24bit vs 16bit, the myth exploded!

[Whazzzup]

 

[skhan007]

So here's a question, that may have already been explored, but I'm curious: If certain members the audiophile community are fond of hi-res music that is reported to be sampled at rates higher than the human ear can perceive, can we see on a spectrum analyzer that there is no sound/data above 20 kHz? When I see threads that describe equipment capable of sampling and producing at higher levels, I'm perplexed. Unless I'm missing something (which is entirely possible, as I'm no audio engineer), a spectrum analyzer should be able to clearly indicate that no sound is produced or audible above a certain limit. Furthermore, what recording equipment is capturing these frequencies in the first place? I don't think studio mics are capable. If we were talking televisions, it would be ridiculous to see a product that produces light at frequencies beyond what is detectable with our eyes (e.g. infrared, ultraviolet, x-ray, etc.). Thoughts?

 

[Slaphead]

So here's a question, that may have already been explored, but I'm curious: If certain members the audiophile community are fond of hi-res music that is reported to be sampled at rates higher than the human ear can perceive, can we see on a spectrum analyzer that there is no sound/data above 20 kHz? When I see threads that describe equipment capable of sampling and producing at higher levels, I'm perplexed. Unless I'm missing something (which is entirely possible, as I'm no audio engineer), a spectrum analyzer should be able to clearly indicate that no sound is produced or audible above a certain limit. Furthermore, what recording equipment is capturing these frequencies in the first place? I don't think studio mics are capable. If we were talking televisions, it would be ridiculous to see a product that produces light at frequencies beyond what is detectable with our eyes (e.g. infrared, ultraviolet, x-ray, etc.). Thoughts?

Well analog magnetic tape is able to capture very high frequencies, well in excess of standard 16/44. And those frequencies can actually be transcribed onto vinyl except they don't last much longer than one or two playthroughs on that medium. The big advantage that digital has over analog mediums is dynamic range - the best studio reel to reel tape machines have around 13 bits of dynamic range, so around 78dB, and as for vinyl it tops out at around 10 bits of dynamic range (so 60dB) even for the best decks. Once you've got a medium that covers the complete human hearing frequency spectrum then dynamic range becomes king, but only to a point, going beyond around 90 dB or so is for the most part meaningless - standard redbook is 96dB without noise shaping.

And if you're talking televisions then all of them come with something that emits light beyond the visible spectrum - the remote control.

 

[bigshot]

Here is a spectrum from a song from an SACD. Note the steep drop off at CD quality sound, a noise floor and spurious noise spikes... all of which are inaudible because of the volume as well as the frequency.

Well analog magnetic tape is able to capture very high frequencies, well in excess of standard 16/44.

That is not true. At 15ips/24 track/2 inch, the format most late period analogue recordings were made on, most commercial analogue recordings just barely come up to 20kHz. They weren't designed to accurately record beyond that. http://www.endino.com/graphs/ None of them match the specs of 16/44.1 for response, noise and distortion. https://en.wikipedia.org/wiki/Comparison_of_analog_and_digital_recording

It was standard practice to apply a low pass filter to remove all frequencies above 20kHz. They weren't wanted.

 

[old tech]

Here is a spectrum from a song from an SACD. Note the steep drop off at CD quality sound, a noise floor and spurious noise spikes... all of which are inaudible because of the volume as well as the frequency.



That is not true. At 15ips/24 track/2 inch, the format most late period analogue recordings were made on, most commercial analogue recordings just barely come up to 20kHz. They weren't designed to accurately record beyond that. http://www.endino.com/graphs/ None of them match the specs of 16/44.1 for response, noise and distortion. https://en.wikipedia.org/wiki/Comparison_of_analog_and_digital_recording

It was standard practice to apply a low pass filter to remove all frequencies above 20kHz. They weren't wanted.

