Why is 16 to 24 bit no big deal - no popups for conversion?

[EvilDragon]

Changing bit depth is NOT upsampling!!!

[lfm]

I think I see where lfm is going with this: that there is some 'right' value for each sample that, after the dithering and truncation process, is lost but might be recoverable on restoration to 24bit or higher. The reality is, unless you are dealing with a highly predictable signal or make heroic assumptions about it, no algorithm is going to be able to recover that missing data.

You do no (further) damage by just filling in with zeroes - because that has no audible effect whatsoever. OTOH, interpolation runs the risk of doing the same as bad dither, where you get artificial periodic signals appearing just above the noise floor.

That's is what I might suspect too - do more harm than good.
Just thinking a better approximation that a previous truncated value would be possible, not entire spectrum om what was there.

But what if you did analog equipment math to what you import, making it sound a slightly bit more pleasant to ears. You can do this with a plugin obviously - and maybe we should, something really subtle.

Just curious if any daw market themselves on this, or any audio editor.

The question arose when I was looking at the documentary Soundcity which I love to watch for inspiration in between. And also bought CD track for this film with the full tracks involved. And it just sounds so boring compared to the film - audio on film is so much better - but sadly not full versions on film.

The newer Sonic Highways though, do full tracks inside the film. I bought the dvd for that where FooFighters recorded in 8 cities and doumented some of musical history of that city as well as some recording sesions or part of. Really interesting.

And that is why I bother everyone with 24-bit detection, if that is the answer why film is sounding so much better. And also this about conversion stuff what it does to our music.

[lfm]

Changing bit depth is NOT upsampling!!!

But you could do as part of that operation 16 to 24 bit - that was my reasoning.
To make a better approximation of which of 256 levels probably better suite final 24-bit sample.

[BertKoor]

Dude, the 16bit audio content you get was probably already dithered. Nothing there under -96dB but (dither) noise... Nothing there to rescue, nothing there to gain. For all real-world purposes 16bit resolution is really good enough.

[Burillo]

it's no big deal for the same reason changing color bit depth upwards is not a big deal - the original information is perfectly preserved, because 24 bit is a strict superset of 16 bit. this is not analogous to upsampling because upsampling by e.g. 2x actually doubles the amount of samples, while changing bit depth upwards just changes individual samples to being capable of representing more values.

imagine a 8-bit picture. if you display it on an 8-bit display, it displays perfectly. if you display it on a 16-bit (65536 colors) display, it'll display perfectly as well. if you display it on a 24-bit (16 million colors) or 32-bit (4 billion colors) or... or... display, it'll still display perfectly, because you're only changing the capability of the medium, but not the data. a 16-bit display will display everything 8-bit display can, and more.

in contrast, upsampling is more analogous to scaling the picture upwards - suddenly, you get holes between pixels, and you need to fill them somehow. that's what you do with interpolation - a better interpolation algorithm is more accurate at filling holes, e.g. finding out what the pixel would have been if your picture was bigger.

[lfm]

Dude, the 16bit audio content you get was probably already dithered. Nothing there under -96dB but (dither) noise... Nothing there to rescue, nothing there to gain. For all real-world purposes 16bit resolution is really good enough.

I guess that is why SACD arised and even vinyl still exist!!!!!

Of course there are things that were lost doing 16-bit reduction.
Dither is used to minimize effect of that reduction.

I think shifting 16 bit 8 steps left, and then add 127 will probably do a better job, than just multiply by 256. Then probably do DC offset removal due to this.

Think about it - when reduced bitdepth - you divide lower 256 levels in two, then let dither decide if truncated to lower bit, or raise to higher bit. Value 129 and up probably raise most of the time, 126 and down truncate. Then randomness of dither noise decide what to do with value 127-128 or so. Noise is probably higher than a single bit step, but as a principle this is what happends. And noise is shaped differently depending on material - various smart algorithms exist.

You continue mixing at 16 bit if you want, I notice clear difference doing it in 24 bit, since now 12 years.

[lfm]

it's no big deal for the same reason changing color bit depth upwards is not a big deal - the original information is perfectly preserved, because 24 bit is a strict superset of 16 bit. this is not analogous to upsampling because upsampling by e.g. 2x actually doubles the amount of samples, while changing bit depth upwards just changes individual samples to being capable of representing more values.

Yes, if you are of the impression that 16-bit is perfect - then you don't loose out on making 24-bit conversion.

My reasoning is built on that something was lost going 16-bit to start with. Is there maybe a way to make better conversion going up in bit depth. I think there might be.

But appreciate the discussions - I was asking since I don't know exactly what daws do and not in this regard.

[Burillo]

You continue mixing at 16 bit if you want, I notice clear difference doing it in 24 bit, since now 12 years.

i'm tired of people confusing processing audio with storing audio. you're not mixing in 24 -bit - your DAW probably processes sound with 64-bit floating point depth.

