Last updated 1/5/94 The good news is that source is now available. Look in /IUMA/mpeg_players for the file mpegaudio.tar.Z We will continue to gather source and executables and hope that some enterprising shareware authors or academics will provide various platforms with real-time players. According to Jared V Boone below, the Xing real-time player for Windows plays only the lower half of each subband of only one of the two channels. By my ears, that's pretty good. Another worthy undertaking would be porting the source to the DSPs increasingly being found on motherboards and add-in cards, such as the Mac AV series' AT&T 3210 or the Turtle Beach MultiSound's Motorola 56001, for real-time full-quality encoding and playback. That would be cool. =) -IUMA staff Here's the latest word on other non-commercial MPEG audio players for Unix workstations. ------------------------------------------------------------ Date: Mon, 3 Jan 1994 12:20:33 -0800 (PST) From: Jared V Boone Subject: Re: MPEG decoder... I'm working on commenting/tidying the code and perhaps I'll add mono decoding too. Then I'll either e-mail it to you or put it in the 'incoming' directory of your IUMA directory on sunsite.unc.edu... Give me a couple of days. Unfortunately, my program DOES NOT decode in real time. But then, Xing's program cheats. It does not decode the entire file, but plays the lower half of the subbands and only one channel of a stereo pair. My program will decode the whole thing, but there's a price to be paid. Decoding 'together.mp2' takes approximately 797 seconds on a Intel 486DX2-66 Windows NT/Visual C++ PC, and 1152 seconds on a Intel 486DX2-66 NetBSD/GCC V2.4 UNIX system. So I guess that's about 3-5 times slower than necessary for real-time playback. I've got some tricks I want to try, but they'll involve a lot of code modification. I also don't think they'll make THAT much difference. We may be asking these processors to do more than they can. > Please do. We are using the Xing encoder. Aware Inc. also makes one, > but it's pricey and not for PCs. What others are there? I don't know of any besides Xing, but the plain fact that the Xing encoder is non-compliant to the MPEG spec bugged me and inspired me to write my own. Hopefully I'll have it together this month. I've got some ideas on modifying the perceptual encoder to improve coding quality. I don't know if the ideas'll lead anywhere, but who knows. I'll keep you posted... - Jared Boone (jboone@patriot.wtfd.orst.edu) ------------------------------------------------------------------ Date: Sun, 2 Jan 1994 22:57:48 -0800 From: Jared V Boone Subject: MPEG decoder... I have an MPEG decoder that I can make available. It is in C and I have succeeded in compiling it under Windows NT Visual C++ and NetBSD 0.9 with GNU V2.4. The code is rather rough, only decodes Layer II, and is rather slow. However, I figure if I release the code to the public, some rocket scientist can make it ran fast... My only conditions are that I am acknowleged and notified when someone uses the code in a freeware/shareware/commercial product. Let me know if you're interested. - Jared Boone, Oregon State University (jboone@instruction.cs.orst.edu) P.S. I'm also working on an encoder. It appears that Xing's encoder is not all that great (sound quality), and also does not conform to the MPEG-I spec. If you'd like, I can keep you posted on this as well... ------------------------------------------------------------ From: "Tobias 'Doping' Bading" Subject: Re: The latest word on MPEG audio players for Sun workstation Date: Wed, 29 Dec 1993 18:04:36 +0100 (MET) Hi folks, Michael Rose wrote: > I downloaded some information from ianc@sunsite.unc.edu about > IUMA audio files in MPEG format. What is the status of an > audio player for MPEG files? > > Mike Ok, here comes the actual status of my decoder: - the decoder is still in work!!! - it works with the GNU C++ 2.5.1 compiler - it runs on SUN SPARCstations - it can decode layer II sound files, all modes - currently the reconstructed PCM samples are written into a file and this file can be fed to an audio device with CD quality, like in Iris Indigo machines. - the needed CPU time on a SPARC 2 is slightly higher than the playtime of a single channel Layer II 32 kHz 96 kbit/s soundfile, like intel.mp2: playtime: 30 Sec, decoding time: 32 Sec (CPU time) Stereo sound files need 2 times the playtime for decoding. => No realtime playing for now!!! :-( Work to do next: - Think about the synthesis filter, which uses almost the whole CPU time. (Meditate how the XING guys do the decoding on 486SX machines in realtime) => write a new filter!!! (who likes a decoder that can't decode directly to an audio device?) I'm interested in any ideas for a fast synthesis filter. - implement layer I (for what?) - write a routine for playback on SUN 8-bit ulaw audio devices - test compatibility (other compilers, machines etc.) That's all for now, but remember: This programm will need some more weeks until it is really usefull. Bye, Tobias ------------------------------------------------------------ From: "Tobias 'Doping' Bading" Subject: Re: MPEG Audio ISO 11172-3 request To: ianc@sunsite.unc.edu Date: Thu, 9 Dec 1993 13:02:03 +0100 (MET) Hello, > Who else is developing/has developed > an MPEG audio player? What can you tell me about your project? I am working on a decoder in c++ for layer I and II for sparc platforms in a normal project at the university. The decoder is only 3 weeks old now, so I am still at the beginning. But with the tables that I just received from Michael Simmons (michael@ecel.uwa.edu.au), I should be able to do some tests of the code I've implemented. Michael is also working on a decoder, but I don't have more information about his work by now. An other student at my university is working on a decoder, too. I'll meet him next week to exchange some experiences. If you'd like to contact him, his address is hb@cs.tu-berlin.de I'll take a look at your archive soon. Bye for now, Tobias --------------------------------------------------------------- Please send any info you may have about MPEG audio players to ianc@sunsite.unc.edu. -The IUMA staff