Why 24/192 is a bad idea?

[LenWhite]

I just discovered this interesting albeit somewhat complex article http://www.enjoythemusic.com/magazine/manufacture/0512/ regarding high resolution audio. The author lists reasons why he believes higher sampling frequency and not bit depth is the answer to better digital file audio currently. He states more DSD-capable systems taking advantage of the 1-bit downloads available on the Internet will become more available.

 

[Ethan Winer]

If the response is within 0.1 dB (or 0.2 dB) and all artifacts are 100 dB down, then the device is transparent.

untouched file, -39dB pre-echo 20 samples prior to music & -39dB pre-echo 60 samples prior to music.

these files also show Winer's rules of transparency are wrong!

I've seen people take liberty with facts before, but never this badly. And never when it's so easy to prove they lied. So I say that artifacts 100 dB down are inaudible, and often when only 80 dB down or even louder. Then you post files with artifacts only 39 dB down, and that somehow proves me wrong? Wow, just wow. Your sheer chutzpah is amazing.

--Ethan

 

[NorthStar]

32/352.8 x2=???

-----64-bit/705.6kHz?

 

[amirm]

I've seen people take liberty with facts before, but never this badly. And never when it's so easy to prove they lied. So I say that artifacts 100 dB down are inaudible, and often when only 80 dB down or even louder.
--Ethan

So if the goal is a transport that is inaudible with respect to any content we ever want to put in it, the number as stated by you is 100 db. Correct? Again, I am not asking what is often good enough. I am asking what in your opinion is inaudible in all cases.

 

[Ethan Winer]

^^^ Yeah, 100 dB is certainly inaudible in all cases, and I'm pretty sure even 90 dB is always inaudible. Of course, I can't speak for all 6 billion people in the world.

Why is this so important? And why do you think that the masking effect is not relevant? What exactly are you trying to get at?

--Ethan

 

[amirm]

^^^ Yeah, 100 dB is certainly inaudible in all cases, and I'm pretty sure even 90 dB is always inaudible. Of course, I can't speak for all 6 billion people in the world.

Well, if 90 db is inaudible also why say 100? Is it because you were in front of pro/industry people? And yes, you do need to speak on behalf of anyone who wants to buy audio equipment if you say something is inaudible.

Why is this so important? And why do you think that the masking effect is not relevant? What exactly are you trying to get at?

--Ethan

I say it for the same reason you said it. People are very interested because it sets the bar above which they can sleep easy that they have purchased a system that does not bring into own distortion/noise to the party. It makes it much easier to interpret measurements as one has a nice reference. As I noted in the other thread, my standard has always been "an honest 16 bit system" which would put us at 96 db. Your 100 db number is a bit better than that so in that regard, we have convergence of views and standards which rarely happens in these discussions.

 

[opus111]

I've seen people take liberty with facts before, but never this badly. And never when it's so easy to prove they lied. So I say that artifacts 100 dB down are inaudible, and often when only 80 dB down or even louder. Then you post files with artifacts only 39 dB down, and that somehow proves me wrong? Wow, just wow. Your sheer chutzpah is amazing.

To me its really curious that you'd call a pre-echo 'an artifact'. Normally artifacts are non-linear effects - like harmonic distortion, or jitter sidebands.

But if time delays added back in to the signal really become 'artifacts' then any FIR digital filter whose impulse response includes values for delayed (positive or negative) samples above 10^-5 (read all FIR filters ever designed) is going to fall foul of your 100dB figure.

So at a stroke, all ADCs and DACs using internal digital filters become non-transparent by definition.

 

[opus111]

An FIR filter is a tapped delay line. It works by applying pre-stored coefficients to the contents of that delay line, forming a sum-of-products over the total number of samples stored as its output, for each input sample.

Thus all digital filtering done by such topologies relies on adding together appropriately weighted delayed samples. If a delayed sample becomes an artifact then under Ethan's criteria that sample can't be multiplied by more than 10^-5 (strictly speaking modulus, it can be + or -) or it will violate his 100dB rule.

 

[opus111]

OK, here's another go then.

Think of the delay line as a train with many carriages. The audio samples come in at the locomotive end, and exit at the last carriage. Each clock tick, they all move one carriage towards the rear, one sample in each carriage. So our train holds a recent history of the audio - typically for an FIR filter in audio hardware, the train could have up to 150 carriages.

An output sample (done for each clock tick) is calculated by the ticket inspector running down the whole length of the train, noting the value of each audio sample present in each carriage and putting that down in his notebook. The notebook also contains a list of coefficients - each carriage is associated with a fixed number, called a 'coefficient'. The output sample is the sum of all these (coefficients X audio sample).

Now in the case of jkeny's files, they were created by using a notebook with only two numbers in - 1 and 0.0112. 1 applied to (I think) the 21st carriage back and 0.0112 applied to the first carriage. So to generate the output samples, the ticket inspector only had to multiply together two pairs of numbers, and in one case that multiplication was by unity. In a general case, an FIR filter will have more than 2 coefficients - the sequence of them typically looks like the sin(x)/x wiggles seen on impulse responses of DACs etc shown in Stereophile reviews.

