Master Thread: Are Measurements Everything or Nothing?

[AndyBell]

I wonder if people conflate digital audio with digital photography and it creates wrong expectations.

I’ve worked as a C++ programmer in digital photography. In the early 2000’s I had a hand in one of the early ‘raw converters,’ as they are called. So I know a bit about it.

In digital photography higher resolution is measurable and generally beneficial. Higher dynamic range is measurable and generally beneficial. Sharper lenses are measurable and better. Comparing images is trivial and seeing differences is easy. And, generally speaking, you do get what you pay for.

But this just doesn’t translate into audio. Hires audio doesn’t guarantee audible improvement. Even better SINAD doesn’t mean you’ll hear better music. And so on.

But I didn’t understand any of this about audio until I came here, and the average audio enthusiast probably doesn’t either - so they equate terms such as hires with the photographic equivalents and assume they mean the same.

Expectation and confirmation bias do the rest…

 

[solderdude]

In the US, it tends to refer to overpriced audio junk, if I understand correctly (which is not always the case).

High-end can be either 'top performing quality build' gear as well as 'audio jewelry'.
The latter may be well measuring as well as not great measuring yet 'lovely sounding' to the owners.
High-end is usually overly and/or very expensive and generally does not have 'run of the mill' looks but uses premium materials.
Then there are cheaply manufactured devices that mimic the expensive ones but are of questionable build quality.

In general very expensive and exclusive with good enough performance to sound nice.
Heavily relying on 'bias' in most cases.

Of course there are devices that measure well, have good support and are build well with an absence of 'snake-oil' which also fall under high-end.
But there is high-end and there is high-end.

 

[antcollinet]

It was better than necessary. Some folks did not like that. You may ask. I do not know why.

For me, that's a Schrödinger answer; I understand it and at the same time I don't understand it.

Speculating - but the normal source of that type of annoyance is when a better than needed device cannibalises the market for something positioned at a higher level - and higher price.

It's a bean counter annoyance.

 

[RandomEar]

I wonder if people conflate digital audio with digital photography and it creates wrong expectations.

I’ve worked as a C++ programmer in digital photography. In the early 2000’s I had a hand in one of the early ‘raw converters,’ as they are called. So I know a bit about it.

In digital photography higher resolution is measurable and generally beneficial. Higher dynamic range is measurable and generally beneficial. Sharper lenses are measurable and better. Comparing images is trivial and seeing differences is easy. And, generally speaking, you do get what you pay for.

But this just doesn’t translate into audio. Hires audio doesn’t guarantee audible improvement. Even better SINAD doesn’t mean you’ll hear better music. And so on.

But I didn’t understand any of this about audio until I came here, and the average audio enthusiast probably doesn’t either - so they equate terms such as hires with the photographic equivalents and assume they mean the same.

Expectation and confirmation bias do the rest…

To be fair, there are such limits even in digital imaging. At least when viewing images and monitors: Assuming an appropriate viewing distance where you can see the whole image/screen, the angular resolution of the eye limits the useful upper resolution of your image. I would have to check the exact number, but as far as I remember, it's below 12 MPix.

That means printed images with arbitrarily higher resolution all appear equally detailed as those at that resolution when viewed from the appropriate distance. For screens, this means somewhere around 6K is the practical limit where you don't gain any visible detail anymore by increasing resolution.

In contrast to HiFi, you could argue that you like to inspect your images or screen with your nose touching the glass or using a magnifying glass, so it's not a hard all-around limit like in audio. But for most practical applications, such a limit does exist. Doesn't keep any company from hyping 8K or even 16K TVs and screens...

 

[AndyBell]

To be fair, there are such limits even in digital imaging. At least when viewing images and monitors: Assuming an appropriate viewing distance where you can see the whole image/screen, the angular resolution of the eye limits the useful upper resolution of your image. I would have to check the exact number, but as far as I remember, it's below 12 MPix.

That means printed images with arbitrarily higher resolution all appear equally detailed as those at that resolution when viewed from the appropriate distance. For screens, this means somewhere around 6K is the practical limit where you don't gain any visible detail anymore by increasing resolution.

In contrast to HiFi, you could argue that you like to inspect your images or screen with your nose touching the glass or using a magnifying glass, so it's not a hard all-around limit like in audio. But for most practical applications, such a limit does exist. Doesn't keep any company from hyping 8K or even 16K TVs and screens...

Agreed. I was more thinking of cameras and lenses.

Of course, if you only view on screen at normal distances or make smallish prints, then many of the advantages of high megapixel cameras are discarded. But even then, the ability to aggressively crop an image increases with more megapixels and sharp lenses, and as super telephoto lenses are expensive, cropping can be a good workaround, if you have enough megapixels.

