Conclusive "Proof" that higher resolution audio sounds different

[RogerD]

I know what an implication is. The issue is with 'profound'. Meyer and Moran (2007), which Your Host has now dismissed as superseded(!), already provided evidence that audible difference can occur between RB and hi rez versions (even apart from obvious mastering difference) -- *depending on how the signals were made and played back*



Meanwhile, the rather less well-controlled forum test results are being touted as 'conclusive' proof of a very definite claim, and an 'inflection point' , (I'm waiting for 'game changer', did I miss it?) by a party that sells 'high end' gear; and it took all of 5 minutes of reading this thread to bump into someone claiming they hear 'night and day' differences between Redbook and hi rez, routinely.

LMAO Sometimes I think people listen to boom boxes.......

 

[Orb]

JohnW is still posting on PFM but he seldom if ever posts outside of the MDAC thread which is his latest DAC. Anytime I have seen him in the past post on other threads, a number of the usual suspects on there, who are of the same mindset as Tim (as he has just expressed), have posted insults & abuse at him. So one of the many things in Tim's post that I don't agree with is his characterisation of PFM - there are a number of very vocal people of the same mindset as he himself.

And sadly they do not realise just how much knowledge and expertise he has, along with a close colleague who is a mathematician and does a lot of the work on coefficiants-etc.
If I remember John identified a bug within one of the ESS Sabre DAC chips while working on the MDAC and spent ages trying to develop a workaround; cannot remember the bug unfortunately.
Cheers
Orb

 

[Phelonious Ponk]

You cannot take the -95db and say ah it is inaudible for jitter...
Fuse blowing in my head so dropping this subject but please search for papers by Agilent/HP/Wolfson/AD/Julian Dunn/etc.
BTW some of the best digital engineers have posted on Pink Fish and at great length (John Westlake being one who spent a lot of time sort of like a blog there on the development of the Audiolab DAC and technical challenges-considerations-etc).
Thanks
Orb

Perhaps not. It's not great performance, of course, but in the context of real music, the likelihood that anything at -95db would be masked is pretty high. And it's a mass market AV receiver running HDMI! Sorry to blow your mind fuse.

Tim

 

[jkeny]

And sadly they do not realise just how much knowledge and expertise he has, along with a close colleague who is a mathematician and does a lot of the work on coefficiants-etc.

Yes, he gets much abuse when he sticks his head above the parapet. Are you speaking of Werner?

If I remember John identified a bug within one of the ESS Sabre DAC chips while working on the MDAC and spent ages trying to develop a workaround; cannot remember the bug unfortunately.
Cheers
Orb

Yes, I can't remember the bug either

 

[jkeny]

Perhaps not. It's not great performance, of course, but in the context of real music, the likelihood that anything at -95db would be masked is pretty high. And it's a mass market AV receiver running HDMI! Sorry to blow your mind fuse.

Tim

It's great to deal with single figure value & then say that they will be masked, makes it all nice & simple, doesn't it?


Mod: post edited.

 

[amirm]

Is anyone taking note of the Y axis here?

One sideband peaking at -95dB; the rest at < -100dB

Oh , such FAILS.

??? I was just providing an example comparison of HDMI vs S/PDIF not any kind of worst case scenario.

Here is one from my own measurements:

The cursors at the bottom show the actual amplitude of those peaks: -72, -83, -84. If the DAC chip designer saw those performance metrics, he would have a heart attack!

This is a reference to my ancient 15 year old Mark Levinson DAC:

So yes, FAIL and then some.

 

[Orb]

Perhaps not. It's not great performance, of course, but in the context of real music, the likelihood that anything at -95db would be masked is pretty high. And it's a mass market AV receiver running HDMI! Sorry to blow your mind fuse.

Tim

And again Tim,
look back to the context...
What was Amir talking about and to whom?
Both were NOT talking about audibility nor real music but how simple DACs are and assumption mass produced manufacturers can build them perfectly.

