copper infrastructure and lack of proper maintenance. 

 

Regards

Alex

(Retired Principal Telecommunications Technical Officer-Telstra Australia )

 

True.  I was doing  VoIP testing over power lines via BPL.  Basically undeployable and that's why it failed.  Smallest amounts of ground noise would cause huge jumps in latency for us.  Chasing all that noise in a distribution grid was just herding cats. 

 

Small world.  I helped support Telstra way back when during my Nortel tenure, system testing PPPoEoA aggregation.

[fas42]

1 hour ago, BrokeLinuxPhile said:

I'd put money on things being physiological/psychological more than a physically measurable phenomenon.  Superstition comes to mind, easy to convince yourself something brings good luck.  Or in this case, better sound.    There could be engineering reasoning behind it although much less likely IMO.  It would be a huge amount of effort to find the root cause either way and would the end result be worth all that effort?  Diminishing returns. 

 

If it sounds good, do it I say.

 

Nope. It's just like the telephone situation now being brought up - a little defect somewhere in the system brings the mechanism to its knees; a circuit can effectively stretch around the world - but it's still a circuit ... that little bit of moisture in a junction box way, way, away - does the damage.

 

Audio people believe in Santa Claus  - toss together expensive, impressive looking gear; and by golly, it's gotta sound good!! Ummm, no ... that "bit of moisture in the junction box", out of sight, doesn't give a damn for how fancy the rig appears - and the SQ will be below par ...

 

Yes, it's tiresome to sort out why a setup is not sounding up to scratch - but throwing money at it, increasing its bling quotient, or technobabble ingredients, is not exactly a smart approach ...

[BrokeLinuxPhile]

But the junction box full of water can always be found using a sound engineering approach.  It's noise, and can be triangulated, found, and repaired.  It's just a question whether or not it's cost effective to fix most of the time.  It's also erosion of quality over time, it used to work well and now not so much. 

 

Here we are talking about a sense of 'sounds better' right from the get go without decay or breaking.  Or explainable noise even coming in from anywhere.  We are talking about restarting a single track from a CD and it sounding better.    Perhaps something is going on inside the machine that could cause that, but it's not the most likely culprit.  There would have to be a functional delta between starting a track fresh and hitting the back arrow button from either a hardware or software design standpoint.  It would cost more to do, take longer to develop, and you are always going to have 2 situations, one sounding better than the other.  Not saying it's impossible, just would not make much sense to do.

 

I'd put more money on physical changes inside the ear, call it fatigue or "calibration" to a new source track.

[Ralf11]

actually it already has a name - Kal trotted it out just a week or so ago

 

remember the "sweet spot" is after neural accommodation, but before listening fatigue

[fas42]

25 minutes ago, BrokeLinuxPhile said:

But the junction box full of water can always be found using a sound engineering approach.  It's noise, and can be triangulated, found, and repaired.  It's just a question whether or not it's cost effective to fix most of the time.  It's also erosion of quality over time, it used to work well and now not so much. 

 

And its works exactly the same in audio. Yes, it's a type of noise, but far, far more subtle - instead of sparkling, lifelike sound you get a flatness, deadness, edginess with treble; the quality of 'naturalness' is missing.

 

Quote

 

Here we are talking about a sense of 'sounds better' right from the get go without decay or breaking.  Or explainable noise even coming in from anywhere.  We are talking about restarting a single track from a CD and it sounding better.    Perhaps something is going on inside the machine that could cause that, but it's not the most likely culprit.  There would have to be a functional delta between starting a track fresh and hitting the back arrow button from either a hardware or software design standpoint.  It would cost more to do, take longer to develop, and you are always going to have 2 situations, one sounding better than the other.  Not saying it's impossible, just would not make much sense to do.

 

I'd put more money on physical changes inside the ear, call it fatigue or "calibration" to a new source track.

