MP3 repacker Options

[psyllium]

Hey does anyone have a copy of the rar file (for either the executable and/or the source) that they could send me? I'm having trouble downloading it.

Thanks

[neomoe]

removed my URL. wrong package. sorry.

This post has been edited by neomoe: Apr 9 2006, 10:32

[Omion]

Hey does anyone have a copy of the rar file (for either the executable and/or the source) that they could send me? I'm having trouble downloading it.

Thanks

I really shouldn't set my hosting computer to go to standby. My site should be working now.

I'm going to purchase a new computer soon to host everything off of. That should help the availability, since I crash my main computer far too often.

[Omion]

1.01 is out.

Squashed a bug when the bit reservoir goes off the beginning of the file (thanks, psyllium!)

The problem was that if, for some reason, the bit reservoir wanted more bytes than were actually in the buffer at any given time, the program would die. This only happens with corrupt mp3 files, but will happen quite a bit if the file was split with a naïve splitter that doesn't redistribute the bit reservoir data.

I have no idea why it was happening on psyllium's files, but it'll just put up a warning that the first frame(s) may be garbage, instead of croaking.

[psyllium]

1.01 is out.

Squashed a bug when the bit reservoir goes off the beginning of the file (thanks, psyllium!)

The problem was that if, for some reason, the bit reservoir wanted more bytes than were actually in the buffer at any given time, the program would die. This only happens with corrupt mp3 files, but will happen quite a bit if the file was split with a naïve splitter that doesn't redistribute the bit reservoir data.

I have no idea why it was happening on psyllium's files, but it'll just put up a warning that the first frame(s) may be garbage, instead of croaking.

Yeah the files were split with mp3splt.

[Omion]

Oops. It looks like I didn't quite squash that bug.

I 'fixed' it by changing the internal offset and length of the frame data to not go off the front of the file. However, I just noticed that I don't actually output the new frame length. This means that if you re-run mp3packer on the output file, it will crash with the same error message. I turned a buffer underrun into a delayed buffer overrun.

Anyway, fixing it may take a bit of time, since it looks like Visual Studio got COMPLETELY screwed up. I'll fix it as soon as I can (it shouldn't be too hard)

[Omion]

Oops. It looks like I didn't quite squash that bug.

1.02 out. Actually fixed the underrun this time. Also fixed another problem when a weird XING tag is present.

[Architectonical]

I'd like to try this, so if you could rehost it, that would be great.

[Antonski]

@Architectonical: Check your PM

[Gnerma]

Could I get a copy too? I could host it for the time being if somebody sends me the file.

Thanks

EDIT - nevermind, the link is live again. Thanks Omion.

EDIT2 - Well it tried it out. It's certainly nifty being that I'm all for efficienty. I'd do a complete run on my whole collection if that was possible

This post has been edited by Gnerma: May 15 2006, 02:51

[Omion]

Oops. I just saw the recent responses. Sorry for the hosting issues. My server is hooked up wirelessly to the router. It's about 8 feet away, but sometimes it just loses the signal altogether. I'll try to get a mirror on my main computer in case it dies again.

@all:
The links shouldn't be broken for more than a few hours at a time, so just keep trying. If that doesn't help, send me a PM and I'll e-mail it to you.

[Omion]

Released 1.03.

mp3packer will now check to see if the frame is silent. If it is, all the frame's data will be cleared (there's no point in having any data if it's going to be thrown out anyway)

This mainly affects FhG encodes, which will pad frames in such a way that mp3packer can't pack them. Actually getting the data would require a "rehuff" type program, which I'm not going to write. So I made a special case for silent frames. It's not going to help in the vast majority of cases, but the entire program shouldn't need to exist in the first place


PS. Hosting issues should be more or less gone. I now host the files off a RAM drive instead of a CompactFlash card. Much faster

[Gabriel]

PS. Hosting issues should be more or less gone. I now host the files off a RAM drive instead of a CompactFlash card. Much faster

What kind of host is it? A diskless one?

[Omion]

What kind of host is it? A diskless one?

Yup. I built my always-on computer to be as quiet as possible. That means passive CPU cooling and no hard drives. The power supply fan is the only moving part in it. But fitting Windows XP on a 1GB CompactFlash card was not easy...

Of course, I didn't realize the amazing slowness of CompactFlash (although I didn't buy the fastest one). The result is that something would time out while files were being read, making for lots of incomplete transfers.

I just installed a trial to a RAM drive program, which I now have everything hosted off of. Of course, the demo will expire in 2 weeks, so I'll have to shell out $50 to get the full version. But it's mighty fast!

[rutra80]

Omion, are you hosting your files on your machine because you want to, or because you have no choice?

This post has been edited by rutra80: May 28 2006, 21:34

[Firon]

I can mirror the files on my server for you, if you want.

[psyllium]

What kind of host is it? A diskless one?

