is it all hype?

[snow lizard]

It's been a few months since this was posted and it appears we need it again:

That video brings up some good points, but there are a few things that are misleading or missing. It's probably a good description for someone wanting to download 24/192 files or something. Not so much for recording and mixing.

Monty's video tries to compare cassette tapes to 9 bit audio or whatever. Doesn't seem relevant or applicable. Analog has no bits.

The Nyquist/Shannon sampling theorem is not wrong, which is a good reason to say that a digital waveform doesn't get reproduced as a stairstep. What it doesn't tell you is that digital signals in fixed point PCM audio are quantized to a discrete "stairstep" value in order to represent the strength or power level of the samples. Quantizing at low resolution such as near the zero crossing of the waveform will result in distortion from data truncation.

At 13 minutes into the video he shows you distortion from quantization error on the spectrum display. Then he demonstrates how to remove the distortion with dither. By 17 minutes he's dismissive about how important it is. "No one ever ruined a great recording by not dithering the final master".

Seems like bad advice. No one ever ruined a great tape recording by not using bias. Bias is necessary for tape to have linear transfer function. Dither is necessary for fixed PCM to have linear transfer function, especially at lower resolution or when processing the signal is involved. It's not a 24/192 download scenario. Listen to your reverb tails.

 

[joey2000]

Reading about people homerecording who can choose between several $1500 mics, plugged in to thousands of dollars of gear. And then they produce there projects in 16bits and 44.1khz, using 16bit/44.1khz im- and exports.
I still don't get it at all. .

I don't get what you don't get, but while I haven't read all the posts in this thread, I get the impression people tried and failed so get through to you, so I won't waste time doing similar.

It's been a few months since this was posted and it appears we need it again:

I'm one of these weird people that prefers reading to videos For anyone else who'd rather, FYI: 24/192 Music Downloads are Very Silly Indeed

the executive summary is "digital recordings do not have stairsteps and, viewed on oscilloscope the output waveform of the digital (after conversion back to analogue) is indistinguishable from the analogue original.

Actually I thought the takeaway (assuming the video and link above are about the same) was when it comes to bit depth and esp sample rate, more does not equal better. Not only is there no discernible diff at some point (easily most points in fact), but ultimately, if anything, going beyond 48 will make a recording worse, not better.

Anyway, back to the mics......

Granted I was just listening on laptop speakers as well, but I saw little diff in general and no correlation - none - to price and quality. Some sounded a little brighter, some less so. That's it. If I listened on top-notch headphones really intently, I suspect I could hear a little bit more diff. I emphasize the word LITTLE. And IMO I have a pretty good ear...but this brings me to my point, which no doubt has been made by others: the diff in mics, as is the case with most things, is often exaggerated, in fact greatly exaggerated, mostly by pretentious fools (who companies that make this stuff LOVE, because they make a mint off of them).

I also pray I never hear even a tiny snippet of those songs again.


PS: that male singer sucks.

 

[Farview]

The video and your link are not speaking on the same subject.

 

[garww]

The subject being shipping boatloads of money to Sweetwater during Microphone Month

 

[42low]

Laughing myself broken.
Can't explain it simpler than this.
No steps? Anyone knows what (sound) bits are then? Anyone knows why analoque doesn't have (sound) bits? You know what difference that brings within the signal curve line? Think, yourself, through all the foggy yelling!!

And if you know/understand that then think about what the influence off less or more bits would be. Confirming my statement. Less bits = more inconsistant space between = less quality and bigger failure risk.

I don't give a f*** if you understand it or not, and what you think about me.
Those who really know how a digital sound is build, and knows what the bits within it are, knows there are steps between those bit points (or vectors if you want to call them that). Negliable, so the digital curve does his job imitating the analogue curve ... but the "steps" are there. (where they lack within the constant analogue curve. No bits, constant = unlimited bits)

Last thing i said about it. I will not react any more. Promised.

 

[Farview]

Monty's video tries to compare cassette tapes to 9 bit audio or whatever. Doesn't seem relevant or applicable. Analog has no bits.

Since bits have to do with dynamic range, you can make the somewhat awkward analogy of the dynamic range of analog being equivalent to the dynamic range of a digital signal with a certain number of bits. In the context of the video it wasn't a very useful point, especially since he didn't explain it.

