The scientific merit of Pono
Sep 22, 2014 at 6:55 PM Post #106 of 318
The real question, which for some reason people who do tests don't seem to be interested in finding out, is where the line of full transparency lies. At what point does the sound no longer improve with any kind of music being tested? I've done the test myself, so I know, but I don't think a lot of audiophiles do. They think more is better forever into eternity.
 
Sep 22, 2014 at 8:46 PM Post #108 of 318
  Yes. Even I can tell the difference between a 128kbps file and a 320kbps one, and I think it's already been established that I don't have golden ears.

heh~!! me too!!!!! i am quite proud of being able to tell the difference between sub-320kbps vs 320kbps :D years of dedication & training man *cue to rocky-esque training montage*
 
most everything else in this hobby that ppl talk about i usually have a hard time hearing eg. most dacs sound about the same to me =(
 
but it wld b pretty neat to have golden ears!! ...tho I bet it wld b hard to survive w/ all the crap sound quality on youtube & most other streaming sites.
 
Sep 22, 2014 at 9:34 PM Post #109 of 318
  The real question, which for some reason people who do tests don't seem to be interested in finding out, is where the line of full transparency lies. At what point does the sound no longer improve with any kind of music being tested? I've done the test myself, so I know, but I don't think a lot of audiophiles do. They think more is better forever into eternity.

 
I actually purchased the Sammy Davis, Jr. The Decca Years CD at Amazon based on your observations from this testing.  
 
While I did not get a pressed CD, Amazon does create a copy on a CD-R.  I did find some sections where I was able to hear artifacts, but at 128kbps mp3 (Lame)!  Though, I did not do an exhaustive test on the entire CD.  I was attempting to play sections with massed strings, as you described this in your post. (Six Bridges to Cross?)  I have been able to successfully ABX a live album with audience applause at 192kbps mp3 files vs FLAC.  I believe it was something from Simon & Garfunkel.  I did not bother testing 256kbps, but at 320kbps I could not identify a difference between the mp3 and the FLAC ripped from the same CD.
 
Sep 22, 2014 at 9:57 PM Post #110 of 318
  I don't have Audacity, but in Music Bee the bitrate of the HD files comes up at just over 3000 kbps, whereas the flac files from my cd are just under or over 1000. Each album also takes up over 1 gig of space, which is 4-5 times that of most albums in flac format, and about 3 times that of an uncompressed wav file, so I'm pretty sure I got what I paid for (or maybe I should say I got what they said I was getting). In the end it was a cheap way to find out that I don't need to spend a bunch more $. And I have no idea why the sound would be different in the way I reported. Since I can't reliably tell them apart, for all I know it was in my head.

You certainly paid for and received the 24/96kHz box.  Whether you got any Hi-Rez files in that box is a different issue.
 
Audacity is free:  http://audacity.sourceforge.net/  There are versions for Windows, MacOS and Linux.  It's a lot smaller download than your albums were, too!
 
I'm no expert in how to use it, perhaps BigShot or Ethan Winer could post a few pointers, but for the purpose of determining whether your files are truly Hi-Rez I can coach you through the steps.
 
Download, install, and start Audacity.  Click past the "Where is Help?" pop up.
 
From the "File" tab choose "Open..." and navigate to and open one of your Hi-Rez files. 
 
While the song is stopped or paused, from the "Analyze" tab choose "Plot Spectrum..."  If a "Too much audio was selected" dialog box pops up just click on "OK."  A "Frequency Analysis" window should pop up.
 
If it's a RedBook file, the frequency plot should go out to 22kHz with meaningful signal out to or very near 20kHz.
 
If it's a true "recorded at 96kHz" file, the frequency plot should go out to 50kHz with meaningful signal out at least to 40kHz or so.
 
If it's a RedBook file hiding in a 96kHz box, the frequency plot should go out to 50kHz but meaningful signal will only go out to 22kHz or so and drop off a cliff after that. 
 
Sep 22, 2014 at 10:12 PM Post #111 of 318
I'm afraid I can't help with that because whenever I get a file that is 24/96, I knock it down to redbook then encode to AAC 256 VBR. I don't mess with high bitrate / high sampling rate files on my music server.
 
Sep 22, 2014 at 10:20 PM Post #112 of 318
I'm just curious how much high frequency information can be found in a 20 year old recording, and especially in a 40 year old recording.  Even assuming recording at 24/96 is valuable, remastering to that level cannot create signal that isn't there in the original.
 
