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: Lame 3.98a7 (Read 6332 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Lame 3.98a7

This afternoon I realized there was a new 3.98 alpha out on Rarewares.

Just finished my listening test with my standard bad samples harp40_1, herding_calls, trumpet, and now eig as well as lead-voice which I found added a tremolo effect with prior 3.98 alphas.

-V2:
trumpet: not annoying to me
herding_calls, eig: slightly annoying to me
harp40_1, lead-voice: annoying to me. The tremolo artefact in lead-voice is still there, pretty serious already in the first 2 seconds

-V1:
trumpet, harp40_1: not annoying to me
herding_calls, eig: slightly annoying to me
lead-voice: annoying to me

-V0:
trumpet, harp40_1: not annoying to me
eig: very slightly annoying to me
herding_calls: slightly annoying to me
lead-voice: annoying to me

--alt-preset 250:
trumpet, harp40_1, lead-voice: not annoying to me
herding_calls: very slightly annoying to me
eig: slightly annoying to me

So with the exception of lead-voice VBR now is quite good with these bad samples. Compared to ABR eig seems to take a little profit from VBR (at least to my ears), whereas herding_calls suffers a bit from VBR.

According to lead-voice there seems to be a bug in the VBR routine (I don't think  this tremolo effect is due to a weakness in the psy moodel).

EDITED:
I just tested ABR 125 with lead-voice to get at a comparable file size of -V0's. ABR 125 is fine.
lame3995o -Q1.7 --lowpass 17

Lame 3.98a7

Reply #1 –
I'm assuming you've ABXed this?

Lame 3.98a7

Reply #2 –
Quote from: Canar on
I'm assuming you've ABXed this?

What I did is: I went into foobar's ABX environment and compared the original against the encoding.
I concentrated on the critical spots which I know pretty well. Most of the time I compared just the A against the B sample, and usually the differences were obvious. In case they weren't I started a real ABX test, and in all the cases I did not rate 'not annoying' things became clear pretty soon.  I never bothered going through a real ABX session.

For the results I rated 'not annoying' this does not mean they aren't transparent to me. I don't know whether or not they are transparent, they are just 'not annoying' to me. With these bad samples I don't ask for more (I'm even content with 'very slightly annoying').
lame3995o -Q1.7 --lowpass 17

Lame 3.98a7

Reply #3 –
Can someone provide a working link for the lead-voice sample please? It seems the one from the original poster doesn't work anymore.

Lame 3.98a7

Reply #4 –
Quote from: robert on
Can someone provide a working link for the lead-voice sample please? It seems the one from the original poster doesn't work anymore.

I will upload lead-voice when I'm at home tonight.
lame3995o -Q1.7 --lowpass 17

Lame 3.98a7

Reply #5 –
I've posted an entry in the upload section with lead-voice: locu's lead-voice sample.
lame3995o -Q1.7 --lowpass 17

Lame 3.98a7

Reply #6 –
Yay, new version out. But I'm disappointed in how little attention this version has created. Earlier, a new alpha would be put to the test immideatly by a big number of people. I'm so curious to how good this version is, but unfortunately I don't have the resources nor the knowledge to test sound quality myself.
 

Lame 3.98a7

Reply #7 –
That's what I feel too, especially as the 3.98 branch is a very important  development. VBR has seriously improved, and some severe weaknesses IMO of 3.97 (Wombat's samples, trumpet) have been fixed too. It's such a pity that it doesn't get the attention it deserves.
lame3995o -Q1.7 --lowpass 17