 

A high likely culprit are power supplies. Loading the CD, and initiating the pick up laser requires significant current draws, meaning that the transformer, etc, waveforms are different from stable playing status - replaying the track is giving all of the power supply circuitry a touch more time to fully restabilise; enough to make a difference.

 

Obviously engineering can make all these behaviours go away. But first the designer has to be aware that these issues can exist; if he doesn't, then it's a toss of the dice whether a particular unit will show signs of audible variations.

 

Having done a bit of experimenting in this area, I can quite easily say most power supplies in audio equipment are not good enough - yes, they nominally do the job, but their misbehaviour is often extremely obvious, in the loss of important SQ.

[fas42]

2 minutes ago, Ralf11 said:

actually it already has a name - Kal trotted it out just a week or so ago

 

remember the "sweet spot" is after neural accommodation, but before listening fatigue

 

The mark of competent playback sound is that it's instantly in the "sweet spot" - and that it never falters to become fatiguing - one of the many reasons to chase this realisable goal.

[BrokeLinuxPhile]

I don't think you are giving the hw designers enough credit here.  Of course they would know about this if it was real, how many people have claimed to have heard this over the years?  Many, this isn't the first reporting of this phenomena by a long shot, i've heard it a bunch of times before this thread.  I refuse to think they have their heads stuck in the sand all this time, they are in tune to the user base. 

 

Surely someone by now would have invented a CD player by now that did not have this problem, CD players have been around in consumer homes for over 30 years now.  An uber high end boutique maker by now would have said costs be damned and fixed it just to say they were the only ones to do it.  That is, if the problem was real, AND it was possible to fix.  Or it's just inherent to the physical properties of the process, and won't ever be fixed because it can't be fixed.

 

Could it be the supply?  Sure.  But I would not call it likely.  On a walmart player yeah, but a player costing more than a grand?  What are you paying all that money for?  It's not all in the transport.

 

[sandyk]

1 hour ago, fas42 said:

replaying the track is giving all of the power supply circuitry a touch more time to fully restabilise; enough to make a difference.

 

Not bloody likely ! 

Even the poorest power supply should have fully recovered even before the short silent period at the beginning of the track has finished ,let alone affect the following audio for even a second or 2..

[fas42]

31 minutes ago, sandyk said:

 

Not bloody likely ! 

Even the poorest power supply should have fully recovered even before the short silent period at the beginning of the track has finished ,let alone affect the following audio for even a second or 2..

 

"Should" is the operative word ... something needs time, or some nudging, to settle into the optimum electrical state for best replay quality; what it is most likely will vary per design, and model - there won't be a unique answer. I recall that I immediately hit pause after play, and left it like that for about 30 secs - the disk was spinning, the laser was on target; it was just hovering on a single read spot - and then I let it off the hook...

 

Yes, I'm certain there will players without issues - my unit that had the behaviour was a well made item from 1986; early days for digital, .

[sandyk]

Frank

 You are grabbing at straws !

This is easily verified by checking the supply rail voltages.

Alex

[firedog]

15 hours ago, Jim Smith said:

My question was an honest request for possible help in an area which is not in my skill-set (computers).  I can make it work, but not at its best without restarting the tracks when they are from different CDs.  It is not all that subtle - I have demo'd it to various listeners and all could hear it easily.

 

Not sure why you cannot offer a helpful suggestion rather than an uninformed diatribe.

 

However, having just seen your system set-up in your profile, I think I may know why after all.   🤔

 

You can have the last word...

 

 

 

Jim-

Not everyone here is nasty for no reason. I hope you don't leave the site b/c of a few posters like Ralf. Others will appreciate your contributions and insight.

[firedog]

On 3/5/2019 at 2:26 AM, Ishmael Slapowitz said:

Honestly, IMO, it is this kind of meshuga that keeps folks away from high end audio in droves.😅

Well both your responses are sort of off topic.

Jim reported his observation. He didn't make some overall claim about audio in general. Why get insulting? If you think the idea is stupid, unscientific, etc. - maybe just don't participate in the thread. It's clear that he wasn't asking for either of the type of responses he got from you.