The idea that since an analog signal doesn't have bits or samples (and therefor a sample rate), it has unlimited bits and samples is erroneous. Since bit depth in digital set the dynamic range, an unlimited number of bits would mean an unlimited dynamic range, which clearly doesn't exist in analog. Since sample rate sets the highest frequency that can be recorded and reproduced, an unlimited sample rate would mean that any frequency can be recorded, which is not true of audio recording equipment. Even lab equipment will have limits on how high a frequency can be recorded.

 

[joey2000]

The video and your link are not speaking on the same subject.

ah my bad then. I'll resist the tempation to edit/delete and go "what are you talking about?"

I will not react any more. Promised.

 

[42low]

Damn. Discussion is going the right direction with smart talk again, triggering me to join it again.

I must offer you my respect Farview. You're one of the rare ones here who knows what he's talking about within the technical side of this subject.

Since bits have to do with dynamic range, you can make the somewhat awkward analogy of the dynamic range of analog being equivalent to the dynamic range of a digital signal with a certain number of bits.

Glad you explain that. That's were i was made a fool of when i said that.
That's what digital sound is develloped for. Imitating analogue as good as possible. A great invention.

The idea that since an analog signal doesn't have bits or samples (and therefor a sample rate), it has unlimited bits and samples is erroneous.

That's correct too. Accept (sorry) for the bit part.
Analogue has no bits. Bits is a 100% digital principle. It's the digital 0 and 1 series. And therefore not continuous.

Everyone who knows how analogue sound works and knows how amp transistors/tubes work know that both the electric stearing as the amplified signal are a constant. And NOT a serie of bits/points/vectors (looking like a line).
Analogue is unlimited in it's signal because it has NO bits but it's 'curve line' is constant, continuous.
No matter how many bits/dots (digital) you have in line, it will never ever equal the quality of a continuous line (analogue).

Since sample rate sets the highest frequency that can be recorded and reproduced, an unlimited sample rate would mean that any frequency can be recorded, which is not true of audio recording equipment.

Sample rates are, digital sound indeed is limited. By it's bits, chosen and set in the settings.
16bit, 24bit, 32bit float.
Analogue isn't limited by bits. 16bit will never equal 24bit quality, let alone unlimited analogue.

Since bit depth in digital set the dynamic range, an unlimited number of bits would mean an unlimited dynamic range, which clearly doesn't exist in analog.

As this "unlimited bits" ain't right it DOES limit the dynamics. It can't go higher/deeper than the settings which is the max. 16bit will never equal 24bit quality, let alone unlimited analogue.

And that's were we're back at what i said about the expensive mics.
As one uses the settings which are the lowest (16bit/44.1khz) for the dynamics as you explain confirming too, and with that limiting those dynamics, what is the purpose then form a wide and deep dynamic (most expensive) microphone?

And what would be the result of sound editting (plugins like reverb, compressor) if the core signal already is limited by those lowest settings.
Not even starting about re-using low dynamics/settings im- and exports over and over again (loss on loss on loss...). Although DAW internally works in 32bit float, a 16 bit import will never become 32bit float. Although DAW internally works in 96khz, a 44.1khz import will never become 96khz.

A $1500 mic recording at lowest settings (16/44.2), using high quality mic on low quality settings not using the (expencive) mics high specs in full would be stupid. It would be money throwing away cause then a cheaper mic with some lesser specs/dynamics would do the same job easily.
In this case it would be no more then an expensive hype.

So my vision and advise to use higher settings (laught away as foolish) isn't that stupid at all.
To bad i will loose anyway. Not on subject, but as "stupid ignorant newbie" against popularity. Thosefewoffwhichidon'tgivenames will hate and bully me either way.

 

[Farview]

You are completely missing that this is an analogy, probably do to language issues. You are also using the word 'unlimited', which means something different than you think it does in the context of what you are saying.

When you say "Analogue is unlimited in it's signal", what it sounds like you are saying is that analog signals have infinite dynamic range and infinite frequency range. That is how that statement reads to native english speakers. That is what started the argument.

After that, there was the argument about the importance of what goes on between samples, which doesn't matter because anything happening between samples would be above nyquist and couldn't exist in the output anyway.

 

[garww]

Many of the DC amps won't go below 0Hz. Shame and Pity.

 

[42low]

You are completely missing that this is an analogy, probably do to language issues. You are also using the word 'unlimited', which means something different than you think it does in the context of what you are saying.

