audio questions

[digitalwanderer]

Why should it matter if it's a blind test or not would be my question....

 

[thehulk]

http://en.wikipedia.org/wiki/Blind_experiment

Blinding is a basic tool to prevent conscious as well as subconscious bias in research. For example, in open taste tests comparing different product brands, consumers usually choose their regular brand. However, in blind taste tests, where the brand identities are concealed, consumers may favor a different brand.

Double-blind describes an especially stringent way of conducting an experiment, usually on human subjects, in an attempt to eliminate subjective bias on the part of both experimental subjects and the experimenters. In most cases, double-blind experiments are held to achieve a higher standard of scientific rigour.

 

[Simon F]

Any of the long-time regulars whom I know are welcome to PM me which ones they think are which, if they want to test their skills I will eventually post the answers in this thread, unless of course 2008 just nails it on the first try.

Until you have one that is recorded from vinyl, played through a valve amp and using gold plated silver speaker leads, you won't have a true representation.

 

[_xxx_]

Albuquerque, would you mind repeating this test with one of the tracks from the records I recommended? Say "Poison Godmachine" from Nevermore or Death "In Human Form" or "Trapped in a corner". Should be easy to find some free snippets online.

These would be much better for the comparison of this kind.

 

[thehulk]

There are always going to be tracks which are difficult to compress and these will always be (comparatively) easily abx'ed.

The point of this exercise is to refute bullshit claims like the following, NOT to find test cases for compression artifacts.

I don't know how many times i have to say this, but even though i have lousy speakers, I DO notice a great magnitude of difference between an mp3 and the same track in wma lossless. I'd like to know why people keep on telling me i can't notice something when I actually do. When people try to tell me that I can't notice a difference when I actually can, for no good reason, it kind of pisses me off. Please don't make some completely retarded incorrect assumption about me.

I mean after all, wma lossless wouldn't exist if the average music listener like me couldn't tell the difference on $50k stereos, when in fact people with $20 headphones can probably tell the difference between an mp3 on myspace and the same track downloaded from music giants, or a cd track ripped in mp 3 vs wma lossless.

 

[Albuquerque]

Until you have one that is recorded from vinyl, played through a valve amp and using gold plated silver speaker leads, you won't have a true representation.

Pfft! Until you sit in the ampitheater and listen to it live, you won't have a true representation

_xxx_ I can look around, but I doubt the samples I'll find online are uncompressed...

 

[_xxx_]

Maybe a friendly user can forward you a few short clips? I would, but I'm offline at home at the moment due to provider switch. Anyway those are pretty extreme in terms of dynamics and would make the potential differences much easier to spot.

 

[Albuquerque]

I have media with extreme dynamic range changes already. I wouldn't have to go out of my way to obtain them either -- say, about 50 yards to my car (at the office, far less distance to the car at home.)

 

[_xxx_]

Not only the dynamic range, but also many tracks and bluntly said "lots of sound" with many "strong" sounding instruments fighting for space in the spectrum. Something that's already mastered with lots of compression, Death Metal stuff or Slayer, Slipknot etc. is usually great for that. Convert that to a 320 kbit mp3 and you can literally hear some stuff missing, like gentle cymbals or certain guitar harmonics. Though only sensitive ears will notice that, but then again I'm a musician and thus have a bit less tolerance.

 

[Albuquerque]

Something that's already mastered with lots of compression, Death Metal stuff or Slayer, Slipknot etc. is usually great for that

What purpose would this test serve? Crap source + lossy compression = crap? Well yeah, crap in, crap out I don't think anyone will argue that using lossy compression against already over-compressed audio is going to do it any favors.

However, I think there's certainly merit in finding a fantastically mastered track with lots of dynamic range and a wide variety of instruments to demonstrate the potential weaknesses of these types of compression. In other words, in my opinion, it's not fair to pin "bad audio" on Lossy Compression when the source was shit to begin with -- but it's more than fair to pin "bad audio" on Lossy Compression when the source was good.

 

[Sonic]

Meh. I'll try to rip some of my vinyl collection of Mozart and Beethoven records to FLAC later this weekend. That might help some of you who wish to test your systems with dynamic range or whatever tests you wish to do. If you have a specific piece you would like me to rip for you I will provided I have it. Most of these were recorded from the San Francisco Symphony throughout the years so I do dare say the source audio is one of decent quality.

And yes, nothing is quite like hearing music live.

