For anyone interested, I have produced compiles of lame.exe, lame_enc.dll and lamedropXPd.exe, all compiled from the LAME 3.90.2 libs with ICL4.5 using Dibrom's compile switches.

Although this lame.exe does not produce absolutely bit identical results when compared to Dibrom's own compile, the file size is within approx. 200bytes in 6Mb. All of the above produce bit identical results as between themselves. The lame_enc.dll will work with the current CDex and allow proper use of the --alt-presets.

These are available at my 'Others' page at Mirror 1.

What advantages would we find using 3.90.2 john33 vs 3.90.2 dibrom.

I use oggdrop... perhaps lamedrop would be useful as well. Does it use the .dll or .exe?

Hmm? Dibrom's version was also compiled with ICL 4.5, if I remember correctly. Assuming you used all the same compile switches, why the output difference? Could it be a processor-related optimization, perhaps?

- M.

Yes -- what's the difference? Why did you do it? Just curious, since doing it & posting about it seems to suggest there's some possible benefit.

can't find the new compilies anywhere..

http://homepage.ntlworld.com/jfe1205/others.html

I too am curious about ur compile & how does it differ

About the difference, it may well be down to processor optimisation.

Why did I do it? I produced the dll and the version of lamedropXPd because they haven't previously existed using these libs, compiler version and compile switches. The dll that's available was a hacked version that I produced to enable the presets in the old CDex that didn't directly support them. While I was at it, I thought I'd add the lame.exe just to complete the set!!

So, it was really just to provide a couple of options using the approved libs, etc., that haven't previously been available. Incidentally, the lamedropXPd supports --alt-preset ABR over the whole range, not just down to 80kbps.

Since 3.90.2 has always been the 'HA approved' library compiled with ICL4.5 and Dibrom's switches, I thought that it was overdue that there were options other than the normal lame.exe available. Just broadening the choice!

Hey John, could you do me a small favor?

Could you make an idiot-proof almost user-friendly version of LAME?

There would only be one mandatory option. The input.wav. Output.mp3 is optional but the assumption is the same filename with extension .mp3.

It will always encode to --alt-preset standard. No other option is available, and you can't choose.

Something like:

lame.exe input.wav

This is for a friend of mine and my brother so they can not make any mistakes at all; and for me too, heheh.. (Might be nice for drag and drop version too, but I think the batch file might do for that.)

This is also what I'll give to other friends when they ask me what encode I use. When they are ready to do more experiments they can always download the regular version.

This is similar to the idiot-proof md5sum.exe. It's permanently set to binary.


If its not too much to ask.

I'll take a look at it, but it's not quite as simple as it may seem!!

I believe APS are already idiot-proof, using a GUI like EAC or lamedrop set to alt presets leads to the result you go for, IMO.

Good idea John!
You make the difference!
The new compiles are my defaults now!
Great tnx pal!

LIF

Why dont you get AudioWizz. OK, its not completely complete yet, but its very easy to use!

e.g.

step 1 - drag wav onto AudioWizz window

step 2 - choose mp3 from the codec options - only mp3 will be available if you only copy lame.exe into the pluggins directory!

step 3 - choose where you want the new file - I will be modifying this option so, if the user does not choose a save location, the new files will be created in the same location as the original file

step 4 - press encode

and thats it!

Excellent Idea. I will use this newest dll with cdex 1.5
Many thanks

I have created a version that does as you ask. It ONLY encodes '--alt-preset standard' and the only parameter on the command line must be the input file name. Usage, therefore, is:

lame input.wav

OK? It's only an ICL6 compile at the moment. I'll fire up my other system and do an ICL4.5 compile a little later and then upload it for you. I'll let you know when it's available.

Thanks John. This is what LAME really needs. Now if only the right people were interested in such a cleanup of its user interface...

The special compile referred to above is now available at my 'Others' page at Mirror 1.

Why would we not use ICL 6 or 7 anyway. What advantage does ICL 4.5 have.
Also seems prudent to use the latest MSVC.

Please educate me.

Thanks,
David

In short, each of the different compilers produces executables that generate slightly different encodings. Since the major tuning was done by Dibrom, at least as far as the --alt-presets are concerned, based on compiles using the Intel 4.5 compiler, that has tended to be the compiler of choice. In reality, whether there is anyone who can truly differentiate between the results in blind tests is a different matter entirely!

Where is it? Can't find it.

BTW,
My intent and goal is to make it as idiot proof as possible. Assume that the user is a complete newbie. Perhaps call the file LAMEAPS.EXE or something. Or MAKETHEBESTMP3.EXE.

If you give people options, the tendency is to use it without understanding what they really do. We have all sorts of people believing they can use the default setting (which is 128 CBR) or insisting that they need -alt-preset INSANE or BRAINDEAD or --quality 10 higher than braindead go to heaven and hell.

If you make a nice newbie friendly compile... they'll see it first. You stuff it with some documentation that says "The best MP3 maker in the werld!" and some other scientific terms and some quotes and some ABX results .... well.... u get the idea.

I'm starting to use MPC now, but everyone who has a hardware MP3 player will still use MP3 and let's help them with an alternative as easier to use than Xing.

Here!