Pro Tools sessions at 44.1k vs. 48 k?

[sfaudiogeek]

How much more memory is required to record 48 tracks at 48k as opposed to 44.1 k? at 24 bit?

[robmix]

Originally Posted by sfaudiogeek

How much more memory is required to record 48 tracks at 48k as opposed to 44.1 k?

At what bit rate ?

[johnjm22]

Prolly 24 bit.

Nobody still records at 16 bit these days!

[Ziggy!!]

Im guessing you are using protools|LE because TDM sets track counts on DSP power.


The question you are asking doesn't have a definitive answer (im having a bad day so you are lucky im not swearing and yelling at you for asking such open ended questions - one of my pet hates ).


Breathe... relax... Breathe.





Lol... but anyway, when you say memory do you mean harddrive space or RAM?

Track counts rely on both RAM and HD speed and space as well as CPU power.

To answer your question, 48khz uses roughly 9% more resources than 44.1khz... roughly. YMMV. There isn't a definative answer, but if your ram is suffering just double it


Good day to you sir!

[XSergeantD]

I did the math a while ago and posted it on my computer monitor in storage, but from my volatile memory....
about 8.2 to 8.8% so you can calculate using 9% to cover any variables
if I remember right, breaking down to make things difficult, it's something like (please correct if this is off):
44.1 - 1.35/sec
48 - 1.46/sec
but don't go math all the way, cause you have a 'fudge factor'. OK it's not a fudge factor - more about the fact of binary than anything else: 1,024 bytes equal 1 KB; 1,024 kilobytes equal 1 MB; and 1,024 megabytes equal 1 GB.

more understandable numbers:
boils down to bout (44.1)5.05MB/min to (48)5.49MB/s on a mono track

this is all 16 bit, but for 44 to 48, a percentage is a percentage. If you need to go to 24 bit, multiply my numbers by 1.5 and then do the percentage calculations - should remain the same



and whats a half a byte?

a nibble.

OK then, [(48/44.1)-1.0]*100 := 8.8435374149659863945578231292517%

* * * * * * * * * * * * * * * * *
and now, the fun math:
x = y.
Then x2 = xy.
Subtract the same thing from both sides:
x2 - y2 = xy - y2.
Dividing by (x-y), obtain
x + y = y.
Since x = y, we see that
2y = y.
We divide by y
Thus 2 = 1

[matt thomas]

thirty nine four hundred and eightyths. Minus a bit.

ooooo, maths

thirteen seventieths as much again, minus a bit

[fremitus]

a more relevant and perhaps incendiary comment would be...

'what's the point? if you're just doing audio 44.1 is BETTER than 48.'

why? cuz it's going to end up at 44.1 and the downsampling is far more deleterious than the simple sample gain at input. up the bit rate and smile at the CLEARLY evident difference in sound. up the sample rate that nominal amount to 48 and watch time slip away as you convert and wonder where all the good bits went...

[djui5]

http://www.gearslutz.com/board/showthread.php3?t=44123

[Ziggy!!]

Originally Posted by fremitus

a more relevant and perhaps incendiary comment would be...

'what's the point? if you're just doing audio 44.1 is BETTER than 48.'

why? cuz it's going to end up at 44.1 and the downsampling is far more deleterious than the simple sample gain at input. up the bit rate and smile at the CLEARLY evident difference in sound. up the sample rate that nominal amount to 48 and watch time slip away as you convert and wonder where all the good bits went...

Well I guess you know that sample rate conversion upsamples to some ridiculous sample rate before downsampling regardless of the initial samplerate, it always ends up at the same upsampled rate first.

[XSergeantD]

guess you're assuming he's printing digitally, but your point is moot if he's laying down to 1/2"
remember what they say about assuming, u make an ASS outta U and MING. but who's MING, and who are they?

Originally Posted by fremitus

'what's the point? if you're just doing audio 44.1 is BETTER than 48.'

why? cuz it's going to end up at 44.1

[dokushoka]

44.1khz was never considered a "professional" standard, it was the consumer one. That is why we had 48khz even from the get go.

48khz has a nice shallow filter. I actually PREFER working at 48khz for most material, and, I actually know a lot of other engineers that do as well.

The real answer is:
If you have good converters, sample rate doesn't matter that much. If your converters are shit, go as high as you can...

[JSVice]

Originally Posted by XSergeantD

* * * * * * * * * * * * * * * * *
and now, the fun math:
x = y.
Then x2 = xy.
Subtract the same thing from both sides:
x2 - y2 = xy - y2.
Dividing by (x-y), obtain
x + y = y.
Since x = y, we see that
2y = y.
We divide by y
Thus 2 = 1

Tsk, tsk. Don't try and sneak that "division by zero" jazz by me!

[JonCraig]

Originally Posted by JSVice

Tsk, tsk. Don't try and sneak that "division by zero" jazz by me!

ahhh... now i get it. (algebra has gotten a little fuzzy...)

--jon

[XSergeantD]

Originally Posted by JSVice

Tsk, tsk. Don't try and sneak that "division by zero" jazz by me!

Damn you. You thwarted my attempt to take over Digi by dividing their bits by zero. I tip my slide rule to you

[nukmusic]

Go 48k/24bit. It did sound a bit smoother on the top end to me.

[JonCraig]

Originally Posted by nukmusic

Go 48k/24bit. It did sound a bit smoother on the top end to me.

go 44.1. i don't like what the SRC does to the top end when you track @ 48.

to each his own, eh?

--jon

[max cooper]

Originally Posted by nukmusic

Go 48k/24bit. It did sound a bit smoother on the top end to me.

I noticed a lot more 'fatness' at 48k. I've been meaning to do the SRC thing myself, but haven't had the opportunity. I'd assume 44.1 - 44.1 is better than 48 - 44.1, I mean it makes sense, but ya never really know until you hear it yourself, right? But that seems to be a pretty standard bit of wisdom.

I'm thinking to have a synth and a drum module running from midi files so that the recordings match as closely as possible; what do you guys think? The other way would be to play some LP's into the Apogee; one at each SR. Maybe that'd be better.

[juicemaster1500]

The gain in sound quality with dsp at 48k far outweighs any problems with sample rate conversion between 48 and 44,1k imo.

[dokushoka]

Or just get some good SRC software and then you're fine.

[robmix]

And I would say just the opposite, avoid the extra math and track at 44.1khz. The SRC methods I've tried were far more destructive to the audio than any benefits I could gain from tracking at 48khz. We're talking about applying an extra conversion just so we can raise the sampling rate 4khz, thus moving the filter point from 22khz to 24khz ? No sir, I don't like it. Now mixing to analog, whole different story. dokushoka pretty much nailed it, if you're hearing a major difference there's good chance your convertors are not of the best quality.

[Bob Olhsson]

Many converter designs are optimized for 48k and don't even have separate filters for 44.1. The result is a lot of common gear that sounds better at 48k because it sounds worse than it needs to at 44.1. When in doubt, I generally do 48k.