 

[ahofer]

The get out clause So basically what you are saying is there "Is no difference". Well that was very helpful. Thanks.

No difference is, by definition, the Null Hypothesis against which any scientific conclusion has to be defended. And it remains undefeated in the case of properly designed audio electronics.

Which doesn’t mean listeners won’t experience different devices differently, it’s just that there’s no evidence the source of that experience is audible sound waves.

 

[RexrothPigeon]

Agreed. I was more thinking of cameras and lenses.

Of course, if you only view on screen at normal distances or make smallish prints, then many of the advantages of high megapixel cameras are discarded. But even then, the ability to aggressively crop an image increases with more megapixels and sharp lenses, and as super telephoto lenses are expensive, cropping can be a good workaround, if you have enough megapixels.

Cameras and lenses would be microphones and recording gear, not playback devices

 

[olieb]

Of course, if you only view on screen at normal distances or make smallish prints, then many of the advantages of high megapixel cameras are discarded.

Sure, but the main point is, that "resolution" does not translate well to audio. [However, taking the right conditions into account, it does nevertheless, as @RandomEar explained.]
In audio you have magnitude and frequency in time.
For pictures this would translate to contrast and colour (and fps for video).
And you will hardly see a difference between 16bit and 24bit (or 32bit) greyscale in a photo. Or between 10 billion colours and 100 billion colours.
The same for 1000fps and 5000fps. Or for the presentation of infrared of 3000nm wavelength.
Electronics performance in respect to audio perception limits is much more advanced than in relation to video visual perception limits.

Edits in green for clarity.

 

[Roland68]

Sure, but the main point is, that "resolution" does not translate well to audio. [However, taking the right conditions into account, it does nevertheless, as @RandomEar explained.]
In audio you have magnitude and frequency in time.
For pictures this would translate to contrast and colour (and fps for video).
And you will hardly see a difference between 16bit and 24bit (or 32bit) greyscale in a photo. Or between 10 billion colours and 100 billion colours.
The same for 1000fps and 5000fps. Or for the presentation of infrared of 3000nm wavelength.
Electronics in respect to audio is much more advanced than in relation to video.

I consider that a rumor.
I work in both fields and collaborate with many manufacturers and developers in both areas, and the development in the photo and video sector over the last 30 years has been many times greater than in the audio sector. This applies to both optics/lenses and electronics in both photo and video.

Resolutions, both optical and electronic, have increased more than tenfold.
Light sensitivity has improved by at least 1000 ISO.
Even with telephoto lenses, you can take sharp handheld shots at 1/8 to 1/2 of a second. Previously, this required 1/500 to 1/1000 of a second.
High-speed cameras take 10,000 pictures per second, an incredible 100 times faster.
And the list goes on.

In the audio sector, we have 30-40 year old amplifiers, DACs, and CD players that are still just as good as, and even better than, many current devices (in terms of sound and measurements), and are only surpassed by a small number of devices, and even then only by a few percent.
Of course, we have more modern circuits, smaller devices, and lower power consumption.

The difference has a simple origin.
In the photo and video sector, major manufacturers have invested millions and billions of dollars/euros directly into the development of electronic components (sensors, special ICs, etc.) and lenses/production facilities, as well as in their own development hardware and software.

In the audio sector, on the other hand, development funds have primarily flowed into the development of ever-newer devices with measurement values that have remained almost stagnant for 30-40 years.
Hyped technologies/ICs, such as IcePower and TPA3251/55 Class D amplifiers, are simply repurposed byproducts from the consumer/automotive sector.

 

[ahofer]

The difference has a simple origin.
In the photo and video sector, major manufacturers have invested millions and billions of dollars/euros directly into the development of electronic components (sensors, special ICs, etc.) and lenses/production facilities, as well as in their own development hardware and software.

In the audio sector, on the other hand, development funds have primarily flowed into the development of ever-newer devices with measurement values that have remained almost stagnant for 30-40 years.
Hyped technologies/ICs, such as IcePower and TPA3251/55 Class D amplifiers, are simply repurposed byproducts from the consumer/automotive sector.

I’m not sure it’s all about investment. You presume that there are gains to be made with investment, and you aren’t counting the investment in theater, surround, and noise-canceling technologies. There may be a reason funds have flowed in that direction.

Image magnification benefits from finer resolution; audio magnification, at least within normal rooms and non-painful sound pressure levels, is a solved problem.

 

[AndyBell]

Sure, but the main point is, that "resolution" does not translate well to audio

And that was my point.

But people who don’t understand this make the connection anyway. And it reinforces their expectations…

Analogies only work if they are accurate. Otherwise they mislead alarmingly…

 

[AndyBell]

In the audio sector, on the other hand, development funds have primarily flowed into the development of ever-newer devices with measurement values that have remained almost stagnant for 30-40 years.

