An Edifying Journey ...

[fas42]

A quick update ... close to zero further progress in getting that signal cleaner mechanism done - various things got in the way, like hardware playing up, and I lost the motivation to pick it up again; there's nothing to stop it happening, apart from a decent injection of enthusiasm ...

 

Last two days of listening to the rig tell a story very familiar to most ... hadn't run it for a bit, and the first day started well; but then it just didn't "feel right" - wasn't in the mood to major checking - and gave it away. Next day, nothing was altered - but it fired up nicely, and sustained. Was in top form, Bev was bouncing around the room to the high energy stuff; 'more difficult' classical recordings were flawless.

 

Why? Could have been simply that I had left it too long unused, and the circuitry needed more conditioning then I was in the mood for, on the first day. Or, maybe one of the subtle tweaks was just slightly better aligned, when I started the next day, and that made all the difference ... it's a tough game - but when it comes together, there's nothing better!!

[fas42]

Still nothing happening ... I've been distracted too much lately by what's going on in the world of creating convincing visual presentations ... sound familiar ? We're on the edge of being able to capture anything that the eye can see, and then present a photographically realistic version of that, to explore; finally, a 'good enough' "digital twin" of the world.

 

Yes, there have always been ways of doing that, to some degree of accuracy. But anyone who has explored what was possible quickly found that the results were often very clumsy, downright poor; or required major expenditure of money and time to get it right. This has changed very significantly in recent months - the process of 'encoding' is still pricey and evolving, but the playback is very, very impressive; and can work as fast as one's head can move, so to speak, on relatively ordinary hardware. Meaning, the ability to explore visual spaces "as if you were there'" is on the horizon ... sorta like, what one can do with auditory captures, .

[fas42]

I deeply despise mobile phones ... because, they never work. If you don't use one constantly, have been well and truly conditioned, by the phone, to soak up The Knowledge, of the magic sequence of pokes, holds and swipes to make something happen - it's almost impossible, to make it happen . I've spent hours, fighting the stupid things, trying to get something trivial to work - quite often, giving up in disgust.

 

I'm in such a battle now, trying to transfer a file between an Apple iPhone and a Windows laptop - the two are physically connected by a cable; which in the old world of computers meant it would be falling off a log simple to get a file from one to the other ... you very silly fool, you!! Don't you realise you're living in world of magic, the world of 2023 - where unless precisely the right incantations are uttered, you shall surely fail ... .

 

I have tried about half a dozen things so far, trying to wrestle the two to cooperate ... which have all, of course, failed. Next step, spending hours with Google, sifting through all the sad stories posted by other poor souls - hopefully, somewhere, somewhere, there is a shred of a clue about how to get this mighty simple thing to take place, .

[fas42]

Hey!!! ... finally got it! Had to send it to iCloud, which required several goes to unlock - hadn't used it enough, so it refused to talk to me, until I jumped through the hoops, several times - downloaded it to one machine, which won't talk directly to the other - so, back up again to the cloud, this time Google Drive - and finally where it needed to be, on the laptop.

 

Okay, interesting artifact: CD of Swing Boogie Woogie - many weeks ago I idly picked up Bev's old iPhone, and recorded a chunk of playback ... and tried to end the recording, but unbeknownst to me, was still recording - went to the kitchen, checked a cupboard for soup, , and finally realised it was still on mic!

 

The point? First job of replay is to deliver the sense of the recording, and the 2nd track that comes on is this,

 

 

Different volumes, etc, but what I'm getting at is that the playback is largely not imposing its personality on what you hear ...

 

And the clip:

[fas42]

Interesting exercise a day or so ago ... bought a handful of CDs from an op shop, $1 each, as I do when opportunity arises. Effectively anonymous labels that were bargain basement when bought brand new, with "invisible" artists, on two: famous Chopin pieces - tick; Broadway hits - this was quite brilliant; top notch stage musical voices, great arrangements; of The Phantom of the Opera recording standard. Then a CD of Bach organ Tocatas, etc - solid, 'heavy' organ sound; not as impressive an instrument as an other recordings, but satisfying enough.

