PDA

View Full Version : Grading of Films



Woebot
28-07-2005, 09:26 PM
I've been editing together a whole lot of big blockbuster movies for a TV channel that will remain nameless, and i discovered that when you bring the quicktimes up on a standard monitor they all look really dark and murky. to grade them i had to really crush the levels, the whole badwidth of RGB is really towards the darker end with just the scantest flashes of white in the top areas.

then it hit me. the way they grade these movies is so that when you watch them in the dark, your eyes become accustomed to the gloom and see it as normal. cinemas trump card is being able to show bright lights, like explosions or for instance light sabers, and have them really bite out in a quite unusually powerful fashion. your eyes are startled by their brilliance in the context of the generally much deeper darker hues.

bloody clever really.

Diggedy Derek
29-07-2005, 05:03 PM
There was a theory in the early days of film that part of the magic of cinema was largely down to the physical effect of exposure to bright lights in dark rooms. The art of cinema was felt, absorbed even. Can't remember what the theory was called though, something in German I think.

Woebot
29-07-2005, 09:29 PM
There was a theory in the early days of film that part of the magic of cinema was largely down to the physical effect of exposure to bright lights in dark rooms. The art of cinema was felt, absorbed even. Can't remember what the theory was called though, something in German I think.

well i couldnt really comment either but maybe you're referring to the "Platos Cave" thing.

i remember reading a piece about how historians of neanderthal man believed that the painted caves (like those at Lascaux http://www.culture.gouv.fr/culture/arcnat/lascaux/en/) were used as sites of initiation. that shamen brought unsuspecting young tribal members down into the dark and then "shocked them into a heightened conciousness" with light shows featuring the wall paintings. supposed to have been one of the vital sparks behind behind the growth of our intelligence.

that and, if you were asking terence mckenna, lsd.

all conjecture and waffle obv.....

blunt
14-09-2005, 09:45 AM
i remember reading a piece about how historians of neanderthal man believed that the painted caves [..] were used as sites of initiation. [...]

I've read something similar that suggested that the reason that neanderthal man was attracted to cave-like spaces in the first place was primarily acoustic - that they were intrigued by the echoing of their own movements and voices (well, vocalisations, at least ;)) - and that this, when coupled with the later emeregnce of cave painting, made for the Urgesamtkunstwerk.

Like you say, Woebot: all conjecture and waffle. But such great waffle!

ome
26-09-2005, 08:57 PM
As far as I understand colour theory and image processing (please correct me):

The major advantage to our eyes over film/video is the range of contrast range that can be perceived in one image (without iris movement)

Reality = 20,000:1
Eye = 600:1
Film = 150:1
Video Grade1 CRT = 50:1
Normal Video = donít know? About 15:1?

---

Films gamma curve (responsiveness to various levels of light at different levels) is like our eyes, we stretch (see more steps of differensation) the blacks and crush (see less variations of brightness) the whites - this is an exponential curve. Reality though is a flat straight line.

In addition to this we adjust areas of contrast to perceive shadow detail This is fairly major hallucination (i.e. we need to hunt in dark etc..). : see attachment for example or go here (http://miafx.com/23/tiki-index.php?page=Peception+Theory)

---
the techy bit

RGB is a colour space i.e. a way of mathematically dividing the pictures colour elements so it can be stored. RGB is uncompressed and the default processing mathematical mode that computers deal with images. Whether the image has a gamma curve or not is totally independent to color space. (Although colour space can use compression that lessens image quality, and some compression codecs use gamma curves to save space but when playing the image via the codec it only shows a linear image). Note Colour space is not the same as colour profile i.e. AdobeRGB.

Bit depth on the other hand is how many levels of variation are recorded. Most applications and monitors work in 8bit (765 levels of contrast variation). Other standards are 10bit (3069 levels) and 16bit (196605 levels).

Film quality can be achieved via a RGB file that does not have a gamma curve (called a cineon look up or LUT) with 16bit images. But this creates files that are practically too big to use, so what most people do is put on a lut or gamma curve that matches ours eyes, this louses the differensation in the highlights that we cant see anyway and then by using a 10bit RGB file, the file MB size becomes manageable. When processing a LUT always get the software to process at 16bit or higher(float depth), other wise the images whites become clipped (truncated to louse visable detail).

---

Note in addition to this, when dealing with monitors, they have gamma curve and create further stretching of the blacks that has to be compensated.

---

Never composite in log(mon-linear) - convert to Lin(linear) - this is because all the pixel maths especially with greys in the matte become incorrect) you can grade in log though.

--
read more on my notepad: miafx.com (http://miafx.com/23/tiki-index.php?page=Colour+Manipulation) & about film (http://miafx.com/23/tiki-index.php?page=Film)

jd_
31-10-2005, 08:27 PM
I remember in a Z'ev interview in that Research Industrial handbook that he was saying that if yu are listening to a prolonged held note your vocal chords try to harmonize with it, so that when there's a suprise and they cut from a sustained ominous tone in a horror/thriller movie to a blast of a much higher pitched sound it has a physically jarring effect on your body that adds a big punch to the experience. I never knew about the light trick before, that's interesting. I wonder if all this is just a result of small discoveries over time becoming trade secrets or if studios actually paid scientists to figure out the sorts of effects they could pull off with the tricks at their disposal.

Woebot
01-11-2005, 09:01 AM
science, ome!

jd_
09-11-2005, 08:13 AM
I read somewhere, although I forget where now (might have been here actually), that now some companies are bypassing focus groups and just doing brain scans on people exposed to their products to try to guage public reactions/adjust their product. That's gotta lead to some pretty spectacular mutations of consumer goods.

There was more to this post that I realised was inspired by a misreading of your response Woebot, so now it doesn't make that much sense, but there's no delete so I just clipped out the most wrong bits. Enjoy!