Just to put my own view across; yes technology (includes build process along with design) has moved along because even £150 or less in cost DACs can have great performance (jitter being inaudible) being as low as possible with just below 120psec CD quality and 10-20psec for 24bit; I am going by figures for products measured/reviewed and were not mass produced manufacturers but audio focused ones.
So more things have improved greatly over the last decade, but that is not the context of Amir's post and the one he responded to.
Anyway these days it is about optimum digital filter used and understanding compromise between fast-slow rolloff/minimum-linear phase for some of the differences between DACs, along to a certain extent the analogue circuitry (be surprised how many have very different noise regarding noise-distortion to dbfs for very higher frequencies such as 20khz compared to 1khz).

I would say we agree a lot that jitter is really not the issue it used to be, especially for traditional interfaces; USB can still throw some curveballs, for some so does LAN, and as Amir shows HDMI is a problem child and one that should be shunned
Thanks
Orb

 

[Orb]

Yes, he gets much abuse when he sticks his head above the parapet. Are you speaking of Werner?
Yes, I can't remember the bug either

Yeah and yeah
Well think it is Werner lol.
Cheers
Orb

 

[Phelonious Ponk]

It's great to deal with single figure value & then say that they will be masked, makes it all nice & simple, doesn't it?


Mod: post edited.

The only relevant response to a single figure is one about that single figure, John, particularly when it is presented as singular evidence of the inferiority of DACs in consumer AVRs. If you have other figures that better reveal the inferiority of such DACs, you can always present them. If anyone here has done the research to uncover opportunities to make audible improvements in DACs, it should be you, as that is your business. Surely in product development you measured and found the weaknesses of DACs that came before your own, so you would know where to apply your efforts to produce an audibly superior product. Show us. That would make a much more interesting thread than telling us we're over simplifying.

Tim

 

[Don Hills]

... HDMI is a problem child and one that should be shunned ...

No. It should be fixed.

 

[Orb]

No. It should be fixed.

For sure but until then it is still a problem child
Please appreciate I was talking about end user and not implementation standards-specification/committees and manufacturers; and context really in this regard is dedicated stereo audio transmission rather than audiovisual.
Thanks
Orb

 

[jkeny]

The only relevant response to a single figure is one about that single figure, John, particularly when it is presented as singular evidence of the inferiority of DACs in consumer AVRs. If you have other figures that better reveal the inferiority of such DACs, you can always present them. If anyone here has done the research to uncover opportunities to make audible improvements in DACs, it should be you, as that is your business. Surely in product development you measured and found the weaknesses of DACs that came before your own, so you would know where to apply your efforts to produce an audibly superior product. Show us. That would make a much more interesting thread than telling us we're over simplifying.

Tim

But surely you don't want me to present measurements when a bit of logic is all you need? The plot that Amir showed of spuria -72, -83, -84dB down from a 12KHz tone. It really doesn't take a great leap of logic to realise that with music as the signal we will have spuriae at all sorts of frequencies. To claim that all these spuriae will all & always be masked by the tones in the music itself is quiet some leap of faith.

 

[Phelonious Ponk]

We could always use component video instead. But then we'd have to worry over whether the video DAC in the AVR is better, or the one in the TV is superior. And if we conclude that the one in the AVR is better, then we'd have to worry over signal lost and noise gained in the (sometimes) long runs from the AVR to the TV. But I'm sure someone out there would be able to supply us with a high-end, free-standing DAC. And by free-standing we mean standing rather expensively in at least two boxes. Minimum. And someone else will be able to supply us with a component video cables as expensive as the AVR itself, and assure us that...oh no, it doesn't show up in measurements...no, no, you won't actually see an improvement in video noise or distortion...but surely you can see how much deeper the sight stage is? No? You might need to upgrade your TV, your BR player and your AVR to see the benefits of my wire...and perhaps a converter in three boxes?

