[Coco] CoCo Video Player
John W. Linville
linville at tuxdriver.com
Thu Mar 17 15:02:29 EDT 2011
On Thu, Mar 17, 2011 at 06:47:41AM -0500, Boisy G. Pitre wrote:
> It's interesting how these feats of hardware have only blossomed of
> late, and it is fun to speculate the impact of such a demonstration 10,
> 20 or even 25 years ago when the CoCo 3 first came out. John's work
> on the video player definitely has that "cool" factor, and you can
> see that by looking at the comments on the YouTube video of his demo.
Someone (William Astle?) described the player as a whistling dog --
amazing that it can whistle at all, but it really can't carry a tune...
> Several key pieces of technology that allowed John's player to work
>were either non-existent or available but out of reach for most of
>us back in the CoCo 3's hey-day:
> 1. The SuperIDE with its large CompactFlash allows John to stream
>video data fast enough to blit it on the screen. Perhaps there were
>SCSI drives back in the late 80s and early 90s that could maintain
>the same data rate, but I'm not sure if they could keep up at the
>same rate as the SuperIDE with the controllers available back then.
True. OTOH, the basic technology might be applied to a smaller video
resolution and a specially crafted video (particularly an animation)
might be used to get something with lower bandwidth requirements.
Hey, we are all speculating...right? :-)
> 2. The video software conversion tools that John used to compress and
>modify the original video were probably not in existence back then,
>although they could have been in theory. Assuming they did exist,
>the hardware that they would run on would be so slow comparatively
>that it would make encoding unfeasible due to the inordinate amount
>of time it would take.
> John I'm speculating a a bit on point 2, so feel free to comment more on it.
The conversion process is lengthy. It could probably be streamlined
by someone more familiar with such conversions. My process involves
busting videos out to a sequence of individual PNG (i.e. bitmap)
files, scaling them to the appropriate resolution, mapping the colors
to the available palette, computing frame-by-frame differences,
encoding the data so the player can understand it, etc. -- much of
it done with standard ffmpeg and ImageMagick utilities. IIRC, the
whole process tends to take roughly twice as long as the playtime of
the original video, even on modern machines.
More importantly, there was no youtube or similar source of videos to
convert. But again, someone motivated to create content specifically
for the technology might have been able to do some short videos.
Anyway, the storage is the main thing -- you won't be doing much
video using 5-1/4" floppies...
> I would like to see John take this project back even further, to
> a CoCo 2. More challenging since it runs at half the speed of the
> CoCo 3 and has nowhere near the color range.
Hah -- is that a challenge? :-)
I did find a guy doing something similar by playing video on a Pac
Man acrcade board. That is even more difficult, since the video
hardware is tile-based. :-)
> Go CoCo!
This, of course, is what it is really about. While it might be cool to
put some video sequences into a game or even to build a laserdisc-style
game completely around video clips, it is true that the video stuff
is really a novelty. I could have done a project that synchronized
lights in my yard to music on a loudspeaker, or that sent a tweet
whenever we run low on Mt. Dew in the fridge. Instead I decided to
do something on the CoCo that most thought couldn't be done (and some
still say it can't be). And, I had a ball. :-)
John W. Linville Someday the world will need a hero, and you
linville at tuxdriver.com might be all we have. Be ready.
More information about the Coco