Matt Sandstrom

the grain was blueish but the overall color was rather warm despite the use of a non b filter. i even had to dial in some extra blue in telecine to get the look i wanted. /matt

if you shoot the entire scene in the highest speed you need so that everything is in slow motion and speed up the normal parts in post it will look pretty much exactly the same as an in camera ramp, since the camera changes shutter angle to maintain exposure, but it will waste a lot of film. /matt

no, it's not that. there was no gain on the scanner and the noise doesn't look like video noise at all, just like huge film grain. it is a video issue in a way though since it didn't show projected, but not in that way. the blue channel is always grainier with tungsten stocks but my guess is that it also has more contrast and thus the scanner doesn't see the same amount of detail. it would be interesting to see how it looks on a spirit or similar. this was done on a flashscan which just like flying spot scanners seems to have a bit of trouble with very dense film. checking the film in a microscope sounds like a good idea. i'll try that. /matt

i looked at the blue channel only and it's quite horrible. the grains are up to 10 pixels in size, while for the green and red they are around one or two pixels. blurring the blue channel quite a bit really helps and doesn't seem to affect sharpness. black balancing doesn't work though since the blue grain is pretty much uniform across the whole range. /matt

ok, the grain does increase in telecine. not in size obviously but it's more visible. it's a kind of blueish noise that really pops when transferred to video, while i didn't see it at all projected. is that what you saw too alex? i'm looking into a post solution to this. i think it's mainly a black balance issue. clips to come. /matt

next to the projector. maybe four or five meters away projecting a two meter wide image. i did look at it up close too though, as well as from afar. i'm curious too. i'm happy that 64t seems to capture the information so that you *can* color correct. the reason nothing happens with kodachrome is because there are no subtle shades. ;-) /matt

i shot 13 rolls of 64t recently and i don't think it's very grainy. it looks a little grainer than kodachrome in some shots but mostly because it also looks sharper plus the extra shadow latitude brings out some grain that's hidden by the deep blacks of k40. it's certainly in the same class as kodachrome and i doubt if many people would notice the difference without a side by side. you saw what you saw, sure, but it's not what i saw so i think you should consider taking another look. i've only seen the footage projected so far though. i'll get back to you with more info after we've scanned it tomorrow. /matt

you can tint the highlights orange in post, as in million dollar baby. don't know if that was a digital effect or done on film, but i think this is the effect you're after? /matt

you can typically get worldwide festival rights even to quite well known tracks for as little as a couple of hundred bucks. /matt

i just had 13 rolls processed by frank (super8.nl) and it turned out great. as for the 85b issue, i shot some exteriors using the built in 85 filter because it wanted it to be a little blue but it turned out normal even though it was an overcast day with a very blueish light. i've heard that kodak did something to the stock between the test period and the release, and maybe it was to make it a bit warmer? i'll post a clip as soon as we've scanned it. /matt

make sure the windows are evenly lit. that way if they don't blow out completely they will still be a uniform white which looks pretty much the same. the easiest way should be to use fresnel light sources, or light every window with several lights from different angles. /matt

yeah, that's what i always do when i shoot anything other than k40 which i know my camera auto exposes correctly (and even then sometimes of course, for more creative control). like santo i gave up on figuring out the notching system a long time ago. every interpretation of the spec seems reasonable but everyone seems to be using their own so nothing is ever compatible. now, the question is just whether this camera has manual exposure. many don't. /matt

i don't know how the tri-x carts are notched but the notch for 160t actually means 100d if there's no filter notch. that's what the spec says and that's how for example canon cameras do it, but i hear some manufacturers handle it differently. that would be the simple solution of course, but he's using a camera that sets the asa automatically. /matt

i already posted these in another thread, but since you asked for more one light setups and i'm kind of proud of these, here goes: a blonde through diffusion over the window. a blonde bounced off foamcore then through diffusion. i just learned that this is called a book light. :-) the walls are white providing ambient fill. /matt

you probably have to cut your own filter notch for that though (easy). most cameras won't let you set them to tungsten if you pop in a daylight cart. or if you have an exposure compensation dial you can set it to underexpose a stop. /matt

sounds like plenty. ;-) (see my night for night topic) i'd probably use the kinos for frontal fill of the actors, simulating moonlight, then the two 1k's as back/cross light from a direction where it's motivated. maybe use the smaller arris as "set lights", augmenting the pattern coming from the available light. then for closeups you can do whatever you like to make it look good as long as you keep it the same style and contrast, it doesn't even have to match. exact angles and the number of light sources are easily forgotten by audiences when you go close so it doesn't have to be real to be realistic. good luck. /matt

i wouldn't call billions of meters plus one the double of billions of meters though, not even if i choose to accept huge roundoff errors... ;-) /matt

look, i don't know the exact math, but if you're doubling your distance from the center of a very large light source, you haven't doubled the distance to its edges. this must make for a slightly different equation, right? /matt

the point is that the original is a jpeg pulled from a dv file on mac, each step adding its own idea of how color should be represented, and no matter how you twist it you can't adjust to any common reference. i could post a dv clip, you could correct it on your broadcast monitor and send it back, that would work. or i could upload the jpeg with a solor sync profile which you could color correct on your calibrated computer monitor, that would also work. otherwise finding a common reference is impossible. when i read this post on a pc at work the images are way too dark, while on my mac they look ok, and if i connect my mac to my tv they get really bright. well, you see what i mean, right? /matt

yes. i put high wattage regular bulbs or photofloods in cheap chinese lanterns from ikea. works great. they provide plenty of light that's very soft and wrapping yet directional. hang a few of them above the audience and if you're after what i think you are that should get you there. /matt

sounds like a job for china balls? /matt

i know what *you* said, but did you completely miss what he *asked* for? as for attacking, you gotta be kidding me? it was a joke. /matt

thanks. the film is 14 minutes. i guess i should mention that on the site since a lot of people think it's a feature. (or maybe that's cool?) ;-) since this is a cinematography site i should probably mention that i directed it but didn't shoot it. johan nordström did. check out his link on the site. sorry if we're hijacking the thread. back to the spotlight thing; i'd be very interested in more opinions on that. /matt

hmm, exactly what "professional" application would one consider buying an arri s for? ;-) /matt

autopoles are great and i've used them a lot lately. either between walls, which is what they're really for i guess, but also hung between two stands. /matt