Quantcast Best Audio Extractor - Page 3 - digitalFAQ.com Forums [Archives]
  #41  
09-03-2004, 02:12 PM
jorel jorel is offline
Invalid Email / Banned / Spammer
 
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
from Beatallica - Metallica singing Beatles music "hey jude"

HEY DUDE

Hey, dude-it'z true not sad
Take a thrash song and make it better
Remembah! That metal iz in your heart
Then you can start to be a fretter

Hey, dude-don't be fuckin' 'fraid
You were made to go be a shreader
The minute you let us under your skin
Then you'll begin to be a fretter

So crank your amp and deal the pain
Hey, dude-you're fuckin' insane!
The riverz run red with blood of poseurs
And don't you know that he'z the fool
Who playz it cool
But needz for hiz beer to be much colder

Hey, dude-nevah turn it down!
You must pound her, I mean Kip Winger
New wave of British heavy metal iz in your heart
And you can start with Diamond Headerz

So let it out! Let it in!
Hey, dude, begin
Don't wait for the Eye of the Beholder
You'll never know when bellz toll for you
Hey, dude, you'll do
Just sling that flying-V 'cross your shoulder

Hey, dude-it'z true not sad
Take a thrash song and make it better
Admit it! Beatallica'z under your skin!
So now begin to be a shreader

:P Beatallica is like eac + ogg....quality over quality
Reply With Quote
Someday, 12:01 PM
admin's Avatar
Site Staff / Ad Manager
 
Join Date: Dec 2002
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
  #42  
09-03-2004, 06:02 PM
kwag kwag is offline
Free Member
 
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
Where the hell you get that jorel

Ok, here's the real test, comparing original WAV file to a 45Kbps Ogg.

WAV: http://www.xiph.org/ogg/vorbis/liste..._minus_one.wav
Ogg: http://www.xiph.org/ogg/vorbis/liste..._minus_one.ogg

Any difference
That's 2.84 times lower in bitrate than a 128Kbps MP3
I can't tell the difference. They sound the same

-kwag
Reply With Quote
  #43  
09-03-2004, 08:46 PM
kwag kwag is offline
Free Member
 
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
Hi guys

I had to do my own exhaustive test, instead of taking someone elses word for it.
So here it is,
I cut a 30 second WAV, extracted from a very complex musical CD.

Here are the files, including the uncompressed WAV file:

1) Uncompressed: http://www.kvcd.net/test-uncompressed.wav [ 5,224KB ]

2) MP3 at 64Kbps CBR: http://www.kvcd.net/test-64kbps-cbr.mp3 [ 238KB ]

3) MP3 at 128Kbps CBR: http://www.kvcd.net/test-128kbps-cbr.mp3 [ 476KB ]

4) MP3 using -alt-preset=standard: http://www.kvcd.net/test-alt-preset-standard.mp3 [ 769KB ]

5) Ogg Vorbis using Q -1.0 (~45Kbps): http://www.kvcd.net/test-Q-minus-1.0-45kbps.ogg [ 165KB ]

6) Ogg Vorbis using Q 0 (~64Kbps): http://www.kvcd.net/test-Q-0.0-64kbps.ogg [ 286KB ]


In this particular case, I can hear a very small difference in the 45Kbps sample.
The difference is so small, that if I hadn't had the original CD, I couldn't have been able to tell the difference.
Now go ahead and compare the MP3s, and you'll definitely hear the difference on the 64Kbps and 128Kbps sample.
On the --alt-preset-standard, it sounds as good as the original, but the file size says it all.
The 64Kbps Ogg sample is just as good as the WAV, but weighting only 286KB in size against the best MP3, which is 769KB

-kwag
Reply With Quote
  #44  
09-04-2004, 03:46 AM
muaddib muaddib is offline
Free Member
 
Join Date: Jun 2002
Location: São Paulo - Brasil
Posts: 879
Thanks: 0
Thanked 0 Times in 0 Posts
Hi karl,

Well, I also did my exhaustive test. By that I mean ABX tests... 10 trials per test.
First I want to say that 45kbps Marc Nelson sample is amazing! And it was completely transparent for me!

Now my test...
First track - Marillion - Anoraknophobia - Between You and Me
Second track - Van Halen - Balance - Don't Tell Me (What Love Can Do)

My goal was to determine what would be the transparent bitrate for me with ogg vorbis with these two tracks. When I say transparent I mean totally transparent. If I can distinguish one tiny little frequency between the source and the encoded file then it’s not transparent.

I spent more than an hour listening to the same sample. My ears are aching!
As you said, I used oggenc.exe version 1.0.1 for the encodes.
And used foobar2000 v0.8.3 for the ABX test.