 

[Bambers]

What purpose would this test serve? Crap source + lossy compression = crap? Well yeah, crap in, crap out I don't think anyone will argue that using lossy compression against already over-compressed audio is going to do it any favors.

However, I think there's certainly merit in finding a fantastically mastered track with lots of dynamic range and a wide variety of instruments to demonstrate the potential weaknesses of these types of compression. In other words, in my opinion, it's not fair to pin "bad audio" on Lossy Compression when the source was shit to begin with -- but it's more than fair to pin "bad audio" on Lossy Compression when the source was good.

Well it doesn't have to be a crap source as such, there's plenty of music styles which are a great deal more 'full' that a fairly slow piano piece, they don't have to be ones where the DR compression has been overapplied.

 

[_xxx_]

What purpose would this test serve? Crap source + lossy compression = crap? Well yeah, crap in, crap out I don't think anyone will argue that using lossy compression against already over-compressed audio is going to do it any favors.

No, it means there is lots of data in there and more gets lost during the conversion. Thus it's much easier to discover stuff that's missing. Also notable is the general loss of highs and higher order harmonics.

 

[Gubbi]

Music with high entropy can have audible differences. An example is rock and roll with lots of distorting guitars, where cymbals start sounding distorted with a distinct hissing "ssss" sound instead of the crisp sound it should have. Stuff like Metallica, Slayer etc.

Another problem is music with transients, - short spikes, where the psycho acoustic model can cause pre-echo and smearing of the sound. Try to listen to Röyksopp or Trentemøller or similar electronia. This is mostly a problem with MP3s encoded with inferior encoders (or "fast" encoding).

Cheers

 

[_xxx_]

Music with high entropy can have audible differences. An example is rock and roll with lots of distorting guitars, where cymbals start sounding distorted with a distinct hissing "ssss" sound instead of the crisp sound it should have. Stuff like Metallica, Slayer etc.

Exactly, thus I recommended the records mentioned previously - I have these as mp3 in my car and that with cymbals and missing harmonics really annoys me.

 

[L233]

Exactly, thus I recommended the records mentioned previously - I have these as mp3 in my car and that with cymbals and missing harmonics really annoys me.

You notice that with all the background noise while driving the car?

Anyway, I'll upload some death metal in a few minutes.

 

[L233]

Right, so _xxx_ wants Death Metal.

44 seconds of "Thousand Years of Oppression" by Amon Amarth, ripped with EAC using the same CD drive as Albuquerque (SH-S203B). The files were compressed from within EAC using LAME 3.98.2. I also used EAC to convert the files back to PCM.

Sample 1: uncompressed
Sample 2: CBR MP3 320 kb/s
Sample 3: VBR MP3 192 kb/s
Sample 4: CBR MP3 128 kb/s


I've uploaded the .zip file (30 mb) to different free file hosters, use whichever works best for you:

http://www.przeslij.net/download.php?file=qtest.zip
http://www.file-upload.net/download-1187230/qtest.zip.html
http://www1.zippyshare.com/v/1379222/file.html


And here's some Thrash Metal. 28 seconds from Destruction's "The Final Curtain". Everything same as above:

http://www.przeslij.net/download.php?file=qtest2.zip
http://www.file-upload.net/download-1187369/qtest2.zip.html
http://www6.zippyshare.com/v/52187944/file.html


... and one more. Prog Rock - 32 seconds "New World Order" by Shadow Gallery:

http://www.przeslij.net/download.php?file=qtest3.zip
http://www.file-upload.net/download-1187551/qtest3.zip.html
http://www.zippyshare.com/v/72733390/file.html


Ok, last one for now. Nevermore - "Sanity Assassin", 34 seconds:

http://www.przeslij.net/download.php?file=qtest4.zip
http://www.file-upload.net/download-1187719/qtest4.zip.html
http://www10.zippyshare.com/v/94848308/file.html

 

[Davros]

For the sake of interest i took Axkfa.wav from page 4
and re-recorded it as wav stereo 48000khz 24bit (sorry didnt realise the original wasnt recorded axactly the same)
using x-fi's Crystaliser which is supposed to make compressed audio sound better
what does everyone think ?
http://media.putfile.com/Crystaliser

 

[Moloch]

It was discovered shortly after the X-Fi came out that Crystaliser is mearly a [FONT=Arial, Helvetica]multi-band compresso[/FONT]r.
http://www.behardware.com/articles/586-6/creative-labs-x-fi.html

 

[Bambers]

We supposed to post answers here?