I know some of you will appreciate the humor and many are skilled at ignoring me. To the few that can't, it's just a bit of fun, boys. Try not to take offense.

Tim

 

[Phelonious Ponk]

But surely you don't want me to present measurements when a bit of logic is all you need? The plot that Amir showed of spuria -72, -83, -84dB down from a 12KHz tone. It really doesn't take a great leap of logic to realise that with music as the signal we will have spuriae at all sorts of frequencies. To claim that all these spuriae will all & always be masked by the tones in the music itself is quiet some leap of faith.

I see your logic, but like Amir's example I see little reason to expect it to rise above the noise floor. Possibly my ignorance. Surely you gathered this kind of data during product development? It would make your point clearly, decisively Why would you not use it?

Tim

 

[jkeny]

I see your logic, but like Amir's example I see little reason to expect it to rise above the noise floor. Possibly my ignorance. Surely you gathered this kind of data during product development? It would make your point clearly, decisively Why would you not use it?

Tim

As far as I know nobody has measured the jitter spuriae resulting from using a music signal as source - for good reasons?

 

[jkeny]

I see your logic, but like Amir's example I see little reason to expect it to rise above the noise floor. Possibly my ignorance.

Ah, so you mean the noise floor will mask it? I think Amir has something to say about the noise floor of RB as not being sufficient for our needs (-96dB) & you are saying that -72dB will be below the noisefloor?

 

[Phelonious Ponk]

As far as I know nobody has measured the jitter spuriae resulting from using a music signal as source - for good reasons?

No DAC manufacturers have measured the performance of their devices doing the job they were intended to do? Really? Why? Is it too difficult? Too expensive? What did you use in product development to measure the jitter performance of your products?

Tim

 

[jkeny]

No DAC manufacturers have measured the performance of their devices doing the job they were intended to do? Really? Why? Is it too difficult? Too expensive? What did you use in product development to measure the jitter performance of your products?

Tim

Tim, you need a defined signal as source in order to accurately measure spuriae at the level that jitter produces.

Here's another thought for you - if the jitter products with Amir's signal produces 3 very noticeable spuriae with a single frequency input signal then real music signals, with many frequencies, will produce a whole grass of spuriae. What does a grass of spuriae = an elevated noise floor & one that modulates with the signal. Now consider what modulating noise floors do to our psychoacoustic perceptions - the finer details become hidden, the start & end of notes become shortened, the dynamic range of the music gets reduced. These & more effects have further reaching consequences to our perception of the music than the simple statement "a rise in noisefloor" might suggest.

 

[amirm]

Lossy encoding will never be acceptable as an archiving format. But as a consumer format, it is *effectively* 'transparent to most listeners, if done right.

In this forum, we discuss what is best. Not what is good enough for "most listeners."

The real ding to audio quality of music for sale, unfortunately has nothing to do with lossy/lossless or 'resolution'. It's the production and mastering.

That's neither here, nor there. Lossy compression degrades the quality of both good and bad production/mastering. At any rate, is not subject of any debate.

 

[Phelonious Ponk]

Tim, you need a defined signal as source in order to accurately measure spuriae at the level that jitter produces.

Here's another thought for you - if the jitter products with Amir's signal produces 3 very noticeable spuriae with a single frequency input signal then real music signals, with many frequencies, will produce a whole grass of spuriae. What does a grass of spuriae = an elevated noise floor & one that modulates with the signal. Now consider what modulating noise floors do to our psychoacoustic perceptions - the finer details become hidden, the start & end of notes become shortened, the dynamic range of the music gets reduced. These & more effects have further reaching consequences to our perception of the music than the simple statement "a rise in noisefloor" might suggest.

Even if it's all at -95db or less?

So your primary objective, as an audiophile DAC designer, is to produce a product with better performance than the average DAC in the average AVR, right? A single measurement, like Amir presented here, doesn't give you nearly enough information, I get that. So in order to develop a better than average product, you must measure a defined signal at at many points up and down the audible range, yes?

Tim