[Burillo]

it's no big deal for the same reason changing color bit depth upwards is not a big deal - the original information is perfectly preserved, because 24 bit is a strict superset of 16 bit. this is not analogous to upsampling because upsampling by e.g. 2x actually doubles the amount of samples, while changing bit depth upwards just changes individual samples to being capable of representing more values.

Yes, if you are of the impression that 16-bit is perfect - then you don't loose out on making 24-bit conversion.

My reasoning is built on that something was lost going 16-bit to start with. Is there maybe a way to make better conversion going up in bit depth. I think there might be.

But appreciate the discussions - I was asking since I don't know exactly what daws do and not in this regard.

but that's a different issue. whether or not something was lost in capturing a recording in 16-bit instead of 24-bit has nothing to do with whether anything will change once you convert it to 24-bit. the answer is, nothing will change. do a null test if you don't believe me.

Yes, if you are of the impression that 16-bit is perfect - then you don't loose out on making 24-bit conversion.

yeah, way to put words in my mouth. straw man much?

[lfm]

but that's a different issue. whether or not something was lost in capturing a recording in 16-bit instead of 24-bit has nothing to do with whether anything will change once you convert it to 24-bit. the answer is, nothing will change. do a null test if you don't believe me.

On the daw you use - I believe you.

Is there maybe a daw that do it differently - that is pretty much the heart of this topic?

Or is there nothing to gain - maybe there isn't.

In the beginning of digital revolution everything was perfect. As time went by, people starting using their ears - and went back to vinyl. Despite there being statics in between and other unwanted noise.

Well, others hearing were degraded such of high in ear listening so they thought mp3 was how things should sound.

And the more were moved to digital domain - there more analog was requested.
Or higher resolution digital.

I saw some expert panels from Pensados Place at some exhibition fair, or whatever it's called. The renowned awarded mixers and producers told how they moved everything to tape first.

Digital synths was a blessing in the sense oscillators didn't wander off - and needed constant retuning the first couple of hours it was running.

And we all know the amount of analog synths that are back with the analog things

It's not like - in the beginning there were nothing but ones and zeros

Any more input, please just throw it here....

[Burillo]

On the daw you use - I believe you.

Is there maybe a daw that do it differently - that is pretty much the heart of this topic?

no, that's not the heart of the topic. the heart of the topic - as in, the original question the OP asked - is about why DAWs don't warn about conversions from 16 bit to 24 bit. and the answer is, the same reason they don't mind playing back 24-bit material but process it in 64-bit floating point: there's no difference to be had there. resampling (up or down) is a destructive process. increasing bit depth is not a destructive process.

In the beginning of digital revolution everything was perfect. As time went by, people starting using their ears - and went back to vinyl. Despite there being statics in between and other unwanted noise.

you and i both know that's not why people "went back to vinyl". "digital revolution" has nothing to do with it.

And the more were moved to digital domain - there more analog was requested.
Or higher resolution digital.

no, that's just marketing. especially the latter part. time and time again studies show people - as in, professionals - can't tell the difference between 44.1/16 and 24/96 mixdown. there's no "there" there.

I saw some expert panels from Pensados Place at some exhibition fair, or whatever it's called. The renowned awarded mixers and producers told how they moved everything to tape first.

Digital synths was a blessing in the sense oscillators didn't wander off - and needed constant retuning the first couple of hours it was running.

And we all know the amount of analog synths that are back with the analog things

It's not like - in the beginning there were nothing but ones and zeros

Any more input, please just throw it here....

all of that has nothing whatsoever to do with the topic at hand.

[cron]

Upsampling refers to sample rate. It's trivial stuff to add points between existing points there, and the sound you get at the end of it would be the same.

However, If you interpolated the values of a 16 bit waveform across the entire 24 bit range, you'd get something akin to super harsh dynamic range expansion. Your loudest stuff will still be at -0db, but the quietest stuff will now be at -144db instead of -96db.

If you make a recording at 24 bit, it doesn't make any difference at all to the 'grid' above -96db when compared to recording at 16 bit. It just means you've got between -96db and -144db to play with too.

multiple neurotic edits for clarity

[Gamma-UT]

In the beginning of digital revolution everything was perfect. As time went by, people starting using their ears - and went back to vinyl. Despite there being statics in between and other unwanted noise.

No it wasn't. The first CD players were pretty terrible. It wasn't until sigma-delta converters became mainstream that the perceived quality of digital at the consumer end really outstripped. I was a relatively late adopter of CDs. Perhaps because of that, I don't see much greater value in vinyl other than it's a medium that doesn't benefit from brickwall limiting and so the mastering vs CD/MP3 is a lot gentler.

[sonicpowa]

Is there maybe a way to make better conversion going up in bit depth. I think there might.

Are you thinking there is a way to somehow restore those once truncated/dithered bits? This is of course impossible.

[whyterabbyt]

Is there maybe a way to make better conversion going up in bit depth. I think there might.

Are you thinking there is a way to somehow restore those once truncated/dithered bits? This is of course impossible.

Meh; of course its possible. The only reason we dont just store everything as 1-bit/44K and interpolate back up to 24-bit is because of the hard-drive manufacturer's propaganda.