Protools 192k

[Digital Sound Lab]

The real question is will this 192 HD or whatever sound good enough and be reliable enough (no errors, bugs and such) to justify the cost of upgrade time and money? I think not now but mabey after the bleeding edge is dulled down and clogulated (spelling?) a little.
I love my PT system now but I'm excited as a mudaphuqa (spelling?)!

[Noiz2]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by Park Seward:
It is a common misconception to think that more samples will more accurately describe a waveform. That is not so. Once a waveform is accurately described, more samples will not describe it more accurately.

A higher sampling rate lets you describe higher frequencies but does not make the lower frequencies have more detail or resolution.
-----------------------------------------------------

First, as illogical as it seems, bits does NOT equal better resolution of anything but the noise in its signal path. More bits ONLY equals dynamic range.
<HR></BLOCKQUOTE>

I would quible about theese two points. In the first while it is true for a simple waveform like a sign wave it is not nec. true for more coplex waveforms. Higher sample rates will more acuratly reflect subtle distortions in even low frequency waveforms. And the second quible is that I think the term "volume resolution" would be a more acurate term than "dynamic range", since more bits will give you better representation of subtle volume changes in the middle of the "dynamic range". You do get more "dynamic range" but you also a bit more "resolution" also.

[Philthy]

Wait, wait, reality check:


MP3


[img]images/icons/wink.gif[/img]

[Jules]

Me too!

I just realised that as I do mostly 1-3 track rock sesssions mostly a 96k standard would be OK and perhaps not max out my drives to much...

Jules

[Nika]

In regards to the people discussing "beat frequencies", or "subtones" created by inter modulation distortion, let me address:

Fortunately the thread on bit depth has about ceased so we can tackle the other aspect of sampling - sample frequency - I suppose.

The argument here is that when two tones are played together in an acoustical space there are two tones that get created. There is a sum tone and a difference tone (A-B and A+B) that the ear can hear. This means that if you play 400Hz and 500Hz in an acoustical space, the ear would be able to hear a 100Hz harmonic and a 900Hz harmonic.

For those that have never experienced this before, it is easy to try. Create a mixer engine with two channels and the signal generator plugin on both channels. Pan them hard left and hard right. Put a sine wave in one at 400Hz and one in the other at 500Hz. If you listen and move around the room you will hear the 100Hz and the 900Hz enter into your system.

The crux of this argument is what happens to the subtone if the two root frequencies are both above the Nyquist limit (~20kHz) and get filtered out. Then there would be no subtones created that we do normally hear. After all, if I'm listening to an orchestra in the 6th row seat there is a lot of ultra HF content bouncing around, and a lot of this content would make it back to our ears as content below 20kHz but is stimulated by frequencies above 20kHz, no? Violins and trumpets and other instruments do put out information up at around 35kHz, so all of the 35kHz frequencies bouncing around would create some real information at around 1kHz, right?

OK, so the other people say that if you record a single microphone in an acoustic space it will end up capturing all of these complex subtones just like the ear does, so it is not a concern.

But the other side says that when close micing any ensemble each microphone is limited to 20kHz, and the "blending" is done in the mixing console. Since the mixing console is also limited to 20kHz, there is not any way for those higher frequencies to blend, creating subtones in our hearing range that we've been missing. The mixing process is inherently not going to give us as accurate of a mix as real acoustic spaces do because we haven't allowed for any way for these high frequency tones to blend, creating audible "enhancing" data below 20kHz.

Here is the reply to this particular validation of higher sampling frequencies:

The creation of these subtones is a characteristic of the ear, and the ear is unable to hear any subtone frequencies caused by inter modulation distortion if EITHER root component is above the audible frequencies of the ear. Thus, a 24kHz and a 25kHz tone will NOT create an aubible 1kHz tone. A 22kHz tone and a 15kHz tone will NOT create an audible 7 kHz tone.

Using the same signal generator plugins in Protools on the same mixer template mentioned above this will be easy to duplicate.

Therefore, there is NOT a benefit to higher sampling frequencies for the sake of allowing "beat frequencies" or "subtones" to be created.

I hope this helps.

Nika.

[Nika]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by Noiz2:


I would quible about theese two points. In the first while it is true for a simple waveform like a sign wave it is not nec. true for more coplex waveforms. Higher sample rates will more acuratly reflect subtle distortions in even low frequency waveforms.<HR></BLOCKQUOTE>

The anti aliasing filter in an A/D converter works much like an ear - brickwalling everything above a fixed point. A notable difference is that a good anti-aliasing is inherently higher than most human ears' frequency range.

So what data is it that you'd like to capture that is higher than the anti-aliasing filter in an A/D, which is higher than the human ear? And why would these subtle changes somehow make a difference to a band limited ear that is being fed everything from the signal that is within it's band?

Nika.

[Nika]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by SmlTwnGuy:



I wouldn't be quite so confident quoting Nika. Bob Katz, yes.<HR></BLOCKQUOTE>

I never write anything with the intent of being quoted. So often, it seems, I am quoted out of context, or from early points in a discussion that address a particular post. I kringe when I see myself quoted for fear that it is either a quote from when I didn't fully understand a concept, and the quote will make an ass of me, or because it will be trying to take something out of context, unknowingly.

So please, do quote Bob Katz instead!

[Nika]

RE: Frederick Umminger's post.

I completely agree. Many effects plugins and such already double sample or more in order to give them more data to play with. This is a common way of designing good quality effects algorithms:

Upsample -> effect -> downsample.

According to a couple of these engineers, however, they say that the only advantage to being fed higher sample rate material is that they don't have to do the up and downsampling themselves, freeing up at least SOME of the DSP that they would require.

This argument is relatively moot because the small amount of DSP that they are not using, suddenly, would get used multiple fold over elsewhere.

They say that there would be no sonic benefit to being fed the higher sampling rate material over base rate.

Nika.

[Park Seward]

Three trumpets playing together in a space and recorded will sound much different than one trumpet playing in a space and recorded three times. Same with vocals. The sound of people singing together in a space will sound better than single tracking each individual voice one at a time.

[Nika]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by Park Seward:
Three trumpets playing together in a space and recorded will sound much different than one trumpet playing in a space and recorded three times. Same with vocals. The sound of people singing together in a space will sound better than single tracking each individual voice one at a time.<HR></BLOCKQUOTE>

Yes, but this is not because the mixing process doesn't allow for intermodulation distortion to occur on individually mic'd tracks.

Going to 192kS/s will not suddenly allow individual tracked tracks to mix together and sound like it was a single mic in a performance space.

Nika