Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[grantis]

Quote:

Regarding the claim that the 002s inability to record 16 simultaneous tracks at higher sample rates was an intentional limitation, please keep in mind that the ADAT format technically does not support sample rates above 48k.

Technically yes...but why no S-MUX?

[Podgorny]

Pro Tools' HD interfaces don't support S/MUX, so why should LE?

I think the real question is, why do you need to record at 96k?

[grantis]

So you can hear all those frequencies at 48k!!!!

[mixwell]

grant richard wrote on Tue, 02 September 2008 17:23

So you can hear all those frequencies at 48k!!!!

More importantly, can your converters analog input stage hear up to 48K and is its transient response good enough to capture all the overtones that are produced in and above the human hearing range?....

I think Human hearing is also about perception, and I think with higher sampling rates, it only makes sense that there are more puzzle pieces snapped into place. Conversion is a FINITE process vs Air & Time [INFINITE] which is converted to electrical VOLTAGE [Amplitude] and PITCH [Frequency] which is also INFINITE and the carries along at the speed of light.

A smooth moving wave form becomes a randomized step ladder when its digitally encoded. So the laws of common thinking dictate that we will notice a higher resolution of captured frequency content. So, the clocking, the analog stage, the power supply, the DSP filtering, all of these things matter to the converters response, but I assure you, if your recording a new version of GiGi Allen, then there's no point in recoding 96K.

There is a TON of stuff I would never record WITHOUT being in 96K, and there is a TON of stuff I would NEVER switch the converter higher than 44.1K. Some converters won't sound different in 96k, they'll sound worse.

[compasspnt]

mixwell wrote on Tue, 02 September 2008 23:35

A smooth moving wave form becomes a randomized step ladder when its digitally encoded.

Hmmm...

[grantis]

mixwell wrote on Tue, 02 September 2008 22:35

grant richard wrote on Tue, 02 September 2008 17:23 So you can hear all those frequencies at 48k!!!! More importantly, can your converters analog input stage hear up to 48K and is its transient response good enough to capture all the overtones that are produced in and above the human hearing range?.... I think Human hearing is also about perception, and I think with higher sampling rates, it only makes sense that there are more puzzle pieces snapped into place. Conversion is a FINITE process vs Air & Time [INFINITE] which is converted to electrical VOLTAGE [Amplitude] and PITCH [Frequency] which is also INFINITE and the carries along at the speed of light. A smooth moving wave form becomes a randomized step ladder when its digitally encoded. So the laws of common thinking dictate that we will notice a higher resolution of captured frequency content. So, the clocking, the analog stage, the power supply, the DSP filtering, all of these things matter to the converters response, but I assure you, if your recording a new version of GiGi Allen, then there's no point in recoding 96K. There is a TON of stuff I would never record WITHOUT being in 96K, and there is a TON of stuff I would NEVER switch the converter higher than 44.1K. Some converters won't sound different in 96k, they'll sound worse.

While I agree, I've found that the increased load on DSP, (or CPU) and the disk space usage makes it not worth the extra .5% of quality that may or not be gained.

Not to mention the time it takes to shoot out sample rates.

[Silvertone]

Some are saying the max is 16bit 48K out the ADAT lightpipe but I'm 99% sure you can do 24bit 48K.

[marcel]

Silvertone wrote on Wed, 03 September 2008 06:38

Some are saying the max is 16bit 48K out the ADAT lightpipe but I'm 99% sure you can do 24bit 48K.

You can.  I do.  Every day.  

[compasspnt]

mixwell wrote on Tue, 02 September 2008 23:35

A smooth moving wave form becomes a randomized step ladder when it's digitally encoded.

Hey Adam,

I have been enjoying your posts recently...many good thoughts.

But this one gives me pause to wonder.

Can you please explain what you mean by this statement?

Regards,

Terry

[JDNelson]

Yeah, it's my understanding that with PCM the pulse-code, pulsing at the sampling rate, modulates a carrier signal in a similar fashion to the same way that radio works.  The stair step analogy is not accurate in conceptualizing digital PCM music, in other words.

