24 Bits Applicable or Non-Applicable in Your Music

::mic-minimal: · Post by **::mic-minimal:** » Thu Oct 02, 2003 1:05 am

I'm curious if I should record everything at 24bit, I'm using Reason, and Live, then turntables, guitars and bass and vocals. I have a large library of samples but none of them are 24bit native, I have alot of storage space
would I benefit from just staying 24bit throughout the whole process?, I'll also be using fruity after the rewire uprade version comes out for vstis in Live. is anybody just doing everything above cd quality, got any tips???

Guest · Post by **Guest** » Thu Oct 02, 2003 3:42 am

I'm in the same boat as you as far as gear and using live with reason, instruments, and fruity (I'm loving the FLrewire beta, it is solid!). I waffle between 16/44.1 and 24/48--I'm tending to try to just do 16 for gigs, as it seems like it makes for better software performance. That may just be faulty logic--ANYONE know if recording and working in 16 bits instead of 24 actually increases Live's performance (like more tracks and effects at 16, or lower possible latency at 16)???? I'd love to have a definite answer on that. Other than that, I try to do really important "studio" type stuff at 24. 24 takes up so much more space, and honestly i find it hard to tell the difference usually. The theory goes that 16 bit is not as capable of the full range of frequecies and dynamics as the human ear is, and 24/48 is beyond the capabilities of the human ear. I mean its gonna end up at 16 anyway, but 24 is there....its a dilemma. I think in the end, the quality of your AD/DA conversion is of more importance than 16 vs 24 bit. A bit off topic, but a quality soundcard with quality converters at 16 will make your instruments sound better than a lesser soundcard at 24.

2 cents

Ryan

Pitch Black · Post by **Pitch Black** » Fri Oct 03, 2003 12:27 am

A bit of anecdotal evidence - when I first made up my LIVE sets for existing songs I went the whole hog - 24bit, stereo all the way. Result: LIVE struggles to do everything I wanted it to do. I found that mono-ing files wherever possible AND running 16bit gave me a LOT more tracks.

Hey, but then I'm on a Tibook 500 so PC owners, ready... aim...

PS one of my favourite tricks for getting pseudo stereo out of a mono file is to insert a simple delay over the channel, 100%wet, no feedback, Delay time Left:minimum (use the milliseconds setting) Delay time Right:2 - 8 milliseconds. Yeah I know phasing and all that, but hey, it ain't broadcast and where I come from, folks LURRVE dat ol' outta phase business.

cheers
paddy

bensuthers · Post by **bensuthers** » Fri Oct 03, 2003 12:50 am

hey paddy! how are you going! is that your work on S-Dub album?

I find it a lot easier to record real instruments using 24 bit; you don't have to fuss so much over record level.

but once you're done, spend the time, normalise, maybe compress or limit a little, and export as 16 bit. (especially when working in live - I mean you're going to fuck them up even more anyway).

as paddy says, you'll get more tracks. well, out of TiBook's anyway

.

Guest · Post by **Guest** » Fri Oct 03, 2003 5:16 pm

I've noticed on my system that I get lower latency when I use a higher sample rate--but it increases the pobability I'll get crackles and pops. My monitors don't reveal the difference between 24 and 16 (or maybe my ears just can't hear the difference), but I stick to 24 bit. If you have plenty of storage and a fairly recent model PC, it shouldn't be a substantial performance hit to use 24 bit. You can always dither down to 16, but there is now way to dither up to 24.

Pitch Black · Post by **Pitch Black** » Sat Oct 04, 2003 9:08 am

Hey Ben!

Yep thats me on the Dub album.... engineering/producing/child minding...

How about a gig in, you know, January?? Maaaate!

I'll instruct my person call your person.....!!!

paddy

kent_sandvik · Post by **kent_sandvik** » Mon Oct 06, 2003 6:56 am

Well, I do now everything in 24-bit, most contemporary computers are fast enough, and there's plenty of disk space, so the additional bits are good when doing the final downsampling for CD or other production use.

If you really listen carefully, you could hear the dynamics missing in 16-bit, assuming you have a 24-bit sound card, of course. And sometimes the dynamics are good for the final mix-down, to hear the tiny details that could be missed otherwise. I noticed that when I used my internal 16-bit card for a while when moving from USB to Firewire audio boxes...

