24bit vs 16bit, the myth exploded!

[OddE]

  Yes, I think that would have to be my approach. So much of the 16vs24 debate seems to focus on the sound science part of it, while purposefully discounting mastering differences. It seems that the mastering variations are perhaps not as prevalent or obvious as I had thought based on my reading. I guess it really does come down to trial and error.

 
-That is because those are two completely different debates; the 16/24 debate - at least here in Sound Science - tends to focus on what either format is capable of and how that capability relates to the capabilities of the most important transducer of all - our ears.
 
The music being distributed using the various formats, on the other hand, is a different issue altogether; if some publisher decides to release an album with, say, 40dB of dynamic range over a -55dBFS noise floor in 24 bits, while also sampling at 96 or 192kHz source material with no content whatsoever above 15kHz - that is an inherent problem with the publisher, not the format.
 
Edit: Ninja'd by RRod; that's what you get for loading the thread, reading at a later time and then eventually getting around to writing an answer...

 

[U2Bono269]

Now that makes sense.
 
And sorry, I didn't realize this was the Sound Science forum. I'm using Tapatalk and it's not very organized. I posted it in this thread because it had the most information. But for the most part, every other thread I read made the same claims, for the most part.

 

[lamode]

  I've been reading as many of the 16bit vs 24bit threads I could find around here, because I'm interested in high resolution audio and whether or not it's worthwhile. I do have 2 albums in 24/96, though I don't yet have the equipment needed for them. I figured this would be the best thread to ask this in.
 
I've seen the argument that there's no difference in practice between a 16bit and 24bit file, assuming the same mastering. A common argument seems to be that those who do hear differences between 16bit and 24bit, and between lossless and 320mp3, are hearing this difference due to mastering and other factors. This makes a lot of sense and I understand it.
 
However, doesn't the fact that a high-resolution file may use different mastering justify it's existence? It would seem to me that if a high-resolution file offers a different master, it's automatically a worthwhile endeavor. Am I missing something, or oversimplifying?

 
There is no technical reason to do it, but the audience for HD material is more discerning so perhaps they remaster some releases with less compression to keep audiophiles happy.

 

[jugate]

   
ok, marketing. I see.
 
But if i have a z906 (logitech) that dont have decoder for DD/DTS HD and play a movie with that codec audio... the system sound exactly that if i have conected the analog imputs and decoding with software?

 
somebody?

 

[RRod]

   
somebody?

 
How exactly do you have things set up? The HD surround formats have the ability to send out a non-HD "core" when connected with something like optical cable that can't handle the full bandwidth, but this all depends on the exact settings of your source and sound card.

 

[bigshot]

Blu-ray disks have multiple audio tracks in various formats. Generally, you set your player to default to the highest quality track your player is able to decode and don't worry about it any more. The only really significant difference in sound quality is 2 channel vs 5.1/7.1.

 

[Greenears]

  I want lossless audio. Whether or not it's "necessary" over mp3 or aac or whatever is irrelevant because it's what I want. I tend to think that a file that has more information in it is more accurate than one with less. I don't care if I can hear it or not. That is my preference.
 
I'm also really not interested in endless ABX testing and fretting over technical information. I want to listen to the music. So anytime a user here quotes test results, or graphs, or frequencies, or any of a billion other things, it misses the point for me.
 
And "controlling for master differences" is simply not the point of my question. I want to find the best sounding masters/transfers/versions (whatever you want to call it) of my favorite albums, regardless of whether they're ripped in FLAC off a CD or downloaded from a hi-res website. But it seems that there's not really an answer to my question.

 
Actually there might be an answer.  There seems to be some sites that run Foobar's DR (dynamic range) plugin on all the tracks in an album and post logs.  Now of course beware, one DR reading vs another does not guarantee quality.  But it does sometimes shed light on which "mastering" is downloadable from which source.  The nice thing is you can do this at home and try to match up to their database.
 
This doesn't have much to do with 16 vs 24 format, they both yield the same results if they are derived from the same mastering.  I've seen it go both ways - I looked at a circa 2000 remaster of Dire Straights (Communique if I remember) in 24 bits on Pono store and I was warned in forums that that is a notoriously bad remaster (in the original 16 same in 24).  I was advised to stick with the original CD as the nicest sound.  Then I looked at AC/DC 2014 release (ok hardly audiophile haha) and was advised that in Europe it was mastered with DR8 and US DR6 (logs are posted).  Apparentlly the HD tracks 24 bit is the DR8 version and _may_ be the only legit place in US you can get that version (other than importing a plastic CD which is too much trouble). 
 
Again examples chosen to show nothing to do with the 24b vs 16b debate - but a fun sidetrack.

 

[bigshot]

There is a contradiction between saying "I want lossless because it has more data and I don't care if compressed audio sounds exactly the same" and then saying "I don't want to be fretting over technical information, I just want to listen to music". If it sounds perfect to your ears, it *is* perfect. I agree that you shouldn't worry about technical information. All you should worry about is what your ears hear. MP3 LAME and AAC are capable of complete transparency. Just listen to the music and don't worry about file sizes.

