Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: MP3 at 128kbps public listening test (Read 53398 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

MP3 at 128kbps public listening test

Reply #150
OT:
Quote
<nitpick mode>
It's Mac, not MAC. MAC is something your ethernet card has, while Mac is short for Macintosh.
</nitpick mode>

Then why do you use "mac" in your signature?
Really, all-caps is just a way of emphasizing...

MP3 at 128kbps public listening test

Reply #151
Good... lots of replies

So, I reckon Gogo is wellcome? In this case, it'll be featured instead of Audition Legacy.

@Guruboolez: I don't think 6 codecs is too much considering some of them will (theoretically) sound bad. And I agree, I don't think there will be more MP3 tests happening, so it's better if this one covers most significative encoders.

Also, I agree with the idea of adding Fatboy. But, in this case, what sample will be replaced?

For those that don't remember the sample list:
http://www.hydrogenaudio.org/forums/index....howtopic=12358&

Regards;

Roberto.

MP3 at 128kbps public listening test

Reply #152
I still think that you should consider new Lame. Perhaps a new release could be available soon...

MP3 at 128kbps public listening test

Reply #153
How "soon"? 

MP3 at 128kbps public listening test

Reply #154
Well, probably between 1 and 2 weeks.

MP3 at 128kbps public listening test

Reply #155
I plan to start my test on the 14th.

So please try to hurry

MP3 at 128kbps public listening test

Reply #156
According to the fact that 3.90.3 is the same as 3.90 for abr encodings, I'd like to see the newest lame encoder in the test. After all, lame 3.90 is more than two years old now. I know it was heavily tested; but then, why not considering this public listening test as a first step for a complete checking? Even if 3.94 is beta... (recommanded version of musepack is still alpha...)

MP3 at 128kbps public listening test

Reply #157
Quote
According to the fact that 3.90.3 is the same as 3.90 for abr encodings, I'd like to see the newest lame encoder in the test. After all, lame 3.90 is more than two years old now. I know it was heavily tested; but then, why not considering this public listening test as a first step for a complete checking? Even if 3.94 is beta... (recommanded version of musepack is still alpha...)

I would like to see whether the technical improvements since 3.90x (substep quantization and all the other quality improvement changes detailed in history.html) will really result in better quality than the time-tested version like 3.90.3  .

MP3 at 128kbps public listening test

Reply #158
Quote
According to the fact that 3.90.3 is the same as 3.90 for abr encodings, I'd like to see the newest lame encoder in the test. After all, lame 3.90 is more than two years old now. I know it was heavily tested; but then, why not considering this public listening test as a first step for a complete checking? Even if 3.94 is beta... (recommanded version of musepack is still alpha...)

Guruboolez, may be you could do a presonal listening test between 3.90.3 and 3.94 like you did for WMA and Vorbis before previous test? This will not only help with test decision, but also provide some usefull information.

-Eugene
The  greatest  programming  project of all took six days;  on the seventh  day  the  programmer  rested.  We've been trying to debug the !@#$%&* thing ever since. Moral: design before you implement.

MP3 at 128kbps public listening test

Reply #159
I think before a build from the 3.94 branch can be considered, all known problems (e.g. this one ) should be resolved, and a short comparison test (perhaps only individual) should be performed.

 

MP3 at 128kbps public listening test

Reply #160
eltoder> if I find time enough and some motivation, why not.
Continuum> I don't see any problem in the mentioned thread with abr/cbr settings ??

MP3 at 128kbps public listening test

Reply #161
Quote
quickly ABXed fsol.flac
with preset 128
(3.90.3 and 3.94b1)

both 14/14 but 3.90.3 is (much) better

-lazka

But I don't know the status of that.

MP3 at 128kbps public listening test

Reply #162
That opinion on quality can be biased, he ought to use HR program for quality comparison.
ruxvilti'a

MP3 at 128kbps public listening test

Reply #163
Well, the test is almost upon us.

I already selected the switches for each encoder. This is what we will have:

-Lame 3.95 --preset 128
-FhG Audition VBR quality 40, only mid/side stereo enabled, Current - Best codec
-iTunes VBR Highest quality, 112kbps base bitrate, joint stereo, smart encoding
-AudioActive CBR 128kbps, high quality
-Gogo 3.12 -b 128 -a
-Xing 1.5 VBR quality normal

That list can be considered soft-frozen, it will only change if some major issue is brought to my attention.