Yup. I built my always-on computer to be as quiet as possible. That means passive CPU cooling and no hard drives. The power supply fan is the only moving part in it. But fitting Windows XP on a 1GB CompactFlash card was not easy...

Of course, I didn't realize the amazing slowness of CompactFlash (although I didn't buy the fastest one). The result is that something would time out while files were being read, making for lots of incomplete transfers.

I just installed a trial to a RAM drive program, which I now have everything hosted off of. Of course, the demo will expire in 2 weeks, so I'll have to shell out $50 to get the full version. But it's mighty fast!

If you use Linux, /dev/shm comes as a mounted ram disk automagically

[Firon]

I assume he's using Windows, since RAMDisk XP is about $50

[kwanbis]

I just installed a trial to a RAM drive program, which I now have everything hosted off of. Of course, the demo will expire in 2 weeks, so I'll have to shell out $50 to get the full version. But it's mighty fast!

did you try http://support.microsoft.com/kb/q257405/

or

http://users.compaqnet.be/cn021945/RAMDisk/ramdiskfree.htm

[Omion]

Omion, are you hosting your files on your machine because you want to, or because you have no choice?

Well, I technically have space on my university server, but I don't know how long it will be there. It's been a few years since I've been in school. I guess I mainly host them myself because I can. Of course, anybody can put up mirrors if they like.

BTW, are people still having problems with downloading? My apache log file is showing there haven't been any restarted transfers today. (it used to take a few tries to get the whole thing)


For those who want the gory details of my setup:
Windows XP (I seriously considered Linux, but I'm still more comfortable with Windows. I might end up installing Ubuntu when the RamDisk trial expires)
Super-slow 1GB CompactFlash card in one of these
Trial to SuperSpeed RamDisk Plus with two HD-backed drives (one for Apache which saves on shutdown, and the other for the data which is thrown out on shutdown)
The whole setup is connected wirelessly to my main computer, where I control it with Remote Desktop.

It's dead silent, and only has one cable coming out of it (power).


@kwanbis:
There are 2 big problems with most ram drives:
1) Most are limited to 32 or 64MB. I have at least 100MB of stuff on my site
2) Most won't save the image on shutdown or load it on startup.
The SuperSpeed program was the only one I found which could do exactly what I wanted (other than Linux, of course )

I played around with having Apache do some very aggressive caching, but that resulted in some bizarre server responses for some reason.

[Firon]

Here's a mirror. http://download.utorrent.com/mp3packer-1.03-123.rar
http://download.utorrent.com/mp3packer-1.03-123_src.rar

This post has been edited by Firon: May 30 2006, 02:14

[psyllium]

Version 1.0.3 of WinMP3Packer has been released here. Includes version 1.03 of mp3packer.exe

[bukem]

i've been playing with mp3packer 1.03 tool for a week without any problems till now. i've got one mp3 in my collection (CBR 192kbps) and every time i try to pack it with mp3packer i get following error in the middle of the process.

Fatal error: exception Invalid_argument("Buffer.sub")

I've checked that mp3 with mp3trim for any errors and found none. please find the "mp3packer -i" output bellow.

INFO:
MPEG1 layer 3
20401 frames
44100 Hz
38.281250 frames per second
532.924082 seconds
12772182 bytes in file (191.729853 kbps)
12771026 bytes in MP3 frames (191.712500 kbps) = current bitrate
96286830 bits of payload data (180.676448 kbps)
12036754 bytes of payload data (180.689962 kbps)
7202 bits wasted from partially-full bytes (0.013514 kbps)
12771190 bytes of MP3 data (191.714962 kbps) = minimum bitrate possible
-164 bytes of padding (-0.002462 kbps)
1156 bytes outside MP3 frames (0.017353 kbps)
Bitrate distribution:
192: 20401,0
Largest frame uses 6864 bits = 858 bytes = 262.762500 kbps
Smallest bitrate for CBR is 192

isn't a negative value for padding a strange thing?

ps. i can upload somewhere the mp3 file i'm talking about for debugging purpose if needed

This post has been edited by bukem: Jun 24 2006, 18:29

[Tonio Roffo]

WinMP3packer fails to install on XP x64 - keeps asking for .NET framework 1.1.x, I have x64 .NET framework installed.

[Omion]

i've been playing with mp3packer 1.03 tool for a week without any problems till now. i've got one mp3 in my collection (CBR 192kbps) and every time i try to pack it with mp3packer i get following error in the middle of the process.

Fatal error: exception Invalid_argument("Buffer.sub")

...
isn't a negative value for padding a strange thing?

ps. i can upload somewhere the mp3 file i'm talking about for debugging purpose if needed

A negative value for padding is indeed a strange thing. And that's probably what went wrong. For some reason a frame references bytes that have already been used in a previous frame. I thought I turned that into a warning, but I guess not.

If the file's smaller than 10MB, you can e-mail it to me. I'll PM you my address.