Or is your ego so fragile you can't deal with that and feel compelled to put in your dismissive 2 cents that contributes nothing to the discussion?

One of the things that "keeps folks away from high end audio in droves" and from this forum are responses like yours.

[firedog]

On 3/6/2019 at 2:47 AM, Ralf11 said:

 

You realize this is the guy with the 10 pages of commonly known stuff inflated into an entire book, right?

 

It's like arguing with a used car salesman about camshaft lift and combustion flame spread velocity.

Either contribute positively to the thread or leave the rest of us alone. Your ad hominem argument is, of course, meaningless to any question at hand. You can think the whole premise of the OP is ridiculous. So what? Do you think Jim came here to get that kind of response? If you don't want to engage in the type of discussion he's asking for you can just ignore the thread.
Or is that too hard for you?

[Miska]

On 3/5/2019 at 2:35 AM, Doak said:

Another good reason, if you need any more, to keep a computer out of your music reproduction system. Any decent dedicated digital music streamer will preload at least a whole song before pressing its own little play button. 

 

Well, first of all each "dedicated digital music streamer" is a computer.

 

But usually a very small short-on-resources one. Have you ever looked how much RAM they have on board? Considered how long it would take for them to transfer entire song for example from Tidal server over the internet, or a hires track within a local network? I have tracks that are several gigabytes in size.

 

[Miska]

1 hour ago, firedog said:

You can think the whole premise of the OP is ridiculous.

 

The part that you start playback, stop and start again is not entirely. But it depends on the player (different players have different file access strategies), OS and the computer in question, and of course how much interference the computer in question has with the actual analog signal reproduction.

 

Difference is that when you start playback, OS disk cache will start pulling in the data into RAM. You stop and it's still there for a while, then you start again and at least part of the data is pre-cached. However, after couple of seconds of playback the OS would have already pulled the file into disk cache anyway. Assuming you have enough spare unused RAM. So any difference disappears at the point where both playback cases reach the same level of caching. Which is roughly couple of seconds into playback.

 

Data itself is of course always the same, no difference there.

[mansr]

Stopping and starting sometimes helps with video streaming using adaptive bit rate that ramps up too slowly. Of course, then the data is different the second time around.

 

The caching Miska mentions is true in principle, but I doubt it has much of an effect in practice. When a file is accessed the OS will pre-cache some amount ahead of the reading position of the application. How much depends on the OS and its settings, but an entire high-res music file will probably not be pre-loaded in one go. As the playback gets close to the end of the cached data, the OS will fetch another chunk off disk. Fiddling around in the first seconds isn't going affect what happens a minute or more into the playback. If disk access is having an adverse effect, there should be dips in quality at regular intervals when playing large files. Then there is the question of how much of an audible effect disk activity has in the first place. I have owned computers with poor built-in sound cards where disk activity was very audible over speakers. In those same computers, installing a $20 PCI sound card made the problem go away. I find it hard to believe that the effect, if any at all, on a USB-attached quality DAC would be anything other than subtle.

[Doak]

4 hours ago, Miska said:

 

The part that you start playback, stop and start again is not entirely. But it depends on the player (different players have different file access strategies), OS and the computer in question, and of course how much interference the computer in question has with the actual analog signal reproduction.

 

Difference is that when you start playback, OS disk cache will start pulling in the data into RAM. You stop and it's still there for a while, then you start again and at least part of the data is pre-cached. However, after couple of seconds of playback the OS would have already pulled the file into disk cache anyway. Assuming you have enough spare unused RAM. So any difference disappears at the point where both playback cases reach the same level of caching. Which is roughly couple of seconds into playback.

 

Data itself is of course always the same, no difference there.

Tracks of several GB are certainly an ultra-extreme example.I doubt that I could find more a very small handful in my large collection.

