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: Multiformat 48 kbps Listening Test (Read 47044 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Multiformat 48 kbps Listening Test

Reply #51 –
Quote from: AstralStorm on
Please add LAME ~128 kbps as a high anchor instead of AAC.
It would be nice to check the codecs against it instead of much better AAC 128kbps.

It shouldn't make a difference. A high-anchor has a specific role to play and that is too sound significantly better than the contenders. If the anchor does not do this, you are getting a more difficult to interpret test. Also, last 128 kbps indicates the difference between iTunes AAC and LAME was neglectable. So it doesn't appear to matter.

We cannot use the same codec with the same setting both as high-anchor and as contender.
"We cannot win against obsession. They care, we don't. They win."

Multiformat 48 kbps Listening Test

Reply #52 –
I'd still suggest to use AAC-LC as a low anchor. I think that mp3 would be a too bad anchor, too far from contenders (quality-wise).

For contenders, once again I am suggesting to add wma std. This way, we would have a clear ranking between wma pro and wma std.

Multiformat 48 kbps Listening Test

Reply #54 –
Since there aren't any existing tests with LC-ACC and MP3 at low bitrates, I really would like to see both of them in this test.  We need to test LC-AAC vs MP3 at low bitrates.
Vorbis-q0-lowpass99
lame3.93.1-q5-V9-k-nspsytune

Multiformat 48 kbps Listening Test

Reply #55 –
If Gabriel already says that MP3 is too bad to serve as low-acnhor for this test, I really don't see the need for it.
"We cannot win against obsession. They care, we don't. They win."

Multiformat 48 kbps Listening Test

Reply #56 –
Quote from: Sebastian Mares on
Gabriel, do you recommend 48 kbps for low anchor, too?

If your low anchor is AAC-LC, then yes I think that 48kbps is fine. In the AAC 48kbps AAC test I think that the low anchor was correctly positionned regarding quality.
If your low anchor is mp3, then I'd suggest something around 64kbps.

Multiformat 48 kbps Listening Test

Reply #58 –
Quote from: gameplaya15143 on
Since there aren't any existing tests with LC-ACC and MP3 at low bitrates, I really would like to see both of them in this test.  We need to test LC-AAC vs MP3 at low bitrates.

Why would we need that? I think we can safely say both will perform crap at 48 kbps because they are not meant to be used at these low bitrates. What if the outcome would be LC-AAC at 48 kbps is slightly better than MP3 at 48 kbps? Both of them will still be tremendously outperformed by HE-AAC and Vorbis (among others) so nobody would seriously consider using MP3 or LC-AAC below 96 kbps.
Every night with my star friends / We eat caviar and drink champagne
Sniffing in the VIP area / We talk about Frank Sinatra
Do you know Frank Sinatra? / He's dead

Multiformat 48 kbps Listening Test

Reply #59 –
MP3@48kbps was too bad as low anchor in Nero HE-AAC test.

@Maurits
Yep. That's just wast of time and resources.
Some people don't recognize that anchor is not contender.

Multiformat 48 kbps Listening Test

Reply #60 –
Quote from: Maurits on
Why would we need that? I think we can safely say both will perform crap at 48 kbps because they are not meant to be used at these low bitrates.

assuming goes against the spirit of TOS #8.

Multiformat 48 kbps Listening Test

Reply #61 –
Anyway, MP3@48kbps is too bad.
The only check we have to do is to quick-compare HE-AAC, LC-AAC and Vorbis.
Two samples should suffice for checking if LC-AAC is suitable as an anchor.
ruxvilti'a

Multiformat 48 kbps Listening Test

Reply #62 –
Quote from: kwanbis on
Quote from: Maurits on

Why would we need that? I think we can safely say both will perform crap at 48 kbps because they are not meant to be used at these low bitrates.

assuming goes against the spirit of TOS #8.

TOS #8 is a very valuable thing, I agree, but surely you can't be serious? Would you ABX Shine @48 against Lame 3.97b2 @256 or is there room for a safe assumption in that case? The quality of MP3@48 is obvious after previous tests where it functioned as a low-anchor.
Every night with my star friends / We eat caviar and drink champagne
Sniffing in the VIP area / We talk about Frank Sinatra
Do you know Frank Sinatra? / He's dead

Multiformat 48 kbps Listening Test

Reply #63 –
I think LC-AAC would be a better low anchor, MP3 at 48 in recent listening tests just sounds too lousy.

Multiformat 48 kbps Listening Test

Reply #64 –
Quote from: Maurits on
TOS #8 is a very valuable thing, I agree, but surely you can't be serious? Would you ABX Shine @48 against Lame 3.97b2 @256 or is there room for a safe assumption in that case? The quality of MP3@48 is obvious after previous tests where it functioned as a low-anchor.


It's the wording. You wrote it so it could also parse as:
"LC-AAC at 48 kbps is crap." We don't know it yet.
ruxvilti'a

Multiformat 48 kbps Listening Test

Reply #65 –
OK, I think we should start discussing settings.

High Anchor: LAME 3.97 Beta 2 / -V5 --vbr-new
Low Anchor: iTunes LC-AAC / 48 kbps VBR

Nero HE-AAC: -br 48000
WMA10+: Create image file with samples (Audio CD), emulate with DAEMON Tools and rip to WMA10+ using WMP 11.
Vorbis: I am going to ask Aoyumi for suggestions
ATRAC3+: Have to install VMware first

Multiformat 48 kbps Listening Test

Reply #66 –
Quote from: Sebastian Mares on
As I said, I am against testing two AAC encoders again. While Nero AAC didn't win the AAC test, it had a slightly better ranking which is why I am going to use it. That's it. No CT. We just had an AAC test that included CT.


CT and Nero are not the same HE-AAC codecs and they in fact act very differently at different bitrates. The fact that you tested Nero and CT in one test, doesn't tell you anything about the result at this bitrate. Actually it would be interesting to see what the difference will be. I would suggest you include CT as well.


Regards,
Goran Tomas

Multiformat 48 kbps Listening Test

Reply #67 –
This is a multiformat test not another AAC contest...So, only one is enough imo.

Multiformat 48 kbps Listening Test

Reply #69 –
Also note that this test was also at 48 kbps, so not a different bit rate.
"We cannot win against obsession. They care, we don't. They win."

Multiformat 48 kbps Listening Test

Reply #70 –
Quote
CT and Nero are not the same HE-AAC codecs and they in fact act very differently at different bitrates. The fact that you tested Nero and CT in one test, doesn't tell you anything about the result at this bitrate. Actually it would be interesting to see what the difference will be. I would suggest you include CT as well.


First of all, as people already mentioned - CT and Nero were tested exactly at 48 kbps which is, in fact, the same bitrate as it will be used in this multiformat listening test. By accepting your proposal you would be ending with testing of the same codecs at the same bit rate all over again?  Why?

Second,

Quote
CT and Nero are not the same HE-AAC codecs and they in fact act very differently at different bitrates.


I would call this a slight TOS #8 violation as there are no ABX proofs and samples for this claim  That aside, I think in order to back-up this,  you would have to to do very careful blind listening test with many people participating.

Multiformat 48 kbps Listening Test

Reply #71 –
Why not using Vorbis@32kb/s as a low anchor? LC-AAC or MP3 at 48 kb/s are useless at this bitrate (aren't used by nearly noone) and so are only useful as a reference. Vorbis@32kb/s seems to be more useful in a real-world scenario, and we can have a rate for this codec that can be useful to someone using it.

Multiformat 48 kbps Listening Test

Reply #73 –
Low anchor is to basically set the scale for the rankings, with the low anchor theoretically being the worst, and the high anchor theoretically being the best.

Is that a correct explanation? Not really sure if it is...

Multiformat 48 kbps Listening Test

Reply #74 –
Yep, that's basically it. So it is never useful to compare an anchor with the contenders, because they are expected to do better or worse than the contenders. If they don't do this, they are not serving their purpose.
"We cannot win against obsession. They care, we don't. They win."