Thanks David, I'll post when I know what went wrong.
Thanks David, I'll post when I know what went wrong.
You may have been asked this numerous times, but what is the point of "re-rating" a stock? Why do some people choose to shoot a stock using an EI that doesn't match its own?
For instance, if you overexpose a fast stock (like 500T), then there should be a decrease in the appearance of grain. What are the downsides of this approach? In what cases would you want to underexpose a stock?
If you want to shoot at a speed that doesn't match the stock you have, then why not use a different stock? Aren't you sacrificing the potential latitude of the stock by "re-rating" it?
Please correct me if I've misused or misunderstood any of the terminology I've tried to present here.
You can't really make a stock less grainy by overexposing it - the large grains, having more surface area, are the first to get exposed, and their size is what determines (mainly) the speed of the stock. What overexposing does is expose the slower, smaller grains that are between the faster, larger grains, thus "tightening" the grain structure, making it seem more solid, so that you don't just see a pattern of swirling large grains as you do with underexposed high-speed stock.
A denser-than-normal print also requires higher printer light numbers to create a print of normal brightness, and generally the higher the numbers, the blacker the blacks are in the print. You can see this by just developing an unexposed roll, essentially a pitch-black image, and then printing it in the 10's, 20's, 30's, 40's... the blacks in the print get more solid, denser, at the higher numbers. So a regular image that prints in the high 30's, for example, will have deeper blacks than an image that prints in the low 20's, and with deeper blacks, you have richer contrast and stronger color saturation.
Now there is a tipping over point because the ends of the characteristic curve are flatter and thus lower in contrast. So if all of your midtone information is at the shoulder of the curve instead of straight-line portion, you can end up with sort of flattened highlights lacking in detail and sharpness, and greyish whites. So rating a stock more than one-stop slower doesn't gain you much and over that, you start to have this slightly flattened out look.
The other reason to rate a stock slower is to give yourself an exposure safety margin. Given that the image looks snappier and richer if a bit overexposed and slightly printed down rather than being underexposed and slightly being printed up, this means that with color negative film, if you are going to make an exposure mistake, generally underexposing is worse than overexposing. So rating a stock 1/3 of a stop slower, for example, isn't so much done to get you a denser negative all the time so much as keep you from ending up accidentally a bit underexposed as often. 1/3 of a stop is within a margin of error in exposing - it's only like a 2-point correction in printing after all.
It's particularly useful to rate a stock slower if you are going to be deliberately underexposing it for dark scenes. In other words, let's say that you decide that the correct look for a moonlit scene is 2-stops under for the highlights. Well, if you were rating the stock 2/3's of a stop slower as a base rating, then those shots that you deliberately were exposing to be 2-stops under to look dark are actually only 1 1/3-stops underexposed on the negative, so you have more leeway in printing and more protection from guessing wrong about how dark to make the scene. Of course what you are actually doing is underexposing partways and printing darker the rest of the way.
Thanks a lot, David! I certainly didn't expect this kind of response!
Your answer spoke quite a bit to the use of printing. Given the prevalence of the DI in a workflow, is there any point of over/underexposing in that situation? After all, can't you, to some degree, correct the exposure and look in post?
And a related question - if you overexpose a stop for tighter grain structure, you would develop normally, ie not pull the footage and correct the exposure during telecine?
David what thoughts do you have about taking a red one into a salt mine - there is 0 humidity down there, but salt is certainly present in the very dry air there. Do you think the electronics and camera are safe as long as I keep it inside a sealed plastic bag and just keep the lens element uncovered? I will be as deep as the Empire State Building is high.
Well, to some extent, there is less need to overexpose a negative if doing a D.I. if the only thing you want is deeper blacks and richer colors, since black level and saturation are adjusted electronically in a D.I.
But the benefits to having a tighter grain structure from overexposing a little bit are still there, plus in general, colorists like more information to work with, not less, and silver halide grains like exposure... and with film negative, you have such a long roll-off into overexposure that your high end is generally going to be OK, so a bigger worry can be shadow detail.
