Dubstepforum.com

Is it worth it? I dont think so, do you?

yeah

more headroom and more to lose when the dithering process begins

Why wouldn't you? Btw does FL work in 16 bit, 24, or 32 bit?

yes it's worth it. this is like a 1997-era discussion.

Yeah, for my final bounce to send off for mastering, I always do it at 24 bit, and yes, you can hear the difference. Maybe not at a loud club, but definitely on monitors or headphones.

contakt321 wrote:Yeah, for my final bounce, I always do it at 24 bit, and yes, you can hear the difference. Maybe not at a loud club, but definitely on monitors or headphones.

so then 32bit would be the most ideal, right?

contakt321 wrote:Yeah, for my final bounce, I always do it at 24 bit, and yes, you can hear the difference. Maybe not at a loud club, but definitely on monitors or headphones.

so then 32bit would be the most ideal, right?

*please excuse the foolish question.

therook wrote:Why wouldn't you? Btw does FL work in 16 bit, 24, or 32 bit?

Yes, it has options for all three when bouncing audio out, and defaults to 32 bit.

I always bounce to 32. I bounce the highest quality I can...makes sense

my understanding is that higher bit-depth makes a difference during tracking and processing, but not at the mastering stage for listening purposes. I hear that 16 bit / 44.1kHz for the final listener is more than adequate. But i heard it in a context that was more focused on samplerate than bit depth. Perhaps 24 or 32 bit does make a difference. I honestly don't grasp what it is (I understand sample rates).

*sits back to hope for a bit-depth analogy of some kind*

gonna leave the math out of this, but basically:

16 bit audio is below the dynamic range that most people can hear; i believe it tops out at 96db. Thus, you're never really achieving full fidelity, and if you don't record things on the way in as hot as possible, you're essentially only recording at 8,12,14 bits-- and we all know how things bitcrushed to 8bits sound.

24 bit audio is just beyond the range of human hearing, which gives you a ton of wiggle room-- you no longer have to smash your recordings or mixes to get full-on dynamic fidelity. Provided you're using a good convertor, transients are punchier, reverbs are cleaner, and there's more depth and detail in a recording.

most programs process audio at 32-bit floating point, which gives you even more dynamic range-- mainly to prevent internal clipping.

w/ that said, 32 bit audio is really kind of useless when it's converted to fixed-point, exported 1's and 0's-- it doesn't offer anything more than 24 bit does when all is said and done, and most apps don't like opening 32-bit files.

once audio is limited and maximised up to 0db, it doesn't really make a practical difference if you're playing 24-bit or 16-bit audio out. at 100db in a club, for the life of me I can't tell the difference at full-res audio, while the difference between full-res and an mp3 can be pretty staggering.

On the whole it's almost always best to avoid 16 bit for everything but the final master*, because unless you're dealing with material with little dynamic range that's peaking at near 0dBFS, you lose 1 bit of depth for every 6dB you go down. For example, if you're saving a mixdown that peaks at -18dB then you're actually only getting 13 effective bits. If you're saving a drum loop with much more dynamic range than a mixdown, the quiet bits (decay, reverb etc) feasibly might be sitting at what, another 18 dB under that, or -36dB? So here you're only getting 10 bits.

Conversely, with 24-bit audio you have an extra -48 dB of headroom (=8 bits more than 16) before you get to the point where you're effectively recording with 16 bits of depth.

32-bit is the odd one out because it's a floating-point format. I wouldn't bounce mixdowns to 32-bit because there's not much point and some audio players don't like it, but it can be good for resampling very quiet stuff because you don't have to worry about headroom at all.

(Worth noting also that pretty much all DAWs with the exception of Pro Tools run in 32- or 64-bit float.)

*MASTER, not premaster. As in, what goes onto CD. Give the mastering engineer a 24-bit mixdown.

Thank you so much, i am beginning to understand it.
Does anyone think i should record at higher than 44.1?