--Kent

Akhros · Post by **Akhros** » Wed Oct 08, 2003 7:07 pm

You shoud ever try to record at higher ratio that you have.
It is called resolution : higher it is the best - then we humans are not dolphins we cannot hear below 20 k and 20000 k start making our hears freeze...
Final range to capture the full audio we sense isn not upper than 16 bits 44.1khz - these khz are how many slices per sec we be able to capture to make the digital heard with no hole. 16 bits are double of 8 bits that capture mono voice.
Well when you are done with your mix on LIVE, your computer mathematical processor is more long calculating at higher resolution but indeed the result should be far better optimizing as you are still in process of an application in the puter could it be a mastering process ; there you use 48 khz 24 bits : but when you go back to burn all digital readers have a 16 bits 44.1 khz configuration.
After, higher resolution is a cost to pay for capturing harmonics ranges imperceptible or deteriored in analog definition you want to match over classical pro instrumentists or orchestra.
For what we do in pop we put so much compression to handle dynamics that there is no need to go for biggest definitions.

Akhros

Guest · Post by **Guest** » Wed Oct 08, 2003 7:36 pm

Well, most grown-up humans can't even hear above 15k, or even above 12k, it depend on hearing loss (for example by using headphones all day long, for a long time, semi-permanent or permanent loss).

But higher resolutions help the computer to handle harmonics that are then dithered down to 16bit CD use. There's also a better presence of dynamics. I think an easy test is to run the same source with 16-bit or 24-bit resolution and compare.

It of course depends if your computer could handle the extra load, most contemporary systems should be able to do it easily. --Kent

Guest · Post by **Guest** » Wed Oct 08, 2003 7:46 pm

Oh, I forgot t to mention that there's the notion of resolution (16 bit versus 24 bit) that has to do how accurately a sample is presented value-wise, and sampling rate, 22k, 44.1k et rest) that has to do with sampling audio witout getting the 'ripple effects' of duplicating waves, Nyquist theorem, i.e. you need to sample with double-sampling rate to get clean sampling signal.

So they are not directly tied together. But you of course need to sample with double the normal hearing range of humans or you get aliasing effects (those high pitch noises you might sometimes encounter with a bad sampling session), hence anything abobe at least 36k is used, but 44.1k is a common value (and yould could of course always get rid of the aliasing with a high-end filter, but it's preferred to sample things clean in order to avoid artifacts from filtering).

It is true that CDs, MP3 files and so on use lower resolution, with the help of dithering the higher values are preserved (one of those funny nature things, add a little bit low-end noise and it evens out abrubt level changes in the bits presented). However, the more accurate bit level, the more dynamics are present, compare for example 255 values for a signal versus 2 raised to 24. That makes a difference, especially when mixing, finding out missing holes in the sound spectrum, overcrowding in the middle range, and so on.

I'm no expert on this, anyway :-). But let the ear do the judgement.

--Kent

Guest · Post by **Guest** » Thu Oct 09, 2003 8:31 pm

ok but what about samples?, I understand that I should be recording my guitar at a higher resolution, but how about the lo-fi drum samples off of vinyl?

pgunders · Post by **pgunders** » Fri Oct 10, 2003 5:15 pm

Well if it's lo-fi I doubt the difference between 16 and 24 bits would be audible. Some artists even like to give their drum samples extra crunch by using old 8 bit samplers. You'll definitely hear the difference between 8 bits and 16.

pgunders

noisetonepause · Post by **noisetonepause** » Sat Oct 11, 2003 1:00 am

High fidelity my arse!

24-bit doesn't make sense when everything you're doing goes through a Behringer mixer into a standard Harman/Kardon amp and KEF speakers, anyways

Seriously, I can't be bothered. Also, I often find that running Reaktor at 22050 KHz gives me a sound that I like much more than running the same ensemble at 44100. But that's just me then...

-Paws

Akhros · Post by **Akhros** » Sat Oct 11, 2003 1:44 pm

Only one way there : straight into analog of audio card then - you'd sample 16 bits linear 44.1khz.
Underating your reaktor synth inside a computer is fun for it makes your sound loosing harmonics more playing on impair.
But it is not a normal sound you have there - if you like mickey voices and that genre of sound go for it - trash is fashionable these days.
The Live pug : bit reductor has same effect anyway - it makes of a good sound pure trash...

Akhros

24 Bits Applicable or Non-Applicable in Your Music

24 Bits Applicable or Non-Applicable in Your Music

We are not deaf :)

Analog vinylistics...