Also, it is my pleasure to inform you that all VBR codecs are in the range of 120-128kbps. So, no issues should arise about bitrate deviation.


About why Xing instead of RealOne:
RealOne and RealPlayer10 beta are two flaming pieces of garbage that refuse to install on my PC. Surprisingly, the installation process goes well and at the end it claims installation succesful, down to the requirement of restarting the PC (#%$&@!). But when I check the installation folder (whether before or after restarting), it's nearly empty and, not surprisingly, double clicking the .exe has no effect.

Since I can't be bothered to deal with such shitty software for an anchor, and I also won't risk a friendship asking a friend to try installing it on his PC, I gave up RealOne altogether and went back to Xing 1.5, that installed and ran perfectly. Real should take some hints from XingTech...


I didn't replace any sample with fatboy because noone suggested what sample to replace. So this test will have no problem cases, except the very evil and very interesting Waiting.

Next test will definitely feature evil samples.


Also, I would like to thank Gabriel Bouvigne and the rest of the (active) Lame team for rushing the 3.95 release on time for this test.

The schedule is kept, if there's no catastrophe until then, it will start on the 14th, next wednesday.

Best regards;

Roberto.

MP3 at 128kbps public listening test

Reply #164
Quote
I didn't replace any sample with fatboy because noone suggested what sample to replace. So this test will have no problem cases, except the very evil and very interesting Waiting.


I would have liked fatboy...
Why no setting up a poll for which sample to replace?

MP3 at 128kbps public listening test

Reply #165
Quote
I would have liked fatboy...
Why no setting up a poll for which sample to replace?

Because it's already too late. :/

When I asked about it (Jan 3rd) there was plenty of time. Now there are only two days.

MP3 at 128kbps public listening test

Reply #166
What about LAME 3.90.3 <-> 3.95? Will that be "only" testing & tuning by those interested in it?

MP3 at 128kbps public listening test

Reply #167
Quote
What about LAME 3.90.3 <-> 3.95? Will that be "only" testing & tuning by those interested in it?

No 3.90.3 anymore. There are already too many codecs as it is.

And since 3.90.3 has already been tested a lot (128kbps extension and 64kbps test), I decided it would be better to try something newer this time.

MP3 at 128kbps public listening test

Reply #168
As the samples are same in 64kbps test and this one,
we can compare 3.95 vs 3.90.3 indirectly.
ruxvilti'a

MP3 at 128kbps public listening test

Reply #169
Erm.. actually, I prefer you don't even try to do that :B

In the 64kbps test Lame was the higher anchor, so it was meant to win. Comparing lame to 64kbps samples, it's obvious it'll receive high scores. OTOH, in this test Lame will be amongst it's peers, so it's score will probably be lower even if 3.95 isn't worse in reality.

An example. The exact same Lame version, compile and settings were used in the 64kbps test and the 128 extension test. Still, at the 128 test Lame got a score of 3.66 points, while in the 64kbps test it got a score of 4.29.

Regards;

Roberto.

MP3 at 128kbps public listening test

Reply #170
Quote
-Gogo 3.12 -b 128 -a

Since Lame is using the optimal 128kbps setting (--preset 128), gogo might as well use -b 128 -a -q 0. Would you allow that?

MP3 at 128kbps public listening test

Reply #171
Quote
Would you allow that?

No problem.

Thanks for the suggestion.

MP3 at 128kbps public listening test

Reply #172
I'd like to point that with 3.95, targetting 128kbps I think that -q1 and -q0 can really improve encoding quality.

On the other hand, they might not be representative of real-world use due to their slowness.

MP3 at 128kbps public listening test

Reply #173
Quote
I'd like to point that with 3.95, targetting 128kbps I think that -q1 and -q0 can really improve encoding quality.

is it possible to use this setting along with aps? How much longer would that take? I mean it says in the ChangeLog that the new release is 10% faster, so what's the deal? I rather wait a bit longer and have a smaller file or a file with better quality...
--alt-presets are there for a reason! These other switches DO NOT work better than it, trust me on this.
LAME + Joint Stereo doesn't destroy 'Stereo'

MP3 at 128kbps public listening test

Reply #174
Quote
is it possible to use this setting along with aps?

I think that this is off-topic and would lead to endless debates.
I was speaking about 128kbps.