48 kHZ or 44.1 kHZ ?

How to do this, that and the other. Share, learn, teach. How did X do that? How can I sound like Y?
Post Reply New Topic
RELATED
PRODUCTS

Post

VELLTONE MUSIC wrote: Wed Mar 20, 2019 8:45 am Yep,If you are deaf...
no, you don't understand what i'm saying, it's got nothing to do with how it sounds, claiming 48 sounds better at 8bit but being the same and not so audible at 24bits makes no technical sense because that's not how digital audio works. If there's a bigger difference at 8bit than at 24bit it means something is being done wrong.

I'm not trying to argue for or against 48k, it might be great, but all that benefit is potentially wasted on poor SRC or additional filtering when resampling to 44.1k.

The problem is, there's no real target medium anymore. YT is 44.1, streaming can be 48, so you shoot yourself in the knee no matter what you pick.
Image

Post

Relax i am kidding :):):)no offense,i'm not deep enough in theory to argue,just saw it and i'm curious why is less noisy at 48 than 41 khz - ok,maybe suggestion about 24bit is wrong,it's laptop default speaker not soundcard,but is audible and i'm interested why is noise reduced with higher freqs?

Post

VELLTONE MUSIC wrote: Wed Mar 20, 2019 1:12 pm Relax i am kidding :):):)no offense,i'm not deep enough in theory to argue,just saw it and i'm curious why is less noisy at 48 than 41 khz - ok,maybe suggestion about 24bit is wrong,it's laptop default speaker not soundcard,but is audible and i'm interested why is noise reduced with higher freqs?
it's not reduced :) it's measurably identical (you can check with iZotope RX)

in fact, you cannot hear the transition from 44.1k to 48k on youtube at all, because youtube by default encodes at 44.1k so the file you're listening to is actually at 44100, so the only thing you can hear at 48k is aliasing from less than stellar SRC. :D
Image

Post

This topic is very interesting and educational,seems abandoned,still have a questions remain -
How bitcrusher is musical fx,if just reduce digital data and produce digital noise?
Why Lofi fx use bit and rate modulation,if there is no difference at different rates?
Cheerz :)

Post

VELLTONE MUSIC wrote: Fri Apr 05, 2019 11:22 am This topic is very interesting and educational,seems abandoned,still have a questions remain -
How bitcrusher is musical fx,if just reduce digital data and produce digital noise?
Why Lofi fx use bit and rate modulation,if there is no difference at different rates?
Cheerz :)
Well firstly, reducing bit depth has no correlation with 'no difference at different rates' Sample rate and bit depth are completely independent things.

Secondly, its not that there's no difference between 48K and 44.1K, its that in both cases, the difference should be undetectable since both have a Nyquist frequency above human hearing capabilities.

Thirdly, both rates are well outside the rate reduction bitcrushers go down to, which are typically down towards the lower end of human perception not the higher end. Just because 44.1K and 48K are indetectable doesnt mean 30K or 20K or 10K or 5K is.
my other modular synth is a bugbrand

Post

VELLTONE MUSIC wrote: Fri Apr 05, 2019 11:22 am This topic is very interesting and educational,seems abandoned
Well, the original question was answered, discussed and beaten to death. Then it drifted off..
VELLTONE MUSIC wrote: Fri Apr 05, 2019 11:22 am still have a questions remain -
How bitcrusher is musical fx,if just reduce digital data and produce digital noise?
Sorry, what do you mean with "digital noise"? I think you mean quantization noise.
What happens in a bit crusher is it introduces truncation noise which is a form of distortion. Truncation or quantization is the rounding of numbers. This adds high harmonics which were not there originally.

Image

Anything that alters a signal can be used creatively as a musical effect. Or maybe I don't understand what the question is....
VELLTONE MUSIC wrote: Fri Apr 05, 2019 11:22 am Why Lofi fx use bit and rate modulation,if there is no difference at different rates?
Who says there's no difference? In what scenarios?

What you can hear is frequencies between 20Hz and 20.000Hz. Any sampling rate higher than 40kHz is sufficient to reproduce it without error. But go lower and you will start to hear changes. Resample to 8kHz for instance (throwing away everything above 4kHz) and you have a usable musical effect. It sounds like you recorded through a telephone. Do it badly (by not counteracting aliasing for example) and it gets more interesting.

What you can hear is sound from 0dB (silence) up to 120dB (pain threshold) but be safe with an average of 83dB for a comportable listening level. 16 bits will give you a dynamical range of 96dB, since each bit represents 6 dB. So you can have peaks 13dB above the average of 83dB and still produce signals barely audible.
There is no benefit (for audio playback) in using more than 16 bits. But if you go lower, you will hear quantization noise. Yes there is quantization noise in 16 bits, but it's just not audible.
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post


Post

never spend much time on pure theory and understanding mathematics behind sound in digital devices(math was weak class,that explain dummy questions hahahah :):):)rely to 'hear' and 'apply' practically something,but the galaxy of digits and specs go above practical 'hearing' of the sound.
Satisfying explanations.Thanks.

Post

Someone may (probably) have already mentioned this, but the standard for video work is 48k. I use it for everything now, but started mainly when I started scoring media. I do find the difference noticeable, and there are many things which either don’t need oversampling, or need much less,when you are working at this slightly higher sample rate. Furthermore, the cpu consumption is almost non existent on my system ( this may be an anomaly as I sometimes will have lower CPU use at even higher sample rates too, and cubase has a notoriously inaccurate “overload meter) so when I need to send out 48k files anyway, it is a no brainer.

Beyond that, I will often run my DAW at the maximum sample rate when doing sound design, as it will make for very pristine sounds that will have very few unwanted artifacts.

The simple version of all this is that, the extra little bit of frequency headroom when using 48k over 44.1k is easily noticeable to my ears (especially with things like hard limiting/ clipping, extreme EQ, and extreme compression), and beneficial enough. It also means that I won’t have an extra (possibly destructive) conversion when I send stuff off for work.
Don't F**K with Mr. Zero.

Post Reply

Return to “Production Techniques”