• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Distortion down to -300 dB, what exactly does that mean physically?

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,836
Likes
16,498
Location
Monument, CO
This of course a ridiculous choice of course for doing reconstruction. One can design either a constant delay, or mixed delay filter of quite finite length for any purpose, while providing reconstruction to any arbitrary level, say, under 1 LSB, without having to resort to sinc functions, etc. Look up "remez exchange" for one well-known way to design a finite length filter that will give, assuming you feed it properly (never assume people know what they are doing) the best filter one can achieve for a given length.

Then you look at the results, and either make the filter longer (if it's not good enough) or shorter (if it's excessively good) or you simply allow yourself to be satisfied with over kill (the usual way to do this).

It's really not hard. If you must do some kind of resampling (say by 14 as you mentioned, although 14 is pretty silly, it would be much more efficient to do 16, Crochiere and Rabiner showed how to do this to any arbitrary cleanliness, efficiently, back in the very early 1980's. I know, I worked for them, some of my work is in the book.

https://www.amazon.com/Multirate-Digital-Signal-Processing-Crochiere/dp/0136051626

If you want to do variable-rate resampling, Proakis shows how to do that to any arbitrary level in one of his early books on digital signal processing.

These are almost hoary-old methods, and they work JUST FINE. There's no need for this "sinc" stuff, you're spouting theory that is both inefficient and unnecessary there. (Yes, the theory is right, but so is the theory of filters and wavelets.)



Well, precisely. Given the basic atmospheric noise that nobody seems to want to remember, there's no such thing as "infinite precision". That's really a very basic point here, basic atmospheric noise due to the particulate nature (i.e. molecules, argon atoms) makes any kind of "infinite precision" nonexistent, timewise, levelwise, you name it.

(Not directed at Geert here)

Furthermore, air propagation at high levels (above 120dB or so, detectable, barely, maybe, at 90dB, by an accurate machine) is NOT LINEAR. So enough with these silly "infinite precision" arguments. If you want to argue about multistage interpolation and decimation, your first job is to prove that Gauss was wrong, and the Fourier Series does not converge in L2 with finite power signals. Good luck with that.

I had a hard time decided to "like" a post that brought up so many painfully tedious grad school memories of wading through all the math... :)

Let's add chaos theory, eh?

I thought sound propagation went appreciably nonlinear up around 140 dB. I am sure you are correct, but a good reminder that memory (mine, anyway) is fallible. I do remember one of my profs saying 1 acoustic watt would blow your ears off...
 

Don Hills

Addicted to Fun and Learning
Joined
Mar 1, 2016
Messages
708
Likes
464
Location
Wellington, New Zealand
On the topic of Rob Watts, where did he say he could hear distortion at -300dB? I believe what he actually said was that he could hear the difference between two filters where the measured differences between the filters was -300 dB. That still makes no sense as a measurement of filter differences though...
 

j_j

Major Contributor
Audio Luminary
Technical Expert
Joined
Oct 10, 2017
Messages
2,267
Likes
4,759
Location
My kitchen or my listening room.
I thought sound propagation went appreciably nonlinear up around 140 dB. I am sure you are correct, but a good reminder that memory (mine, anyway) is fallible. I do remember one of my profs saying 1 acoustic watt would blow your ears off...

It all depends on what "appreciable" means, of course. When can one measure it, when do weird things like mass flow happen?
 
OP
voodooless

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,229
Likes
17,811
Location
Netherlands
On the topic of Rob Watts, where did he say he could hear distortion at -300dB? I believe what he actually said was that he could hear the difference between two filters where the measured differences between the filters was -300 dB. That still makes no sense as a measurement of filter differences though...

Yes, yes, multiple people brought that up already. But that's not what the video is about, it's about small-signal distortion, and that's what he calls it.

In the digital domain, it's even more strange. Once you convert back to integer (assuming you'd use floating point to actually get to the -300 dB), the best you can do is about 192 dB of theoretical dynamic range (assuming 32 bit), and you can't even expand this virtually with noise shaping, because we're already talking high frequency.
 

peterzuid

Member
Joined
Apr 10, 2020
Messages
51
Likes
3
Do you know how to determine if it sounds better?
JJ is one of our luminaires with signal processing one of his core specialties. So please don't try to explain things to him this way. It is liable to get your nose bloodied. What you just wrote is just the basics of how such a filter works which by the way, results in high latencies. Best to not watch any videos with it unless you can delay the video by the same amount.
perfectly aware this processing leads to high latencies and I do not experience this as any issue and fyi, there is a bypass mode for video.
 

Geert

Major Contributor
Joined
Mar 20, 2020
Messages
1,938
Likes
3,526
Then this is where we don't share the same opinion.
Looks like you can't distinguish an opinion from true knowledge. If I'm wrong on that, than please explain exactly where J-J's theory goes wrong and share the technical details on this superior sinc function implementation. I already left some questions to be answered.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,338
Likes
3,278
Location
.de
300 dB--an impossibly large ratio for any two things in real life.
It's approximately the ratio of our planet's circumference to the diameter of a helium atom. Or the daylight illumination received by a 50x50 km area compared to a single photon (which is close to what modern image sensors can pick up).
 

