I was tring to convert wav's to mp3's with frontah using Lame 3.98a9 and the setting -V 2 --vbr-old and I got this error
"Assertion failed: thr >= 0, file psymodel.c, line 670"
Edit: tried with -V 2 --vbr-new and got the same results
I was tring to convert wav's to mp3's with frontah using Lame 3.98a9 and the setting -V 2 --vbr-old and I got this error
"Assertion failed: thr >= 0, file psymodel.c, line 670"
Edit: tried with -V 2 --vbr-new and got the same results
There's a link with a working version in another topic (http://www.hydrogenaudio.org/forums/index.php?showtopic=51129&hl=) about this issue.
Edit: Muahahahahahaha! I was first!
http://www.hydrogenaudio.org/forums/index....&hl=3.98a9# (http://www.hydrogenaudio.org/forums/index.php?showtopic=51129&hl=3.98a9#)
Edit: Damn. Beat while I was searching.
Thx, sorry I'm at work and didn't have much time to search
Rarewares (http://rarewares.org/mp3.html) has updated 3.98a9 compiles which may fix this.
Rarewares (http://rarewares.org/mp3.html) has updated 3.98a9 compiles which may fix this.
Nope, they don't! I could probably 'fix' it by commenting out the offending 'assert', but I'm sure that would not really be the right thing to do. Hopefully, Robert will fix it in the CVS momentarily.
Ah, sorry, John! I simply assumed that because I thought that they were re-uploaded after a day.
Well, I just tried commenting out the offending 'asserts' and although it completes encoding, '-V 2 --vbr-new' produces what is almost CBR at 319kbps!! So, clearly something has been broken.
Haha, this is a pickle! Well, let's hope that the good LAME developers figure something out soon.
Well, I don't have newer sources than what's in CVS already. John33 are you sure you have latest sources? Did you make a clean rebuild?
Well, I don't have newer sources than what's in CVS already. John33 are you sure you have latest sources? Did you make a clean rebuild?
Yep. Just double checked everything and building via the VC6 IDE and the VC8 IDE, both using ICL9.1 yields the same error message.
Hmmm, a straight VC8 compile encodes cleanly, but the ICL version fails. I'll look at this more closely tomorrow.
Still the same error with latest CVS code, compiled by MinGW, no extra optimization.
10 minutes ago I checked in a bug fix, so you may try again later.
Alright.
Now it's OK.
BTW, how can I optimize compilation of lame with MinGW as compiler? Should I change Makefile options or do something else, like get a copy of nasm?
10 minutes ago I checked in a bug fix, so you may try again later.
Thanks a lot Robert, all seems OK now.
The only problem I have now is that Rarewares is down, so I can't upload!!
If anyone is desperate, you can download
alpha10 from: http://homepage.ntlworld.com/jfe1205/LAME/ (http://homepage.ntlworld.com/jfe1205/LAME/) until I can get back to Rarewares.
10 minutes ago I checked in a bug fix, so you may try again later.
As that will be alpha10 on Rarewares and as I tested the dll based alpha9 version from the link you provided:
Should I expect another behavior so that it's worth testing again? Or is it just a technical bug fix for the assertion error?
It's more of a technical fix, you don't need to test the same samples again.
Rarewares is back up and the files are now in their usual place.