The Look of Midnight in Paris and To Rome With Love

[smthopr]

Oh, that's a relief. I got it right.

So how does it work? Do they correct scene per scene? I.e., they pick a frame from a scene, fiddle with it, and then batch-process the whole scene based on that one frame?

Would you say that colour-correction software is more or less powerful than its still-images counterparts?

In motion grading one generally works on a clip by clip basis. That is, one camera angle or cut at a time. When clips match right out of the camera, the grade can be saved and applied to all such clips. Often, filters, lenses do not match, and the clips must be matched by eye, one by one. Heavy ND filters, which are required for daylight shooting on modern digital cinema cameras rarely come close to matching and require careful grading to unify the scene.

Photoshop has more controls than most grading software. There are things I can do in photoshop that I can't do in grading software...yet. But most photographers don't use these techniques in photoshop either

[Atina]

You two, jjj and smthopr, will seriously get me hooked on this colour-correction thing!

So I have two questions: would "a seasoned professional" who is not the guy who colour-corrected Woody's two films be able to come close to their look and when seeing these images, would you be able to somehow define the look, to tell what you think was corrected in the images and how?

You, smthopr, still owe me the explanation about why you hate the images from the screenshots.

[smthopr]

You two, jjj and smthopr, will seriously get me hooked on this colour-correction thing!

So I have two questions: would "a seasoned professional" who is not the guy who colour-corrected Woody's two films be able to come close to their look and when seeing these images, would you be able to somehow define the look, to tell what you think was corrected in the images and how?

You, smthopr, still owe me the explanation about why you hate the images from the screenshots.

There was one particular link where the stills didn't look like the movie, but I don't recall which

[jjj]

You two, jjj and smthopr, will seriously get me hooked on this colour-correction thing!

So I have two questions: would "a seasoned professional" who is not the guy who colour-corrected Woody's two films be able to come close to their look and when seeing these images, would you be able to somehow define the look, to tell what you think was corrected in the images and how?

As there is no one way to do anything, it would be next to impossible to state exactly how something is done. People can guess at how they would try to do a certain look. Also it must be remembered that the look starts with how scene is shot, dressed and lit.

[Atina]

BTW, LOG recording in movies is done to record the entire dynamic range of the sensor (ie. very low contrast) while allocating more tonal gradations to the mid-tones and few tonal gradations to the highlights, which are more difficult to notice in the final color corrected image.  It's really a compression scheme to allow recording 10 or 12 bit files instead of 16 bit.

What does "very low contrast" in the parentheses mean?

By the way, obviously, in colour correction one can select an area of the frame and only work with that. Of course, the area shape and size can change so what does a colourist then do? This isn't that much of a problem, is it?

[Tim Lookingbill]

What does "very low contrast" in the parentheses mean?

By the way, obviously, in colour correction one can select an area of the frame and only work with that. Of course, the area shape and size can change so what does a colourist then do? This isn't that much of a problem, is it?

Drag & drop one of the low contrast LOG rendered images and view its histogram in your editor of choice where you'll see its highlights and black point are pushed in far away from the left/right max. Also note the tapered shaped of the data on both ends of this squeezed histogram suggests there is a lot of discrete steps of varying tonality and color detail where as if it looked chopped off on the ends ("L"vs "A" shape) would indicate a substantial amount of data on the ends (dynamic range) either got thrown away "clipped" or wasn't captured by the sensor.

This is output referred analysis of the data of course which the histogram reflects where as sensors record light linearly but that's not important to distinguish anyway.