Is the term resolution' applicable in audio?

[MattHooper]

Personally if I ever read a subjective review refer to resolution or timber as anything other than their technical definitions I stop reading that review. There are so many words and phrases that can convey meaning that we really don't need to use one audio term that is well defined as a slang term to describe something subjectively.

I get this.

The term has been way overused, and often in very silly and dubious ways. So yeah, it can certainly be a flag for BS.

I also think, even though it is in principle a sensible term, even there it is often oversold.

For instance, I think most sonic information in recordings is resolved through any number of different types of loudspeakers, from a decent cheap speakers and upwards.

Differences tend to be more changes in presentation of the detail rather than revelation of more detail.

(I think the term resolution can still be applied reasonably in some cases… but again, I think it’s often sloppy employed).

 

[Gitfiddle]

I get this.

The term has been way overused, and often in very silly and dubious ways. So yeah, it can certainly be a flag for BS.

I also think, even though it is in principle a sensible term, even there it is often oversold.

For instance, I think most sonic information in recordings is resolved through any number of different types of loudspeakers, from a decent cheap speakers and upwards.

Differences tend to be more changes in presentation of the detail rather than revelation of more detail.

(I think the term resolution can still be applied reasonably in some cases… but again, I think it’s often sloppy employed).

I simply think the phrasing and meaning should be generic between all audio terminology. Words and phrases should mean the same thing on every forum.

 

[dedobot]

Resolution - sample rate.
Color depth - sample depth.
FPS/screen refresh rate - bitrate.*

p.s.
*The last is debatable.

 

[MattHooper]

I simply think the phrasing and meaning should be generic between all audio terminology. Words and phrases should mean the same thing on every forum.

A common nomenclature can be useful. But so can creativity with language.
If that wasn’t the case you wouldn’t be able to read any reports of subjective experiences by writers, or works of fiction without constantly being befuddled.

That said, I don’t see anything confusing about the term “ resolution.” So if somebody is talking about a system with
“ high resolution” they are virtually always used in the sense of “ resolving or reproducing even the tiniest/finest/subtle details in a recording.”

Of course, this can be used in the context of bogus claims, just as terms like “ more dynamic” are sometimes used to describe the effects of tweaks that are implausible.
But that doesn’t mean that “ dynamic” doesn’t mean anything or isn’t useful.

 

[Blumlein 88]

Reminds me of the cliche in Stereophile. There is more there there. That's when have resolution.

 

[krabapple]

Resolution - sample rate.
Color depth - sample depth.
FPS/screen refresh rate - bitrate.*

p.s.
*The last is debatable.

I presume you are refering to video?

sample rate doesn't apply there -- pixels do

And for audio, sample rate != resolution. Audio resolution is determined by bit depth

 

[Roland68]

The word "resolution" (Auflösung) does not originate from photography and has various meanings. In German, for example:
- Photography = image resolution (belongs to physics)
- Economics, e.g., dissolution of a company
- Law, e.g., dissolution of a contract
- Solving a riddle
- Dissolution of a chemical compound, dissolution of a material
- Literature, e.g., crime novel, who is the murderer?
- Various areas in mathematics
- Poetry/Verse theory

Also used in physics/digital technology, for example:
- Resolution in digital technology
- Resolution in measurement technology
- Resolution in optics (not to be confused with image resolution)
- Frequency resolution

In measurement technology, resolution is defined as the smallest possible perceptible or measurable difference between two measured values.

Of course, this also applies to audio/music playback in several areas.
- The simplest way to illustrate this is through transparency, i.e., how much of the detail is reproduced exactly as it was recorded.
- Compression of the recording also reduces the possible resolution, for example.

 

[krabapple]

In measurement technology, resolution is defined as the smallest possible perceptible or measurable difference between two measured values.

And that is its correct meaning in audio, too -- the measured value being level.

Of course, this also applies to audio/music playback in several areas.
- The simplest way to illustrate this is through transparency, i.e., how much of the detail is reproduced exactly as it was recorded.

That is audio review poetry. In the end it still equates to minimal level difference above noise.

- Compression of the recording also reduces the possible resolution, for example.

You mean, dynamic range compression.

 

[Berwhale]

My understanding is that 'Western' music is based on a 12 tone chromatic scale using a 7 note heptonic structure with semi-tones. So music based on a 24 tone chromatic scale using quarter tones must have twice the resolution in the frequency domain, so this must be 'Hi-Res' and therefore superior...

 

[Avp1]

I presume you are refering to video?

sample rate doesn't apply there -- pixels do

And for audio, sample rate != resolution. Audio resolution is determined by bit depth

Sample rate defines how well you can recover phase. Thus is sample rate is low, you will have a lot of phase noise for anything that is close to Nyquist limit. Also frequencies above that limit will be reflected back (due to anti-aliasing filter not being ideal) and thus decrease S/N ratio within audible range.

