Lowpass filters cutting off 20khz-48khz?

Practical tips for recording, editing, and mastering.

Moderators: MattKingUSA, khz

thebutant
Established Member
Posts: 169
Joined: Sun Nov 18, 2012 12:59 pm
Has thanked: 50 times
Been thanked: 8 times

Lowpass filters cutting off 20khz-48khz?

Post by thebutant »

I record and edit in 48khz.
But every lowpass filter I have (let's take Calf as an example) have a max frequency of 20khz.
So when a lowpass filter is all open, when all sound is supposed to run through, when it's practically bypassed - does it nevertheless cut all frequencies from 20000 hz to 48000 hz?
jonetsu
Established Member
Posts: 2036
Joined: Sat Jun 11, 2016 12:05 am
Has thanked: 10 times
Been thanked: 22 times

Re: Lowpass filters cutting off 20khz-48khz?

Post by jonetsu »

They're not about the same thing. One is the sample rate, eg. how many samples per second, 44K1kHz, 48kHz, 96kHz, etc, of the audio material is sampled digitally. The other is about the audio frequency range contained in the audio material itself. An audio filter will filter the audio frequencies, not the sampling rate. As far as I know there's no such thing as a sampling rate filter. It is set and stays the same for a project or all projects.
zoco
Established Member
Posts: 177
Joined: Fri Jan 25, 2019 4:04 pm

Re: Lowpass filters cutting off 20khz-48khz?

Post by zoco »

thebutant wrote:But every lowpass filter I have (let's take Calf as an example) have a max frequency of 20khz.
So when a lowpass filter is all open, when all sound is supposed to run through, when it's practically bypassed - does it nevertheless cut all frequencies from 20000 hz to 48000 hz?
I hope so for you. Otherwise the dogs, cats, wolves and foxes in your surroundings would go crazy. :D
https://en.wikipedia.org/wiki/Hearing_range

Jonetsu explained it already for real.
merlyn
Established Member
Posts: 1392
Joined: Thu Oct 11, 2018 4:13 pm
Has thanked: 168 times
Been thanked: 247 times

Re: Lowpass filters cutting off 20khz-48khz?

Post by merlyn »

The minimum number of samples per cycle required to reconstruct a sine wave is two.
NyquistLimit.png
NyquistLimit.png (1.19 KiB) Viewed 7633 times
So the maximum frequency that can be reproduced is half the sampling rate.
zoco
Established Member
Posts: 177
Joined: Fri Jan 25, 2019 4:04 pm

Re: Lowpass filters cutting off 20khz-48khz?

Post by zoco »

merlyn wrote:So the maximum frequency that can be reproduced is half the sampling rate.
I know about that but it has always wondered me.
Why then is it that hardware, software and electronics designers are stepping into 48kHz nowadays for electronics and other gear from the given reason it gives better result then 44.1kHz? Were 44.1kHz within this statement, reproducing half the samplerate which is a 22.05kHz frequency, which already covers more then the hearing range?
Wouldn't this be totally useless within this nyquist theory? Something i have never read a decent and reasonable explanation for in the many discussions everywere going on for many many years now or one of the many articles about this subject, and from that never understood it.
merlyn
Established Member
Posts: 1392
Joined: Thu Oct 11, 2018 4:13 pm
Has thanked: 168 times
Been thanked: 247 times

Re: Lowpass filters cutting off 20khz-48khz?

Post by merlyn »

48 kHz isn't new. DAT worked on 48kHz. 48kHz was the professional standard and 44.1kHz was the consumer standard.

It's more to do with the sampling side, the A to D. If you sample a frequency higher than the Nyquist frequency you get aliasing. So everything above the Nyquist frequency has to be filtered out before the signal is sampled.

Filters don't have a brickwall cutoff, so for the signal to be zero at the Nyquist frequency the filter will have to start filtering before that. 48 kHz gives more room for the slope of the filter.
thebutant
Established Member
Posts: 169
Joined: Sun Nov 18, 2012 12:59 pm
Has thanked: 50 times
Been thanked: 8 times

Re: Lowpass filters cutting off 20khz-48khz?

Post by thebutant »

jonetsu wrote:They're not about the same thing. One is the sample rate, eg. how many samples per second, 44K1kHz, 48kHz, 96kHz, etc, of the audio material is sampled digitally. The other is about the audio frequency range contained in the audio material itself. An audio filter will filter the audio frequencies, not the sampling rate. As far as I know there's no such thing as a sampling rate filter. It is set and stays the same for a project or all projects.
Oh my god, this is embarassing.
Well, I know these things.
It's one of those times when I get stuck to an idea, and don't think straight at all.
Thanks for your polite answers! :)

And no, I don't have any lowpass filters that cut off samplerate.
I'm also perfectly happy not to. :wink:
jonetsu
Established Member
Posts: 2036
Joined: Sat Jun 11, 2016 12:05 am
Has thanked: 10 times
Been thanked: 22 times

Re: Lowpass filters cutting off 20khz-48khz?

