>Being a VFX nerd for projected features, most of my work gets scanned and compâ€™ed or poked and jiggled and etc before being filmed out again to become part of someone's film â€“ or outputted to join the workflow at the DI step.
>As such, when I am shooting stock tests, etc, I try to look at the material printed first, and then also look at neg scans, since, as we all know, neg latitude is greater than print stock latitude, so there is plenty of information on the neg that won't get seen in projection - either in the shadows or at the highlights, depending on where you print. In the best of all possible worlds, I put the meat of the picture on the same part of the curve that the DP is putting his movie so that the least amount of messing is necessary at various density levels - the less that has to be done in post to make the curve match the rest of the movie, the more likely (in my mind) that the footage will look like it belongs in the movie. Obviously, lots of shots have to be shot differently for a variety of reasons, and often on different stocks - but conceptually, I try to stay true to what the DP is doing with the rest of his movie.
>Now, for those of you who work primarily for SD broadcast finish, where the final output has markedly less dynamic range than your neg, if you are using digital cameras as reference tools on set, do you find that the dynamic range of your digital camera is giving you a good idea of what your dynamic range would be in a 'straightforward" TK where you are not seriously boosting or crushing blacks or seriously messing with curves?
>It seems to me that the inherent "shortcomings" of the digital camera's dynamic range would serve as a "contrast viewing glass" of sorts to help dial in exposures...with the obvious caveat that (at least as far as my tests are implying) the chip is digging into the shadows more than the film at the same ratings but is losing highlight information LONG before the neg does.
>I have finally bit the bullet and bought my own D70 and am working on some stock tests to dial in a reasonable correlation between what I see on the digital camera vs what I am use to seeing on Pola 3000 B&W film vs. what I see on a piece of scanned neg vs. what I see on projected print.
>To the degree that I can develop a quick "cheat sheet" of how to correlate this new tool to what I've been doing for the last couple of decades, I can get comfortable using digi as a confirmation device instead of being distracted by yet another bunch of settings to adjust.
>The last thing I need while shooting is another set of variables to mess with.
This was one reason I opted for the D70, by the way, which allows me to set ASA's in 1/3 stops - when the sun is going down or the actress is aging in front of my eyes, I would rather be able to leave the shutter speed and f number set where they are on my taking camera and adj the asa so I can easily see if I have accidentally jogged one of the parameters - another reason I am using "ancient" manual focus AIS prime lenses - f2 is f2 is f2 - even if I accidentally jog the wrong
LA based writer of overly long posts
>Working mostly on commercials I find a digital camera as a reference tool invaluable (I use the D70 also).
>There are several aspects to it :
>- It functions as a "Polaroid" to quickly record contrast levels, etc.
>- It convinces clients looking at a videotap quickly that that reflective object highlight is not going to record as bright as they see it on the tap etc.
>- If I'm going to dial in an extreme look in TK I can quickly simulate it on the day in Photoshop (I do it in Photoshop, using levels, color balance and saturation, recording each step in Actions, so I can quickly apply a look on the day since time is always precious), dial it into a specific setup and tweak my lighting to make things fall into place.
>- Often I'm not available for TK, so I can tweak each set-up and send stills along
>- It's great for tech scouts: the image sensor roughly matches the size of a 35mm frame and thus it can function as a director's viewfinder in a pinch.
>You can preview depth of field even and match lenses.
>Florian Stadler, D.P, L.A.
>I keep a Nikon D1x on set, which I believe is similar to a D70, and find it more useful and accurate than my Polaroid. Without doing scientific tests, I find the dynamic range and image information seen on my laptop LCD is a good match to film projected dailies (more or less normal processing/printing).
>You may want to check to see if your camera has a "tone curve" setting-- if so, be sure to take it off "auto" (auto "Tone Curve" adds contrast to lo-con images, and flattens hi-con). This is very important to do for consistency!
>You might find it interesting to shoot a grey scale, and calibrate your histogram display to reflected readings.
>Finally, even though I've set my camera to Adobe RGB color space, I always convert my images to sRGB before emailing them to anybody, as that is a more often a default color space. Of course, emailing a photo to an unknown computer is no more accurate than describing a look via telephone, but I still do it!
>At telecine, the colorist sometimes says, "Here's what *your film* looks like before I've coloured it, or done anything", which actually means, "Here's what *my telecine* looks like before I've done anything". As far as telecine, if the image looks halfway decent on my digital camera, then I know it will look great after telecine colouring. Of course, that was usually true with polaroids as well!
>With polaroids, I would often take 2 shots: the first overexposed +1 to check my shadows, and another under exposed to check my highlights. I don't find it necessary to do the same with digital; either the camera is better, or perhaps I no longer care about little details like blown highlights!
class="style7">>It convinces clients looking at a videotap quickly that that reflective >object highlight is not going to record as bright as they see it on the tap >etc.
>I use a D100 for the same purpose but find that the big limitation is the dynamic range of these cameras!
>The highlights usually do clip. very much like the video assist monitor. I usually favour the highlights (or use auto exposure) to get all the info in...then Photoshop away...
>I don't find it that useful to set the stills camera ISO and shutter speed to match what I am doing.
>As we all well know, digital can't come close to the dynamic range of neg.