?

Log in

No account? Create an account
Recent Entries Friends Archive Profile Tags My wildlife photography
 
 
 
 
 
 
David Attenborough, Mark Carwardine, Simon King, and George Monbiot spell out the case against the proposed badger cull in the UK: that the evidence gained has pointed out such a cull will be, at best, ineffective, and at worst, make the problem worse, at a cost of perhaps half the UK's badger population. Rather, they say, deploy an oral vaccine for badgers - considerably more effective, far more humane, and likely to be cheaper too. The Badger Trust has a Q&A on bovine tuberculosis (PDF), for further information.

Brian May has set up a petition over here - but obviously, opposition oughtn't amount to just that. Your MPs should hear as well, and the word should be spread - make the issue noticed.

Green Bytes apparently picked up Tens Complement last year, and are now releasing a pukka implementation of ZFS for OS X (64-bit Intel only, Snow Leopard or later, with a recommendation of being able to spare 1GB RAM per 1TB storage): ZEVO Community Edition, as noted here.

The BBC Radiophonic Workshop is to reopen!

Pinboard ("social bookmarking for introverts") looks sort of tempting - nicely fully featured, clean design, and no advertising or other junk, just a one-time signup fee of $9.89. Check out the testimonial quotes on the front page, too, particularly the Economist's. =:) If you're reading this, do you use Pinboard, Pinterest, Delicious, or similar services? How useful do you find them?

And I'll just leave this here..

 
 
 
 
 
 
So that's what you've been working on. Very interesting! Certainly looks useful for an intermediary format during editing, and will eventually be fast enough on specific hardware for direct rendering.
Eventually editing, definitely, as once you're out of pixelspace, lots of possibilities open up - for now, though, it'll be at the end of the pipeline, where the encode and decode speeds aren't so important, so much as the great bandwidth savings in having a single master file going out to all the recipients. As things stand, folks like S+M have to prepare anything up to 200 variants, depending on the framerate, aspect ratio, frame size, audio format, etc. That takes time, and even more bandwidth.. and the kinds of volumes of bits these guys push around the world every day are indeed eye-popping. =:D

And yes, in the fullness of time, dedicated hardware assistance'll be able to help push the speed up - we're very much in the "proof of concept" stage at the moment, but, the principles are vindicated, and the future of the format is promising. ^_^
And I'll just leave this here..

Sounds interesting! Hmm, I wonder why this sounds vaguely familiar. ^^
*grin* I've been wanting to speak of it more, but it was only quite recently that the core patent filing was made, so we've now got some degree of protection. And indeed, I'm in the process of preparing a draft rendering that'll be part of a demo for CVMP in December, showing the intermediate stage, without the normal full rendering, blended with the final result, so people can immediately see as they pass by that there's actually something going on. With any luck, I'll be able to go along - ought to be quite a networking opportunity, and fun. ^_^ (Pity it's not in a more exotic locale, but still, London's a fairly nifty city.. !)
Sounds intriguing! I wish I could be there to see that demo in action. :)
Oh, I should be able to link to it around showtime. ^_^ Rather looking forward to seeing what S+M can do with the footage.. !
Yay. ^^

I'm really quite intrigued by the whole thing, too. Even with a video codec that deals in vectors rather than pixels, how do you cope with the fact that both CCD sensors and displays are inherently pixelated?
The fundamentals aren't actually that strange - you literally contourise the image. It's a fairly time-consuming process at the moment, but entirely feasible, with a lot of different ways of going about it (and therein lies part of the secret sauce =:). Perform it well, and you wind up with just a huge pile of contours, which you can then manipulate in interesting ways rather more difficult to achieve in pixelspace, before reconstituting them. Scale as you wish, then rebuild to whatever size you want. (It's quite amusing to see how large images can be magnified - of course, you don't ever wind up with more detail than was in the original, but this approach does lend itself to retaining features without going pixellated)

Hence the pipeline: from pixels to contours, then at the other end, back to pixels, just in time for it to hit the display, or an existing broadcast chain.
Ah, cool! :)

Heh, contourization — I still remember being fascinated by how you could actually do that the first time I saw it being done. Come to think of it, I should still have my copy of Computer imaging recipes in C somewhere; not ever a particularly useful book for me, but I found it quite interesting anyway. Presumably the state of the art has advanced quite a bit since its publication, of course. ^_~


Oh man, leave the badgers alone!

That vector-based codec sounds pretty awesome. I always did like vectors and polygons a lot more than pixels. :D