Apologies for the length of the posting. I hope I've answered all the questions...
> When photographing high key sets with dslr set on auto balance there may > well be changes between shots. Best to set a custom balance or at least a > pre set flash colour temperature. As I said, I use the preset tungsten WB when working in a tungsten environment. I discovered early on that AWB is pointless in a setting where there is no white. It's logical that AWB makes for inconsistent, constantly changing colours in this sort of environment. A preset WB followed by tweaking in CaptureOne is the best option. > What needs to be done is to pull the lighting back so that a white is > exposed as such and not blown out. I wish I had the luxury of this. I have to photograph what's there, and there's usually no opportunity for lighting changes, given that most productions are ready about 30 seconds prior to guests arriving. Sure, when time allows I do get the crew to adjust levels but it's often not possible. > What color neg have you been using? When I have shot color in the theatre, I > have found the ASA to be a limiting factor because of the slow tungsten > balanced stock. Also, is it simply a case of the neg capturing what you see > or are there tweaks (i.e. filteration) being applied at the printing stage > to compensate for deficiencies? It depends on the job - ie what the overall lighting looks like. For lighting containing subtle shades of reds and pinks I tended to use NPL as it separates the colours better. Sometimes Reala too but it goes a bit crazy where reds are involved, though its dynamic range is superb. I would sometimes shoot unfiltered and adjust in the printing - again depends on the job. Low ASA dealt with by using a tripod! >> The sensor sees only black and white (after filtering), as far as I can make >> out. It's the software that then translates the captured digits into color. >> If the software that's doing the translating can be tweaked (and it can), >> you will get different results. I think there are actually two problems. Now I think about it, you're right in that it's the firmware and software that are interpreting the colours (and therefore causing the problems with colour interpretation in some instances). Then there's the problem with 'blooming' as Tim described to me on email, which is leakage of electrical charge to adjacent sensors where high intensity light sources are involved. I hope you don't mind me quoting you Tim. > What "scientific" tests have you done to conclude the above? Please share > them as I am personally interested. It may be that the 1Ds is actually > giving you more accurate color reproduction! It just so happens the results > aren't as appealing as the film originated images because they have had > filteration applied to them at the printing stage. No 'scientific' tests as yet. Just a direct comparison with neg/dig. Neg held the whole range of colours on the set and you could differentiate between where reds overlay blues, with the detail of the gobo's showing clearly. Dig merged the reds and blues together to give the usual purple/pink dig appearance with no differentiation between different light sources, and the gobo patterns were invisible. Blooming also occurred even on the set where it wasn't that bright. I tried all sorts of exposures, and even where it was underexposed I still got the same problems with the dig images, but obviously not so much blooming. > You can also have your camera profiled or yourself tweak the profile that > Camera Raw II uses for your camera so that it gives you results that are > closer to the film originated output (that's before you do any tweaks in > Photoshop). Not familiar with Capture One (?) that most Canon owners prefer. > Maybe that gives you similar options. Yes it does. It's an excellent bit of software in principle, but it still has loads of bugs. > A good useable solution is to get a decent ring flash to puff a tiny bit of > quality light into the shadows so as to smooth exposure transition all the > way through to highlights. Just a tiny amount to fill the frame without a > hot-spot, and it wont show hardly - if at all. Maybe not for use on a > working film set, but very useful elsewhere. I always have a Metz tucked under my arm for any close-ups. Works a treat (not tucked under arm at this point obviously)... > The 1Ds, in common with most DSLRs (except I believe the Fuji S3) is > prone to blown highlights, so why not work in spot metering mode and > constantly meter from the highlights with the camera set in manual mode? > I don't mean lights, but the brightest area of tone in the image. In > keeping with film, set the max area that you want detail in to + 2 1/3 > stops and let the shadows sort themselves. The 1Ds can easily be setup > to flash the highlights on the display, but personally I would watch and > trust the histogram. As described above, it's not only the highlights that are the problem. And as mentioned yesterday, I'm dealing with large venues where I need a medium that can cope with a brightly lit, heavily saturated stage area and a darkish rest of the room, all in one frame. I do use the highlights custom function, and the histogram. It doesn't alter the fact though that it just hasn't got the dynamic range that neg has, and has the problems differentiating saturated colours in this sort of environment. > One final thought. By exposing for the highlights, you should move away > from most of the light halo problems. Perhaps you can use C1 to output 2 > versions of each image (light and dark) and use one the the shareware > pluggins to combine the 2 images to give the maximum dynamic range? > (sorry do not know names) Good idea. Any more info on this would be great. And finally, I should add that I shot another production job last night, this time no set but a 'theme' instead. The lighting, whilst saturated, wasn't concentrated on any surface in particular and therefore no areas of much higher intensity. It reminded me just how fantastic the 1Ds is in the right situation. I think I perhaps the answer lies in the fact that where film used to be good across the board, dig has limitations, and I just need to grill my clients a bit more about what I'm going to find when I get there. Many thanks to all who've contributed. I've learnt lots more (and got my grey cells ticking over). Jo =============================================================== GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE
