The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
High Samplerates aren't about frequency capture - it's about data per second capture
Old 3 weeks ago
  #31
Quote:
Originally Posted by BrainCheck123 View Post
but if the songs great, why not have it recorded as amazingly as possible?
Definitely! But if you want to make amazing recordings it can be helpful to have your facts straight. In this case:
It’s not the number of steps but the electronics needed to convert between these steps and the audio.
Old 3 weeks ago
  #32
Lives for gear
I usually don't pitch in on topics like this but just my 2c.
To capture a waveform and its phase and to reproduce it perfectly at least two samples are needed.
The place in time where these samples are taken does not matter, its the value of two samples that determines the phase.*
Just because the digital hardware in an interface is capable of 96 kS/s or higher, doesn’t mean the analogue stages will record or play that signal cleanly.

Just think about it, most analog equipement is trying it's best to get rid of these frequencies.
It’s quite common for ultrasonic content to cause intermodulation distortion into the audible range.
Perhaps some just like the sound of it, as some like to degrade the signal like old analog equipment (and tape) does.
Whatever works for you.


* If I generate an sinewave of 2.7kHz in 44.1kS/s and then shift one side one sample (0.023ms out of phase), re-sample to 6kS/s (distance between samples is 0.17ms) it will capture the phase diference with no problem.
1
Share
Old 3 weeks ago
  #33
Gear Addict
 

Quote:
Originally Posted by thedberg View Post
Definitely! But if you want to make amazing recordings it can be helpful to have your facts straight. In this case:
It’s not the number of steps but the electronics needed to convert between these steps and the audio.
I gues what I'm saying is, if you run a single sine wave or pure sound source, any signal shape, which is what all these graphics in here are... well that's adequate. But acoustic sound isn't one single sinewave. they are cascading soundwaves and reflections - millions of them, and their shape irregular and modulating and then you have the sonic complexity of the phasing phenomenon.

ie, if there is a cymbal in a room, it's not outputting its frequency range as one continuous cycle, every single frequency in that cymbals spectrum and the room reflection sound, has pitch tones and harmonic content that is modulating at rates that are individual of each other and thus have different timing.

The more data you can capture of that complexity the better..


there's a difference between a single source sinewave, and a trillion sinwaves with different timing frequency length timing/speed, and then some what of a patterned modulation timing (in the case of a cymbal teeter-tottering) and then attack/decay modulation - but there is an uncountable amount of this happening - cascades of sound.

The higher the sample rate, the deeper into this cascade you can reach which reproduces a more believable realistic image. If you are not working with this type of depth, you won't hear a difference between 44.1 and 88.2
Old 3 weeks ago
  #34
Quote:
Originally Posted by BrainCheck123 View Post
I gues what I'm saying is, if you run a single sine wave or pure sound source, any signal shape, which is what all these graphics in here are... well that's adequate. But acoustic sound isn't one single sinewave. they are cascading soundwaves and reflections - millions of them, and their shape irregular and modulating and then you have the sonic complexity of the phasing phenomenon.

ie, if there is a cymbal in a room, it's not outputting its frequency range as one continuous cycle, every single frequency in that cymbals spectrum and the room reflection sound, has pitch tones and harmonic content that is modulating at rates that are individual of each other and thus have different timing.

The more data you can capture of that complexity the better..


there's a difference between a single source sinewave, and a trillion sinwaves with different timing frequency length timing/speed, and then some what of a patterned modulation timing (in the case of a cymbal teeter-tottering) and then attack/decay modulation - but there is an uncountable amount of this happening - cascades of sound.

The higher the sample rate, the deeper into this cascade you can reach which reproduces a more believable realistic image. If you are not working with this type of depth, you won't hear a difference between 44.1 and 88.2
Well, well. You seem firmly intent on not learning how digital audio works. With that perspective, it's no point in the rest of us trying to explain.

