[RRod]

  However the gobbledygook employed is baffling.
Take this for example,
"Assume that this is at 44khz (that frequency is actually irrelevant for the rest of this)."
followed by this
"taking it back to 44100."
leaves me floundering to understand it.

 
The two 44100s are unrelated. He simply had to pick a sample rate at which to make the initial waveform. He could also have made the waveform natively at 22MHz and then used a sinc filter for bandlimiting. The take-home is that the delay of 1 sample at 22MHz is maintained when going down to 44.1kHz. Thus 44.1 can resolve peak location at much finer steps than 1/44100 seconds.

 

[FFBookman]

those xiph. org links have done more to hurt audio quality than anything else i can think of  in the last 15 years.
 
don't be suckered in by DSP programmers telling you about sound, telling you what you can and can't hear, what is and isn't important to your well being.
 
these people know nothing of great sound, they are ignoramuses to their own ears.
 
it's called perceptual coding, and seemingly intelligent people under the age of 30 live by it..... it goes something like "if you can't prove to me that i miss it than i don't. so screw you, the new digital way is better.  shutup."
 
most of them have never even heard higher quality audio, or they are the opposite - they have such high end stuff that they shoot down everyone without it as being clueless and less than them.
 
but in the juicy middle, the standard has dropped from 1400k to 256k and that's BS.
 
we should be going the other direction, just like TV, monitors, photos, GPS, sensors, textiles, printing, and everything else in the world. if you are doing a re-creation, max resolution possible is best.

 

[Music Alchemist]

 
In all your posts on this forum, not once have you presented proof to substantiate your claims of being able to distinguish between Red Book and hi-res. Are you willing to do so? Because I can help you in the process.

 

[FFBookman]

In all your posts on this forum, not once have you presented proof to substantiate your claims of being able to distinguish between Red Book and hi-res. Are you willing to do so? Because I can help you in the process. 

you don't believe quality exists, not my problem.  
 
you will find ways to discount any test I might take, and i will find ways to discount any test you might propose.  
 
since my main argument is with the testing procedures itself, i won't be sitting for your tests.
 
i do the test often by downsampling my own digital mixes using different dithering algorithms and then listening to the results on various playback systems.
 
i can't think of any way to prove to you that 4600k > 1000k if you don't already believe that to be true.

 

[Music Alchemist]

  you don't believe quality exists, not my problem.  
 
you will find ways to discount any test I might take, and i will find ways to discount any test you might propose.  
 
since my main argument is with the testing procedures itself, i won't be sitting for your tests.
 
i do the test often by downsampling my own digital mixes using different dithering algorithms and then listening to the results on various playback systems.
 
i can't think of any way to prove to you that 4600k > 1000k if you don't already believe that to be true.

 
If you are hearing a real difference, you would be able to pass an ABX test. If you don't actually hear a real difference, you would fail the ABX test. It's simple, and the only way to verify these things.

 

[FFBookman]

   
If you are hearing a real difference, you would be able to pass an ABX test. If you don't actually hear a real difference, you would fail the ABX test. It's simple, and the only way to verify these things.

Your belief that an ABX test is the only way to verify these things is very wrong.
 
If you rely on a garbage test to find the truth, the truth you find will be garbage.

 

[FFBookman]

btw i suggest the entire led zeppelin catalog in 24/96, recently released. if you don't believe it exists, keep your head in the sand. 
 
i'm sure someone is happy to sell you 10% files for the price of 100% songs.
 
i love how people claim Hi Rez is too expensive -- apple and that model sells you 10% of the song data in a 5mb container for $0.99.  
 
pono/hdtracks/prostudiomasters sells you the entire song in a 50mb+ container for ~ $2.00.  that's more than 2x the data for double the cost.

 

[FFBookman]

re: what is taking so long - from the horses mouth:
 
Director of Sony Music's re-issue department (the back catalog curator) said in 2015:

'We went to 24/88 or 24/96 in the late 90's for our own archival purposes, whether or not it ever was going to be a consumer format. What most people don't realize is that only about 30% of the entire catalog of analog tape was digitized the first time around (80's-90's), and it was done at 16/44 with less than ideal converters. This is all the labels, not just us. So besides having to remaster those releases, we have to address the rest of the content still on analog only, and of course the priority is new releases from our current artists.'
 
He also mentioned how long it actually takes to digitize an old analog tape properly - you have to get the tape, bake the tape, find similar hardware to load the tape onto (difficult these days), setup similar EQ and mix parameters as written in the tape log (very difficult these days), recreate to the best of your ability the original mix on modern equipment, consider whether to fix/clean up or not, use engineers that know tape and vintage gear, then digitize at 24bit and master for all the modern formats. This is an expensive session, resulting in nothing more than another copy of an existing product. Not exactly lucrative stuff for the labels or the artists.

