44.1khz, 48khz, 96khz, or 192khz? Opinions? Answers?

I don't think the human ears can hear much of a difference. However anything higher then 44.1khz is used for the head room. So better recording, better/easier mix/ better song. No it wasn't necessary but it will help.
 
It wasnt necessary but an absolutely fantastic decision!

In my opinion, absolutely everything before the master engineer should be recorded and mixed down at 24 bit 96 kHz.

The human ear can't "hear" above 20khz. But there is soo much energy above that we perceive that adds so much motion to a track. 44.1 is reaallly on the fence for "lossless" back in the cd days it was fine. But with the beauty of digital why lose quality if you don't have too.

Also enough CPU is important....
 
what gear did you upgrade exactly.....
did u not have those setting in your daw...??

Can you guys explain why exactly its important to mix and export in the way you explained....
-why 24 - 96... why not go as high as you can
 
Basically it's like this, the higher the bit depth and samplerate the easier it is to make a loud mix without clipping.The drawback is that 96khz is double 48khz so the filesize will be twice as big.If you make beats that are like 4 minutes long on average you're looking at a 150-400mb project file.[If you use samples in beat].

samplerate doesn't matter much for headroom, it's strength is in the flexibility of stretching.Higher samplerates make it possible to avoid doing the duplicate method on chops that would normally much longer if you wanted it precisely how you heard it in your mind.Bit depth is the one that matters for loudness headroom.
Higher samplerates also use more cpu, because there is more information to process.

1ghz machine shouldn't go above 44.1khz no matter what brand.
2ghz can get away with 96khz at 6ms delay if soundcard's good.
3ghz of any brand that is younger than 5 years will handle 192khz no issues 2-4ms if you need that.I dont see you needing 192khz that only makes a difference in those live performances the orchestra fellas so it'd make their instruments stand out more.

that was a lot of info but well :p
 
Beyond Don said:
I don't think the human ears can hear much of a difference. However anything higher then 44.1khz is used for the head room. So better recording, better/easier mix/ better song. No it wasn't necessary but it will help.
Click to expand...

wow, sorry, but you are wrong:

sample rate has no impact on headroom/dynamic range whatsoever. Sample rate is exactly the same as frame rate in video: the more frames you capture per second the greater the definition of the actual resulting image on playback; this is because you have capture more in between moments (like those high definition/high speed camera shots you see in Mythbusters).

For audio, this means that we are capturing more of the individual points in time, which means our slices are smaller. .However, as noted in the video in this thread it really makes no difference in the end

As for anything bigger than 44.1kHz being able to be heard as a difference I doubt it unless you are a freak of nature with an expanded frequency range. teh nominal frequency range of a 44.1kHz sampling rate is 22.05kHz more or less, allowing for fold-over at the halfway point, this might be reduced to a working range of 20kHz, which is what all the textbooks say is that of human hearing.

the reasons for higher sampling rates are down to synchronisation requirements for different parts of the media industry
- broadcast tv requires 48kHz:
-- it synchs with 29.97, 24, 25, 30 fps video easily - most pro audio gear uses this or multiples
- 88.2kHz
-- double cd audio sampling rate
-- some pro audio gear where the destination is 44.1kHz CD audio
- 96kHz double broadcast audio rate
-- found in DVD audio, BD-Rom audio, HDDVD Audio
- 192kHz 4 times broadcast audio rate
-- found in DVD audio, BD-Rom audio, HDDVD Audio
 
Last edited:
I think theres the beginning of a trend toward higher sample rates, with the vast majority of sales being digital download so you're starting to see people releasing their music at 24/96, because why not? Hardrive space is less and less of an issue and it sounds better. Hopefully there'll start to be a swing this way after people get sick a low bitrate mp3s! hahaaaa
 
grongk said:
I think theres the beginning of a trend toward higher sample rates, with the vast majority of sales being digital download so you're starting to see people releasing their music at 24/96, because why not? Hardrive space is less and less of an issue and it sounds better. Hopefully there'll start to be a swing this way after people get sick a low bitrate mp3s! hahaaaa
Click to expand...
Again, it doesn't really matter.

I can't even go to 96 without my computer giving up.

I'll tell you that Mike Strange, mixing engineer of Eminem, works in 44.1khz for two reasons.

1) It makes no difference.
2) He's often working on records with a hundred or more tracks. Things can get very slow. Adding on top of that, Luis Resto does overdubs after they mix the thing (Ik, dey be crazy) so that's even more stuff piled on after.

