[xnor]

Thanks, that sample seems to be of higher quality. So far my ABX results aren't looking that good. 

 
 
edit: 10 sec samples up at netload, rapidshare (So others can ABX too, also renamed both files so anyone can guess which is the mp3. Just select both files in foobar2000, right-click, Utilities - ABX Two Tracks ...)

 

[Hybrys]

Quote:

Thanks, that sample seems to be of higher quality. So far my ABX results aren't looking that good. 

 
 
edit: 10 sec samples up at netload, rapidshare (So others can ABX too, also renamed both files so anyone can guess which is the mp3. Just select both files in foobar2000, right-click, Utilities - ABX Two Tracks ...)

They would need the ABX component from Foobar's website aswell.
 
I'd be interested in seeing others' ABX logs aswell.

 

[zenpunk]

Ok. I have done the test and I am not sure what to conclude but in short I believe there must be problem with the way the files were normalised or/with the encoding as the two files sound different in an obvious way. The a sound  fuller and has more presence. I stopped before the 25 attempt because the difference was just too obvious on my system . I ve done similarly tests months ago when I only had the STX and I coundn't tell the dirrence between 320 and Flac. The only new equipment is the Concerto, which a very detailed and enjoyable amp. Could that explain the difference? I am busy this week but will test further with my own rips to find out if there is actually audible differences between 320 and lossles. Might have to eat my own words

 

[xnor]

Oh I nearly forgot about these samples..
 
zenpunk, I'll take a closer look at the samples I uploaded if there is something wrong with them, but there shouldn't be an obvious difference.
Also please post the ABX log so we can see that you really heard an obvious difference, thanks.

 

[chinesekiwi]

Blah, people need to frequent the Sound science forum more!
Too many false statements!
 
1) lossy doesn't just affect the higher frequencies. It 'swiss-cheeses' the middle frequencies as well, particularly mp3.
As much as I like the theory of ABX, the way it's conducted by most people is flawed. I would rather trust things such spectrograms to prove my point.
 
I'll quote myself from another thread. There is minimal difference between 320 AAC and FLAC. AAC at ~192kbps is better than VBR -V0 / 320kbps CBR mp3.
Vorbis -q6 = ~192kbps.
 
[size=medium]

Quote:

Quote:

[size=inherit]

Originally Posted by Chef /img/forum/go_quote.gif They're all so bloated it isn't going to matter at all. Kiwi, please show me the 'science' (ie the study) that suggests this ogg q-6 is better than mp3 320... apart from you know, being a smaller file obviously.

[/size]

Spectrograms of Oasis' 'Roll With It' LAME 3.98.4 VBR -V0 mp3 vs. AoTuV 5.7 Ogg Vorbis @-q6. Viewed using the freeware spectrogram analyser Spectro. The -q6 Vorbis example is converted to WAV to work with Spectro, otherwise it is 100% representative of the Ogg Vorbis -q6 sample. I've used a GIF to illustrate the point more easily.
I chose this song as this song is brickwalled due to the Loudness War thus good for comparing the loss of data via lossy formats. All pictures were saved in .PNG format to ensure picture quality and integrity (.PNG = lossless).

and the original FLAC to compare:

aa

Not only does Ogg Vorbis at -q6 represent more data, it represents it more accurately. A good clear example is the two pairs on red lines near the middle of the spectrogram clearly showing in the Vorbis sample but not the mp3 sample.

Also to note from Spectrogram Reading - What are spectrograms? to read spectrograms properly.

Quote:

[size=inherit]

....the horizontal axis shows positive time toward the right, and the colors represent the most important acoustic peaks for a given time frame, with red representing the highest energies, then in decreasing order of importance, orange, yellow, green, cyan, blue... ....and white areas [Black in Spectro] below a threshold decibel level.

[/size]

​

[/size]

 

[xnor]

Since when is it possible to hear differences between two files by looking at a spectrogram?

 

[Ham Sandwich]

Quote:

Since when is it possible to hear differences between two files by looking at a spectrogram?

Since I got the Fourier transform module implanted in my brain I have been able to ABX lossy vs. lossless audio with near 100% accuracy.  There are still some samples though that give me trouble. 

 

[sidewinder]

I have a Wharfedale Diamond 10.2 with Asus Xonar STX setup and to me it's very clear the difference between 320kB MP3 rips versus FLAC or WAV. It was quite inaudible with my Realtek ALC889a onboard chip though. You gain more transparancy and dynamics. Just to note, in the past I had premium computer speakers like the Altec Lansing FX6021 and Harman/Kardon Soundsticks II. With those setups I didn't hear much of a difference, besides a little more clarity. But with my current setup, it's there really.
 
On the other hand, with my portable setup, I can't hear any difference

 

[xnor]

@sandwich: *chuckles*
 
@sidewinder: So what are you waiting for, fire up foobar and show us the results! 

 

[Hybrys]

Quote:

@sidewinder: So what are you waiting for, fire up foobar and show us the results! 

This!  ABX those samples (or I can send you the link to the full song) and post your results!

 

[chinesekiwi]

Quote:

Since when is it possible to hear differences between two files by looking at a spectrogram?

Eliminates all the variables that ABX offers.

 

[Hybrys]

Quote:

 
Eliminates all the variables that ABX offers.

It also removes the entire point of ABXing.  People aren't asking if a difference exists, they're asking if it is audible to them.  You can show them graphs and comparisons all you like, but it isn't going to tell them what they hear.

 

[xnor]

Quote:

Eliminates all the variables that ABX offers.

Including the variable that is called hearing, which is required to be able to actually listen to those tracks. Nice...
 

 

[Lil' Knight]

 
Quote:

Too many false statements!
 

Indeed. Including post #50.
 

 

[chinesekiwi]

Quote:

 
Indeed. Including post #50.
 

 
How, mine's scientific proof.....
mp3, AAC and Ogg Vorbis all use different compression algorithms and thus why AAC and Ogg Vorbis represent more data and the data with more accuracy (thus better sound quality) than mp3.
There is a reason why AAC is industry standard and not mp3.
It's easily replicated..... not hard.