Topic: 24 bit > 16 bit

[Petrus]

Well, in theory with a 96 kHz samples per second recording you get 96/2 = 48000 Hz upper frequency limit, in practice with the pre AD brickwall filters about 45 kHz upper limit. But, like I tried to point out, as practically all microphones, ALL reproducers and ALL HUMANS cut off at around 20000 Hz at the latest, there is no use, point, need nor any sense trying to record something that does not even enter the recording chain. And if it enters, does not get out. And, if by some freak phenomenon, would get out, only bats would hear it.

There is no additional benefits to recording at 96 kHz sampling rate exept higher cut off limit. It does not reveal any "hidden detail" or "unveil" the sound. Just that also the frequences we can not hear can be recorded (if the mic were good enough, and it is not).

---------
Against stupidity and ignorance the Gods themselves contend in vain.

[Nick's Picks]

all true....

but when all is said and done...a *carefully* recorded 16bit capture will always be very, very satisfying to the ear.
still....., headroom is an excellent thing to have.

[Petrus]

24 bits give usable headroom and safety margin, 96 Kz sampling adds only 100% to the file size.

24/96 has higher quality than the analog window (what we can get to the recorder), no need for that (96 Hz sampling, that is)
16/44.1 gives better quality than the acoustic window (noise floors, practical dynamic range, loudspeakers & amps, hearing limits)

It is the acoustic window which defines total fidelity, not even 16/44.1 recording done carefully (or downsampled from slightly careless 24/44.1).

nuf said.

[JasonSobel]

nuf said.

actually, not really.  As has been mentioned earlier in this thread (by myself and others), there are other potential benefits for recording at 96 kHz regardless of the frequency response.  the higher sampling rate allows the recording to better capture timing and spacial information, because we can hear different sounds less than 1/48000th of a second apart.  96kHz allows the recording to more accurately define exactly when a sound occurs, which can produce a more realistic and accurate soundstage.

so, while I agree with Petrus that there is no need to record sound up to 45 kHz, as that is way beyond anything we could ever hear (or anything our microphones could ever capture), there is more to the story and other reasons why one might choose to record at 96 kHz.

[Digital Quality]

Greater dynamic range and higher resolution are two ways of saying the same thing.

[chunga1]

so if we can only hear up 45kHz. Why do they even offer these settings? For more clarity and headroom ??whats does "headroom" and "brickwall" mean in recording terms?

tim

[Nicola Fankhauser]

hi

so if we can only hear up 45kHz. Why do they even offer these settings? For more clarity and headroom ??whats does "headroom" and "brickwall" mean in recording terms?

please search in the forums for the terms "headroom" and "brickwall", and re-read this whole thread to get an understanding of the issues at hand...

regards
nicola

[chunga1]

searched for "brickwall" "headroom" no answers just people using the term....unless I'm just not understanding the context in which it's being used...

tim

[Digital Quality]

There is also an issue of transients - the first strike of a string or drum. Higher frequencies allow us to capture transients more accurately.

I'm also not a fan of running the AD conservatively for 24 bit recording. SNR numbers for amplifiers are almost always given at or near maximum gain because that's where the amp performance is best. Less gain = lower SNR in the analog stages. Running the AD at anything less than full scale is also sub-optimum, less bits = higher quantization noise power.

BTW - quantization noise is also random, highly uncorrelated with the input signal.

Brickwall is pretty straight forward - it happens when you exceed the rail to rail voltage on internal stages of the amplifier. The resulting output signal is clipped. Headroom is a little more nebulous and can mean a lot of things. In the analog realm it could just mean that you increased the rail voltage on an internal stage and now you can handle a hotter signal. In the digital sense it could mean that you have 24 bits available and realizing you will only keep 16 bits of resolution in the end you can justify running conservatively not worrying about running the AD at full scale.

[Petrus]

Fast transients are also nothing more than high frequency components of the audio signal. If the component is more than 20kHz we can not notice it and that's it. Higher frequencies CAN NOT be hidden in "transients" or  "timing or spatial" information. They do not exist to human ears.

Headroom means the extra leeway we get with 24 bits, no need to record as "hot" to avoid hiss, if sudden peak appears in the signal the headroom can absorb it without clipping. An analogy: you have a 3 gallon (16 bit) bucket with water running into it at steady flow and water running out at same speed keeping it at 2 gallon level. You can not splosh 2 gallons in there without the bucket runnung over. With a 6 gallon (24 bit) bucket you can have that same 2 gallons there, but add suddenly that 2 gallons without spilling anything...

