Author Archives: Multimedia Mike

Audio Museum

I expect myriad museums on any touring vacation. However, I did not expect to see a highly relevant museum before I even left the country for my trip. At San Francisco International airport, there is a History of Audio museum exhibit sponsored by Dolby in terminal 3. Countless people have to walk by it on their way to and from flights but few people probably ever get a chance to study it (nor do they care, I imagine). Fortunately, I had loads of time to spare and made plenty of notes about things I wanted to look up later.

The exhibit features many artifacts from the history of audio recording. Two artifacts that stand out for the codec nerd are the hardware AC-1 encoder and decoder (2 separate units). AC-1? That’s right, AC-1. We all know of AC-3. The only reference I have ever seen for AC-2 is in the master Microsoft audio codec list (AC-2 has a wave format assignment of 0x0030). This is the first I have ever heard of AC-1. The specific hardware was the Scientific-Atlanta 97085 AC-1 enabled satellite receiver and the Dolby Model DP80 encoder, both from the year 1985.

Perhaps the strangest artifact was the Highway Hi-fi— an in-dash phonograph player that handles 45 RPM records. The exhibit didn’t actually have one of these specimens; the best they could do was an old advertisement featuring Lawrence Welk.

Further along on portable personal entertainment, the exhibit had a Toshiba portable MP3/AAC player claiming to be from the year 1997. This surprised me because I always thought that the Diamond Rio was the first. Perhaps it was only the first in the U.S. and the Toshiba model was available in Japan. There was also a representative sample of the legendary (which must have been their adjective, copied into my notes) Nagra portable.

I saw something called a Cartrivision cassette with a Fisher-Price logo on it, leading me to believe that FP was at one time into consumer electronics. Maybe, maybe not. They might not have actually made them, just branded them.

There were representative samples of gaming systems and games from various epochs. The last gen was well-represented since it’s so recent. Other random games showcased were King Arthur’s World (SNES) as well as 2 TurboGrafx-CD games. I took down the title of one as The Atlas [Atlus?]: Ron Voyage to which I can find no reference on the internet. The other one was Japanese and looked like an SNK fighting game but had a Hudson bee logo.

Getting into the movie side of things, my notes indicate that one of the exhibits had a remarkable replica of a Star Wars Y-Wing starfighter. That sort of thing always stands out for me. It turns out that Apocalypse Now was the first movie in 5.1 surround sound.

Cinema Digital Sound (CDS) was discussed and Dick Tracy was mentioned as the first film to use it.

The exhibit had a laptop with a static screenshot of a piece of software known as Dolby Show Manager. From the look of it, the software is used in digital theaters to schedule the main movie feature as well as commercials, trailers, everything.

Finally, perhaps the most well-specified — yet still mysterious — artifact I encountered was a hard drive that contains the first “DD-Cinema film”. I’m not sure what DD-Cinema is exactly. The HD listed a date of 2004 and had the following specs: “San Jose version, 2.39:1 unsqueezed 1920×803 letterbox, P3 color space, 5.1 surround EX ePCM, encrypted 100 Mbps, packaged for sys 1.0.4”. That’s all I know. It didn’t even indicate what the movie was.

The only other notable bit of multimedia from my trip was at the Beethoven Haus in Bonn, Germany. There is a studio with many specially configured PCs for studying Beethoven’s music. A firefox web browser is the primary vehicle for this. At some points, it would launch an external program for playing RealAudio or MP3 files. What did it use? Why, it’s Media Player Classic, v6.4.8.2. Particularly nifty, though, was being able to follow along with the music on original sheet music showcased through Flash. It all got me to thinking about music notation in the context of information coding theory.

Gone Fishin’

Vacation approaches. I’m heading out to Berlin.de soon for LinuxTag 2007. If you happen to be at the conference, do stop by the free multimedia booth to meet the crew. Oh, and be sure to attend my presentation, “FFmpeg: Past, Present, and Future”, which will be a fairly high-level overview of FFmpeg, its surrounding culture, the challenges it has overcome, and the challenges it still faces.

Answer to FAQ: Maybe the presentation will be recorded and placed online; that’s really not my department. Enough people seem to be interested that this will probably happen one way or another.

Samples Needed: E-AC3

One of FFmpeg’s Summer of Code projects is a decoder implementation for E-AC3. Thus far, the only samples available are rips from HD-DVDs. However, according to the E-AC3 mentor/student pair, these streams are not especially interesting for building a full-featured decoder; they don’t really flex the new features presented in the E-AC3 (a.k.a. a/52b) coding standard.

Thus, I’m wondering if anyone out there has the capability to encode new samples, preferably ones that make sure of the new bells and whistles? Or if anyone even knows what software can do this, that would be useful knowledge as well.

It would be particularly arduous to have to first write our own encoder so that we could also write the decoder.

Anonymous Drop Box

A correspondent who understandably wishes to remain anonymous is perhaps the first to come forth and point out what others have undoubtedly noticed:

Those naughty websites have a knack for posting samples with lots of errors … Maybe it would be easier to have a “clip bin” that’s anonymous, where you can submit clips ..that break the library.

We already do have the anonymous FTP at mplayerhq.hu for uploading samples. IP addresses are logged somewhere, but big deal. The bigger issue is that you sort of need to ping a development list somewhere in order to make them aware of the problem.

If you are determined enough, you can always go through anonymous proxy services to upload the sample and then use anonymous remailers or email services to alert the list or someone associated with the list.

Just part of our… umm… commitment to quality.

Mr. Anonymous mentions that the second part could be avoided by having an automated script that checks for new material in the upload directory and logs the information somewhere. I’ll file that away for when I finally get the autobuild/test system operational.