When you say "Analogue is unlimited in it's signal", what it sounds like you are saying is that analog signals have infinite dynamic range and infinite frequency range. That is how that statement reads to native english speakers. That is what started the argument.

After that, there was the argument about the importance of what goes on between samples, which doesn't matter because anything happening between samples would be above nyquist and couldn't exist in the output anyway.

At first, signals outside the hearing range being editted do can effect the signals within the hearing range. It gives the effect that there's 'more', deeper, clearer. So going outside nyquist when still producing does have effect on the end result.

And second, when i use the term 'unlimited' i'm pointing at the 'bit range' of an analoque signal. The line which makes the curves. There digital is limited while it ain't a line but serie off dots. With that i'm not point at 'unlimited outside hearing range' or whatever.
And as the analoque signal is much more detailed than a by bits limited digital signal an analoque signal is deeper, and so the edits (reverb, compressor, etc) are deeper too. And with that deeper you preserve more clearness.

To be clear about that, i'm not trying to say that sounds outside the hearing range should be in the end result. But before that, while still producing, those 'outside' sounds will have effect towards that end result (within hearing range).

Another supplementary argument.
If 16bit/44.1khz would be more than enough and higher is impossible, why do DAW's work within 32bit-float and 96khz inside then? What would be the purpose of that then if it doesn't have any effect? That's not implemented only to burden your system.
And if it does have effect, that effect is gone immediately if setup lower at 16bit/44.1khz. Because they will never become 32/96 as what's not there will never be there again (as we all know). The point i'm trying to make clear.

So i stick with the opinion that one doesn't use the opportunities from both system/DAW as $1500 mic in full if system set at 16bit/44.1khz.
If one wants to use $1500 mic specs in full, he should settup his system/DAW to use full so 32bit-float and 96khz, to eventually use the possibilities of the best (deepest and clearest) dynamics within it's project.
Otherwise it would be the same as driving a ferrari at no more than 40mph in it's first gear, city only. To do that a toyota in it's third gear rides better, and cheaper too. (yes, i know how ferrari drives. And lamborghini too. And porsche. And toyota. Not that i owned all of those ).

Oh man, i'm going to stop again. Indeed i think you (Farview) and i are talking beside each other (still thanks again for your patience and interesting conversation ). And some others already showed they don't know what there talking about at all.

 

[42low]

One short try.

Take a 16bit 44.1khz sound.
Bring that back to 8 bit and 22 khz (both about half).
And listen to the difference. It's worse, isn't it?

Bringing down 32 bit 96khz back to 16 bit and 44.1khz (both about half too, actually even more).
That's as much breaking down the quality as in the first example.

And editting a worse sound effects the result even more then worse.

Please react on this.

 

[Bobbsy]

Laughing myself broken.
Can't explain it simpler than this.
No steps? Anyone knows what (sound) bits are then? Anyone knows why analoque doesn't have (sound) bits? You know what difference that brings within the signal curve line? Think, yourself, through all the foggy yelling!!

And if you know/understand that then think about what the influence off less or more bits would be. Confirming my statement. Less bits = more inconsistant space between = less quality and bigger failure risk.

I don't give a f*** if you understand it or not, and what you think about me.
Those who really know how a digital sound is build, and knows what the bits within it are, knows there are steps between those bit points (or vectors if you want to call them that). Negliable, so the digital curve does his job imitating the analogue curve ... but the "steps" are there. (where they lack within the constant analogue curve. No bits, constant = unlimited bits)

Last thing i said about it. I will not react any more. Promised.

No steps. Just samples at whatever rate you've chosen. When reproducing a digital recording, your equipment doesn't just draw a line between samples. I follows an algorithm which creates the single curve that can recreate the analogue audio wave that was sampled to take the samples. Look at the original analogue and the playback from digital on an oscilloscope and the shapes are identical.

Moving to a sample rate of 96kHz gives more samples but they are still reassembled by an algorithm. All moving to 96 does is let you theoretically record sounds up to 48kHz, useful for dogs I guess. In passing I'll mention some contact I had with people using special mics and extremely high sample rates to record bat "sonar". An interesting exercise but nothing that helped in music recording.

As for bit depth, 16 bits has a dynamic range that lets you go below the noise floor of all but the best analogue gear. 24 bit is a standard but all you're usually recording is a bit more analogue noise.

 

[snow lizard]

8 bit, 22 kHz would make a smaller file size (or bit rate if the signal was to be transmitted). You'd also find that most microphones would be able to easily exceed the range of the format in both dynamics and frequency.