Spkrdctr

Major Contributor
Joined
Apr 22, 2021
Messages
2,212
Likes
2,934
Perfect is the enemy of good. It is also the enemy of reality.

Yes, there are people here who will confuse "accurate" with "must sound good", and there are dogmatic views about how things must be, and yes, my personal favorite, the dogmatic view that "accurate signals = accurate sound", which totally ignores what happens on the way to the brain making an interpretation. HOWEVER <-- Highly emphasized <-- There is a big difference between "sounds better" whatever that means, and "sounds different", or should I say "has any chance of sounding different". The delusional audiophile mantra, "everything matters" is simply poppycock. Everything does not matter. A whole lot of stuff does not matter, and that is where we are with your posts.

Thank you. I have been saying this with many posts and hinting at it in others. This should be pinned for the next ten years. 99% of the audio enthusiasts do NOT understand this. Nor, do they understand taking a flat speaker and placing it in a room which will give 10db swings at various frequencies. The average guy wants to buy a tested flat response speaker because "it is relatively flat and perfect" and put it in the room with zero set up, zero EQ and zero room treatments. But are happy with the knowledge in their own mind that they have the "best" speakers because they tested relatively flat. That is crazy to me.

But, in the end, this was a GREAT POST!
 

audio2design

Major Contributor
Joined
Nov 29, 2020
Messages
1,769
Likes
1,830
@Spkrdctr, actually what you wrote and what I intended are not really the same thing and even your post somewhat illustrates what I was indicating. My posts was mainly targeted not at joe blow audiophile as you described, but the faux technical audiophile who does not realize the shallowness of their knowledge.

Take the simple concept of a flat response achieved with EQ, placement, room treatments, etc. It's not really flat as the direction of the sound has a significant impact on the frequency response of what gets picked up by our auditory system. Add in some nasty artifacts due to 2 channel speakers and head transfer functions to that as well. So what is "flat" ?

Is lowest noise always the best? Just like in electronic sensor systems, adding noise to audio can allow us to detect details that we cannot detect without. There could even be correlation between preference (which can improve with added noise), and ability to extract detail.

Other aspects of reproduction viewed as artifacts or "errors" can product a more pleasing and even arguably more accurate representation of the original performance once it reaches the brain. Electrical signals are simple, sound fields are immensely complex. We can state authoritatively when a change to the electrical transmission is inaudible. What we can't state authoritatively at all times is not only what the subjective result is when changes are sufficient to be audible, but what the objective results is inside the all important brain.
 

j_j

Major Contributor
Audio Luminary
Technical Expert
Joined
Oct 10, 2017
Messages
2,267
Likes
4,759
Location
My kitchen or my listening room.
@Spkrdctr

Take the simple concept of a flat response achieved with EQ, placement, room treatments, etc. It's not really flat as the direction of the sound has a significant impact on the frequency response of what gets picked up by our auditory system.

Like I said to a troll who apparently got into it with a moderator before they were invited to depart, how do you even define flat? Is it the direct response? Is it the sum of direct plus room response? What window do you use to measure this? Do you measure near-field, far field, or what?

And that's before we consider HRTF, reflections, or diffuse vs. direct sensation.

Once you learn a bit more, you start to realize that "flat" is a very hard thing to define, especially if your speaker does not have a good power response as well as direct response.
 

peterzuid

Member
Joined
Apr 10, 2020
Messages
51
Likes
3
Why do we need "infinite precision and timing"?
How would the sinc function not being "infinite precise and timed" manifest itself in measurable parameters?
Are there no hearing thersholds that apply to these parameters?
I did not make the statement we need infinite precision and timing did I ?
In case you have an idea, please share
I'm sure there are and it's my understanding Rob Watts is searching for them.
 

Geert

Major Contributor
Joined
Mar 20, 2020
Messages
1,938
Likes
3,526
Archimago published an article regarding the use of a bazillion tap filter: Archimago's Musings: Mega-taps upsampling: Remastero's Pan Galactic Gargle Blaster (PGGB) software. (Broadly on audiophile software & the noise boogeyman.)

In the article he uses 'Remastero', a software to apply the filter offline because it takes so much CPU and time to calculate that applying it in real time isn't feasible. This approach goes much further than what Watts is attempting.

Guess what the outcome is .... "Alas, I see/hear no evidence of any "magic" here and whatever special processing is being done seems to be low-level and inaudible". Also a DeltaWave analysis didn't show any meaningful.

So maybe you can organise a blind test using tracks processed via Remastero to see if this whole idea makes sense.
 

peterzuid

Member
Joined
Apr 10, 2020
Messages
51
Likes
3
Looks like you can't distinguish an opinion from true knowledge. If I'm wrong on that, than please explain exactly where J-J's theory goes wrong and share the technical details on this superior sinc function implementation. I already left some questions to be answered.
Can you ? This is indeed a very hard one, an opinion may last forever, true knowledge is always limited, and science will create new insights as time passes by.
 