It also depends on the recorder. For example, the 8 track recorder the Beatles used for the White Album had a frequency response of 30 - 15 khz +/- 2db. That was a cutting edge recorder back in 1968. Yet some Beatles fans worry about frequency extension beyond 20khz, go figure that one out.

As for dynamic range, while LP records have a range equivalent to around 10 - 13 bits digital, it is not that important for most pop/rock music as these recordings rarely have a range of more than 50db. It is mainly with classical music, particularly full orchestral productions, that the 96 db + advantages of CDs and SACDs become more apparent.

 

[bigshot]

All correct. The most dynamic LPs don't have more that about 45-50 dB dynamic range. And if they get up that high, they get very hard to listen to. You keep having to jump up to adjust the volume.

People tend to think "more is better". That just isn't true. Enough is better.

 

[theaudiologist1]

All correct. The most dynamic LPs don't have more that about 45-50 dB dynamic range. And if they get up that high, they get very hard to listen to. You keep having to jump up to adjust the volume.

People tend to think "more is better". That just isn't true. Enough is better.

Agreed. It's annoying to adjust the volumes on my classical collection.

 

[bigshot]

I've found BIS is the worst label for excessive dynamics.

 

[Slaphead]

That is not true. At 15ips/24 track/2 inch, the format most late period analogue recordings were made on, most commercial analogue recordings just barely come up to 20kHz. They weren't designed to accurately record beyond that. http://www.endino.com/graphs/ None of them match the specs of 16/44.1 for response, noise and distortion. https://en.wikipedia.org/wiki/Comparison_of_analog_and_digital_recording

It was standard practice to apply a low pass filter to remove all frequencies above 20kHz. They weren't wanted.

Ok, yes you're right for standard pro audio recording equipment at the time, However I was more thinking of what we were doing back in university in the 80's when we needed to record some ultrasonic frequencies, think 50 - 60 Khz for some other department that was studying aerodynamic flutter on a micro scale. We eventually used a jury rigged VHS VCR with a bit of a turbo on the tape speed - I think around 8x the normal video recording speed, maybe less, and could get a solid 50Khz, and higher with a bit of fiddling.

I often wonder why analog audio recording never adopted the helical scanning technique used in video recorders and then later for DAT (as well as a backup medium for computers) as it provided for a far greater dynamic range and frequency response.

 

[theaudiologist1]

Is DSD a horrible outdated meme format, considering it only has 2 quantization levels and 6dB of dynamic range? People claim DSD64 is garbage and noisy and that DSD is only good at 128fs+, but I couldn't tell a difference.

 

[chef8489]

Is DSD a horrible outdated meme format, considering it only has 2 quantization levels and 6dB of dynamic range? People claim DSD64 is garbage and noisy and that DSD is only good at 128fs+

There pretty much is no such thing as pure dsd. Just about every dsd album was converted to pcn at one stage or multiple stages.

 

[theaudiologist1]

There pretty much is no such thing as pure dsd. Just about every dsd album was converted to pcn at one stage or multiple stages.

Pretty much. That's why it's a meme format. In theory I liked DSD more than PCM, but what's the point if DSD doesn't even exist? It's like trying to promote a video format that's 32K but it has to go through millions of processes of being converted to 4K or 1080p and THEN upsmapled to 32K again for mastering since there is no way no master 32K video.

 

[sander99]

I often wonder why analog audio recording never adopted the helical scanning technique used in video recorders and then later for DAT (as well as a backup medium for computers) as it provided for a far greater dynamic range and frequency response.

"HiFi Video" recorders did. They recorded frequency modulated audio with rotating heads. But at that time digital audio was already upcoming.

 

[bigshot]

16/44.1 is all human ears can hear. Beyond that you’re recording for bats.

 

[theaudiologist1]

16/44.1 is all human ears can hear. Beyond that you’re recording for bats.

I know DSD doen't sound better, but I'm actually wondering if it sounds worse due to the noise in the higher frequencies.