 

[Another Bob]

I consider (most) modern electronics to be, for all intents and purposes, essentially perfect. So I don't really care whether resolution more properly applies to sample rate, bit depth, or something else in that realm. But I do consider the OP's question to be an interesting one. I, like I imagine many of us, have had the experience that certain speakers seems to resolve details better than most. (For me it was the best example was the Quad electrostatics.) Was that subjective experience truly a matter of resolution, or is it explainable through things like frequency response and directivity/reflections? Was it the source material? Was it all in my head? I've never been in a position to do a controlled test of the issue, nor have I read about anyone else trying to.

 

[voodooless]

Sample rate defines how well you can recover phase.

No!

Thus is sample rate is low, you will have a lot of phase noise for anything that is close to Nyquist limit

Phase noise is a property of clocks. It has no meaning here.

Also frequencies above that limit will be reflected back (due to anti-aliasing filter not being ideal) and thus decrease S/N ratio within audible range.

Frequencies about that limit are filtered out.

 

[krabapple]

Sample rate defines how well you can recover phase. Thus is sample rate is low, you will have a lot of phase noise for anything that is close to Nyquist limit. Also frequencies above that limit will be reflected back (due to anti-aliasing filter not being ideal) and thus decrease S/N ratio within audible range.

You wouldn't use a sample rate too low for your purpose. In the case of audio, that means capturing the audible band, which means a minimum of ~44 kHz.

When you can audibly discern a difference in resolution (let's say S/N) by going from 44 to 48, 88, 96 kHz etc. let me know. Let's assume competent filter design, converters, etc.

Because if you can't hear it, audio 'resolution' has no meaning.

 

[Avp1]

No!

Phase noise is a property of clocks. It has no meaning here.

Frequencies about that limit are filtered out.

Nyquist theorem applies to signals of infinite length. Now consider a signal of one cycle at 10kHz. You cannot correctly recover timing of it (phase) with 48kHz sampling rate, which in your assumption should be sufficient. You need to refresh your calculus knowledge.

 

[Avp1]

You wouldn't use a sample rate too low for your purpose. In the case of audio, that means capturing the audible band, which means a minimum of ~44 kHz.

When you can audibly discern a difference in resolution (let's say S/N) by going from 44 to 48, 88, 96 kHz etc. let me know. Let's assume competent filter

design, converters, etc.

Because if you can't hear it, audio 'resolution' has no meaning.

If you are far sighted, you won't see the resolution of the picture on the screen. But when you put on glasses...

 

[Chrispy]

Generally when an "audiophile" uses "resolution" I run away as it is likely nonsense.

 

[Johnkeizer]

I suspect when most audio people use the term resolution they really mean being able to hear little details in the recording.
I usually find loudspeakers to be the main culprit in muddying sound so small details are obscured.

 

[voodooless]

Nyquist theorem applies to signals of infinite length. Now consider a signal of one cycle at 10kHz. You cannot correctly recover timing of it (phase) with 48kHz sampling rate, which in your assumption should be sufficient. You need to refresh your calculus knowledge.

A one-cycle 10 kHz burst is not band-limited because the abrupt start/stop spreads energy across a wide spectrum. Nyquist applies to band-limited signals, so this is not a counterexample, it just violates the assumptions of the theorem.

You’re describing missing bandwidth and calling this phase resolution.

 

[Avp1]

A one-cycle 10 kHz burst is not band-limited because the abrupt start/stop spreads energy across a wide spectrum. Nyquist applies to band-limited signals, so this is not a counterexample, it just violates the assumptions of the theorem.

You’re describing missing bandwidth and calling this phase resolution.

Real audio sources are NEVER band limited. Attempt to band limit them causes adding distortion and noise. This I call loss of resolution. The same actually applies to pictures where sharp object border becomes diffused if you limit number of pixels.

 

[voodooless]

Real audio sources are NEVER band limited.

Sure they are. Not a single microphone has infinite bandwidth. Any ADC or resampling process does the band limiting. It’s a prerequisite for the sampling theorem to work. Your ears are also band limited, which is why none of this matters: you cannot hear any of the tones above 20 kHz (and in your case significantly below that). So no need to capture them.

Attempt to band limit them causes adding distortion and noise.

Well yeah, they distort in that they can only capture within a limited bandwidth. That is inherent. As for noise… no. It might add some ringing, but not noise

This I call loss of resolution.

Yes, frequency resolution: you can encode more frequencies. Not phase or timing resolution! Those are governed by the bit depth, not sample rate!

The same actually applies to pictures where sharp object border becomes diffused if you limit number of pixels.

Yes, it adds frequency resolution. You’re not getting more colors, do you?

Even in video and photography, band limiting is generally applied by using an anti-aliasing low pass filter.