Geert

Major Contributor
Joined
Mar 20, 2020
Messages
1,938
Likes
3,526
Can you ? This is indeed a very hard one, an opinion may last forever, true knowledge is always limited, and science will create new insights.
Science evolves because it builds on true knowledge, it doesn't make it obsolete. The Nyquist–Shannon sampling theorem dates from early 1900. It's still valid. Opinion's on the other hand are like *ssholes, .... you know the saying I guess.

But I'm no longer going to waste any time on you. 'Peter Zuid' was a Dutch resistance leader in the Second World War. Your alias clearly gives away why you are here.
 
Last edited:

peterzuid

Member
Joined
Apr 10, 2020
Messages
51
Likes
3
Science evolves because it builds on true knowledge, it doesn't make it obsolete. The Nyquist–Shannon sampling theorem dates from early 1900. It's still valid. Opinion's on the other hand are like *ssholes, .... you know the saying I guess.

But I'm no longer going to waste any time on you. 'Peter Zuid' was a Dutch resistance leader in the Second World War. Your alias clearly gives away why you are here.

Wow !, you certainly trying to create an atmosphere here!
Never said true knowledge is obsolete. But, let's face it, difficult to judge.
j.j can state he has true knowledge and so can Rob Watts.
These seem to be two worlds apart but I don't like the idea of excluding one of the other.
For me there's no point in taking sides.
I like the vision of Rob Watts, if his design is according j.j "overkill" then so what ?
Don't think it's a problem in case someone else has a different opinion.
Time will tell how it works out...........
 
Last edited:

Frank Dernie

Master Contributor
Forum Donor
Joined
Mar 24, 2016
Messages
6,445
Likes
15,781
Location
Oxfordshire
Wow !, you certainly trying to create an atmosphere here!
Never said true knowledge is obsolete. But, let's face it, difficult to judge.
j.j can state he has true knowledge and so can Rob Watts.
These seem to be two worlds apart but I don't like the idea of excluding one of the other.
For me there's no point in taking sides.
I like the vision of Rob Watts, if his design is according j.j "overkill" then so what ?
Don't think it's a problem in case someone else has a different opinion.
Time will tell how it works out...........
IMO engineering a product to a higher level of performance than needed is a waste of resource.
In my field it is a trap plenty of young or inexperienced engineers fall into and also IME the best engineers I know are the ones who do not do it.
I can see several efforts from engineers achieving improvements in theoretical performance which are pointless based on what we know about the actual audible requirements.
In a way this isn't much of a big deal if the manufacturing cost isn't increased much - ie if done in software - but I remember a new engineer coming into our DO to detail a gearbox design from the chief designer who toleranced the parts such that some couldn't be made at all, some would have been 20x times more per part than normal. Luckily he just got a bollocking and a lesson in manufacturing and tolerancing but without an experienced and knowledgeable senior the 'box would have been many times more expensive than needed and no better at all functionally or reliability wise.
HQPlayer is very clever software but not needed IMO.
Chord make very good converters (I own one) but the performance is better than my ears and has been for years.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,386
Likes
24,752
Location
Alfred, NY
I'm sure there are and it's my understanding Rob Watts is searching for them.

He's searching for revenue.

Nonexistent problems are the hardest ones to solve. And often the most profitable, especially if morals are mere inconveniences.
 

peterzuid

Member
Joined
Apr 10, 2020
Messages
51
Likes
3
IMO engineering a product to a higher level of performance than needed is a waste of resource.
In my field it is a trap plenty of young or inexperienced engineers fall into and also IME the best engineers I know are the ones who do not do it.
I can see several efforts from engineers achieving improvements in theoretical performance which are pointless based on what we know about the actual audible requirements.
In a way this isn't much of a big deal if the manufacturing cost isn't increased much - ie if done in software - but I remember a new engineer coming into our DO to detail a gearbox design from the chief designer who toleranced the parts such that some couldn't be made at all, some would have been 20x times more per part than normal. Luckily he just got a bollocking and a lesson in manufacturing and tolerancing but without an experienced and knowledgeable senior the 'box would have been many times more expensive than needed and no better at all functionally or reliability wise.
HQPlayer is very clever software but not needed IMO.
Chord make very good converters (I own one) but the performance is better than my ears and has been for years.

Quality is the conscious acceptance of acceptable deviations. All about managing the production cost and thus staying competitive.
From a marketing point of view making a difference is key though.
In audio there are lot of hollow marketing phrases.
Chord takes a technical different approach by this M-scaler oversampling device as far as I know. Tolerances on the printed circuit boards and specs of mounted devices will be just enough though I guess.
 

peterzuid

Member
Joined
Apr 10, 2020
Messages
51
Likes
3
He's searching for revenue.

Nonexistent problems are the hardest ones to solve. And often the most profitable, especially if morals are mere inconveniences.
I am pretty sure he is attacking something he experiences as a problem and puts all his effort in solving it. Otherwise he must have had many acting lessons.
 
Top Bottom