Why is digital audio so complicated? Where did it all go wrong?

[Hifi_swlon]

I would love to take part i something like that, to see if a raspberry Pi DAC/Amp versus a Naim Statement or whatever could be picked out blind, I find it fascinating.  Anyway, drifted off my own thread!

Another interesting bias is the reliability of our memories in these fields : https://www.youtube.com/watch?v=XgRlrBl-7Yg

Yes, I just don't think you get these situations in digital imaging - not to the same extent anyway - and I think the primary reason is you can compare things side by side at the same time, whereas audio doesn't allow that. And also its considered that the measurement tools are adequately accurate. That doesn't get us round which people 'prefer' but it does make it obvious which one is closer to achieving the desired goal.

Thanks for the video btw, I'm only 5 mins in and am going to have to do the rest tonight but am hooked!

[Antoine]

What makes it all complicated and confusing is discussions like these and the want to be able to fully understand and explain it all. No offense!

Personally I can often only understand part of it, explaining it to others is even harder. While I still would like to understand it all I accept that I, many others and often even manufacturers can't explain in all cases what's going on exactly. We can often at best hypothesize about what's going on. The result is what counts to me.

Thinking we're there and understand it all in our science (which is not the same as engineering) is naive. Again no offense. In a hundred years they'll laugh at our level of scientific understanding, another 100 years later, the same will happen to that generation etc. etc. Scientists often are, in a way, the most stupid (or even corrupt) people on earth, denying/not seeing whatever's right in front of their noses. Creative thinking 'outside the box' is something -many- (so not all) scientists of our age are unable to do, their educations made sure of that. (Unfortunately I've seen it up close/been part of it for some time)

In this hobby I personally like to learn (and thus benefit) from others who've done the painstaking work of experimentation. The good work will always find it's way to the surface, accepted science and even widely adopted engineering principles will no doubt catch up later.

[Hifi_swlon]

What makes it all complicated and confusing is discussions like these and the want to be able to fully understand and explain it all. No offense!

None taken, I love this site and a bit of banter. That said, I can't see how discussing and trying to understand it can make it more complicated or confusing…. to me hearing thoughts/opinions/experiences of others is learning - and a lot quicker/cheaper in this case than trying to do it by experimentation myself

It's a complex one I know. Personally, I think there is some truth to these things, but also a healthy dose of bias - thats the conclusion Im coming to. I also think many times it could be a case of a box that reclocks or regenerates, or a new cable, or a new power supply, may add an improvement indirectly somehow, rather than because of what its actually designed to do. I don't know.

I just can't convince myself that so many things plague a digital data stream, and yet I can't convince myself it isn't the case because I hear differences,and so many others also report hearing them. On Monday night for example I rebooted my mac mini, and since then it sounds better as far as I can tell/perceive, more detailed, less 'clouded'. I've experienced this previously too. Did this actually happen - the sound improved - or did I imagine it because I watched it reboot and have experienced a difference previously? Would I have noticed in the same way if someone else had rebooted the mac? I have no way of knowing.

Thats why I'm trying to understand the digital data signal integrity clock noise thing better, because I find it hugely interesting and extremely frustrating at the same time.

[Bartype]

I can only admit it's true, digital audio seems to make playback easier but it isn't true when it comes really to it. I am sing a self build silent PC installed with Windows 2012 server and Jriver. At first try i had build in a hifi grade USB card (JCat) and connected the PC to the Devialet this way. But the sound was lacking, while with Air there was so much more.

In my search for something to make the Devialet sounding even better i am considering products like the Melco's, these are said sounding better then air. But these Melco's seem still lacking in being easily controlled. If i understand it correctly, you still have to feed a Melco from a PC/Mac, exactly the one thing i want to remove from the link.

So for now the PC with windows 2012, Jriver and Air via ethernet have to do. At least it seems there is no white noise issue with this configuration so that's something!

[Antoine]

@Hifi_swlon,

Of course you are right. What I meant was that IMO "meta discussions" like this one will rarely provide satisfying answers. There's simply no ready answer that answers it all. These discussions generally also attract people with opinions but no experience clouding the discussion and creating confusion.

It is indeed a complex one. Personally I'd rather focus on a specific subject and read, read, read about it. Did I already mention reading about it? Then, before advancing or changing to a different subject, if there's still questions left I'd ask them to peers who are experienced on the subject. Of course there's the own experimentation in between. It's time consuming hard work, well kind of, especially if you don't enjoy it. In the beginning it can clearly be overwhelming since there's so many different, subjective opinions and experiences and "things to do'/possibilities. At the same time there's little hard and solid truths. So yes, there's a steep learning curve and one can avoid that by simply buying a turn key solution.

Like I said I have accepted that I won't always understand what's going on exactly or get the answers I'm seeking. I'm happy though that people are sharing the results of their experimentations. There's things I would have never tried several years back and still don't always make any real sense but after trying I was convinced "it did work" or hold truth in practice. We'll get there, manufacturers/designers of commerical hardware/solutions are also reading these forums. In a few years many of the things 'those tweakers' do today will have become broad and commonly accepted engineering principles or at least inspired those changes.



