New Data Showing No Difference Between 16 and 24 bits audio


I won't bother to read that, I already know people can't hear the difference between 16 bit audio and 24 bit audio. That is not the real reason for 24 bit audio.

The reason for 24 bit audio is that it is necessary to be able to generate 16 bit audio. For 16 bit audio to be good enough, it must use the entire dynamic range of 16 bits. It is very difficult to set the gains before recording to guarantee you will get the maximum peak right at full scale. You either get it substantially lower than full scale, or it clips.

So you record at 24 bits so you have lots of headroom, then in post processing you adjust the gain so that the highest peak just touches the maximum value of the 16 bit range. This way you get the full 16 bit dynamic range without clipping.

Then people got lazy and decided just to deliver the 24 bit recording without re-scaling it to 16 bits, and let the user adjust the scale with the volume control.
 
Last edited by a moderator:
A 'study' on a website called http://archimago.blogspot.jp/- whilst interesting enough, really isn't concrete proof of anything ;) so certain nothing has been concluded forever from that... so whilst 16 is enough for the end user, we can still look forward to the pointless debate continuing for years to come.
 
The argument is essentially its better to have your audio data at 24bits even if you possibly cannot tell the difference and since youve got it in 24bit anyway why not listen to it that way?
 
Interesting article, shame they didn't get a slightly larger / more diverse group to take the test. Hearing tends to deteriorate with age so it would've been interesting to have had a few younger people involved.

As mentioned in the body text it would be good to see a similar experiment for sample rates, it is said that 44.1kHz is enough for anyone but many people swear that they can appreciate the benefits of 96kHz.
 
High frequency hearing tends to depreciate with age, but I've heard nothing suggesting the responsiveness of those cilia in our ears still alive and working decreases with age, so I don't see a problem with having an older set of respondents from that point of view. If this were a 44.1kHz/96kHz test then the age of the respondents would likely be more important, but I don't think you can criticise it as it stands for that reason.

You can criticise it for having a tiny sample size, and for only using three test cases, meaning statistically it's insignificant, but it's not completely without worth to now be able to say that I don't know whether I can get anything extra out of a 24-bit recording than a 16-bit one (other than I can glean from my understanding of that giph.org piece), but this group of self-described audiophiles can't, so any money I spend on being able to reproduce 24-bit audio reliably is likely money misspent.

Edit: ziph.org, not giph.org - there's a link down there if you don't know what I mean vvvv
 
Last edited by a moderator:
A 'study' on a website called http://archimago.blogspot.jp/-whilst interesting enough, really isn't concrete proof of anything ;)
Why, all your results need to come from Nature to be valid ?

Not sure why you say its not a proof of anything. So if users cannot tell any difference between 16 and 24 bits tracks, why does not it prove anything ? What else would you like to see ?
 
A 'study' on a website called http://archimago.blogspot.jp/-whilst interesting enough, really isn't concrete proof of anything ;)
Why, all your results need to come from Nature to be valid ?

Not sure why you say its not a proof of anything. So if users cannot tell any difference between 16 and 24 bits tracks, why does not it prove anything ? What else would you like to see ?
I could create a blog, pretend to run a study, post fake findings, then share it on this forum, and make a bold post about how it will end arguments forever, but of course it really proves nothing. It is not that I believe the site you posted is a big hoax, but I do not think you are going to end arguments with what you have posted. Maybe all the people that voted for the 24 bit sample really did hear the difference, and everyone who voted for the 16 bit sample had no idea. There are some many if buts and maybe...

Interesting read, yes. Enough to stop further people claiming 24 bit is a better listening experience, no! ;)
 
Years of game system advertising has taught me that if there are more bits the better it is. Also the best and tastiest bits are Bacon Bits.
 
Last edited by a moderator:
A 'study' on a website called http://archimago.blogspot.jp/-whilst interesting enough, really isn't concrete proof of anything ;)
Why, all your results need to come from Nature to be valid ?

Not sure why you say its not a proof of anything. So if users cannot tell any difference between 16 and 24 bits tracks, why does not it prove anything ? What else would you like to see ?
I could create a blog, pretend to run a study, post fake findings, then share it on this forum, and make a bold post about how it will end arguments forever, but of course it really proves nothing. It is not that I believe the site you posted is a big hoax, but I do not think you are going to end arguments with what you have posted. Maybe all the people that voted for the 24 bit sample really did hear the difference, and everyone who voted for the 16 bit sample had no idea. There are some many if buts and maybe...

Interesting read, yes. Enough to stop further people claiming 24 bit is a better listening experience, no! ;)
Yeah, while its evidence that people cannot tell the difference; it isn't proof. It needs to be repeated with a larger and more diverse sample and perhaps a tighter methodology.
 
Years of game system advertising has taught me that if there are more bits the better it is. Also the best and tastiest bits are Bacon Bits.
I read that exactly in crow's voice...
 
A 'study' on a website called http://archimago.blogspot.jp/-whilst interesting enough, really isn't concrete proof of anything ;)
Why, all your results need to come from Nature to be valid ?

Not sure why you say its not a proof of anything. So if users cannot tell any difference between 16 and 24 bits tracks, why does not it prove anything ? What else would you like to see ?
I could create a blog, pretend to run a study, post fake findings, then share it on this forum, and make a bold post about how it will end arguments forever, but of course it really proves nothing. It is not that I believe the site you posted is a big hoax, but I do not think you are going to end arguments with what you have posted. Maybe all the people that voted for the 24 bit sample really did hear the difference, and everyone who voted for the 16 bit sample had no idea. There are some many if buts and maybe...

Interesting read, yes. Enough to stop further people claiming 24 bit is a better listening experience, no! ;)
How about from Xiph.org:p
 
Back
Top