Brickwall filter means a very steep lowpass filter, which is needed to prevent high frequences reaching the A/D converter in digital systems (a 44.1 kHz sound signal that would reach the 44.1 kHz converter would spell disaster) . Brickwall, because we want also quite high frequences there (max half the sampling rate), but absolutelly nothing at or near the sampling frequency. So the frequences get higher and higher, and at certain point hit the wall... Same thing happens after D/A conversion; the signal is full of high frequency artefacts, we need to smooth out the waveform back to what it was before sending it to the A/D. So we brickwall filter everything above 20 kHz out of the signal to make the staircase look like a wave...

And the high limit of hearing for young persons is 20000 Hz (20 kHz), the lower limit is more vaque, depening on the loudness level it is somewhere between 8 and 20 Hz. Also the lowest sounds are also felt, not only heard.

[chunga1]

perfect thank you very much Petrus.....

[JasonSobel]

Higher frequencies CAN NOT be hidden in "transients" or  "timing or spatial" information. They do not exist to human ears.

we're not saying that the higher sampling rates are needed to capture frequencies BEYOND human hearing.  we are in agreement there.  but you simply cannot deny that *when* a note is played is as important as *what* note is played.  if it didn't matter when the drummer hit the drums or when the guitar player struck a chord, then music wouldn't exist as we know it.  the higher sampling rate better defines *when* a musical note occurs.

[boojum]

Higher frequencies CAN NOT be hidden in "transients" or  "timing or spatial" information. They do not exist to human ears.

we're not saying that the higher sampling rates are needed to capture frequencies BEYOND human hearing.  we are in agreement there.  but you simply cannot deny that *when* a note is played is as important as *what* note is played.  if it didn't matter when the drummer hit the drums or when the guitar player struck a chord, then music wouldn't exist as we know it.  the higher sampling rate better defines *when* a musical note occurs.

JS - assuming that you are correct, can this information be dithered down to 16/44.1???  If it cannot the point is moot.   

[JasonSobel]

Higher frequencies CAN NOT be hidden in "transients" or  "timing or spatial" information. They do not exist to human ears.

we're not saying that the higher sampling rates are needed to capture frequencies BEYOND human hearing.  we are in agreement there.  but you simply cannot deny that *when* a note is played is as important as *what* note is played.  if it didn't matter when the drummer hit the drums or when the guitar player struck a chord, then music wouldn't exist as we know it.  the higher sampling rate better defines *when* a musical note occurs.

JS - assuming that you are correct, can this information be dithered down to 16/44.1???  If it cannot the point is moot.   

yeah, I guess this whole thread is about 24 bit > 16 bit, and not just the benefits of 24 bit / 96 kHz recording in general.
the idea is that we can hear time differences less than 1/48000th (or 1/44100th) of a second.  so, by definition, a 44.1 kHz recording will not be able to define exactly when a note occurs as accurately as a 96kHz recording.  I'll be honest that I don't know all the technical ins and outs of this, but I'd guess that there's no way to bring that time accuracy back down to 44.1kHz.

[Roving Sign]

Higher frequencies CAN NOT be hidden in "transients" or  "timing or spatial" information. They do not exist to human ears.

we're not saying that the higher sampling rates are needed to capture frequencies BEYOND human hearing.  we are in agreement there.  but you simply cannot deny that *when* a note is played is as important as *what* note is played.  if it didn't matter when the drummer hit the drums or when the guitar player struck a chord, then music wouldn't exist as we know it.  the higher sampling rate better defines *when* a musical note occurs.

JS - assuming that you are correct, can this information be dithered down to 16/44.1???  If it cannot the point is moot.   

yeah, I guess this whole thread is about 24 bit > 16 bit, and not just the benefits of 24 bit / 96 kHz recording in general.
the idea is that we can hear time differences less than 1/48000th (or 1/44100th) of a second.  so, by definition, a 44.1 kHz recording will not be able to define exactly when a note occurs as accurately as a 96kHz recording.  I'll be honest that I don't know all the technical ins and outs of this, but I'd guess that there's no way to bring that time accuracy back down to 44.1kHz.

JS - I was having similar thoughts about using 16/44 sources to matrix with - just how close can I align them?

It would seem that higher sample rates would allow finer grain control of time alingment...

sorry to contribute to the hijack...