 

[Music Alchemist]

  Your belief that an ABX test is the only way to verify these things is very wrong.
 
If you rely on a garbage test to find the truth, the truth you find will be garbage.

 
Okay, then what other methods can prove that you are hearing a genuine difference?
 
And about downloading certain HD albums...they sound different because they are a different master! I have converted the 24-bit files to 16-bit to compare them. They sound exactly the same when you do a proper comparison.

 

[FFBookman]

   
Okay, then what other methods can prove that you are hearing a genuine difference?
 
And about downloading certain HD albums...they sound different because they are a different master! I have converted the 24-bit files to 16-bit to compare them. They sound exactly the same when you do a proper comparison.

Material and familiarity with that material is a huge part of it.  If it's a (usually modern) artist with loudness problems, it's really hard to tell. I don't listen to most stuff from 2000-now because of the loudness issues.   The other part is gear - if you don't have a proper DAP or digital setup to render 24bit fully, you are choking the playback.
 
For instance - you mentioned Linkin Park in the thread - I doubt anyone could tell 24bit vs 16bit with music like that.  Most of it is fake source, or faked with ridiculous editing, and the tones are so far removed from reality of what the instrument sounds like that there's no realism to lose in the first place. There's no room, no vibe, no texture. It's bandwidth limiting pretending to be dynamic.
 
EDM, hip-hop, modern pop - none of it benefits from 24bit right now, because they aren't using bandwidth anyway. They are used to fitting it into 256k and making it as loud as possible, there is no reality and detail there to hear in the first place.
 
Go back to things recorded in the 70's, the 60's, even the 80's and into the early 90's - to me it's obvious to hear yet nearly impossible to capture.  It's very easy to get your favorite zepellin or stevie wonder album at 24bit and find the used cd for $10. 
 
There's just MORE.  The amount of signal coming out the speaker, projecting the soundstage into my room - there is more of it at 24bit and thus it has more detail and more power. That is why 1.4Mbs is still > 256k and assuming it's not up sampled, 5.8Mbs is still > 1.4Mbs.
 
Scream mastering and remastering all you want, but everything needs to be mastered to get to it's final format, that's the definition of mastering. Whether it's been 'remixed' or not is not usually known, it's private. If the thing is a new transfer coming from an old tape it has to be a new mix, you have to bring the tape up to a proper mix -it's not a saved session file in pro-tools somewhere. so of course it's remixed. 
 
As far as your test of bought 24bit material down sampled to 16bit, sounds valid to me. Maybe you just don't hear it. No biggie to me. You save money in the long run. Just know that it's from a combination of your material, your player, and your listening abilities, not because it doesn't exist.
 
I wonder, have you tried different dithers and turning dither off when you do your test?  In my experience the dithers do sound different (for instance I'm not a fan of Pro-Tools' 'sound shaping' dither.

 

[Music Alchemist]

  Material and familiarity with that material is a huge part of it.  If it's a (usually modern) artist with loudness problems, it's really hard to tell. I don't listen to most stuff from 2000-now because of the loudness issues.   The other part is gear - if you don't have a proper DAP or digital setup to render 24bit fully, you are choking the playback.
 
For instance - you mentioned Linkin Park in the thread - I doubt anyone could tell 24bit vs 16bit with music like that.  Most of it is fake source, or faked with ridiculous editing, and the tones are so far removed from reality of what the instrument sounds like that there's no realism to lose in the first place. There's no room, no vibe, no texture. It's bandwidth limiting pretending to be dynamic.
 
EDM, hip-hop, modern pop - none of it benefits from 24bit right now, because they aren't using bandwidth anyway. They are used to fitting it into 256k and making it as loud as possible, there is no reality and detail there to hear in the first place.
 
Go back to things recorded in the 70's, the 60's, even the 80's and into the early 90's - to me it's obvious to hear yet nearly impossible to capture.  It's very easy to get your favorite zepellin or stevie wonder album at 24bit and find the used cd for $10. 
 
There's just MORE.  The amount of signal coming out the speaker, projecting the soundstage into my room - there is more of it at 24bit and thus it has more detail and more power. That is why 1.4Mbs is still > 256k and assuming it's not up sampled, 5.8Mbs is still > 1.4Mbs.
 