 

Then, the 'recognised' item - a Shania Twain album ... and did this ever show how critical it is to get jitter under control - took about 6 reloads of the disk to finally get the 'musicality' to a good enough place. The arrangements of her songs use lots and lots of depth and other effects - and the slightest lacking in the transfer of the signal to the speakers is, ummm, glaringly obvious . No, not actual "glare", but there's an edge to the SQ which means that you can't, er, immerse yourself in the sound, . This of course shows how essential it is for me to finally get that optical waveform cleaner working - but as stated before, I'm distracted currently by playing with software that creates illusions in the visual sphere, . ... So, down the track ...

[fas42]

On 11/16/2023 at 8:55 AM, fas42 said:

Still nothing happening ... I've been distracted too much lately by what's going on in the world of creating convincing visual presentations ... sound familiar ? We're on the edge of being able to capture anything that the eye can see, and then present a photographically realistic version of that, to explore; finally, a 'good enough' "digital twin" of the world.

 

Yes, there have always been ways of doing that, to some degree of accuracy. But anyone who has explored what was possible quickly found that the results were often very clumsy, downright poor; or required major expenditure of money and time to get it right. This has changed very significantly in recent months - the process of 'encoding' is still pricey and evolving, but the playback is very, very impressive; and can work as fast as one's head can move, so to speak, on relatively ordinary hardware. Meaning, the ability to explore visual spaces "as if you were there'" is on the horizon ... sorta like, what one can do with auditory captures, .

 

This is a fascinating world to follow, right now - the pace of development is quite astounding; in a month, the difference in what's now being done is remarkable ... compare this to the feeble speed of audio development, .

 

Right now, enter a text description of a scene, or a "view of the world", and a 3D, pretty damn good version of that will be available to explore once the software's stopped grinding - the paper's called, LucidDreamer ... .

 

Why is it happening so fast? The code to play with is available for everyone; all the young, bright things at universities are going hell for leather, pushing the boundaries - so far away from the old, 'closed', commercial world of audio.

[Racerxnet]

On 12/2/2023 at 5:52 PM, fas42 said:

 

Then, the 'recognised' item - a Shania Twain album ... and did this ever show how critical it is to get jitter under control - took about 6 reloads of the disk to finally get the 'musicality' to a good enough place.

Open and close the loader drawer 6 times and you will be in heaven. Got it!

[kumakuma]

16 minutes ago, fas42 said:

 

This is a fascinating world to follow, right now - the pace of development is quite astounding; in a month, the difference in what's now being done is remarkable ... compare this to the feeble speed of audio development, .

 

Right now, enter a text description of a scene, or a "view of the world", and a 3D, pretty damn good version of that will be available to explore once the software's stopped grinding - the paper's called, LucidDreamer ... .

 

Why is it happening so fast? The code to play with is available for everyone; all the young, bright things at universities are going hell for leather, pushing the boundaries - so far away from the old, 'closed', commercial world of audio.

 

So you believe that 99.9% of audio hardware is unable to reproduce what's on a CD but video hardware will soon be able to provide us with a photographically realistic, digital twin of the world?

 

 

 

 

 

 

 

 

 

[fas42]

1 hour ago, Racerxnet said:

Open and close the loader drawer 6 times and you will be in heaven. Got it!

 

That's why there are never simple answers! The CD transport is only a cheap DVD player, and it produces too much, classic, jitter when the disk is not seated as well as it can be - which the speakers are too sensitive to ...

 

Solutions:

 

a) Acquire a very well engineered CD transport - costs plenty of money

 

b) Buy much better active speakers which are more able to handle jitter noise; say, the Dutch & Dutch might be "good enough" - costs plenty of money

 

c) Insert a jitter "cleaner" on the line between transport and speakers; will allow other sources to be input - current project, takes time, on the ToDo list ...

 

d)  "Fix up" either DVD player or speakers to solve jitter distortion - takes time, the components are not worth putting the effort in

 

e) Finally, just play with the CD seating, to get it "good enough" for whatever you want to listen to - for the moment, it fills the gap,

 

Depending on how someone wants the hobby of listening to music to progress, choose any of the above ...

[fas42]

1 hour ago, kumakuma said:

 

So you believe that 99.9% of audio hardware is unable to reproduce what's on a CD but video hardware will soon be able to provide us with a photographically realistic, digital twin of the world?