As a rule of thumb, record in the highest sample and bit rate you can and then convert down as and when required and yeah, 32 bit floating-point is generally best used in your DAW as the playback bit rate because it means that if/when you clip on the master fader, it won't distort (due to the floating point, it basically means that the decimal point moves when required)... but as Sharm said, when you reduce it down to 24/16 on export then obvs it will clip unless you do something about it :3


Dub Fiend

Sharmaji wrote:at 100db in a club, for the life of me I can't tell the difference at full-res audio

especially cos in these scenarios I am wearing earplugs anyway....

so do you guys work in 24 bit the entire time for your whole project? Or do you keep it at 16 to save resources, then switch to 24 when you bounce clips/the whole track to audio?

I work in 24 bit, my samples are mostly from dvds at 24bit, 48,000hz, any bounces I make from synths are also at that res, my soundcards playback is set to 48,000, same with the settings in reason and ableton. Ableton becomes grinding with those rates by the time youve got 30/40 tracks of audio going, I freeze and flatten everything because running 10 high demand synths each with 4 gliss eqs on high quality mode with reverbs etc, it aint having it. But I believe it sounds better as a result, I can hear the clarity and depth in the 24/48 bounces compared to 16/44. Might just be the encoding or whatever, I dunno, but to me its a more translucent, expansive sound, its very subtle but its there, certainly on my limited setup. And thats just 1 bounce, if Id used 16/44 res through the samples in the track, Im sure the difference would be more blatant.

My experience in music and graphic design within a pc is where ever possible retain as high a res as you can. If cd quality is the highest anyones going to hear it then fine, thats what they get at the end but you want to be working in the most quality possible through the process. If I design an album cover I'll do it at about 18x18 inches, full print res, even if its just going on a cd cover thats going to be about 5x5, A. You never know where its going to end up, B. it cant hurt as long as you are aware of how the final product will be presented. Artwork now days has got to work at a thumbnail size, tunes have to stand up at 192kbs mp3 size, but your art would be seriously limited if you worked in those formats from the word go. When you start stacking up layer upon layer of sound or image little differences that were barely noticeable alone start stacking up to something quite noticeable in the end product.

^damn you must have a sick setup. i have an amd quadcore and i start hitting 50% in ableton after like 10 tracks and 10 plugins or so

ruckus49 wrote:^damn you must have a sick setup. i have an amd quadcore and i start hitting 50% in ableton after like 10 tracks and 10 plugins or so

That'll be RAM more than anything else my friend... I've got 8Gb of RAM and I can run 30+ instances of Massive + effects at a time (I've got a 2.66GHz dual core processor overclocked to 3.0GHz too but you've obvs got a better processor than me and yet you're coming up short) Invest in some good RAM (and a better motherboard if you only have 2 RAM slots) but you don't even have to go as high as 8Gb... Upgrading from 1/2Gb to 4Gb (2 sticks of 2Gb) will make an incredible difference to the load you can put your PC under Make sure that you are running Windows 7 though (if you're on a PC that is) because Vista and XP were made in a time where accumulating that amount of RAM wasn't possible, so unless you have the x64 versions of Vista/XP then it will limit the RAM size to 3Gb regardless of actual RAM :3 Hope this helps you out squire


Dub Fiend

I've only ever produced in 16/44, figured the end medium is going to be either 16/44 wav or even worse mp3.

However, my thinking is, if most of the samples I'm using come from 16 bit sources, and anything I bounce down and reload back into my project at 16 bit is not turned up later on, then shouldn't 16 bit be fine? Sure if I was bouncing shit down and then deciding hey I wanna turn that up 6dB later on it would matter but I dont so no biggie?

Then upon 'master' bouncing, just bounce a 24 to give mastering as much resolution as possible... all my synths and 'quiet' sounds will still be bounced at the full 24 bit resolution anyway right? Someone tell me if this is stupid.

For recording however, 24 bit seems totally logical with regards to headroom etc.