Any better?

 

[jkeny]

To me its really curious that you'd call a pre-echo 'an artifact'. Normally artifacts are non-linear effects - like harmonic distortion, or jitter sidebands.

But if time delays added back in to the signal really become 'artifacts' then any FIR digital filter whose impulse response includes values for delayed (positive or negative) samples above 10^-5 (read all FIR filters ever designed) is going to fall foul of your 100dB figure.

So at a stroke, all ADCs and DACs using internal digital filters become non-transparent by definition.

Thanks Opus & great explanation!
Ethan is already aware that these are not distortions & in his own words has said so before here http://www.gearslutz.com/board/6219245-post23.html

Technically, "distortion" is any change. So you could use either term and not be wrong. However, I reserve the term "distortion" for situations where there's non-linearity. Phase shift is linear, so I prefer not to call it distortion. But that's just me.

--Ethan

So I guess it's educational to again quote his own words back to him in reference to himself

I've seen people take liberty with facts before, but never this badly. And never when it's so easy to prove they lied. So I say that artifacts 100 dB down are inaudible, and often when only 80 dB down or even louder. Then you post files with artifacts only 39 dB down, and that somehow proves me wrong? Wow, just wow. Your sheer chutzpah is amazing.

All I have to add is to re-iterate his words
Wow, just wow!

 

[Ron Party]

So I guess it's educational to again quote his own words back to him in reference to himself

All I have to add is to re-iterate his words
Wow, just wow!

Thanks for the scientific explanation.

 

[jkeny]

Thanks for the scientific explanation.

Do you not understand what Winer said - it's simple enough?

 

[Phelonious Ponk]

Thanks Opus & great explanation!
Ethan is already aware that these are not distortions & in his own words has said so before here http://www.gearslutz.com/board/6219245-post23.html

So I guess it's educational to again quote his own words back to him in reference to himself

All I have to add is to re-iterate his words
Wow, just wow!

In my opinion, it is not only harsh to infer that inconsistently splitting the fine hairs between distortions, artifacts, noise and linearity constitutes a "lie," it is a hell of a lot more disengenuous than the inconsistently split hair ever was. But I'll give you the benefit of the doubt on the original one and assume that your posting of artifacts - or distortions - only 39 dB down in a conversation about 80 - 100 dB margins was just a screw-up.

FWIW, I call them all distortions. Now, go find a post in which I dealt with noise as a separate issue, because of it's distinctly separate behavior and call me a liar.

Tim

 

[Ron Party]

Do you not understand what Winer said - it's simple enough?

Oh yes, I understood what he posted. I also understood your post, which was nothing more than words to the effect of "I'm right, your wrong." What I don't understand is what it is which you are right about nor, for that matter, what it is about which he is incorrect. I also understand that is the capital A audiophile way: attack a person rather than the substance of the post. Fortunately here at WBF we try to focus on the post, not the poster, and one will be admonished to stick to this simple rule. So do you have anything of substance to add, e.g., reliable, repeatable, demonstrative proof, not opinion, that artifacts 100 dB down are audible?

 

[opus111]

So do you have anything of substance to add, e.g., reliable, repeatable, demonstrative proof, not opinion, that artifacts 100 dB down are audible?

I am wondering - how is this question relevant here? I see no suggestion that jkeny's claiming anything at all about audibility of artifacts. Rather he's been exploring audibility of pre-echo which is an entirely separate issue.

 

[Ron Party]

opus111, thank you for your reply. I concur with that which you posted. It is an entirely separate issue and should be treated as such.

 

[opus111]

The issue of linear distortions (as opposed to non-linear distortions, aka artifacts) is covered in the criterion for passband flatness, a separate criterion. So then if 0.2dB total passband flatness is the criterion (there so far is no clarity on this particular point) then a -39dB pre-echo falls within that criterion.

 

[jkeny]

Guys, can we not discuss this on two threads please. I started a separate thread to address these issues & not pollute this one!

[Edit: from treitz3] - here's the link to the separate thread for reference>>> http://www.whatsbestforum.com/showthread.php?6612-Ethan-Winer-s-definition-of-Audio-transparency

 

[Phelonious Ponk]

Can't say I care for that thread. It strikes me as a thinly-veiled attack on Ethan. I understand that Ethan is a bit black and white in his views, that he doesn't mince words, that he can be rather demanding of those who attempt to attack his point of view without a well-formed and supported one of their own. I'm sure it's irritating; enough to make you just wait for him to make a mistake. But if you came to the discussion with a bit more substance, you wouldn't need to start threads to dance in a circle with glee when his noisefloor is off by a bit. You might actually have a viable position of your own. You might even find yourself occasionally agreeing with Ethan.

Tim

 

[Soundproof]

Thinly veiled? Completely unveiled, I'd say.

Audiophiles are their own worst enemies, often. To outsiders, we must appear exquisitely uncool. Nattering nabobs of negativity, indeed.