Post by jonetsu »

thebutant wrote:It's one of those times when I get stuck to an idea, and don't think straight at all.
Also happens to me, way too often ! :wink:
zoco
Established Member
Posts: 177
Joined: Fri Jan 25, 2019 4:04 pm

Re: Lowpass filters cutting off 20khz-48khz?

Post by zoco »

merlyn wrote:48 kHz isn't new. DAT worked on 48kHz. 48kHz was the professional standard and 44.1kHz was the consumer standard.

It's more to do with the sampling side, the A to D. If you sample a frequency higher than the Nyquist frequency you get aliasing. So everything above the Nyquist frequency has to be filtered out before the signal is sampled.

Filters don't have a brickwall cutoff, so for the signal to be zero at the Nyquist frequency the filter will have to start filtering before that. 48 kHz gives more room for the slope of the filter.
I know about that. This theory about more room even goes beyond 48kHz for some people.
But that is not the general and irrefutable statement. Especially not among the 44.1 believers. Mostly it is discusses as not true by at least a third off the people discussing it who many times also refer to the Nyquist theory. Some examples of such discussions, http://alturl.com/dn7rc , http://alturl.com/ehz8i , http://alturl.com/dyh7r
Eventually to me it seems to have some contradiction in it at both sides which keeps it unclear to me.
merlyn
Established Member
Posts: 1392
Joined: Thu Oct 11, 2018 4:13 pm
Has thanked: 168 times
Been thanked: 247 times

Re: Lowpass filters cutting off 20khz-48khz?

Post by merlyn »

zoco wrote:This theory about more room even goes beyond 48kHz for some people.
It's not a theory. It's a fact of audio engineering.
zoco wrote:Eventually to me it seems to have some contradiction in it at both sides which keeps it unclear to me.
That's a shame. Study harder. :)
User avatar
Michael Willis
Established Member
Posts: 1431
Joined: Mon Oct 03, 2016 3:27 pm
Location: Rocky Mountains, North America
Has thanked: 67 times
Been thanked: 155 times
Contact:

Re: Lowpass filters cutting off 20khz-48khz?

Post by Michael Willis »

I use a session sample rate of 48kHz instead of 44.1kHz simply because it may help my interface to perform better.
zoco
Established Member
Posts: 177
Joined: Fri Jan 25, 2019 4:04 pm

Re: Lowpass filters cutting off 20khz-48khz?

Post by zoco »

merlyn wrote:
zoco wrote:Eventually to me it seems to have some contradiction in it at both sides which keeps it unclear to me.
That's a shame. Study harder. :)
It is not that much about studying. It is more about the contradiction in those discussions which make it unclear. To me it is strange that the Nyquist theory is used both for and against. Even on this forum the statements in contradiction can be found i just found out searching for it.
merlyn wrote:
zoco wrote:This theory about more room even goes beyond 48kHz for some people.
It's not a theory. It's a fact of audio engineering.
Perhaps it is a better idea to clear this up for those many other people discussing otherwise in the years going on discussions without any end.
For me it is clear. I know myself what i think about this subject and what i do with it. You do not have to ask as i will not answer.
Death
Established Member
Posts: 372
Joined: Sun Oct 11, 2015 1:43 pm
Been thanked: 32 times

Re: Lowpass filters cutting off 20khz-48khz?

Post by Death »

Sorry, I haven't read all the other replies here so don't know what's been said. But here's a tip on this..

Humans can generally only hear from about 20hz-20khz (And that's in a very good scenario. Most of us can't even hear that whole range. The top end starts to disappear more and more as we leave our teenage years and onwards).

When capturing sound, you need the sample rate to be approximately twice the frequency range you wish to accurately capture. So, to capture up to 20khz, 44.1khz is basically enough (although there's a little more to it and you can go down the rabbit hole with this stuff if you wish..). 48khz will capture a slightly larger range, but you really don't need to be hearing anything above 20khz, if your ears are even capable in the first place.. It's standard practice to cut out anything outside of the 20hz-20khz range. Some of us even cut out more than that at times ;) Anyway, that's why your filters & EQ's work that way.

Hope that helps!
User avatar
sysrqer
Established Member
Posts: 2516
Joined: Thu Nov 14, 2013 11:47 pm
Has thanked: 319 times
Been thanked: 147 times
Contact:

Re: Lowpass filters cutting off 20khz-48khz?

Post by sysrqer »

zoco wrote: Perhaps it is a better idea to clear this up for those many other people discussing otherwise in the years going on discussions without any end.
For me it is clear. I know myself what i think about this subject and what i do with it. You do not have to ask as i will not answer.
Doubting the validity of scientific and mathematical data is different from choosing what you will do with that data. There really isn't much doubt on the issue, just subjectivity about whether you find benefit from it.
CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Re: Lowpass filters cutting off 20khz-48khz?

