This is aoTuV beta 5.7 (http://www.geocities.jp/aoyoume/aotuv/) vorbis encoder with some parts of Lancer (http://homepage3.nifty.com/blacksword/index.htm) project.
Patches: [attachment=5523:patch.7z] [attachment=5537:patch64.7z]
No OpenMP support; some routines were changed from aoTuV beta 5 and remain unaccelerated.
You can download MSVS 2008 compiles with SSE, SSE2 and SSE3 support. SSE2/SSE3 versions are ~45% faster than "P4 Only" build from Rarewares and yet ~45% slower than original Lancer.
Tested on Core Duo E6850, nice speedup indeed.
Is it me, or is there a bug with --downmix implementation? I'm getting wierd noises when trying to downmix to mono.
I mde mono.wav (44.1/24/1ch) out of stereo.wav (44.1/16/2ch) and encoded with the following settings:
oggenc2.exe -q1 --downmix stereo.wav -o 1.ogg
oggenc2.exe -q1 mono.wav -o 2.ogg
foobar2000 bit comparison:
All tracks decoded fine, no differences found.
Comparing:
"C:\0\1.ogg"
"C:\0\2.ogg"
No differences in decoded data found.
The file i used was actually stereo. The noise is very audible, doesn't require any ABXing.
Here's what i do: oggenc2_SSE3.exe -q0 oggtest.wav -o oggtest.ogg --downmix
Here's a source file, [attachment=5514:oggtest.wav]
And here's the result: [attachment=5515:oggtest.ogg] - this file sounds severely distorted.
When i do the same with oggenc from rarewares: oggenc2.exe -q0 oggtest.wav -o oggtest2.ogg --downmix
In this case everything sounds fine: [attachment=5516:oggtest2.ogg]
I'm running Windows Vista on Core 2 Duo E6850.
I tested encoding of stereo files... -q4: it encodes correctly; -q3.5: incorrectly.
Thanks for the report, I'll see what is wrong.
Problem with quality setting less than 4.0 was fixed. Thanks to Gregory for bug report.
Problem with quality setting less than 4.0 was fixed. Thanks to Gregory for bug report.
Can you send me the updated patch please? Thanks in advance. I'll try to get this tomorrow.
Can you send me the updated patch please? Thanks in advance.
Updated files were already attached to the first post.
I'll try to get this tomorrow.
Excellent!
minor fix in info.c (vorbis library): ENCODE_VENDOR_STRING should be without "\n" at the end.
Patch was updated.
Can you send me the updated patch please? Thanks in advance.
Updated files were already attached to the first post.
I'll try to get this tomorrow.
Excellent!
Sorry, I didn't get to it and, as I'm away over the weekend, It'll be the early part of next week now.
I'm new to the various builds of Ogg Vorbis encoders. I would like to know what is the aoTuV 5.7 compile with the fastest encoding speed for a processor that supports up to SSE2 in a Windows system.
Am I also correct in understanding that john33's compiles are currently the fastest (because of the use of an Intel compiler)? If yes, are there compiles of john33's that are available for processors that support up to SSE2 and Windows?
I also would like to know if the current recommended Ogg Vorbis encoder in HA's Vorbis wiki refers to the aoTuV Beta 5 2006-10-24 version or does it mean the latest in the 5.x series, i.e. Beta 5.7 (2009-03-03).
I'm not really used to operating using the command line; I use dbPoweramp and just replace its default encoders (oggenc2.exe) with the latest ones from rarewares (I currently use the 5.7 P4 encoder).
Am I also correct in understanding that john33's compiles are currently the fastest (because of the use of an Intel compiler)? If yes, are there compiles of john33's that are available for processors that support up to SSE2 and Windows?
You mean http://www.hydrogenaudio.org/forums/index....st&p=670871 (http://www.hydrogenaudio.org/forums/index.php?s=&showtopic=74345&view=findpost&p=670871) ?
Thanks, anyway!
But I would be happy to see here accelerated libvorbis 1.3.1 also
You mean http://www.hydrogenaudio.org/forums/index....st&p=670871 (http://www.hydrogenaudio.org/forums/index.php?s=&showtopic=74345&view=findpost&p=670871) ?
I tried this, but it created zero-filesized .OGG files. Does it require SSE3? My processor supports only up to SSE 2.
I am currently using this: (oggenc2.7z - http://www.hydrogenaudio.org/forums/index....st&p=668288 (http://www.hydrogenaudio.org/forums/index.php?showtopic=76272&view=findpost&p=668288) ) in dBpoweramp.