 

 

Where the audio hardware fails is that it messes up the fine detail - just like any, current visual attempts to mimic real life, if you know where to look, it's pretty obvious that the creators of the clip have 'faked' it. We humans are very sensitive to that "fine detail" ... you've got to get it, right.

 

The video hardware is not the issue - there are graphic cards around which are capable of completely fooling us, if they are fed the "right stuff". And the latter is the tricky bit, 'encoding' exactly the right type of detail, with minimal flaws, to create convincing, visual illusions. And that's software ... having the algorithms finely tuned to capture, create precisely what's needed, and no more.

 

In audio, that detail is already there. In the recording. In this case, it's a hardware issue to solve achieving sufficient accuracy, so that the ears hear that detail well enough to recreate, in your head, what happened in front of the microphones, etc.

[kumakuma]

4 hours ago, fas42 said:

The video hardware is not the issue - there are graphic cards around which are capable of completely fooling us, if they are fed the "right stuff".

 

Specific examples of hardware that does this, please.

[fas42]

5 minutes ago, kumakuma said:

 

Specific examples of hardware that does this, please.

 

NVIDIA rule the roost in this world - if you want to be creative, some company's card, that plugs into a decent PC box, that has one of their chips, and 16 Gig of video RAM will be necessary, as a bare starting point. To get serious, 24 Gig memory cards, which currently start at AU$3,000 would be required - this is what movie studios use. At the mad end of the scale, you can pay AU$30,000 for the absolute "best of the best" - which just looks like any other video card. These cards chew power, lots of it, so the PC has to have a muscular power supply.

 

But this is not necessary for replay - I have an "ancient" GeForce GTX 980 card, with 4 Gig video RAM, worth perhaps a few bucks on the second hand market - this, with the right viewing software, gives completely fluid motion through a small, photorealistic 3D scene. The next steps, which are being vigorously investigated, are all about software - streaming Levels Of Detail data on the fly; so that, say, you can be far above the earth, and steadily "fly in" to your backyard; and finally have a single grouping of flowers in a garden bed filling the screen - looking as good as a photo. Yes, Google Earth on steroids ... it's getting the visual data together, and some server of that, to pump it out as required, that will hinder realism; not the hardware at your end.

[kumakuma]

14 minutes ago, fas42 said:

 

NVIDIA rule the roost in this world - if you want to be creative, some company's card, that plugs into a decent PC box, that has one of their chips, and 16 Gig of video RAM will be necessary, as a bare starting point. To get serious, 24 Gig memory cards, which currently start at AU$3,000 would be required - this is what movie studios use. At the mad end of the scale, you can pay AU$30,000 for the absolute "best of the best" - which just looks like any other video card. These cards chew power, lots of it, so the PC has to have a muscular power supply.

 

But this is not necessary for replay - I have an "ancient" GeForce GTX 980 card, with 4 Gig video RAM, worth perhaps a few bucks on the second hand market - this, with the right viewing software, gives completely fluid motion through a small, photorealistic 3D scene. The next steps, which are being vigorously investigated, are all about software - streaming Levels Of Detail data on the fly; so that, say, you can be far above the earth, and steadily "fly in" to your backyard; and finally have a single grouping of flowers in a garden bed filling the screen - looking as good as a photo. Yes, Google Earth on steroids ... it's getting the visual data together, and some server of that, to pump it out as required, that will hinder realism; not the hardware at your end.

 

I have a much better video card than this. What software and data can I use to see this magical effect?

[fas42]

Some examples, here, https://niujinshuchong.github.io/mip-splatting-demo/

 

Note, these are all standard research sets of data being used; you will come across these scenes again and again, because that is how progress is judged, amongst those interested. So, no careful manicuring to remove the "ugly bits" ...

 

https://poly.cam/tools/gaussian-splatting is one website where you can use their computers to work on your data; and see the effort of others, on a variety of scenes. Like all new techniques, a huge range of quality, from amateurs "trying things".

 

[Confused]

7 hours ago, Racerxnet said:

Open and close the loader drawer 6 times and you will be in heaven. Got it!

You need to refer to this:

 

[krass]

19 hours ago, fas42 said:

 

That's why there are never simple answers! The CD transport is only a cheap DVD player, and it produces too much, classic, jitter when the disk is not seated as well as it can be - which the speakers are too sensitive to ...