Have a good day
13
Share
Old 3 weeks ago
  #35
Lives for gear
I would like to react (one more time) because it took me some time to grasp this to.

Quote:
Originally Posted by BrainCheck123 View Post
But acoustic sound isn't one single sinewave. they are cascading soundwaves and reflections - millions of them, and their shape irregular and modulating and then you have the sonic complexity of the phasing phenomenon.
Correct, and this can be reproduced using (multiple) sine waves (what digital audio does).

Quote:
Originally Posted by BrainCheck123 View Post
ie, if there is a cymbal in a room, it's not outputting its frequency range as one continuous cycle, every single frequency in that cymbals spectrum and the room reflection sound, has pitch tones and harmonic content that is modulating at rates that are individual of each other and thus have different timing.
Don't forget, temperature, humidity, airflow... (No need to registrate them.)
It all influences how the (complex) sound wave reaches the microphone (couple of mm membrane) which translates it to an electric signal.

Quote:
Originally Posted by BrainCheck123 View Post
The more data you can capture of that complexity the better..
Why? Only data that is needed for a perfect capture matters.

Quote:
Originally Posted by BrainCheck123 View Post
there's a difference between a single source sinewave, and a trillion sinwaves with different timing frequency length timing/speed, and then some what of a patterned modulation timing (in the case of a cymbal teeter-tottering) and then attack/decay modulation - but there is an uncountable amount of this happening - cascades of sound.
Correct, the first is one and the latter are many.

Quote:
Originally Posted by BrainCheck123 View Post
The higher the sample rate, the deeper into this cascade you can reach which reproduces a more believable realistic image.
This is what hi-res audio likes you to believe.
Numbers sell.

Quote:
Originally Posted by BrainCheck123 View Post
If you are not working with this type of depth, you won't hear a difference between 44.1 and 88.2
What are you hearing and are you really only hearing with your ears?
2
Share
Old 3 weeks ago
  #36
Lives for gear
 
ponzi's Avatar
Quote:
Originally Posted by Sizeofanocean View Post
Here is the original paper by the way:

http://lavryengineering.com/pdfs/lav...ing-theory.pdf

I love reading those back from time to time
What I read here is that he is saying 192 is not needed and that 60 khz is enough to minimize fir filter pre-ringing. So, does not look like an unqualified statement that 44.1 is the upper limit of what can help.
Old 3 weeks ago
  #37
Here for the gear
 

Quote:
Originally Posted by ponzi View Post
What I read here is that he is saying 192 is not needed and that 60 khz is enough to minimize fir filter pre-ringing. So, does not look like an unqualified statement that 44.1 is the upper limit of what can help.
...which no one here claimed :-)
1
Share
Old 3 weeks ago
  #38
Gear Head
 

It is staggering how people who don't know what they're talking about ... talk about it anyway.
Brain Check is an apt nym.
4
Share
Old 3 weeks ago
  #39
Gear Guru
 

Quote:
Originally Posted by BrainCheck123 View Post
I gues what I'm saying is, if you run a single sine wave or pure sound source, any signal shape, which is what all these graphics in here are... well that's adequate. But acoustic sound isn't one single sinewave. they are cascading soundwaves and reflections - millions of them, and their shape irregular and modulating and then you have the sonic complexity of the phasing phenomenon.
Any complex signal, which is what you're talking about above, can be viewed as made up of many different sine waves. Doesn't matter if it's from one or a million sources. At the end of the day you have only one membrane (ear drum or mic) that moves according to a sound pressure wave, and that wave is the sum of what's happening at that point in the room acoustically. It isn't a million different movements, you end up with ONE analog signal oscillating between positive and negative.

One.

https://en.wikipedia.org/wiki/Nyquis...m#Introduction

and

https://en.wikipedia.org/wiki/Whitta...lation_formula

But something tells me you're not here to learn.
8
Share
Old 3 weeks ago
  #40
Tui
Gear Guru
 
Tui's Avatar
Quote:
Originally Posted by MacMacMac View Post
It is staggering how people who don't know what they're talking about ... talk about it anyway.
... And if you fail at digital audio, you can always become a politician!