But maybe that’s because audio equipment already transmits everything the human ear can hear - there’s nothing extra worth resolving.

In photography there’s always the opportunity to make a bigger print and so extra resolution or better AI driven upsampling is valuable. I sold a seascape to a UK resident who wanted to print it as a large, wall sized mural. The camera I used and the upscaling software I used made that possible and it looked great.

But you can’t really make music ‘bigger’ in that way…

 

[Roland68]

I’m not sure it’s all about investment. You presume that there are gains to be made with investment, and you aren’t counting the investment in theater, surround, and noise-canceling technologies. There may be a reason funds have flowed in that direction.

Image magnification benefits from finer resolution; audio magnification, at least within normal rooms and non-painful sound pressure levels, is a solved problem.

You're absolutely wrong there. I believe that investments can foster innovation and technological development, especially when someone has a vision and an idea. Sigma (lenses are a good example of this). In the first 10 years of digital SLR photography, Sigma had the worst possible reputation for lenses, and the optical quality and production variance were abysmal.

Then the son took over the company and invested a huge sum in Zeiss development and production technology, turning the company upside down.
In 2014, just two years later, Sigma introduced the 50mm f/1.4 DG HSM Art, which at the time was the only lens on the market that matched the image quality of the Zeiss Otus.
Anyone familiar with lens development knows that such a process usually takes several years, or even a decade, and isn't guaranteed to be successful.
And even such an investment isn't guaranteed to be successful, and it could have gone wrong for Sigma.

Just because someone buys a Klippel development system doesn't mean they can build good loudspeakers.

There are many areas where the audio industry could be much further along, and that's probably a gross understatement.
A simple example:
Well-functioning 2.1/2.2 solutions that anyone can handle and that are affordable.

How ridiculous is it that, although this technology has existed in the digital realm for over 30 years and has been integrated into AV and standalone systems ever since, it isn't readily available for standard desktop and hi-fi systems?
The entire manufacturing industry should be utterly ashamed that they haven't managed to solve such a simple task satisfactorily, affordably, or comprehensively in 30 years.

And that's just one of many examples.
A manufacturer like SMSL has to resort to using really poor-quality ICs from soundbars.

 

[ahofer]

You're absolutely wrong there. I believe that investments can foster innovation and technological development, especially when someone has a vision and an idea. Sigma (lenses are a good example of this). In the first 10 years of digital SLR photography, Sigma had the worst possible reputation for lenses, and the optical quality and production variance were abysmal.

Well, I'm not "absolutely wrong", that's silly. I'm raising the possibility that investment would not bring as much useful innovation in one area as another. Whether I'm right about audio is open to debate, but it's a fairly established fact that there are variations in productivity of investment across fields. Neither of us is making a falsifiable claim apart from the latter.

I would further limit my claim to the idea that investment in audio *electronics* (and particularly digital resolution) would yield much for audio quality, since that is what this whole 'resolution' topic seems to be focused on.

 

[TTT15]

Any place that ignores or denies science and maths or isn‘t open minded or is biased in only one direction isn’t science. It’s ideology.

 

[sonitus mirus]

Any place that ignores or denies science and maths or isn‘t open minded or is biased in only one direction isn’t science. It’s ideology.

Science isn’t about being perfectly unbiased. It’s about being testable and willing to change.

 

[solderdude]

Any place that ignores or denies science and maths or isn‘t open minded or is biased in only one direction isn’t science. It’s ideology.

Any place that ignores science, measurements and math and chooses to be 'open minded' and thinks they are not affected by 'bias' has an ideology that is founded on personal perception rather than 'science'. Yet they use 'math/engineering/science based' devices to get there (as well as many things that aren't scientifically explainable) and usually 'believe' things work in another way than what 'science minded' people say it does.

Totally fine if it works for them ... it's just not an ideology/theories that are based on actual engineering and science but rather on personal perception and convictions.

Besides science requires an open mind to find out ways or invent something or to prove things and not just make an observation and assume that observation is correct and draw a conclusion that seems to work for the observer.

 

[bmc0]

But you can’t really make music ‘bigger’ in that way…

Digital audio and digital images are both signals sampled at discrete points; the latter is just sampled in two (spatial) dimensions instead of one (temporal) dimension. Sampling rate is then samples per degree of the visual field rather than samples per second.
If one ensures standard viewing conditions such that the spatial sampling rate from the viewer's perspective is known—like temporal sampling rate is in digital audio—maximum perceptually useful image resolution is more clearly defined.

(Edited to hopefully clarify the intended scope)

 

[ahofer]

 

[ahofer]

This kind of captures a lot of the dialogue in this thread (not aimed at the last few posts).