Available here: http://www.monkeysaudio.com/download.html

Version 3.99 (April 29, 2004)

1. Changed: Decoding engine better at handling corrupt streams / loss of internet connection while playing.
2. Changed: Simplified assembly code building for 3rd party developers.
3. NEW: Improved entropy coder for increased compression.
4. Changed: Removed RKAU support. (since it is no longer commonly used)

Strange. I've deleted the previous version af Monkey, installed the 3.99F one, and the encoder (MAC.exe) is still the 3.98 beta one (date= 30/05/2003). The GUI is the 3.99 version, but the encoder isn't.
Could someone confirm this?

It's shown as 3.98 here, too.

Edit:shown as 3.97

I can confirm that my installation of 3.99 is has 3.98 of MAC.exe.

I have just created an ape file with the GUI in insane mode. Foobar 0.7.7b will not play this "ape: undefined". Does foobar not play 3.98 files, or are these 3.99 files?

I could confirm this.
Encoded with the CLI component of foobar2000, the ape files are OK, shown as 3.98 in metadata.
But encoded through the new GUI, the files are unplayable with fb2k 0.81, but are labeled 3.99.

Weird...

http://www.monkeysaudio.com/cgi-bin/YaBB/Y...&num=1083260164

I noticed MACDll.dll in system32 is updated to 3.99 while its frontend isn't.

Edit: I meant its console frontend.

i surely enjoy .flac.

What a wonderful insight into your personal preferences! THANK-YOU!

I like choclate.

anyway, let's hope the updated release will be up and about soon.

edit: lol, Roberto beat me to it ;-)

Is Monkey's Audio open source now? I saw this on the homepage.

It's been for more than two years

3.99 versions:
Console frontend : http://www.saunalahti.fi/cse/temp/MAC.zip
foobar decoder: http://www.saunalahti.fi/cse/temp/foo_ape.zip
foobar encoder: http://www.saunalahti.fi/cse/temp/foo_monkey.zip

Haha I was afraid of that. My bad.

Wow that was fast Case, thanks.

Decoder seems to work fine on 3.99 files here. On insane files there is a long pause before playback starts, but I'm attributing that to the nature of insane apes.

We'll be updating the DirectShow filter shortly

Thanks Case! Good to have you back.

@rjamorim: Zero tolerance style.
Hehehehehehehe
How about encoding time compared to others lossless codecs?

He asked for it :B



Let's wait until Hans Heijden or Speek update their comparisions.

http://flac.sourceforge.net/comparison.html

http://www.monkeysaudio.com/comparison.html (waaay outdated)

He Case,
Nice to see you back
And thanks for the files.

Flac isn't up-to-date either.

As far as I can see, so far no comparision is featuring Monkey's Audio 3.99.

I know, but the flac page is at least more up to date than ape's... missed Speeks comparison though.
I thought SoNiX's question was more of a general nature, and not so much pointed at the brand new release...

Just a quick test of 3.99. Sorry, I didn't have time to compare to another encoder (e.g. FLAC)

System
------
Athlon XP 2400 at 2055mhz, 137mhz FSB (274mhz)
768mb PC2100 RAM
80Gb Barracuda IV Ultra DMA mode 5
Windows XP
Monkeys Audio GUI 3.99.

Encoding results
----------------
Album: AC/DC - If You Want Blood You've Got It (Metal)
Original WAV's: 534.9mb
Insane - 380.42mb, 7:55 encode time, 6:40 decode time
Extra High - 381.20mb, 2:19 encode time, 2:31 decode time
High - 382.42mb, 1:20 encode time, 1:33 decode time
Normal - 383.03mb, 1:12 encode time, 1:26 decode time
Fast - 386.54mb, 0:55 encode time, 1:08 decode time

Note: Insane files won't play gaplessly with Foobar2000. There are long pauses before the start of each track.

Yes things are beginning to pile up, things to update (am I missing anything?):
Monkey's: 3.97 -> 3.99
OptimFROG: 4.506 -> 4.509
Shorten: 3.5.1 -> 3.5.2 (bugfix only)
RAR: 3.11 -> 3.30 (dunno if improved)
TTA: 2.0 -> 3.0
Apple lossless
RealAudio lossless
WavPack 4.0 stable would be nice too, say, tomorrow?

Give me a few days, meanwhile Speek is always faster with updating!

Edit: Thanks Yaztromo, insane mode appears similar to 3.98 alpha 1 results, see here.