Post by CrocoDuck »

Death wrote:Humans can generally only hear from about 20hz-20khz (And that's in a very good scenario. Most of us can't even hear that whole range. The top end starts to disappear more and more as we leave our teenage years and onwards).
I am a good example of guy deaf to anything above 16 kHz.
Death wrote:but you really don't need to be hearing anything above 20khz, if your ears are even capable in the first place..
Plus: if your equipment was able to record it, and your speaker/headphone are able to reproduce it. Which might very well be, as it might very well not, or maybe not without artefacts.
Death wrote:(although there's a little more to it and you can go down the rabbit hole with this stuff if you wish..)
That's some good rabbit hole. Few months ago there has been a seminar at Salford university about that. I will leave the description here. Unfortunately I could not attend.
*SEMINAR* tomorrow is definitely one to attend if you're interested in/ studying digital audio. Get to grips with sampling and find out whether 48kHz is enough!

'Nyquist doesn't work! How can we do better?'
Professor Jamie Angus-Whiteoak, University of Salford.

Sampling is a critical process in digital audio. Sampling is the process of converting a continuous time signal that exists for all time values, into one that exists only at discrete time values. Continuous time signals must be converted to discrete time ones in order to provide a list or sequence of signal values that can be processed or stored.

In audio this is usually achieved by taking snapshots, or samples, of the signal at, regular time intervals Ts and therefore at a constant sampling frequency Fs. The most obvious application of sampling is at the analogue to digital interface, where the analogue signal must be sampled, so that it can be quantized into a digital word. Likewise at the digital to analogue output the sampled signal must be converted back, or reconstructed, into a continuous time one. However, sampling, or re-sampling, also occurs implicitly in any process that changes the sample rate of the incoming digital signal. In these sample rate change processes the sampled signal must be converted to a continuous, or very high sample rate, signal and then re-sampled at the new output sample rate.

Thus sampling processes may occur both within an audio processing system, as well as at its analogue inputs and outputs. Sampling can be, in principle, “lossless” in that, under the right conditions, a finite bandwidth continuous signal can be sampled and reconstructed without error, as shown by Shannon, Nyquist, Whitaker, and Kotel’nikov.

This Seminar concerns itself with the sampling process. The reason for this is that modern sampling approaches are being used in some modern high-resolution audio formats such as MQA. There has also been research that seems to show that wider bandwidth, and thus higher sampling rate, material may be preferred over 44.1kHz or 48kHz sampled material. High-resolution audio is concerned with obtaining the highest possible audio quality and modern sampling approaches allow better reconstruction of the audio waveform with finite length filters compared to the traditional approach.

This seminar will look at modern theories of sampling, and explain them, in as much of a non-mathematical way as is possible.

We will first examine the traditional sampling process and the classical theory behind it. Then the modern techniques will be discussed. It will show that sampled audio using modern techniques, when properly reconstructed, preserves most of the original signal.

BIOGRAPHY: Professor Jamie Andrea Shyla Angus-Whiteoak

Jamie Angus-Whiteoak Is Professor of Audio Technology at Salford University. She was one of the progenitors of the UK’s first music technology course at York in 1986. Her interest in audio started with a visit to WOR radio and television in NYC when she was 11. After this she was hooked, and spent much of her free time studying audio, radio, synthesizers, and loudspeakers, and even managed to build some! She studied electronics at Kent (UK) doing her BSc and PhD there from 1974 to 1980. During her PhD study she became interested in A/D conversion and worked on a sigma-delta approach, but had to give it up to concentrate on her Thesis topic of designing a DSP processor. After her PhD she joined Standard Telecommunications Laboratories, who invented optical fibres and PCM. There she worked on integrated optics, speech coding, speech synthesis, and recognition in the early 80s and has been active in audio and acoustic research from then. She joined the University of York and helped develop the first integrated masters (Meng) in electronic engineering, and create the UK’s first Music Technology course. She is the inventor of; modulated, wideband, and absorbing diffusers, direct processing of Super Audio CD signals, and one of the first 4-channel digital tape recorders. She has done work on signal processing, analogue circuits, and numerous other audio technology topics.

She teaches audio and video signal processing, Psychoacoustics, Sound reproduction and audio and video coding. She has co-written two textbooks and has authored, or co-authored over 200 journal and conference papers and 4 patents. She is currently investigating environmentally friendly audio technology.

She was awarded an AES fellowship and the IOA Peter Barnett Memorial prize for her contributions to audio and acoustics.

For relaxation she likes playing drums and dancing, but not at the same time.'

Wednesday 13th February, 2pm - 3pm. Newton 238/9
Post Reply