Can you hear the difference between 44.1 and 48k?

[KG10]

In theory it should be none at all since human hearing typically has an upper limit of 20k.

But then I've seen claims online that it can actually pick up frequencys up to 50k.

 

[SeeMoreDigital]

Was the digital source (not analogue) encoded at a higher sample-rate of say 96 or 192kHz prior to being reduced to 44.1 and 48kHz?

 

[KG10]

Was the digital source (not analogue) encoded at a higher sample-rate of say 96 or 192kHz prior to being reduced to 44.1 and 48kHz?

Good question, although for this I'm mainly talking about generally.

 

[Marplot]

I would also suggest most of my 44.1 come from loudly mastered CDs while the 48 come from DVD/BR that are not treated in such a ham fisted way. As a result, yes they sound different but it is due to mastering and not 4kHz

 

[SeeMoreDigital]

Good question, although for this I'm mainly talking about generally.

Unfortunately it's not a good idea to use generalities when asking such questions...

When conducting such audio tests it's good practice to start with a 'known source' that's been encoded at the highest possible sample-rate and bit-depth (preferably one that's been captured, mixed and mastered digitally).

Then you create your 44.1kHz and 48.0kHz 'comparison encodes' from the known source for testing.

 

[KG10]

Unfortunately it's not a good idea to use generalities when asking such questions...

When conducting such audio tests it's good practice to start with a 'known source' that's been encoded at the highest possible sample-rate and bit-depth (preferably one that's been captured, mixed and mastered digitally).

Then you create your 44.1kHz and 48.0kHz 'comparison encodes' from the known source for testing.

Those are good points there.

 

[par4ken]

We are kind of splitting hairs with this poll. You are more likely to hear a difference between 44.1K and 96K or 192K. Theoretically 44.1K should be more than enough to pass everything that we can actually hear. In practice those old first generation CD players sounded horrible. The advent of oversampling improved the sound quality greatly. Any noticeable difference is likely to be equipement and program dependant.

I'm all for improving sound quality in ways that are measurable i.e. wider frequency response, lower distortion. Only if those improvements make things sound worse is there a problem, in that case we must be measuring the wrong parameter.

So I'm voting "Yes but only a little" to cover all the bases. I admit that "little" difference could translate into no noticeable difference much/most of the time.

 

[marpow]

When HDTracks sells a 16bit/44.1Khz vs a 24bit/44.1Khz, yes I can hear the difference, the 24bit being the better.
But that was not Post 1 question.
So based on the question I voted No.

 

[Steve Bruzonsky]

Maybe yes Maybe no too many factors to give a clearer answer.

 

[splinter]

When HDTracks sells a 16bit/44.1Khz vs a 24bit/44.1Khz, yes I can hear the difference, the 24bit being the better.

AFAIK, the difference between 16 and 24 Bit happens below -96dB, so for once you would have to listen at really loud volumes, and as many mics and preamps produce noise above -96dB, you can most likely only hear noise at a better resolution.

Regarding the original question, I wanted to "hear" high res (which I cannot, because my listening threshold lies somewhere shortly above 16 kHz). So I produced a differential track from the 48 kHz and 192 kHz versions of Paul Young's "Come Back and Stay". Which, if we can believe the advertising, was sourced from the original 2-inch master tapes, so we could expect the full range of analog recordings.

The result is a signal with significant information between 22 and 30 kHz, the peak volume being around -82.5 dB (at 22 kHz, way lower beyond). I normalized the signal and then transposed it so that 20 kHz are 1 kHz in the transposed track.

The transposition creates a lot of artifacts, so this is nothing more than an approximation, but I was just curious what I am missing. What was left was peak signals from snare hits and noise in a repeating pattern which I assume is tape hiss.

So that is the difference hi res makes in this specific example: ultrasonic snare peaks at 22 kHz at -82 dB volume, and even lower volume tape hiss between 22 and 30 kHz.

I can only urge everyone to make such experiments yourself. It is not so hard if you already are able to rip audio to WAV and use Audacity for the audio transformations.

 

[newslane]

Unfortunately it's not a good idea to use generalities when asking such questions...

When conducting such audio tests it's good practice to start with a 'known source' that's been encoded at the highest possible sample-rate and bit-depth (preferably one that's been captured, mixed and mastered digitally).

Then you create your 44.1kHz and 48.0kHz 'comparison encodes' from the known source for testing.

In addition, it helps to focus on one instrument when comparing.