I think I actually used to work in 96 on my old set up because my audio interface decided what sample rate to use and I didn't know how to change that. My Pro Tools instructor at MMI said "Excuse me? You are using 96khz? Why are you doing that? That's a waste of time."
 
As BC said the samplerate do change the quality of the sound.
Lower settings can't represent the higher frequencies as good, therefore the waveform "distort" (not sonically speaking) and aliase, introducing harmonics that can reduce the definition and clarity of the highend.
Think of a sine-waveform at 10 kHz, you have around 4 samplepoints to represent it. It won't be a perfect sine. Sure interpolation will do its job and try recreating it, but I don't think it will be the perfect sineshape.
I mean if I've understood it right...

Though in the end the track will be at 44.1 kHz most of the times anyway.
So I guess it's more about your choice of workflow.
If you want to produce in high definition, that's great. Personally I prefer to work at 44.1 kHz so I know I actually hear the finished quality of the sound (except bitdepth and bitrate of course, since I work in 32-bit and with no dataloss that an MP3 would have).
But I really look forward to the day when a higher samplerate becomes the new standard.
 
upper limit of human hearing is 20 000 Hz,
upper limit of human pitch perception is about 6 000 Hz
upper limit of the piano keyboard is 4 186.01 Hz;
- it's harmonics are therefore 8 372.02, 12 558.03, 16 744.04, 20 930.05, 25 116.06, 29 302.07, 33 488.08
- we will not hear anything above 16 744.04 and would be lucky if we captured 20930.05 using our current technology

the sample rate itself will not affect the clarity of your upper freqs: anything above about 6kHz is simply not heard as a direct frequency but as colour on other sounds

higher sample rates are chosen for their ability to synch to other sources

if we were talking about doubling bit depth it would be an entirely different ball game
 
I agree with bandcoach he knows what he's talking about. I've read up on the nyquist theory about the human hearing and all that. Now although I can't physically hear the extra freq. in my 96khz sessions I do hear a lot of extra clarity especially when I'm recording my guitars. Using samples on the other hand I can see how it wouldn't matter too much because most of them have already been professionally recorded and processed.
 
If you really want to get into this, there are probably about a thousand long-winded discussions on this on Gearslutz. None exactly conclusive.
 
Band coach and everyone else already gave you the scientific answer. The simple answer is, what's your end product gonna be? Is it's "music" use 44.1khz, if it's film/tv/video use 48khz for sync reasons.
 
Eventually 96 is gonna be standard but right now the only thing you need 96 for is video.

Don't use 96 for music (yet)
 
No problem for 1xFS for diffusion. Both 44.1KHz and 48 KHz are OK. But when producing, the audio is processed and 2xFS or higher sampling rates reduce most digital artefacts such as aliasing.
 
Last edited:
Lots of really great info here. Impressive. That being said, I noticed a lot of recommendations for 96khz and 48khz but only one mention of 88.2. Here in Chicago, most of the "big-time" studios are using 88.2, because it is exactly two times what has become the standard rate for digital audio playback, 44.1. If you use 88.2, the final product sounds much better when down sampled to 44.1 because it's not sample conversion; it's just down sampling. The intricacies of this are far too technical for me to understand (let alone explain!), but from what my ears tell me, 88.2/24 -> 44.1/16 sounds MUCH better than 96/24 -> 44.1/16 or 48/24 -. 44.1/16, and just about every high-level engineer I've worked with here in Chicago agrees.

Most of them say that the absolute best method would be to track at the highest sample rate possible, at 24-bit depth, and then mix down to a high-quality analog source. Then, run the analog master through some high-quality AD/DAs at 44.1/16. Unfortunately, most of us don't have access to a 1/2" stereo analog deck (or the tape required to use it), so digital bouncing is the best option.
 
Hahahah while I do enjoy my share of Scotch, this is absolutely true. Google it. You'll find articles from Sound on Sound and other reputable sources that say the exact same thing: there's a HUGE difference between down-sampling and sample conversion. :)
 
down sampling and sample rate conversion are the same thing; the difference lies in the following:

sample rate conversion means create a new set of samples form the existing sample rate at a new sample rate.

down sampling from 88.2kHz to 44.1kHz means drop every other sample - it is still a conversion from one sample rate to another, just a simpler conversion

so to my mind, your ears are lying to you: you have engaged feel good factor based on a magic number not on true evidence
 
You must log in or register to reply here.
Back
Top