Sep 23, 2014 at 12:11 AM Post #113 of 318
There isn't much high frequency information in modern recordings either. They don't want lots of high frequencies. That will end up distorting in homes with stereos not designed to play it.
 
Sep 23, 2014 at 1:17 AM Post #114 of 318
  There isn't much high frequency information in modern recordings either. They don't want lots of high frequencies. That will end up distorting in homes with stereos not designed to play it.


Aren't those extra high frequencies the entire point of Hi-Rez files?  I'll agree we probably can't hear them, but the high frequencies are certainly there in a proper Hi-Rez file.  I too wonder what all that excess high frequency energy could do to marginal amps or tweeters.   Not to mention your poor pets!
 
Correct Hi-Rez file:
 

 
The same Hi-Rez file downsampled to RedBook but placed in a Hi-Rez container:
 

 
As I said before though, I'm curious whether 20 or 40 year old albums were recorded with that much high frequency energy.  If yes, how on earth did they do it?  If not, then what's the point of releasing them in Hi-Rez other than to scam money out of people?
 
Sep 23, 2014 at 1:45 AM Post #115 of 318
a lot of the high frequency info is noise.
 
Sep 23, 2014 at 2:00 AM Post #116 of 318
You want to see some noise?
 

 
I can't imagine that  peak out at 43kHz is a harmonic. 
 
And to add to the irony, here is what the same song looks like in Audacity:
 


Not much compression, is there?  Someone took some care when recording and mastering this.
 
Sep 23, 2014 at 2:35 AM Post #117 of 318
Just crank the volume and get a free ear piercing.
 
Sep 23, 2014 at 10:26 AM Post #118 of 318
  The real question, which for some reason people who do tests don't seem to be interested in finding out, is where the line of full transparency lies. At what point does the sound no longer improve with any kind of music being tested? I've done the test myself, so I know, but I don't think a lot of audiophiles do. They think more is better forever into eternity.

 
For me, it's not that do/don't care, it's that I don't want to spend the time (maybe that means I don't care enough
rolleyes.gif
)
 
In my experience, it's difficult to find tracks or portions of tracks that allow me to prove to myself where the line of transparency lies.  So, in other words, it's easy to find sections that are transparent (even at relatively low bit rates), but every once in a while I find a portion of a track that really sounds different when encoded to AAC from a lossless copy.  It's usually a section that displays a change in audio imaging/positioning of instruments (or lack thereof).
 
But, none of this really matters, as storage is basically free now (4 TB HDD for $150.00 ... wow) so why not just rip your CDs as lossless files and forget about it?  Even when I am out and about, network bandwidth is also basically free (my mobile routinely clocks in at 20+ Mbps data streaming to the phone).  So, I stream those same lossless rips from home to my Android phone using BubbleUPnP Server as 44.1/16 FLAC and just forget about it.
 
I'm not at all disputing the sonic issues of transparency between lossless Redbook and AAC/256 or LAME/320 - I listen to these, as well, and they sound perfectly great.
 
Sep 23, 2014 at 11:03 AM Post #119 of 318
My journey for identifying transparency was a long process.  
 
I had no reason not to believe all the hype, as the industry is flooded with anecdotal evidence to suggest that sound quality can be improved from burn-in, expensive cables and interconnects, high resolution files, SACD, etc.
 
Always curious, I wanted to see if I could successfully ABX between lossy and lossless formats.  I read through the steps to do this with Foobar, and when I was unable to tell a difference, I had to make sure I was actually setting up the test correctly.   Once I was confident that my test procedure was good, I ripped my CD collection to Lame VBR 0 mp3, following guidelines to make sure I was encoding correctly for the best results. (used Exact Audio Copy and, later, dbPowerAmp)
 
As I was listening to a song, if I thought I heard an artifact or something odd about the playback, I would grab the CD, rip to FLAC, and ABX the files in Foobar using the steps to ensure the tracks were volume leveled.  What I found, every time, was that whatever I was hearing in the mp3 file, I was also hearing the same thing on the CD.  Initially this would occur quite frequently, though I could never identify a difference between my mp3 files and the FLAC rips.  My skepticism eventually waned to the point where I just didn't bother checking anymore.
 
I decided that even if there were a few sections in a small number of songs where I might be able to barely hear a difference quickly switching between files in an ABX test, I most likely would never be able to tell in a normal listening environment.  So now I just listen to music ripped to a well encoded mp3.  I'd prefer to use AAC, but my solution with Google Music play trumps this, as AAC is re-encoded to mp3 by Google when uploading.
 

Users who are viewing this thread

Back
Top