Scream mastering and remastering all you want, but everything needs to be mastered to get to it's final format, that's the definition of mastering. Whether it's been 'remixed' or not is not usually known, it's private. If the thing is a new transfer coming from an old tape it has to be a new mix, you have to bring the tape up to a proper mix -it's not a saved session file in pro-tools somewhere. so of course it's remixed. 
 
As far as your test of bought 24bit material down sampled to 16bit, sounds valid to me. Maybe you just don't hear it. No biggie to me. You save money in the long run. Just know that it's from a combination of your material, your player, and your listening abilities, not because it doesn't exist.
 
I wonder, have you tried different dithers and turning dither off when you do your test?  In my experience the dithers do sound different (for instance I'm not a fan of Pro-Tools' 'sound shaping' dither.

 
I completely agree that it can sound different when you get an album from two different sources. That's why it's essential to convert the files yourself to isolate the variables and ensure you are listening to the same master of a recording. At the least, you have to take the 24-bit file and convert it to lossless 16-bit / 44.1 kHz. Did you do that beforehand? I recommend using dBpoweramp. It does a perfect conversion. For the record, I listen to just about all types of music. Anyway, all this is beside the point of my question. What is your alternative to an ABX test for the purpose of proving you can hear a difference?

 

[FFBookman]

   
I completely agree that it can sound different when you get an album from two different sources. That's why it's essential to convert the files yourself to isolate the variables and ensure you are listening to the same master of a recording. At the least, you have to take the 24-bit file and convert it to lossless 16-bit / 44.1 kHz. Did you do that beforehand? I recommend using dBpoweramp. It does a perfect conversion. For the record, I listen to just about all types of music. Anyway, all this is beside the point of my question. What is your alternative to an ABX test for the purpose of proving you can hear a difference?

This is the thing - you can continue to downsample further and/or go more lossy and you probably won't be able to tell the difference easily and confidently.  I've failed listening tests all the way down to 4bit, trying to figure which was 8bit lossless, 192k MP3, 16bit wav, etc.
 
This is why MP3 is so successful. Most people "can't tell" or "don't care" with their conscious mind. That doesn't mean it doesn't exist.  It means our detection and repeatable identification abilities are lacking.
 
If you need to save space downsampling is one way. Lossy is another. These days they are often times both used one after the other to get music to the listener.
 
My whole point is you are only reducing for your enjoyment or convenience. You are knowingly degrading the files whether you think it's perceptible degradation or not. Degradation used to be very necessary to be digital and mobile. Now not so much.
 
My debates online are with people who say hi-res quality doesn't exist or is completely unnecessary, and therefore anyone who claims it's existence are idiots, or worse, scam artists.
 
Instead of "why can't I hear it?" or "Help me experience this" the internet is overrun with "snake oil!" "scam alert!" and "science says its not true" arguments. All bull.
 
This line of thinking is helping to kill music. Look around and listen to what's going on. No one talking, no one dancing, no one happy, constant mp3 trickle into their inner ear. No air, no life, all robot sex. OK not all of us, but you know I'm right. Lossy over compressed fake music is doing no one any good.

 

[FFBookman]

   
I completely agree that it can sound different when you get an album from two different sources. That's why it's essential to convert the files yourself to isolate the variables and ensure you are listening to the same master of a recording. At the least, you have to take the 24-bit file and convert it to lossless 16-bit / 44.1 kHz. Did you do that beforehand? I recommend using dBpoweramp. It does a perfect conversion. For the record, I listen to just about all types of music. Anyway, all this is beside the point of my question. What is your alternative to an ABX test for the purpose of proving you can hear a difference?

re: listening tests - I'm working on some formats myself, and have been reading up on other testing styles in use. There are a couple that have been published. This testing of people is not my field so i'm digesting their nomenclature, methods, and lessons learned in hopes of publishing mine some day.
 
Here's an overview of the one I'm cooking up:
New Listening Test – A Proposal | WFNK.COM
 
I was also thrilled to learn that Ayre audio, the hotshots that make $15k DAC/amp combos (and also the ponoplayer) have their own listening test model and rely on it almost exclusively to build their products. Their test is a much more involved listening session in open air with more than a snippet, actually more than a full song, before switching to a second sample. Their engineers are told to always go with the better sounding circuit regardless of what the specs show.
 
The proof of this model working is in their products, universally lauded for their pure and engaging sound quality. They ship audio products with very few dials, switches, or specs. The work has already been done to ensure it sounds good and customers trust the process of listening to every component.

 

[TJ Elite]

  …with more than a snippet, actually more than a full song, before switching to a second sample.

That's the way to do it.

 

[RRod]

Last I checked, ABX didn't have a time limit.