2
Share
Old 3 weeks ago
  #41
Quote:
Originally Posted by BrainCheck123 View Post
I gues what I'm saying is, if you run a single sine wave or pure sound source, any signal shape, which is what all these graphics in here are... well that's adequate. But acoustic sound isn't one single sinewave. they are cascading soundwaves and reflections - millions of them, and their shape irregular and modulating and then you have the sonic complexity of the phasing phenomenon.

ie, if there is a cymbal in a room, it's not outputting its frequency range as one continuous cycle, every single frequency in that cymbals spectrum and the room reflection sound, has pitch tones and harmonic content that is modulating at rates that are individual of each other and thus have different timing.

The more data you can capture of that complexity the better..


there's a difference between a single source sinewave, and a trillion sinwaves with different timing frequency length timing/speed, and then some what of a patterned modulation timing (in the case of a cymbal teeter-tottering) and then attack/decay modulation - but there is an uncountable amount of this happening - cascades of sound.
Not this.


Quote:
Originally Posted by mattiasnyc View Post
Any complex signal, which is what you're talking about above, can be viewed as made up of many different sine waves. Doesn't matter if it's from one or a million sources. At the end of the day you have only one membrane (ear drum or mic) that moves according to a sound pressure wave, and that wave is the sum of what's happening at that point in the room acoustically. It isn't a million different movements, you end up with ONE analog signal oscillating between positive and negative.
This!

Humans are only capable of receiving and interpreting at most, two auditory waveforms: One with the left ear and one with the right.
Old 3 weeks ago
  #42
Tui
Gear Guru
 
Tui's Avatar
The cold, nasty sound of digital, IMO, is largely caused by pre- and post-ringing of digital filters. The effect is somewhat similar to a small amount of reverb being slapped on all audio which causes the perceived soundstage to move backward. Listen to the same album - must be recorded analogue, of cause - on vinyl and CD, and you'll know immediately what I mean. The effect isn't so bad with classical music since there tends to be a lot of room acoustics on such recording, however, with popular music the glassy, somewhat distant sound really gets on my nerves.

If I was an audiophile - got no time for that, sadly - I'd throw all digital gear out the window, no kidding.
Old 3 weeks ago
  #43
Gear Addict
 

Quote:
Originally Posted by mattiasnyc View Post
Any complex signal, which is what you're talking about above, can be viewed as made up of many different sine waves. Doesn't matter if it's from one or a million sources. At the end of the day you have only one membrane (ear drum or mic) that moves according to a sound pressure wave, and that wave is the sum of what's happening at that point in the room acoustically. It isn't a million different movements, you end up with ONE analog signal oscillating between positive and negative.

One.

https://en.wikipedia.org/wiki/Nyquis...m#Introduction

and

https://en.wikipedia.org/wiki/Whitta...lation_formula

But something tells me you're not here to learn.
Right


It's the millions of waves are all contained into one, but that doesn't mean that one file doesn't contain the "markings" of all these individual waves. If you take most samples, you will be revealing the "ghost" of everything that was happening between the lower sample rates points of samples.

For instance, with masking, all the frequencies that are masked, are actually still there doing something because every sound wave has periods of quietness that are so quick we cannot perceive them. doesn't mean they don't have an affect though.


You need very quick sample rates to capture this magic. It's why we can hear better with our ears in situation where one sound is masking another because our ears in real life aren't limited by rate bandwidth our ears do a better job in the real environment hearing a quite sound through a loud sound.


Maybe I'm onto something, maybe we should all listen.
Old 3 weeks ago
  #44
Gear Addict
 

you're essentially beginning to capture air data at a higher sample rate.
Old 3 weeks ago
  #45
Gear Addict
 

Quote:
Originally Posted by nspaas View Post
Not this.




This!