16 bit 44.1 kHz done with proper TPDF dither can have a dynamic range of around 110 dB and capture frequencies close to 20 kHz. This exceeds the range of many professional microphones.

 

[Farview]

At first, signals outside the hearing range being editted do can effect the signals within the hearing range. It gives the effect that there's 'more', deeper, clearer. So going outside nyquist when still producing does have effect on the end result.

That has been debated, especially since most instruments don't have any energy much beyond 20khz, most mics won't pick it up and most playback systems can't reproduce it. In theory, I can see how it's possible, in practice it doesn't happen.

And second, when i use the term 'unlimited' i'm pointing at the 'bit range' of an analoque signal. The line which makes the curves. There digital is limited while it ain't a line but serie off dots. With that i'm not point at 'unlimited outside hearing range' or whatever.

Bit range sounds like bit depth, which is what determines the dynamic range. Your description sounds like sample rate. The number of dots would be the samples, how loud the dots are would be the bit depth.

Even with analog tape, there are a finite amount of rust particles to magnetize, any change smaller than two particles next to each other will not be recorded or played back, so it isn't unlimited or infinite.

And as the analoque signal is much more detailed than a by bits limited digital signal an analoque signal is deeper, and so the edits (reverb, compressor, etc) are deeper too. And with that deeper you preserve more clearness.[/except for the really high noise floor...(compared to 24 bit)

edit- Now that I'm reading everything, I think you are using the term 'bit' when you actually mean sample. Or you are using 'bit' in it's non computer meaning, like "I'll have a bit of ice cream with my cake".

 

[Farview]

One short try.

Take a 16bit 44.1khz sound.
Bring that back to 8 bit and 22 khz (both about half).
And listen to the difference. It's worse, isn't it?

Bringing down 32 bit 96khz back to 16 bit and 44.1khz (both about half too, actually even more).
That's as much breaking down the quality as in the first example.

And editting a worse sound effects the result even more then worse.

Please react on this.

The problem with your scenario is that 22k sample rate will cut off the top octave of what we can hear. Going from 96k to 44.1k, you won't lose any useful frequency range. It's like having two steaks with 8oz of meat, but one has another 8 oz of fat on it. If you cut them both in half, one will still have 8oz of meat, but the other will only have 4oz. Cutting off extra stuff you don't want or need doesn't affect the meal. Cutting off stuff you do want does.

 

[snow lizard]

Many of the DC amps won't go below 0Hz. Shame and Pity.

Well yeah, but where's the passion?

Where's the LOVE that made you who you are?

 

[miroslav]

Even with analog tape, there are a finite amount of rust particles to magnetize, any change smaller than two particles next to each other will not be recorded or played back, so it isn't unlimited or infinite.

I wondered when tape would be finally mentioned...because all this talk of analog recording is basically about tape recording VS digital recording...there's realistically no other analog recording format.
When talking about the range and quality of analog VS digital...even the most highest quality tape recording only approaches the ranges of digital...and any lower quality tape recording is easily beaten by even the lowest quality digital that we have today.

This analog VS digital recording discussion then quickly becomes an academic debate in most cases. For those who have the luxury and finances to afford the absolute highest quality tape systems, there might be a more even comparison of range and quality, though digital these days surpasses even that in frequency and dynamic range capability.
For those who don't have the tape gear....at least pretty decent tape gear...then what's the point of talking about analog recording?

So then what's the real argument here?
It seems more about personal preferences for the tonal differences or what have you, than it is about technical capability or audio quality.
The problem is when misconceptions lead to absolute beliefs based on mostly myths that have been debunked by many...many who have much deeper technical backgrounds than most here...yet here we're still doing this analog VS digital stuff....?...really?

And I say all that with equal love of both analog/tape and digital technologies...as I use both regularly...so no bias on my part...but I can't say that for everyone here, as there appears a bias because their knowledge and experience comes from "ancient" audio history when analog was king (well, when you go back 20-30 years in audio technology...it's ancient).
Digital of the last 10-15 years totally rocks... ...I just love recording to tape before dumping it to digital.

 

[garww]

Well yeah, but where's the passion?

Where's the LOVE that made you who you are?

Some of us know there exits a Bizarro Audio World. Hello, for now : ) So, isn't 0Hz about the same as 0 Celsius ? hahah

 

[Farview]

I think it would be the same as 0 kelvin.