@Bartype, A PC with similar specs and the right software configuration, using USB should IME clearly outperform AIR. The devil is in the details and with USB a lot comes down to the power supply and avoiding ground loops. Audiophile Optimizer is a "must have" on W2012. I'm already off topic so will keep my comments limited to this but it seems there's still work to be done in your situation. Whether you want to do that is up to you of course.

[Hifi_swlon]

….i am considering products like the Melco's, these are said sounding better then air. But these Melco's seem still lacking in being easily controlled. If i understand it correctly, you still have to feed a Melco from a PC/Mac, exactly the one thing i want to remove from the link.

I'm not one of them, but there are a few Melco users here. My understanding is it doesn't require a computer to be connected- you can use a tablet to directly control it.

[Krisp]

You initially asked why it is so complicated and than you asked whether the digital path is reliable and than the discussion went off.
First we have to separate between things that can be verified and things that cannot.
Lets keep with the assumption that the digital path from the source (we start here with the CD or file and not the instrument in the recording studio) to the output of the DAC is error free (as long as the CD has no read errors). This is assured by transport protocols otherwise computers will not be able to work as a predictable machine.

Than we have the analog path from the DAC to the speakers through the listening room into our ears. Here we have all the various influences that are experimented with for decades.

At the end of the path there is our human music perception and this has other variations which are not constant over time. There are psychological influences (e.g. stress), emotional / hormone influences, chemical /medical influences (drugs or alcohol), fatigue or aging influences and more that I am convinced of make absolute comparison of music playback more or less impossible.

So the best is to get a reasonable gear which does technical the best to reproduce the source signal like our Devialet Experts, use Dirac for room correction and just enjoy the music.
You are getting older each day and your ears loose frequencies. Enjoy it as long as you can because when the time comes the ears will get so decayed that you might not be able to enjoy music listening anymore.
So save your time and money. Listening to music is an enjoyable hobby. Searching for perfect sound will never satisfy you, since you will never find the end.


Cheers,

Krisp

[Antoine]

Lets keep with the assumption that the digital path from the source (we start here with the CD or file and not the instrument in the recording studio) to the output of the DAC is error free (as long as the CD has no read errors). This is assured by transport protocols otherwise computers will not be able to work as a predictable machine.

Let's not as it is not true. The digital path is at best error free from the source up to the input of the DAC. Better said, the D to A conversion itself is not lossless.

https://en.m.wikipedia.org/wiki/Digital-..._converter

I agree that the transport of the bit stream into for example an USB receiver inside the DAC machine is easy to do right, as in bit-perfect.

[Hifi_swlon]

OK, Krisp, Antoine, excellent - this is what I'm getting at.

So, is it true that the digital path from source to the input of the DAC is error free? (Antoine suggests this is 'at best' error free). I mean if there are no dropouts can we assume this, or is it not that simple in audio data?

Assuming error free transmission, how can it be made to be 'better' that having arrived error free?  I think this is the bit I'm really hazy on.

I understand there are a myriad of things involved on the D>A that all effect the final analogue sound.  But the starting point of that process - the final data the DAC uses in its buffers or whatever before it starts its conversion work - how can that be bettered than arriving error free and within the tolerances of the buffer memory of the DAC? Here I'm lost.  Is it that the more work the DAC has to do to 'control' this data, the more noise it generates? If so, why? And is this what we're talking about with all these server tweaks - re-clocking etc - that's the issue?  Are there any articles etc written on this?

And yeah, Krisp, I agree with most of what you said re listening and enjoying and the quest.  I believe I've been on that quest a few times, but even though I will still tinker, I am just trying to enjoy the music (in my system the room EQ and speaker choice seems to overpower any other tweaks I make anyway) but the real quest now is for knowledge and understanding.

[Antoine]

Krisp suggested it to be error free from source to DAC OUTput. I'm saying it is at best error free up to the D to A conversion. And with error free I mean bit perfect while we're in the digital domain.

The bitstream signal representing the "ones and zero's" is an analog signal. With it travels noise. The signal also distorts while "in travel". The noise and distortion influences the D to A process. Noise, for example in the ground plane, also influences clocks inside the DAC, causing jitter.

Of course no one can explain it all in just a few sentences and I'm retiring from this topic. I have to be honest and admit I don't enjoy these types of discussions. The language is a bit of a barrier for me too. Also I'm enjoying listening to my system now and typing on an iPad which is not ideal. Going back to fully passive mode now!

John Swenson is one of the gifted men who has written plenty or articles on these matters, so I'll just refer to those as a nice starting point. Of course lots more can be found on the net.

http://www.audiostream.com/content/qa-jo...at-digital
http://www.audiostream.com/content/qa-jo...-just-bits
http://www.audiostream.com/content/qa-jo...fect-sound
http://uptoneaudio.com/pages/j-swenson-tech-corner