Moving up to 96K/24bit
- Thread starter ido1957
- Start date
Farview
Well-known member
Not double the resolution, just double the highest frequency that you can record. Bit depth is what would be resolution.
If it sounds better to you, go for it. I've found that different interfaces will.sound better or worse at different sample rates. This is more a function of the interface desig than the sample rate. All things being equal, the sample rate should not change the sound of the recording at all.
If it sounds better to you, go for it. I've found that different interfaces will.sound better or worse at different sample rates. This is more a function of the interface desig than the sample rate. All things being equal, the sample rate should not change the sound of the recording at all.
+ 1. Doubling the sample rate is somewhat analogous to running tape at 30ips. Since any decent tape machine would "3dB" at about 30kHz at 15ips you should get up to 60kHz. But that's not why they did (do) it! The fast boys eschew noise reduction and 30ips gives a bit less noise.
But just as super 44.1/48 sample rates can cause issues with certain AIs, they were optimized for standard rates, so too does 30ips cause hardware issues since heads are optimized for 15ips and the LF response is compromised. Of course the higher speed uses a dietyawful lot of tape! In the same way your hard drives will stuff much quicker but, unlike good tape, them's cheap!
Very few people have repro systems that are at the same time quiet enough and powerful enough to do 16bits justice (tho' 24 bits makes abundant sense for tracking|) IMHO the case for higher than 44.1/48kHz has not yet been made. (unless you want to entertain small mammals).
Dave.
R
RAMI
Guest
It's really not. Any difference you think you hear is most probably psycho-acoustics.
Chili
Site Moderator
I started out at 96/24 and moved down to 44.1/24 aafter a year or so. The workload on plugs and cpu wasn't worth any discernible increase in sound quality. Honestly, I couldn't tell a difference. But I realized a huge gain in computer and dsp performance and the number of plug instances I could run when running at 44.1k sample rate.
miroslav
Cosmic Cowboy
Yeah....I kinda did that same thing.
When I was doing all my FX/processing OTB (even though I was playing back out from the DAW)...I was using 88.2/24
It was never a problem.
However, as I started doing more VSTi stuff and using more ITB FX/processing (even though I still OTB playing back from the DAW)...I found that going 48/24 took a load of the CPU.
I was never concerned about file sizes, as HD space is pretty cheap these days.
That said...I've been getting ready to move to a new DAW system, and I dunno....I may at least try 88.2 again just to see if the new DAW can handle it along with the increase in plug use.
Is there any difference....?
Well...AFA comparing just the pure recording/playback of a 41/48 VS 88.2/96 or higher...yeah, there is a slightly better clarity/openness, but it is very slight. The other thing reason a lot of people DO use higher rates is actually because of the plugs, since certain types of FX/processing sounds better at higher rates...but again, depends on the music, the plugs, the overall production etc...so I think you would need to be at the highest quality levels from the first track up to the final mix to make higher rates worthwhile.
IOW...t
Hmm? I have a 160 quid AI that returns a rec/play noise floor of -102dBFS (at 44.1kHz 24 bits). Since anything you connected to said AI would be worse than that I suggest such an advantage, if it exists is purely academic?
Dave.
Farview
Well-known member
From the book -"Moving to high sample rates automatically provides a signal-to-noise advantage, so 16 bits at 96kHz is 3.4 dB quieter than at 44.1, sonically equivalent to about 16 1/2 bits. Noise-shaping at high sample rates can allow shorter wordlength files with very low psychoacoustic noise floor - the noise can be made extremely low and flat in the audible band and the shaping moved above 20kHz. In fact, 16-bit noise shaped dither at 96kHz can sound as good as 24-bit/44.1, as I discovered one day when I accidentally left 16-bit dither on while working at 96kHz."
As I understand the text above, the perceived lower noise floor / perceived increase in dynamic range is only attributed to the fact that at high sample rates in a dithered system, significant portions of the required dithering noise can reside in the higher, inaudible frequency range. In other words the dithering noise level within the audible band drops because it is moved up the spectrum out of our hearing range..
