Quick question which is the “good” fastenc encoder?, I found fastenc 1.01 from here .

fastenc 1.02 from
here

I found 1.01 and 1.02 have the same output file and speed using:
-no switches (cbr 128)
-vbr 90

Binaries are the same size but do not compare

edit: same goes for -mq and -hq. Seems 1.02 is a hacked 1.01

**OLD POST**

Use the latest ones, as they have some extra tuning in the algorithms and thus may sound better!

**OLD POST**

**EDIT**

Fastenc version are essentially the same, with just little bugfixes and fixes to the EXE frontend. So no extra tuning at all thus sound the same!

Thanks to rjamorim for the information.

**EDIT**

Just don't use high quality setting with it. It's broken. But then it's not really fast anyway if you use high quality setting.

If you use the fastencc.exe command line program, don't use the normal (fast) setting, because it's got the stereo separation bug. The high quality setting is ok.

ff123

LOL!

Which is it now?

JohnV xor ff123...

Try encoding mc_sich_short.flac, which can be found on my samples page at:

http://ff123.net/samples.html

The stereo separation bug should be very apparent on this sample.

JohnV found this sample.

ff123

The medium quality setting

What the?..
Fastenc is so bad.....

Huga.. You're right. No problem even for me to tell the difference.

But maybe JohnV still has a point... It's not worthy the name fastenc for the high quality setting so why not use lame instead

Well, the thing is that the commandline fastencc is totally useless software..

I remembered a bit wrong; Like ff123 said, it's the fast (default) mode which has this horrible stereo separation bug. BUT, high quality mode has horrible dropout problems (for example track01b.wav), so hq-mode can be considered very buggy also..

It's better to keep away from the fastencc alltogether, and not recommed it to anybody, and yes, this applies to fastencc 1.02 as well!

And a bit clarification to the mc_sich sample: Actually I didn't "find" it since it's from minidisc.org, but I noticed that this sample shows the stereo separation bug of fastencc's fast(default) mode very clearly.

sheeesh... I thought either you all have gone mad or -I- have...
fastenc completely sucks.

fastenc isn't that bad. It's not in Xing/Blade suckness category.

Fastenc in fastencc 1.02 is horrible. Clearly in Xing/Blade category, because the stereo separation can be totally destroyed similarly at all bitrates, upto 320kbps...

Fastencc 1.02 does not have the latest fixed fastenc codec...

rjamorim:

Well, are you aware that all Fastencc versions are derived from the same libraries, leaked by a guy that broke his NDA agreement? (Unless I’m terribly mistaken)

(I know that, because a very knowledgeable member of this community and friend of mine is the one that got these libraries, and sent them to another guy so that this guy would wrap them in an exe file)

He still has the libraries BTW.

WHAT?!?!
Oh my god, this is good sh@t to know!

rjamorim: I did not know that. Where do you get your information from, whoever it is bloody good stuff! rjamorim super ninja spy !

So basically all versions of Fastenc are exactly the same?

**
Edited my first post on this thread, to contain information provided by rjamorim. No more misinformation!
**

Heh. I have my contacts.





I guess so, with little bugfixes/modifications to the EXE frontend.

Must find "xxxxxxx" now!!!

Thanks for all that information rjamorim!

All the same...all terrible.

1.) High-quality mode is "broken" somehow and takes forever (1.5x on PIII 1.4Ghz) to boot.
2.) Fast mode features an almost complete collapse of the stereo image
3.) Uses 22.05Khz sampling rate until 96kbps (I think this can be corrected with switches)
4.) Artifacting is highly salient at bitrates where this is not a problem with LAME (especially on fast transients such as low brass, hi-hats).

Even the default mode in LAME sounds better than FastEnc and will give you decent encoding speed if you so desire. But you know what they say, "good things come to those who wait". Who needs speed anyway? 4-6x in --aps is good enough for me. (And the sound quality is great)

O.K after all the explanations, one more quick (: question what is recommend for 128kbps (mp3) encoding, transcoding from movie audio (DVD AC3).

No, no. Fastenc is the codec name. All versions of this hacked fastencc.exe use old buggy fastenc codec libraries.

But not all fastenc codec versions are buggy. Latest fastenc version in new versions of MMJB, Cool Edit, SoundForge etc. do NOT have this stereo collapse problem.

OK yeah, fastenc does suck.

I stopped using it (in MMJ) a couple years ago in favor of LAME 3.70 (I think it was). I would still prefer a fastenc encode over a Xing or blade at 128kbps despite the stereo callapse. I end up with way too many Xing/blades from P2P usually with only 128CBRs available.



A LAME ABR commandline like --alt-preset 135 or so would get you a bitrate around what you want. If you really want CBR for some reason, --alt-preset cbr 128 or just the default LAME commanline will get you your cbr.

Sorry I meant the Fastenc command-line version was essentially the same.

All the others are bundled in software packages, the way it supposed to be . ff123 site has all the information about the bugs in the fastenc, so if you want to know more head over there .

One more thing then...
Fastenc must be Fraunhofer commercial encoder then and MP3Enc/L3Enc their consumer encoder. Am I right to say that?

Is Nero another? Hmm, I still haven't used any of my trial encodes