I just have some lossless albums on my notebooks, encoded with 3.98 at -c2000 (= normal). A fresh encoding with 3.99F shows slight improvements in bitrate. Here are the results for three discs :



Note that all these discs are classical music (Vivaldi & Haydn Concertos and contempory chamber music for the last one), and average bitrate is generally more friendly than with other kind of music (~41% on average - tested on > 300 CD).

Seems to be worth to upgrade, though I didn't test encoding and decoding speed.

I have tested 7 files with maximum compression.

1: 10 CC - The wall street shuffle.wav
2: 10,000 Maniacs - Noahs dove.wav
3: 3 Doors down - When I'm gone.wav
4: Abba - Dancing queen.wav
5: Air - Cherry blossom girl.wav
6: Bob Marley - Is this love.wav
7: DJ Food - Strange taste.wav

Or maybe a Release Candidate the day after tomorrow..

Heh, I gave the new version a quick test on ~25 random tunes:



So that's about 6mb saved for each gb of ape's by moving to the new extra high, but is that worth the 12% higher decode time?

It'd be interesting to know what -insane does at the moment. I know it's experimental still, but it saved about 300kb in a gig of waves and adds 3s to seek times

Has anybody else tried playing 3.97 files with the 3.99 Winamp plugin? I get a noticable skip between tracks which is fixed by reinstalling 3.97 and it's version of the Winamp plugin..

Just to clarify, this is playing 3.97 -extra high files, not the new insane files which suffer megalag.. hmmz..

I frequently get checksum errors when seeking quickly through ape files at normal compression and above. Fast compression doesn't seem to have this problem though.



Is this a known problem?

Still no pipe support for mac.exe, pitty.

edit:
oh one more thing, why is the decoding so terribly slow ?

hmm.... i coundn't get it work, it just says "i/o error" everytime i try to start the compression...

Short answer: no

The following was moved to off-topic the last time this came up:

http://www.hydrogenaudio.org/forums/index....=17&t=17485&st=

Offhand comments:

It seems a bit strange to call this a point release (point .02 actually) when the new encoder breaks the old decoder.

Thank you Case for those quick foobar updates. Perhaps you have not completely lost interest, or you just like Monkey's a bunch.

It would be nice if the mac dll could be obtained without downloading and installing the entire 3 mb package.

I disagree. And I agree with Garf's point on this subject. It IS open source, considering the sources are... open!

Update on my website, including Monkey's 3.99, see the www link below.
Using Monkey's interface for now, because the 3.99 mac.exe is still missing.

Apple and Real lossless still to do.

Please comment on missing/broken links, errors, or whatever I may have overlooked!

for all those who don't see a link below ...
look here.

comments:
you might want to define what the %-values on the y-axis stand for (though it's pretty obious, but still...)

Done, added that X/Y-axis talk on the main page. Not in the graphs, don't want to waste space over there!
Also Apple lossless has just been added, and indeed its compression is close to FLAC.
I ditched BONK for that, but under "ALL (old)" the previous graphs are kept for comparison.

Edit: BTW, yet again there is a link below But seriously, is there a chance Digga can't see the PM/email/www buttons? Mods?

hrm. it seems that a certain person... *someone in the audiece asks: 'is it Digga?', but is booed down immediately by the crowd* ...has indeed the ability to NOT see special types of buttons, which are deeply hidden in the board structure...
I'm just realy used to hyperlinks and didn't look at these little buttons... my bad.

edit: typos

my result

original:
49,575,980 bytes (100 %)
2,116 kbps

3.99 extra high:
31,090,936 bytes (62.713 %)
1,328 kbps
comp. time (about): 10.5 sec.

3.99 insaine:
31,931,644 bytes (64.409 %)
1,363 kbps
comp. time (about): 35 sec.

why dose "insane" make bigger file than "extra high" ?

Please keep in mind insane is an experimental mode.

OK. i'll keep in mind. thanks.

add result at same sample (TOM WAITS in the neiborhood).



3.98 extra high:
31,168,364 bytes (62.869 %)
1,331 kbps
comp. time (about): 10.1 sec.

in this sample, 3.98 faster than 3.99, but 3.99 smaller than 3.98.

i wll use new 3.99 CUI version. many thanks for Case.

Matt has updated 3.99

Can anyone tell me how to transcode my 3.97 files to 3.99 while preserving the tags?

You can use foobar2000's foo_ape together with foo_monkey.

Is it OK that the foobar2000 plugin produces slightly different files than the current 3.99 encoder? They both decode to the same WAV, though.

Edit: Sorry, the WAV files are not the same. I will compress a WAV with Monkey's Audio and decode it back. Anyway, as I said, when comparing the decoded WAVs from 3.97 with the ones from 3.99, they do not match.

Tags?

What does foobar2000's "bit compare" tell you?