The advantage of having more shadow information is that you can essentially "crush" the blacks to make them deeper, stronger, and yet still control the degree of visibility in the shadows. The same principle works with digital photography.
I've found that in general, the principles that allow a good photochemical post or a good optical printer dupe still work well for a D.I. -- get a good exposure, a healthily dense negative.
Now you can go too far of course, start to end up with such dense highlights (on a negative, the bright areas of the image are the densist, darkest) that you can get noise in the whites with some telecines / scanners. For some reason, for example, I've tended to see noise in whites more often in D.I.'s that use Spirit Datacines recording to HD tape than I do with film scanners recording to RGB data files. Don't know whether it is an HDCAM-SR issue or a Spirit issue.
And some DP's like a thinner negative -- they don't like hard blacks and the snappier, contrastier look of dense negative printed down, they like the rougher texture (i.e. graininess) and softer look of thinner negatives. But they have to be very careful to expose (or underexpose in this case) consistently (one-stop under may get them the look they want, but accidentally underexposing another half-stop on top of that, for example, may fall them off of a cliff, there's less room for error).
A bit of overexposure leading to extra density can also save you if there is a bit of mild radiation fogging, etc. to the negative, if not using fresh stock -- printing down may compensate a bit for a higher base fog density. Also, over time, a denser negative will survive the aging process better in a vault.
But as you say, electronic color-correct can compensate somewhat for faded color or lifted blacks. Noise reduction can compensate somewhat for excessive grain. But you run the risk of the image looking a bit "processed" when you over-rely on post to solve these problems. Even with a D.I., it's still a "garbage in, garbage out" process -- the best shots come through the process the best, and the fastest and easiest, and thus maybe the cheapest. So there is still an incentive to delivering a well-exposed and consistenly exposed negative, unless you have artistic reasons for not doing that.
You have to test though because some D.I. facilities seem to prefer different densities than others. Also, you have to be honest about the way you expose. In some old American Cinematographer article, some DP's will boast that they can rate a 500 ASA stock at 1000 ASA and still be printing in the high 30's. To me, that just suggests that they are not really exposing at 1000 ASA, the way they use a meter or the way they expose for highlights or shadows are leading them to expose more than normal.
So you have to also think about the way you tend to expose. If you are prone to underexposing, going for a darker image, you may want to give yourself more leeway in post by rating the stock slower, and giving yourself more of a safety net if you expose incorrectly. On the other hand, some DP's are prone to overexpose all the time and don't need that exposure protection.
Last edited by David Mullen ASC; 11-13-2010 at 05:57 PM.
Now if the whole image is going to go through a D.I. anyway, it probably matters less whether you pull-process to compensate for the overexposure if it's only by one-stop... you can do it if you want to, but it may not be worth the extra cost or mild delay in processing (often pull and push-processing orders are saved for the end of the day).
The overexposed image, processed normally and printed down, gets this snappier, harder look but some people prefer to pull-process to compensate for the overexposure, which gets them the same grain reduction look but softer contrast and colors, a creamier look. Again, a lot of this can be done in the D.I. by playing with the gamma, so if a D.I. is intended, what really matters is delivering a negative with the grain structure you like and the amount of shadow and highlight detail you need to work with -- from there, you can manipulate the gamma, black level, and saturation to create any number of looks.
I have a question about zooming in/blowing up part of an image. I recently shot a project on the Red with the new chip. The director has a few shots that she would like to make tighter-- an insert of a letter that wasn't close enough to read easily, a group shot where she wants to single-out one character more, and a few others. She's gotten different info from different people on this subject.
She's been told that zooming in 50% is ok, since we shot 4K and will be delivering an HD finished project. This doesn't seem right to me. It seems that losing resolution is losing resolution-- and it's going to look like 50% less resolution, not to mention artifacts and noise issues.
Am I missing something or is she getting bad information?
|« Previous Thread | Next Thread »|