 

Solutions:

 

a) Acquire a very well engineered CD transport - costs plenty of money

 

b) Buy much better active speakers which are more able to handle jitter noise; say, the Dutch & Dutch might be "good enough" - costs plenty of money

 

c) Insert a jitter "cleaner" on the line between transport and speakers; will allow other sources to be input - current project, takes time, on the ToDo list ...

 

d)  "Fix up" either DVD player or speakers to solve jitter distortion - takes time, the components are not worth putting the effort in

 

e) Finally, just play with the CD seating, to get it "good enough" for whatever you want to listen to - for the moment, it fills the gap,

 

Depending on how someone wants the hobby of listening to music to progress, choose any of the above ...

 

so you believe that the little spinning platter is responsible for clocking the data that comes out of the optical head of a CD transport ?

Oh dear.. What nonsense.

[fas42]

13 hours ago, Confused said:

You need to refer to this:

 

 

Yes, I had escaped from the "jitter nonsense" up to that point - how? Because I always used gear which had the transport in the same chassis as the DAC - no connection issues could arise ... but then, the active speakers route turned up, was too good a deal to ignore, and demonstrated its potential immediately. I looked towards the "usual suspects" for bottlenecks in the SQ, found them, sorted out some fix or workaround - but what was clear, in hindsight, was that there was still too much variability in SQ, for some specific CD, on different days - why? The process of elimination determined that with no other variables in play, that unloading and reloading the CD was the causal link - there was, is a very faint chance that some variation in mains power supply noise is involved, but far more likely is that the optical waveform is the main culprit - I had already determined that the speakers were sensitive to precisely how the Toslink cable was organised; stabilising the optical signal immediately before it enters the speaker cabinet is an obvious path to investigate.

[fas42]

51 minutes ago, krass said:

 

so you believe that the little spinning platter is responsible for clocking the data that comes out of the optical head of a CD transport ?

Oh dear.. What nonsense.

 

The nonsense is the belief that variations in the level of 'effort' by the read mechanism, entailing quite different patterns of current draw from the power supplies, will have zero impact on the behaviour of the circuitry responsible for generating the optical signal. In an ideal world, or in a very well engineered component, electrical noise always stays in the 'box' where it originates, and is not a nuisance - meanwhile, in cost conscious items, or where the engineers don't believe it matters, there are no such safeguards ...

[Racerxnet]

21 hours ago, fas42 said:

 

That's why there are never simple answers! The CD transport is only a cheap DVD player, and it produces too much, classic, jitter when the disk is not seated as well as it can be - which the speakers are too sensitive to ...

 

Solutions:

 

a) Acquire a very well engineered CD transport - costs plenty of money

 

b) Buy much better active speakers which are more able to handle jitter noise; say, the Dutch & Dutch might be "good enough" - costs plenty of money

 

c) Insert a jitter "cleaner" on the line between transport and speakers; will allow other sources to be input - current project, takes time, on the ToDo list ...

 

d)  "Fix up" either DVD player or speakers to solve jitter distortion - takes time, the components are not worth putting the effort in

 

e) Finally, just play with the CD seating, to get it "good enough" for whatever you want to listen to - for the moment, it fills the gap,

 

Depending on how someone wants the hobby of listening to music to progress, choose any of the above ...

You can cut through all the baloney and use a computer with a SPDIF converter. These things are at a reasonable cost and have better clocking than the cheap player. I have a battery powered Jack Kenney SPDIF converter you can have from me. It uses a modified M2TECH. You pay shipping.  Rip/catalogue/play from a PC. Maybe not as fun as puttering around, but it is light years ahead of the player you have jitter wise. Foobar with georgia reborn is a nice setup.

[fas42]

Thanks for those good thoughts, @Racerxnet !

 

However, what I'm really aiming to do is have a jitter cleaner that can handle the most rubbishy input, from the worst of the worst jitter generators - the idea then is that you can hook it up to anything that spits out digital data, no matter how crappy the waveform is, and still get good SQ; it makes it a general solution, rather than one specific to what I have right now. Also, it will in ultimate form handle any input format or transport method, switchable over say 6 channels - anything in the house can be fed to the box.