If you are referring to files that have been oversampled to the max and saved for playback, that too is a very specialized situation.

[Miska]

19 minutes ago, Doak said:

Tracks of several GB are certainly an ultra-extreme example.I doubt that I could find more a very small handful in my large collection.

 

Doesn't change anything. In any case it is easy to monitor network activity to those streamers to see what kind of traffic patterns they have.

 

Most streamers run Linux as OS and many use open source MPD as the player software. So you can inspect the source code to see how it works.

 

19 minutes ago, Doak said:

If you are referring to files that have been oversampled to the max and saved for playback, that too is a very specialized situation.

 

No, I'm not.

[Ishmael Slapowitz]

The vinyl guys who computer 'philes often like to make fun up are laughing their tuchus off right now..😎

[The Computer Audiophile]

5 minutes ago, Ishmael Slapowitz said:

The vinyl guys who computer 'philes often like to make fun up are laughing their tuchus off right now..😎

I think I just saw one fall down after getting zapped by his anti-static gun :~)

[Ralf11]

10 hours ago, firedog said:

Either contribute positively to the thread or leave the rest of us alone. Your ad hominem argument is, of course, meaningless to any question at hand. 

 

I commented on his book with which I am familiar.  Have YOU read it??

 

I made no ad hominem argument at all.  i suggest you look it up

 

you can suck up to him all you want

 

what happened to the no bitching injunction??

[fas42]

14 hours ago, sandyk said:

Frank

 You are grabbing at straws !

This is easily verified by checking the supply rail voltages.

Alex

 

Alex, I'm afraid it is the silly little things that matter - checking supply voltages for the precise nature of the noise that exists on them, something that will constantly change, depending upon what the unit and its electrical surroundings are doing, is going to be difficult to do, and merely tells you whether there is a chance of some sort of interference effect being in the pictue ... what really needs to be solved is that the circuitry that takes care of the audio signal is not affected by what's going in its environment, outside of what it sees as the usual input.

 

If the sound quality changes, unexpectedly, then there is some cause. I would like to be able to punch a time card, for the periods spent over the years tracking down "what's gone wrong" when a rig has lost its current optimum SQ. Quite often it's just stupid little things - I make lots of highly temporary alterations to things, to check whether some factor matters; and sometimes these "come undone", partially. I rarely make something 'permanent', unless I'm 100% certain that doing the mod is absolutely key to the best sound.

[Dutch]

2 hours ago, Ralf11 said:

I made no ad hominem argument at all.

 

It couldn’t be more ad hominem..you started your disrespecful rant with “this is the guy with”.

[fas42]

11 hours ago, Miska said:

 

Difference is that when you start playback, OS disk cache will start pulling in the data into RAM. You stop and it's still there for a while, then you start again and at least part of the data is pre-cached. However, after couple of seconds of playback the OS would have already pulled the file into disk cache anyway. Assuming you have enough spare unused RAM. So any difference disappears at the point where both playback cases reach the same level of caching. Which is roughly couple of seconds into playback.

 

 

 

An exercise I did some time ago was to organise a good media player for my HP laptop. The usual suspects were hopeless - foobar amongst them. Media Monkey turned out right for me - and it was pretty obvious why: watching the resource usage monitor showed how much sharper MM was in not using CPU cycles, and how the access to the drive was done in short spurts - its footprint on the m/c was as little as possible, as a function of time.

[Miska]

8 minutes ago, fas42 said:

An exercise I did some time ago was to organise a good media player for my HP laptop. The usual suspects were hopeless - foobar amongst them. Media Monkey turned out right for me - and it was pretty obvious why: watching the resource usage monitor showed how much sharper MM was in not using CPU cycles, and how the access to the drive was done in short spurts - its footprint on the m/c was as little as possible, as a function of time.

 

Never used that one, I only know my own software and how it accesses files. Of course I burn quite a lot of CPU and GPU time for doing upsampling, room correction and such. Usually steady around ~50 - 60%.