[Barish]

compasspnt wrote on Wed, 03 September 2008 19:02

mixwell wrote on Tue, 02 September 2008 23:35 A smooth moving wave form becomes a randomized step ladder when it's digitally encoded. Hey Adam, I have been enjoying your posts recently...many good thoughts. But this one gives me pause to wonder. Can you please explain what you mean by this statement? Regards, Terry

I think he is talking about that diagram where a sine wave is sliced vertically and horizontally, and then laddered in the crossing boxes, in order to explain to the novices how and where the analog-to-digital sampling is done along the analog wave.

Normally, there's no such result. It's just an indicative graphic thing. It all becomes one-amplitude square wave toggling between 0V and 5V once the sampling to digital is achieved. No stepladders, no cliffhangers.

And it is never random. If it were random, you couldn't possibly put them back together in the DA ever again.

It's all in order. Very strict order.

M.

[Tomas Danko]

compasspnt wrote on Wed, 03 September 2008 19:02

mixwell wrote on Tue, 02 September 2008 23:35 A smooth moving wave form becomes a randomized step ladder when it's digitally encoded. Hey Adam, I have been enjoying your posts recently...many good thoughts. But this one gives me pause to wonder. Can you please explain what you mean by this statement? Regards, Terry

Adam is talking about a very common misconception regarding how digital audio works. The "steps" are just a temporary state to store the audio data but the output does not resemble a step ladder after all the reconstruction filters etc. The waveform will never look like those numbers, although that can be quite difficult to wrap one's head around.

Mr. Paul Frindle has written some great primers on this, if you Google.

Here at PSW there are numerous threads dissecting this myth. Jon Hodgson has contributed a great deal, for example his posts in http://recforums.prosoundweb.com/index.php/mv/msg/23895/0//2 699/ which is a very big thread with a lot of noise but halfway through you can find some gold from Jon.

For the record, this does in no way take away from Adam's big pool of knowledge and experience. Most audio people I know believe the same thing about digitized audio, but it doesn't make any difference when it comes to talent and professionalism. Adam, you're one of the better contributors to this forum if you ask me.

It's just that I can't seem to ignore bringing up this subject whenever someone mentions it. Call me a digital audio evangelist, perhaps.

We need to bust this myth once and for all.

[mixwell]

compasspnt wrote on Wed, 03 September 2008 13:02

mixwell wrote on Tue, 02 September 2008 23:35 A smooth moving wave form becomes a randomized step ladder when it's digitally encoded. Hey Adam, I have been enjoying your posts recently...many good thoughts. But this one gives me pause to wonder. Can you please explain what you mean by this statement? Regards, Terry

I guess I was lumping in F-F-T quantizing amplitude values with clocking.....MY BAD.....although the clock is a square wave...no?

It's measured by how GOOD [whatever that means] it can repeat. By my estimation, there's no crystal, or digital clock that will measure the actual frequency content the same way your microphones do.

Papi can hit the ball as close to the Green Monster as he can, but its still a foul if it goes beyond that yellow line no matter how you slice it!

The Big Ben Clock just about makes the square wave [0,1,0,1] a sine wave [at the corners], so I only make that reference in regard to the clock positioning accuracy with respect to the actual time that exists apart from the time that is accounted for by the clock. Look fellas: I'm not saying I claim to know all of this technical crap that has nothing to do with music, I just observe things by experimenting, and thats what leads me to my conclusions, backed up by my prestigious Associates Degree in Audio Production [Kidding].....

[tom eaton]

mixwell wrote on Thu, 04 September 2008 16:37

It's measured by how GOOD [whatever that means] it can repeat. By my estimation, there's no crystal, or digital clock that will measure the actual frequency content the same way your microphones do.

Microphones are transducers, first... changing energy from one form to another. So no, until you get to a speaker, there is no other piece of the chain that is similar.  Regardless, crystals and digital clocks have nothing to do with "measuring frequency content."  The sampling rate and filter design determine the bandwidth of the recorded signal.

Quote:

The Big Ben Clock just about makes the square wave [0,1,0,1] a sine wave [at the corners], so I only make that reference in regard to the clock positioning accuracy with respect to the actual time that exists apart from the time that is accounted for by the clock.

Correct me if I'm wrong but I believe the clock data is derived from zero crossings... and the signal is more like -1,1,-1,1.

tom

[mixwell]

Quote:

Microphones are transducers, first... changing energy from one form to another. So no, until you get to a speaker, there is no other piece of the chain that is similar.

Hey Tom,

I am of the opinion that converters and transducers are one and the same, even if its not technically and scientifically correct to say so.