NOTE:
Now the latest CML compiled binary is included in the project rev11 boundle.
From now on, the latest version be put on the top.
ChangeLog:
rev11
fix 22050 output crash
all printf change to fprintf(stderr) [setup.c, bitallos.cpp, bitallo3.cpp]
ICL9 compile with speed increase (with a different name "hmp3enc.exe")
rev10:
Length bug found by level fixed
New compile with asm enabled, several small bug-fixes.
-V0 -X should give correct vbr-header now.
Edit: Uploaded the wrong file. Please Redownload. This time is OK
Edit2: second compile obsolete
The third compile. All asm optimization used. Slitly undeceivable speed increase can be expected.
Edit: This compile is recommended.
Edit2: Use "-U2" option to achieve better speed via P3 optimizations
Project files to compile
The whole thing here.
Problem: pow34.asm not used (don't know how to compile)
TODO: gapless plaback support (somebody can do this?)
Project files to compile[a href="index.php?act=findpost&pid=313615"][{POST_SNAPBACK}][/a]
Can you please provice VC6 project files too?
rjamorim,
I cannot provide VC6 project file. But I think it's not too difficult for you to creat one with all these sources in the uploaded package.
Regards,
Edit: My friend CML made a VC6 compile. But the result .exe file is bigger, while the speed is more slower than that compiled with VC7. I have no idea why.
BTW, the asm file need MASM to compile (pow34.asm as an exeption).
Thanks to my friend CML, who changed the asm code that cannot be compiled directly by MASM. Now all code can be compiled by VS7.
And some unused file (.c and .asm) is eliminated to keep project clean.
The whole package is uploaded here.
Helix mp3 encoder binary rev4
Changelog:
more detailed switch description with -Help
Helix mp3 encoder binary rev5
Changelog:
1). -hf switch description in -Help
2). other minor tweaks
Note: can encode high frequencies in low bitrate or low vbr_scale. (not a good idea, I know, will change to the default in next rev.).
rev6
This is the correct hf restricted version.
rev7 (piping supported)
just use command like
stdin %d -V50 -X -U2
with foobar2000 cli diskwriter ("%s" replaced with "stdin").
Latest project/source files
Edit: now obsolete
rev8 binary
Changelog
1) support stdout now
2) using "-" instead of "stdin" with previous rev7
You can do test with
mp3enc - - < 001.wav > 001.mp3
Comparative results (plots) for Real 5.0 rev6 at 96 kbps
See this test (http://www.hydrogenaudio.org/forums/index.php?showtopic=35531&view=findpost&p=314613).
hmm, tried to download rev7 and rev8 but downloads are corrupt here. FF and WinRAR. anybody else experience this?
hmm, tried to download rev7 and rev8 but downloads are corrupt here. FF and WinRAR. anybody else experience this?
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=314617")
Works fine here...(FF & winRAR too). [a href="http://kurtnoise.free.fr/helix_mp3enc-rev8.rar]here is a link mirror[/url] if you want.
No problems here.
Get them at http://www.neilpopham.pwp.blueyonder.co.uk...p/mp3enc-r7.rar (http://www.neilpopham.pwp.blueyonder.co.uk/temp/mp3enc-r7.rar) and http://www.neilpopham.pwp.blueyonder.co.uk...p/mp3enc-r8.rar (http://www.neilpopham.pwp.blueyonder.co.uk/temp/mp3enc-r8.rar) if you want.
I see kurtnoise beat me to it. Rev7 here as well.
thanks guys, that did the trick.
rev9 (Project & Binary included)
A lot of small tweaks
Add -EC switch to show parameters used with the encoding
Thank you.
I want someone to up-load the binary to the Rarewares.
Will it take on magical powers then?
Cool!
Edit: There seems little point in releasing it to the masses until it stops being revised twice a day.
Following the requests for VC6 project files for this, you can d/l a complete cml version 9 source with VC6 project files in an 'hmp3' sub-directory, plus an executable compiled with VC6/ICL8.1 (but without any specific optimisations) from: http://homepage.ntlworld.com/jfe1205/helix_mp3enc_cml9.zip (http://homepage.ntlworld.com/jfe1205/helix_mp3enc_cml9.zip) BTW, the executable generated is called 'hmp3.exe' just to differentiate it from the other binaries. You can call it what you like!!
Although I haven't tried, I am assuming that all likely optimisations for speed are already taken care of by the extensive asm code.
As already mentioned, not much point in putting this on Rarewares until it stabilises.
There's some known bugs with this encoder. It cannot take 22050khz & 11025khz files reported by Raffles. Can someone fix this issues?
Rev9 is quite clean and stable, I hope someone more specialized in audio science can take over the job.
,it's a mp3 encoder not a mp3 decoder.
CML's rev10
Length bug found by level fixed
New cml10 VC6/ICL8.1 compile with the default changed to mpeg1 to correct crashing when trying to encode 11025Hz and 22050Hz samples available here: http://homepage.ntlworld.com/jfe1205/helix_mp3enc_cml10.zip (http://homepage.ntlworld.com/jfe1205/helix_mp3enc_cml10.zip)
CML's rev10
Length bug found by level fixed
[a href="index.php?act=findpost&pid=315015"][{POST_SNAPBACK}][/a]
mp3enc.zip is corrupted.
680e71afc101914ecd2714bab998446d *mp3enc.zip (115.694 bytes)
Edit:
Very strange...
I read Gambit's reply below, but fact is that I actually use Firefox (1.04, Moox optimized compile). Retried the download in IE, now it's correct:
5cd595b28f54eba5cbf772b3ac5a12ea *mp3enca.zip (115.724 bytes)
It seems that the first time the last 30 bytes of the files were truncated.
Thanks Gambit,
I'll test this with some more browsers and OSs and report later in another forum section, in order to prevent this thread from being contaminated too much by offtopic discussions.
It worked just fine for me
CML's rev10
Length bug found by level fixed
[a href="index.php?act=findpost&pid=315015"][{POST_SNAPBACK}][/a]
mp3enc.zip is corrupted.
680e71afc101914ecd2714bab998446d *mp3enc.zip
[a href="index.php?act=findpost&pid=315199"][{POST_SNAPBACK}][/a]
It's a problem with the forum software and downloading attachments in IE. Firefox will download the file correctly.
EDIT: It's fixed now. IE downloads work now correctly too.
VC6/ICL9.0 compile, source and VC6 project files for cmlr11 here: http://homepage.ntlworld.com/jfe1205/helix_mp3enc_cml11.zip (http://homepage.ntlworld.com/jfe1205/helix_mp3enc_cml11.zip)
Enig123, could you please add somehow the word "Helix" to the name of this topic? It will be much easier to find, using forum's search engine.
I totally agree, but I have no right to change it.
Please Admin do this for us.
Fixed.
CML rev12 (Only for test)
* Add a switch -SBT that can set the short_block_threshold (default is 700). This value can range to negative values.
** Short block detection logic changed a little (decide by each channel, more prone to short-blocks).
Caution: This is for testing purpose ONLY, and may cause efficiency or quality degradation.
Any further testing is welcome. For the problem sample level offered http://www.hydrogenaudio.org/forums/index....c=35531&st=100# (http://www.hydrogenaudio.org/forums/index.php?showtopic=35531&st=100#) , can someone (level) find a -SBT value to let the distortion gone?
Helix mp3 encoder x64 binaries