 

Cheers!

[fas42]

From The Purpose of ...thread,

 

Quote

how do you find time to actually listen to & enjoy music ? Quite astonishing……

 

Just now, switched on these speakers ... and it happened to still have that CD in from last time, the 'difficult' one . It "kicked arse" from the first second; Bev and I were bogeying from the energy, the rich, deep bass backing, the fullness of the sound - not a single complaint ...

 

A sample,

 

 

Poundin' the rafters, with not a single hair out of place ... this is what putting the time in optimising achieves; having tracks like this deliver, wait for it , Big Sound ...

[fas42]

On 12/7/2023 at 5:44 PM, fas42 said:

Some examples, here, https://niujinshuchong.github.io/mip-splatting-demo/

 

Note, these are all standard research sets of data being used; you will come across these scenes again and again, because that is how progress is judged, amongst those interested. So, no careful manicuring to remove the "ugly bits" ...

 

https://poly.cam/tools/gaussian-splatting is one website where you can use their computers to work on your data; and see the effort of others, on a variety of scenes. Like all new techniques, a huge range of quality, from amateurs "trying things".

 

 

I'm enjoying following what's happening in the "digital twin" visual world; the speed at which improvements are occurring puts audio to shame - just released, the latest demos, done by the Google research people; getting ever closer to making photorealistic inspection of whatever, part of the everyday world - the following works quite well on my 'creaking' graphics card:

 

https://smerf-3d.github.io/select_quality?scene=berlin

 

Simplest way to move around is using the gaming keys: WASD, QE, etc

[fas42]

Well, the audio rig is getting a good rest at the moment, ... my head is deeply immersed, heh, in the visual world, still - progress is occurring at breakneck speed, all from the academic world ... the very good thing is that all the latest ideas are published openly, unlike, possibly, the medical arena , and many times usable code is ready to play with soon after, to see how well everything jells.

 

What is so intriguing are the parallels with the audio world - it is possible to mimic some 'real' environment so well, right now, that the illusion of "being there" is not broken, at least via the normal visual methods; just expand the 'projection' to encompass you, add 3D glasses, and, job done. But, it's a capture of what was there, at a specific time - sound familiar?

 

Research is going like crazy, say to expand in the 4th dimension; that is, the time element - you could watch it "as a movie", as a starting point. And to intelligently guess what is missing - only have the front of something, as a single photograph; AI kicks in, and fills in what the behind looks like, a "best guess". What are the limits? Processing power, lots of it, to encode the visual data; and the storage and streaming mechanisms to feed it back to you, seamlessly. Will happen faster than people might envisage could be possible - there is already one project out there, to have people capture what is local to them, and send it to a central repository, to be part of an all encompassing 'database'; quality is too low, currently, to take seriously - but it will evolve ...

 

 

[fas42]

I'm losing forward momentum on the "digital twin" in the visual sphere - just like in audio world, there is so much fine tuning of precisely how things are processed, and getting it exactly right means the difference between being engaging to look at, versus obviously defective ... hmmm? But this means a lot of ongoing, focused attention - and I need a break from this level of thinking.

 

Kicked off the Edifiers rig yesterday, first time in many weeks ... how was it? Again, what was impressive is that from a cold start, after many weeks not in use, it delivered - big, full sound, that could go as loud as you wanted, with no artifacts obvious; threw up a deep, completely out of speakers soundscape, which could be enjoyed completely for the musical content. But ... the old gremlin of going through a flat spot at about the end of the second CD was there, and required plenty of disk reseating in the player, to reduce jitter induced distortion as much as possible. An explanation is that full resolution of what's on the album is not heard when first switched on, and the balance is "just right", to the ears; then the electronics, etc, dig a bit deeper as everything stabilises more fully, and the remaining weaknesses become apparent.

 

Overall, a solid tick. This is the best system I've fiddled with, in so many areas - previous ones required long, looooong warming up to deliver; could sound very ordinary until many hours or even days passed - and it can go high intensity levels, effortlessly. The negatives? Obviously this ongoing need to improve the link between source player and the speakers, so that the nonsense of finding optimum physical alignment in the drive can be forgotten; and, it has only come very close to the best that some of of the older rigs were able to achieve, in terms of the speakers being 100% invisible.