So head up for the test:

Quality 0:
I got a 10/10 score with the two tracks.
It was not difficult to distinguish the two files.
BTW, a 10/10 score means that I have 0.1% of chance to be guessing the tracks.
(to get a 10/10 you really have to be hearing differences between the files)

Quality 1:
Again I got a 10/10 score with the two tracks.
The Van Hallen track was a bit more difficult, but clearly different.
The Marillion track was easy again.

From this point I decided to stop testing Van Hallen (that was more difficult to distinguish) and concentrate only with the Marillion track. (my ears were starting to complain).

Quality 2:
Again I got a 10/10 score, but only test the Marillion track.

Quality 3 – Quality 4 – Quality 5 – Quality 6:
All them I was able to distinguish 10/10.
Q5 and Q6 were really difficult to tell the differences, but there ARE differences.

Quality 7:
That was the point where I could not distinguish the encode from the source.
I started the test but stopped at 2/5 score… I was really guessing and was not able to tell any difference.

I also test MP3 Lame –alt-preset standard and could not tell the difference from the source.
Them I tested Lame –alt-preset 192 and again I could not distinguish the encode from the source.

So my final conclusion is that for achieve transparency with this particular Marillion track, I need higher bitrates with OGG (q7 = 224kbps) than with MP3 (alt-preset 192 = 192kbps).

This DOES NOT prove that MP3 is better than OGG. We have many different styles of music, and the results could be completely different with them.
And as I said before, transparency is a subjective matter. And I think that OGG could be much more efficient than MP3 at lower bitrates… But at higher bitrates, reaching for transparency with this particular track, MP3 wins.

Conclusion of the conclusion… for now, I’ll stick to MP3. It’s a great codec and is compatible with all players.


Here is the log from the last test (-q 6):
Code:
foo_abx v1.2 report
foobar2000 v0.8.3
2004/09/04 04:13:53

File A: file://E:\Music\Ripping\teste vorbis\Marillion - 01 - Between You and Me.wav
File B: file://E:\Music\Ripping\teste vorbis\Marillion - 01 - Between You and Me_q6.ogg

04:14:06 : Test started.
04:14:54 : 01/01  50.0%
04:15:06 : 02/02  25.0%
04:15:16 : 03/03  12.5%
04:15:21 : 04/04  6.3%
04:15:30 : 05/05  3.1%
04:15:36 : 06/06  1.6%
04:15:41 : 07/07  0.8%
04:15:59 : 08/08  0.4%
04:16:05 : 09/09  0.2%
04:16:33 : 10/10  0.1%
04:16:39 : Test finished.

 ---------- 
Total: 10/10 (0.1%)
PS: If any one wants the samples tested, I’ll post them here.
Reply With Quote
  #45  
09-04-2004, 10:05 AM
kwag kwag is offline
Free Member
 
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
Hi muaddib,

Great tests
I'll do some more testing on more mixed music material.
I do think that low bitrate Oggs reproduces higher frequencies much better than MP3s.
But as you say, maybe on high bitrates, the MP3 reproduces the audio more transparently, and Ogg has some "unnatural" or artifact reproduction.
I have to make more tests with Ogg, because the compression is just too awesome.
BTW, could you cut a small WAV from "Marillion - Anoraknophobia - Between You and Me"
I'd like to test that, and I don't have that CD.

-kwag
Reply With Quote
  #46  
09-04-2004, 12:45 PM
jorel jorel is offline
Invalid Email / Banned / Spammer
 
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
muaddib and his jbls!

...yes, send the source (waves) too...we need to test "everything".
this thread is encreasing in good taste!
Reply With Quote
  #47  
09-04-2004, 01:47 PM
kwag kwag is offline
Free Member
 
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
Here's another site with more comparisons.
The uncompressed test file is VERY complex in frequency spectrum, so it's a good source for encodings: http://www.xciv.org/~meta/audio-shootout/

-kwag
Reply With Quote
  #48  
09-04-2004, 10:40 PM
jorel jorel is offline
Invalid Email / Banned / Spammer
 
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
hy boys...i had one idea (is too rare)

i have one gentle giant cd- the power and the glory.one music from this cd have a magnific introduction with great basses, strings and solos.
i can cut some seconds (sound forge as wave)and send to do tests.
another great option is from santana-abraxas(remastered).the first music have very cool effects(basses and strings)....is another option.
another great is king krimson- 3 of a perfect pair.
.....or i playing acoustic guitar (with a hammer-i love metal....hammer is metal)!
cool?...or lame?
if "lame", choose another source (maybe i have)!
Reply With Quote
  #49  
09-04-2004, 11:28 PM
kwag kwag is offline
Free Member
 
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
How about some 30 second WAV cuts jorel
The ones with "strings" and high frequencies are the best to test, because on low frequencies, just about all Codecs do a good job. It's on the high frequency spectrum where they choke, specially 128Kbps MP3's, that sound like "tin cans" or "under water" effects

But I agree that MP3's over 192Kbps are just like the original (to the ears).
But I like the Ogg better, because it seems it hides the artifacts WAY better than other Codecs, and even when there are small differences compared to the original, it still sounds excelent, where a MP3, WMA, AAC just sound completely horrible at 45Kbps

-kwag
Reply With Quote
  #50  
09-04-2004, 11:55 PM
jorel jorel is offline
Invalid Email / Banned / Spammer
 
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by kwag
How about some 30 second WAV cuts jorel
The ones with "strings" and high frequencies are the best to test, because on low frequencies, just about all Codecs do a good job.
ok i will send to your mail ! (the basses from gentle giant have magnific harmonics too- seems sintesized with vcf... are really cool)

Quote:
Originally Posted by kwag
It's on the high frequency spectrum where they choke, specially 128Kbps MP3's, that sound like "tin cans" or "under water" effects
true!

Quote:
Originally Posted by kwag
But I agree that MP3's over 192Kbps are just like the original (to the ears).
But I like the Ogg better, because it seems it hides the artifacts WAY better than other Codecs, and even when there are small differences compared to the original, it still sounds excelent, where a MP3, WMA, AAC just sound completely horrible at 45Kbps

-kwag
true too! my mp3 are in 192 or more!
i'm witness that in low bitrates, vorbis is winner.
for high bitrates deserve more tests(i never did)...we will do it!
Reply With Quote
  #51  
09-05-2004, 02:25 AM
muaddib muaddib is offline
Free Member
 
Join Date: Jun 2002
Location: São Paulo - Brasil
Posts: 879
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by kwag
I'll do some more testing on more mixed music material.
Great kwag! Try to make some ABX tests... Foobar2000 (or a direct link to the special installer) has a good, and really easy to use, tool to make this kind of test.

Quote:
I do think that low bitrate Oggs reproduces higher frequencies much better than MP3s.
And I agree with you!

Quote:
BTW, could you cut a small WAV from "Marillion - Anoraknophobia - Between You and Me"
I'd like to test that, and I don't have that CD.
Before post the samples, I need to say that today afternoon I tried to duplicate the test of last night with these samples. Until -q5 was the same, but I could not pass through -q6 (like last night).
I'm not sure why this happened... may be the louder noises from the day had distracted or confused me. Or maybe the "save selection" function from Adobe Audition (with I used to trim the wav) is not lossless... I don't know.

Here are the samples (if left click don't work, try right click->save as):
-> wav sample - 2.88MB RAR

-> ogg -q0 - 194KB
-> ogg -q2 - 281KB
-> ogg -q4 - 382KB
-> ogg -q5 - 489KB
-> ogg -q6 - 585KB
-> ogg -q7 - 676KB

-> mp3 alt-preset 192 - 625KB
-> mp3 alt-preset standard - 657KB

If you are going to make the ABX test... pay attention to the high frequencies at the beginning of the drums. That's where I find most of the differences.
Reply With Quote
  #52  
09-05-2004, 02:39 AM
jorel jorel is offline
Invalid Email / Banned / Spammer
 
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by muaddib
is not lossless... I don't know.
:P are lossless or not?!?!?...just listen!

thanks muaddib!
i got the foobar( fubá? )and all samples to listen in the morning...my kid is sleeping!
Reply With Quote
  #53  
09-05-2004, 02:47 AM
Boulder Boulder is offline
Free Member
 
Join Date: Sep 2002
Location: Lahti, Finland
Posts: 1,652
Thanks: 0
Thanked 0 Times in 0 Posts
Has anybody tested MusePack (.mpc) at low bitrates? Sounds better than MP3 at avg 128kbps. I tried searching for a low-bitrate listening test at hydrogenaudio.org but it looks like there's yet to be one. There's discussion about a dial-up bitrate listening test but the test isn't probably done yet.
Reply With Quote
  #54  
09-05-2004, 02:56 AM
muaddib muaddib is offline
Free Member
 
Join Date: Jun 2002
Location: São Paulo - Brasil
Posts: 879
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by kwag
But I like the Ogg better, because it seems it hides the artifacts WAY better than other Codecs, and even when there are small differences compared to the original, it still sounds excelent, where a MP3, WMA, AAC just sound completely horrible at 45Kbps
I totally agree with you... again.
When I say that I can distinguish the encoded file from the source, in no way I'm telling that the encode was bad. Far from it! OGG sound quality at low bitrates is amazing! Thought I must say that (for my taste) I found 45kbps and 64kbps just too low. At that kbps the music sounds kind of artificial to my ears... But that feeling disappeared at around -q2 (~96kbps).

I noticed that the recomended encoder from HA for that low bitrate is differente from the one we are using (the standard Xiph.Org encoder). Look here. May be we can squeeze a bit more quality with that one at low bitrates... I'll check that later.
Reply With Quote
  #55  
09-05-2004, 03:10 AM
muaddib muaddib is offline
Free Member
 
Join Date: Jun 2002
Location: São Paulo - Brasil
Posts: 879
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by Boulder
Has anybody tested MusePack (.mpc) at low bitrates? Sounds better than MP3 at avg 128kbps. I tried searching for a low-bitrate listening test at hydrogenaudio.org but it looks like there's yet to be one. There's discussion about a dial-up bitrate listening test but the test isn't probably done yet.
http://www.hydrogenaudio.org/forums/...howtopic=24311
It's done. And looks like the winner is Nero HE-AAC.
Here are the complete results of the dial-up bitrate (32kbps ).

And here are the results of the 128kbps test. OGG Vorbis (aoTuV encoder) was the winner.
Followed by MPC in a close second place.
Reply With Quote
  #56  
09-05-2004, 03:14 AM
Boulder Boulder is offline
Free Member
 
Join Date: Sep 2002
Location: Lahti, Finland
Posts: 1,652
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by muaddib
http://www.hydrogenaudio.org/forums/...howtopic=24311
It's done. And looks like the winner is Nero HE-AAC.
Here are the complete results of the dial-up bitrate (32kbps ).
Hehe, now I know what a newbie feels like I don't understand how the search didn't find that page even though I used the words dial-up bitrate listening test in it
Reply With Quote
  #57  
09-05-2004, 08:20 AM
rds_correia rds_correia is offline
Free Member
 
Join Date: Apr 2003
Location: Chinese Democracy starts now!
Posts: 2,563
Thanks: 1
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by Boulder
Hehe, now I know what a newbie feels like
__________________
Rui
Reply With Quote
  #58  
09-05-2004, 03:13 PM
jorel jorel is offline
Invalid Email / Banned / Spammer
 
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
marillion samples from muaddib

q0- loose extreme trebles and show simbilances

q4- encrease the middles and add "air" in trebles(a little artificial)

q7- really better(not perfect but good)...loose extreme trebles(just a few)

alt-preset_192- loose basses...hey, i love basses..no more comments for this sample!

alt-preset_standard- losse basses too....no comments!

source- of course, better general atmosphere!

about basses-we can listen more harmonics if the basses stay "proud" then in that parameter all oggs are better than mp3 ...for my taste of course!

ok.....using the source is good to compare...the general results are great, i only wrote my bored opinion!
maybe i change my username to Bore-L

thanks muaddib, great samples! i could "see the music in the sky" too! can you send the full wave?

@ Kwag
i will send my samples, they have more details in bass-mid-treble!
Reply With Quote
  #59  
09-05-2004, 05:15 PM
kwag kwag is offline
Free Member
 
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
I really don't know what weed those who did the test (AAC) at Hydrogen audio were smoking, but my results are EXTREMELY different

I can't encode at 32Kbps with Ogg, because the command line encoder only lets me encode at 40Kbps minimum.
So I decided to encode a test clip, at 40Kbps, both with Nero's AAC encoder and with Ogg encoder.

I don't think this is a subjective comparison, because ANYONE can clearly hear the worse cip.

Hear for yourself:

The original, which is the same WAV I previously uploaded is this:
http://www.kvcd.net/test-uncompressed.wav

AAC sample: http://www.kvcd.net/test-aac-40kbps.mp4

Ogg sample: http://www.kvcd.net/test-ogg-40kbps.ogg

Now, if anybody says that the AAC sounds better than the Ogg, please let me know, because it means I must see a doctor



-kwag
Reply With Quote
  #60  
09-05-2004, 10:36 PM
jorel jorel is offline
Invalid Email / Banned / Spammer
 
Join Date: Aug 2002
Location: Brasil - MG - third stone from the sun
Posts: 5,570
Thanks: 0
Thanked 0 Times in 0 Posts
@ Kwag

you (we) don't need a doctor,
without doubts, ogg is much better in your samples
Reply With Quote
Reply




Similar Threads
Thread Thread Starter Forum Replies Last Post
Easy CD-DA Extractor 6.5 vdk_au Audio Conversion 7 02-15-2004 10:34 AM
Audio Codecs: LSAMP (leon's simple audio mastering program) jorel Audio Conversion 0 11-18-2003 11:10 PM
Audio Codecs: Audacity - cool and free audio editor jorel Audio Conversion 0 10-23-2003 12:19 AM
Audio Codecs: AVI2WAV Extractor jorel Audio Conversion 0 09-30-2003 03:45 PM
Audio Codecs: Befa - Time signal removal Band Eliminate Filter for Audio jorel Audio Conversion 0 07-24-2003 06:53 PM

Thread Tools



 
All times are GMT -5. The time now is 05:42 PM  —  vBulletin © Jelsoft Enterprises Ltd