Some of you know what my husband does for a living. But none of you REALLY knows, and I say that with confidence because: I don't really know what he does for a living.
When I first got to grad school in '93, I realized right away that I had no idea what they were talking about it. Not just the professors--the students, too. For nearly a year, I sat in my interminable three-hour seminars and made lists of words under the heading "Look Up When You Get Home." Unfortunately, none of the words were in the dictionary. It took me years to wield any of this "theory-speak" with authority, and only in context. I couldn't give you a definition of "epistemological" to save my life.
And then I met David. The first time I heard him having a conversation with a co-worker, his mysterious idiom was a bit of a turn-on. What can I say, I'm one of those girls who gets all twitterpated over guys who talk math.
Eventually, David started a blog, which is linked from mine, of course. But I'd wager that if any of you have ever clicked on it, you unclicked in a hurry, since the whole thing seems to be written in code. But you're in for a rare and educational treat: Today I've chosen a few morsels to deconstruct for you. Without further ado, your education in video software engineering. You're welcome.
STUFF DAVID HAS WRITTEN IN HIS BLOG, WHICH HAS MANY MORE VISITORS THAN MINE:
As CineForm is wrapper agnostic, we needed to store it differently so that it can't be lost. CineForm is the name of their company. What they sell is a little bit like rootbeer barrels, but they are deeply ambivalent about being wrapped.
The demosaic algorithm is the process that converts the one value per pixel of a raw bayer sensor to the more useful three color primaries per pixel like RGB (see Wikipedia article on demosaicing.) Well, what are you waiting for? Find that article! How can you go another minute without knowing about demosaicing? The raw bayer sensor is getting a headache from your recalcitrance. The rest is simple economics: which would you rather have, one value or three color primaries?
CineForm has become a very high-quality native acquisition format, making the issue of transcoding moot. Kind of like what happened in "Jesse's Girl."
CineForm is the company most guilty of using the term Visually Lossless (VL)--we may even have coined it. It saddens me to admit that until this moment, I didn't know the extent to which my husband was guilty. Now I learn he's been bandying "Visually Lossless" up and down the west coast, not caring a bit who hears. Discuss among yourselves--I'll give you a topic: Is Lossless the same as Found?
There is a point in most high bit-rate image compression technologies when the amount of loss is insignificant; this occurs when the compression error falls well below the inherent noise floor of the imaging device. The inherent noise floor is the bane of my existence. Noise on the floor, ceiling, walls...my bit-rate is totally compressed.
Lossy--sounds bad doesn't it? But even visually lossless compression is not mathematically identical to the original. To reliably achieve image compression above about 2:1, a lossy compression is required. If that isn't a rhetorical question, well then, yes, I have to admit--it sounds bad. Lossy, come home!
Linear compression eats shadow noise - that may be perceived as a good thing. Unless you're the shadow noise. Then you probably perceive it as a bad thing.
The problem arose when I did state that there can be a subtle issue for chroma keying when using any 4:2:0 24p signal encoded into 60i. That saucy thing! But I see his point--chroma keying can be so subtle, sometimes you forget you were ever encoding it. You forget everything, in fact--your name, your age, how to get home...luckily, 60i is one of those rare natural phenomena that might not occur in our lifetime.
24p editing on a 60i timeline is far more common than it should be, and just like interlace video, it should be outlawed. I just knew he was hiding something. The horror! An anti-Interlacist, living under my roof!