Humans are only capable of receiving and interpreting at most, two auditory waveforms: One with the left ear and one with the right.


I'm not sure that's true. We can separate waveforms by frequency and timbre.
Old 3 weeks ago
  #46
Gear Guru
 

Quote:
Originally Posted by BrainCheck123 View Post
Right


It's the millions of waves are all contained into one, but that doesn't mean that one file doesn't contain the "markings" of all these individual waves. If you take most samples, you will be revealing the "ghost" of everything that was happening between the lower sample rates points of samples.
You're just talking about higher frequencies, nothing else. It's amplitude over time of one waveform, complex or not. Once you "contain" "millions of waves" "into one" that's what you have... one waveform. One measurement of amplitude over time.

Increase # of samples = increased range of frequencies we can capture.

You're just wrong conceptually as far as I can tell.
5
Share
Old 3 weeks ago
  #47
Gear Addict
 

Quote:
Originally Posted by mattiasnyc View Post
You're just talking about higher frequencies, nothing else. It's amplitude over time of one waveform, complex or not. Once you "contain" "millions of waves" "into one" that's what you have... one waveform. One measurement of amplitude over time.

Increase # of samples = increased range of frequencies we can capture.

You're just wrong conceptually as far as I can tell.
Actually I don't think that's correct, that's how we see it and analyze it, but if you open up that waveform in a spectrogram, this actually a better visualization of sound. You can begin to start to see the cascades.

The faster the sample rate, the more complex image can be captured, the differences are thus a greater differential.

The waveform we see, is simply a visually useable summary of amplitude and time.
Old 3 weeks ago
  #48
Gear Guru
 

Quote:
Originally Posted by BrainCheck123 View Post
Actually I don't think that's correct, that's how we see it and analyze it, but if you open up that waveform in a spectrogram, this actually a better visualization of sound. You can begin to start to see the cascades.
Quote:
Originally Posted by BrainCheck123 View Post
The waveform we see, is simply a visually useable summary of amplitude and time.
That's contrary to what you said first. But you're right: The waveform is amplitude over time... ONE amplitude over time. Not many. That is what we need to capture. And we do.

Read the theorem.

You're just thinking about this the wrong way.
4
Share
Old 3 weeks ago
  #49
Gear Addict
 

you're 16 channels of AD/DA are clocked, and therefor, there is going to be a predictable pattern of error, no matter how "sufficient" a typical samplerate is.

If you are recording 16 channels of drum in a reflective room, a higher samplerate will essentially be doing a better job of "scanning" the slight differences and variables of overlapping cascading sounds of which are pretty well an infinite and are continuous in the sense that, there is - always - more sound, between the sound and there are layers of this happening offset from each other, and then more layers, and more layers. A slower sample rate will simply not have as much reach into these micros - but these micros make up what we know as real. the closer we get to real, the better.
Old 3 weeks ago
  #50
Aaarrgghh. This thread is driving me crazy!!

Why, oh why, do people keep spreading this hogwash when you have had these things clearly explained to you again and again?! It’s not like we are standing on uncharted territory; sampling was described before any of us where even born (unless we have some 90 year olds here).

Anyway, I’ll try again:
WE KNOW HOW MANY SAMPLES ARE NEEDED RECORD AND PLAY BACK AUDIO.

If we hear differences between sample rates it due to the analog parts in the converters (or other digital processing than pure record and play back).

Lord, give me strength....
7
Share
Old 3 weeks ago
  #51
Gear Addict
 

Quote:
Originally Posted by mattiasnyc View Post
That's contrary to what you said first. But you're right: The waveform is amplitude over time... ONE amplitude over time. Not many. That is what we need to capture. And we do.

Read the theorem.

You're just thinking about this the wrong way.

It's one waveform containing many... that is exactly what you are hearing, but obviously not seeing in a waveform that is a summary of amplitude of a variable of frequencies.
Old 3 weeks ago
  #52
Gear Addict
 

I'm actually theoretically right,

the argument is whether it makes an audible difference. Your guys' point is and has been 'what is sufficient'. I'm not talking about sufficiency.