Note that the effect only takes place in a dithered system where there is no quantization "noise" (quantization distortion).
Also note that he was talking about the perceived noise floor of a 16 bit dithered system. The noise floor of a 24 bit recording is -144db. How much quieter does it need to be. None of this silliness does anything about the noise floor of what you are actually recording, which in the best of circumstances can only realistically be -110 or so.
Of course you can always argue that lots of people covet the sound of tape, which had a snr of 60-70db and if it's played back on vinyl the snr goes down significantly.
Chasing signal to noise down below -100dbfs is pretty pointless. Especially when most things will get limited to the point where they have a crest factor of less than 10db.
Bobbsy
Boring Old Git
Just a thought...
Unless you're doing audio for video where 48kHz is the standard, I'd probably use 44.1. You won't hear the difference but it'll involve a sample rate conversion when you decide to burn to CD.
ido1957
9K Gold Member
Yes - I should have said 44.1....
In the studio the 96K sounds better on some of the plugins optimized at that rate, like Nebula. It's crazy talk but it improves the quality of the silence. The problem is the final mix file leaves the studio in some form and it all kind of goes back to listening standard and a lot of the perceived improvements are lost. We went back to 44 or 48 for video.
TKeefe | Free Listening on SoundCloud
TKeefe | Free Listening on SoundCloud
Pinky
and The Brain...
48khz/24bit is really the standard. When you investigate the numbers/science and accept what out ears can actually perceive, even this is more than enough. As long as the final product sounds good at 16bit/44.1khz everything that gets you there should be done with least amount of overhead necessary.
Pinky
and The Brain...
In all this I find it interesting to consider FM (and stereo) sound.
For decades this has had an upper limit of 15kHz which implies a sampling rate well under 44.1kHz and indeed stereo is sampled at 38kHz. And yet throughout those decades I do not recall any criticism of the system, even from recording engineers who had access to tape with a response to well over 20kHz and vinyl that reached 18kHz, maybe a bit higher.
Then, FM tuners/receivers threw out low level crap, a mx of 19 and 38kHz switching signals. I must have read scores of reviews of this gear over the years and yet I do not recall anyone noticing the artifacts? That they were there was beyond doubt because they would cluck with your tape machine badly unless it had the required filters fitted.
Until the advent of the CD, FM stereo was the highest sound quality readily available in the home. Top receivers had a dynamic range approaching 70dB and distortion better than 1% at peak deviation. Live broadcasts from BBC 3 would be at this sort of quality level. Both the DR and distortion are way better than tape or vinyl.
Dave.
For decades this has had an upper limit of 15kHz which implies a sampling rate well under 44.1kHz and indeed stereo is sampled at 38kHz. And yet throughout those decades I do not recall any criticism of the system, even from recording engineers who had access to tape with a response to well over 20kHz and vinyl that reached 18kHz, maybe a bit higher.
Then, FM tuners/receivers threw out low level crap, a mx of 19 and 38kHz switching signals. I must have read scores of reviews of this gear over the years and yet I do not recall anyone noticing the artifacts? That they were there was beyond doubt because they would cluck with your tape machine badly unless it had the required filters fitted.
Until the advent of the CD, FM stereo was the highest sound quality readily available in the home. Top receivers had a dynamic range approaching 70dB and distortion better than 1% at peak deviation. Live broadcasts from BBC 3 would be at this sort of quality level. Both the DR and distortion are way better than tape or vinyl.
Dave.
Bobbsy
Boring Old Git
Well, the two compelling reasons I can think of to use 48kHz are:
-if you're working with audio for video (where 48k is the standard) so you can avoid a sample rate conversion,
or,
-if you use a rubbish sound card (like a Soundblaster) which works natively only at 48kHz and does an on-the-fly conversion--badly--for other sample rates. But if you use a sound card that bad, you probably don't care anyway.