 

[dprimary]

  Then I'm confused. I've seen similar sentiments in numerous threads on this topic: that perceived differences between mp3/lossless or 16/24 are due to comparing different masters. Logically, it doesn't seem likely that an album would be made into mp3, redbook, and 24bit versions without some sort of adjustment to take the format into account, right?

A MP3 or AAC release would have different masters, you make minor adjustments to maximize the quality of the final format. For example master for iTunes means the mastering engineer produced a separate master just for AAC encoding most likely listening to the original master through and encoder and processing it to sound as good as they can get it. So a mastered for iTunes track should not be exactly the same as the CD track. 
 
A 16bit release and 24 bit release should have the same 24 bit master, when you dither a 24bit master down to 16 all that you are essentially doing is raising the noise floor. It think it was someone at Harman that said 24bit gives you foot room. 16 bit has a noise level -96dB while 24bit give about 24dB more so the noise is roughly -120dB down, which is limited by physics to get much better. An extremely silent studio will be about 20dB SPL you would playing the music at an ear splitting116dB SPL before you to start to hear the noise floor of the 16bit file. All you do for a 16bit master is dither down the 24bit master.

 

[Don Hills]

 ... All you do for a 16bit master is dither down the 24bit master.

 
I wish that's all they did.

 

[dprimary]

   
I wish that's all they did.

Unfortunately they will "remaster" old albums to sell "better high rez" (AKA louder and brighter) sound.

 

[bigshot]

  A MP3 or AAC release would have different masters, you make minor adjustments to maximize the quality of the final format.

 
Not if you roll your own. You end up with something that is indistinguishable from the original format.

 

[analogsurviver]

  Damn right. I bought an iDSD Micro and honestly my onboard Realtek chip truly sounds much better.

I don't know why, but iDSD Micro + Sony MDR-XB1000 are not an improvement over integrated laptop audio.
 
I have a Fostex TH900 on the way. Hopefully that will open up a whole new world, but if it doesn't, this baby will go on sale.

Did you manage to get micro sounding better than the Realtek onboard chip ?
 
I had quite "a-not-so-pleasent-ball-with-my-PC" before my nano started to sound as it should. PC related, registry leaning etc etc. And it takes a few days to configure whatever software you are using to your liking/preference - ALWAYS taking in account of the limitations of your hardware.
 
It is only fair to any DAC to clearly state with which hardware and software ( and how configurd ...) it is being used - it can do nothing on its own.

 

[Greenears]

  A MP3 or AAC release would have different masters, you make minor adjustments to maximize the quality of the final format. For example master for iTunes means ....

 
I don't agree with that.  There s plenty of science on this thread that shows modern MP3 320 or AAC 256 the differences to lossless are inaudible so no need to remaster anything due to format.  The reason they are doing this push to remaster is to optimize for typical listening conditions of iTunes users (which I have to assume is mobile earbuds).  Which sucks if you are not typical.
 
It would be nice if they start releasing with multiple masters for different conditions, put it as parameters in the EQ.

 

[U2Bono269]

   
Actually there might be an answer.  There seems to be some sites that run Foobar's DR (dynamic range) plugin on all the tracks in an album and post logs.  Now of course beware, one DR reading vs another does not guarantee quality.  But it does sometimes shed light on which "mastering" is downloadable from which source.  The nice thing is you can do this at home and try to match up to their database.
 
This doesn't have much to do with 16 vs 24 format, they both yield the same results if they are derived from the same mastering.  I've seen it go both ways - I looked at a circa 2000 remaster of Dire Straights (Communique if I remember) in 24 bits on Pono store and I was warned in forums that that is a notoriously bad remaster (in the original 16 same in 24).  I was advised to stick with the original CD as the nicest sound.  Then I looked at AC/DC 2014 release (ok hardly audiophile haha) and was advised that in Europe it was mastered with DR8 and US DR6 (logs are posted).  Apparentlly the HD tracks 24 bit is the DR8 version and _may_ be the only legit place in US you can get that version (other than importing a plastic CD which is too much trouble). 
 
Again examples chosen to show nothing to do with the 24b vs 16b debate - but a fun sidetrack.

 
Thanks! That's kind of what I was looking to know, and is very helpful to me.
 
 

  There is a contradiction between saying "I want lossless because it has more data and I don't care if compressed audio sounds exactly the same" and then saying "I don't want to be fretting over technical information, I just want to listen to music". If it sounds perfect to your ears, it *is* perfect. I agree that you shouldn't worry about technical information. All you should worry about is what your ears hear. MP3 LAME and AAC are capable of complete transparency. Just listen to the music and don't worry about file sizes.

 
I think that lossless does sound better to me than compressed audio, so there's no contradiction there. My question was related to 24bit audio specifically, and how much of it is related to mastering or the conversion process. I'm still learning all of this stuff, so I'm probably not as eloquent as I'd like to be, so sorry if there's any confusion.
 
I guess I was just wondering if what you guys do, i.e., making your own exact comparisons, is the same thing they do in a professional setting. Enough people claim to hear a difference that it makes me think that looking at it from a strictly scientific point of view is very limiting. After all, there is little-to-no transparency from labels or engineers, so how do we know for sure?
 
This has been good though Lots to think about and learn about.