The difference is heard more in a real room with a real sound source across multiple phasing mics.




lets get Lavry in here.
Old 3 weeks ago
  #53
Gear Guru
 

Quote:
Originally Posted by BrainCheck123 View Post
It's one waveform containing many... that is exactly what you are hearing, but obviously not seeing in a waveform that is a summary of amplitude of a variable of frequencies.
Yes - ONE waveform.

And we need two samples to correctly describe ONE waveform that contains no frequencies at and above half the sample rate. We don't need more.

FFS, read the theorem...
3
Share
Old 3 weeks ago
  #54
Gear Addict
 

Quote:
Originally Posted by mattiasnyc View Post
Yes - ONE waveform.

And we need two samples to correctly describe ONE waveform that contains no frequencies at and above half the sample rate. We don't need more.

FFS, read the theorem...

I think you are confusing wave cycle and waveform?


and again this doesn't account for multiple waveforms, hitting multiple AD/DA channe;s with different micro macro micro timing differences. the sound between the sound between the sound between the sound.
Old 3 weeks ago
  #55
Lives for gear
 
zvukofor's Avatar
And, BTW, and to make it more confused, modern converters works at several MHz range, not 44.1 kHz... those dudes call it “oversampling”...
1
Share
Old 3 weeks ago
  #56
Gear Addict
 

I think this needs to be tested better.
Old 3 weeks ago
  #57
Gear Guru
 

Quote:
Originally Posted by BrainCheck123 View Post
I think you are confusing wave cycle and waveform?
Absolutely not. You might be, I'm not.

Quote:
Originally Posted by BrainCheck123 View Post
and again this doesn't account for multiple waveforms, hitting multiple AD/DA channe;s with different micro macro micro timing differences. the sound between the sound between the sound between the sound.
Did you learn this from Deepak Chopra?
5
Share
Old 3 weeks ago
  #58
Gear Addict
 

what if we do sine sweeps in a reverberant room, but multipe sine sweeps offset in a very small timing difference, like .0001. and then have an additional layer of harmonizing sinesweeps that start a 3rd or 5th up, and are put into the room at a lower volume.

And then momentary pulses of different noise types into the room while these sinesweeps are running. multiple speaker sources and positions, multiple mic positions, multiple AD/DA channel.
Old 3 weeks ago
  #59
Gear Guru
 

Quote:
Originally Posted by BrainCheck123 View Post
what if we do sine sweeps in a reverberant room, but multipe sine sweeps offset in a very small timing difference, like .0001. and then have an additional layer of harmonizing sinesweeps that start a 3rd or 5th up, and are put into the room at a lower volume.

And then momentary pulses of different noise types into the room while these sinesweeps are running.
Then again: All of the above add together in the analog realm.

THEN analog gets converted... AFTER it's all been summed together into ONE analog waveform.

Quote:
Originally Posted by BrainCheck123 View Post
I think this needs to be tested better.
It's 2019. We've had digital audio for decades. You think you just figured this out???

Read the theorem.

Last edited by mattiasnyc; 3 weeks ago at 11:16 PM..
4
Share
Old 3 weeks ago
  #60
Gear Addict
 

Quote:
Originally Posted by mattiasnyc View Post
Then again: All of the above add together in the analog realm.

THEN analog gets converted... AFTER it's all been summed together into ONE analog waveform.



It's 2019. We've had digital audio for decades. You think you just figured this out???

Read the theorem.

right sure but we don't only work with one summary of a waveform, we work with many summaries. and they will sum differently and process differently and have a different phase relationship depending on the provided data and how frequent it was.

it's undeniable. Faster data rate capture captures more accurate detailed data over the variable of time and time we don't even understand fully.


they key to teleportation may be in extremely high frequency sample rates if you think about it.
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump