sorry, I'm poor english
Source WAV file is test.wav of RMAA 4.2...
mp3 -> lameDropXP 3.93 (--alt insane)
ogg -> oggDropXP (q=10)
Results (http://hifipc.net/Comparison.zip)
ogg is +0.5db gain in 20hz...
mp3 is very nearing source file..
Seems to be useless, because these graphs don't tell you anything about the quality. I doubt you can hear any difference. And please remember that Ogg Vorbis wasn't optimized that much at high bitrates yet.
Natural: graphs aren't taken very seriously here. Using graphs to judge quality is considered very dangerous and erroneous.
if you want my take, see here (in Korean) (http://www.avbest.com/community/board.htm?code=usermini&function=read&no=1996&keyword=&srh=&page=2).
I'm Gum-Eun-Hae.
yeah, and there is a reason for that. graphs show how music looks but our ear judge what they hear, not what the (can't) see!
No..
mp3 is encord near source...
I heard ogg is gain Bass sometimes..
No..
mp3 is encord near source...
I heard ogg is gain Bass sometimes..
Ah, but can you hear it yourself?
Well, the rules here are to provide short samples. Then people can try encoding it themselves.
[span style='font-size:21pt;line-height:100%']Graph is very important.[/span]
files (http://hifipc.net/test.zip)
Graph is very important.
no, ears are more important.
sori-neun gwi-ro deud-neun geo-ji, noon-eu-ro bo-shi-neun geo-ga anibnida.
how about:
don't feed the troll ?
If you're not able to offer any evidence that you can hear the differences, well... please ****
okay, won't feed it
how about:
don't feed the troll ?
Benjamin: I'm not sure if he (Natural) recognizes me, but I recognize him. From what I gather he's not a troll, he's not malevolent or twisted, in fact he intends to do good, if I'm not mistaken.
Natural: How about a real sample of music, not noise? That will help a lot more.
Haha... *lol* Funny thread...
(I agree with Garf... Graphs shows nothing about the quality in psychoacustial compressed formats, except perhaps frequency cutoffs... But that's it... It has been told thousands of times already... )
ROFLMAO @ Garf
but why is he still claiming what garf called bullshit if he was proven wrong ?
[span style='font-size:21pt;line-height:100%']Graph is very important.[/span]
hmmm maybe you should compress your beautiful graphs with JPG or PNG. These formats give excellent quality and take far less space than MP3s or Ogg Vorbis files Have phun watching your music
[span style='font-size:21pt;line-height:100%']Graph is very important.[/span]
OK, and now, assume it :
[span style='font-size:14pt;line-height:100%']LAME --alt-preset extreme : [/span]
http://membres.lycos.fr/guruboolez/AUDIO/t..._02/mp3-wav.gif (http://membres.lycos.fr/guruboolez/AUDIO/test_02/mp3-wav.gif)
[span style='font-size:14pt;line-height:100%']Ogg Vorbis 1.0 -q7[/span]
http://membres.lycos.fr/guruboolez/AUDIO/t...02/ogg7-wav.gif (http://membres.lycos.fr/guruboolez/AUDIO/test_02/ogg7-wav.gif)
Mp3 is very nice on pre-echo. Very high resolution, isn't it ?
[don't flame me too ; theses graphs were made in order to illustrate (and only illustrate) a very tiring listening test]
a total worthless test
I dont take a pivtures and conveter it to raw pcm and an then make a playback test to hear wich lossy pictures formas is better then the order by my ears.
Why should i do the opposite.
...and you all are forgotting something:
...if you convert from mp3 to jpg...
*or*
...you convert from ogg to jpg...
...you are using two different lossy format, and thus *transcoding*...
...and transcoding is *death*
...you can't expect not to loose quality while transcoding
...or is it? lmao!
[Edit]: Sorry, I couldn't resist
GIF is like the ADPCM of digital images.
GIF is lossless, but only supports up to 256 colors. And since most screenshots of nice graphs are in 16-bit, 24-bit or 32-bit color, there is a loss in color precision.
But no phony artifacts.
I noticed how the graphics of the music were animated... is this what we call a musicvideo?
Wow.. I've never seen such colorful posts before...
"...or is it? lmao!"
LMAO too ...
And liked the post by Garf, righton..
ooh look at all the pretty colors!
That graphs are interesting, but in practice mostly useless. Note that those kind of tests are designed to stress 'dumb' linear devices, not psychoacoustic encoders. For one side, in real world, those codecs in some cases will perform worse with real music than with these tests signals, and in other cases better, so the information you can extract from the tests is not useful. Also, in case of these type of encoders, this results are not indicative of how well will the codecs sound in practice with real music, because they 'adapt' intelligently to the music and human hearing. For this particular settings, I'd say they will sound very good in most cases, independently from the results you get from RMAA.
Even if the codecs were 'dumb' devices and the results were indicative of the quality, the differences between both encoders in the test are negligible and very likely would not be audible in practice.
Graphs and spectrograms can be very useful, ofcourse, but also given: In the right context.
As Garf pointed out: Spectrograms are NOT useful with Psychoacoustic codec.
There's nothing to prove, nothing to discuss.
If you believe otherwize you need education.
(btw. excellent clarification Garf )
I'm testing Fhg-lame...
It's interesting...
Fhg -> encoder by Cool edit Pro 2.0 320kbps CBR
lame -> lameDropXP 3.93a 320 CBR(insane)
decoder -> lamedropXP
fhg is high distortion....
I'm poor english..
files (http://hifipc.net/Comparison(lame-fhg).zip)
Natural, please stop using graphs for comparing psychoacoustical encoders.
You have been told they are not useful.
You have been told it even in Korean.
And also, I just noticed it: C-Media? Not exactly the best soundcard, is it?
Your tests are simply wrong.
Use your ears, not your eyes, to judge quality of sound.
He doesn't WANT to understand it
I've lost all hope. And now I'm getting the impression that Natural IS a troll. *sigh* I'm easy to troll, right? : )
hmmm maybe you should compress your beautiful graphs with JPG or PNG.
I think u should use png, cause its lossless
but u can go with a lossy format like jpg, cause I doubt u would pass in a
blid test anyway... B)
Bullshit.
Show me a clip where Ogg fails on bass resolution.
Just to be an ass
It _had_ trouble with bass on at least one clip I know about. Try one of the betas on this clip (http://www.dtek.chalmers.se/~d99papa/files/bassrumbledemo.wav). But since RC2 it's fine.
Just to be an ass
It _had_ trouble with bass on at least one clip I know about. Try one of the betas on this clip (http://www.dtek.chalmers.se/~d99papa/files/bassrumbledemo.wav). But since RC2 it's fine.
Ok, ok, show me a clip where 1.0 fails on bass...
RC2 had bass fixes.