Karim D. Ghantous

I still haven't found out the stock that Ron used. I'll post below if I do. Thanks for all your comments, as I love to geek out on film. David, do you miss those Agfa stocks? I recall a series of ads for Agfa in AC, back in the '90s. I only have a few issues but one of them features the DP for 'Civil Wars'. So back then DPs had three distinct choices for stock: Kodak, Fuji, Agfa. I wonder if we will eventually get a company to take Fuji's place.

Ed Lachman said on the recent Kodakery podcast that some productions would release movies on Fuji print stock, as it was a bit cheaper. The result was that hues and tones would sometimes be a bit off (presumably because most of these productions were using Kodak camera negative film). That certainly had something to do with the '70s look.

Firstly, here is an interview with Ron by the ASC: https://ascmag.com/podcasts/twin-peaks-pilot-1990-ron-garcia-asc FWIW, IMDB says that Fuji stock was used for season 1, and Kodak stock for season 2. I have to assume that 500T stock was used throughout (that's what I would do!). Ron says, This list of Fuji stocks says that the F-Series was not released until 1988. However, the older AX 500T was already available to Ron before Twin Peaks. I'm guessing he shot that, and not the F-500T. https://en.wikipedia.org/wiki/List_of_motion_picture_film_stocks#Fuji

FWIW, based on what I have seen of raw scans, '219 handles push processing really well. Even in 16mm. You'd be surprised.

Forgive me if I am repeating myself from a previous conversation, but back in '85, a press photographer I read about pushed Fuji Pro 400 colour negative film by three stops, giving him ISO 3200. The results were pretty damned good, considering. I also hear that Ektachrome P800/1600 was a 400 speed film that could be pushed to 3200 if needed. Black & white film can be underexposed by five stops and give useable results. From what I've seen, 5219 can give you a good ASA 3200. I wonder what you think about that! It's funny, but digital cameras never offered good competition for high ISO until the Nikon D3, then the D700 (now superseded by cameras with smaller sensors). And forget using digital cinema cameras at 3200, unless you're using a Dragon sensor or better.

I have seen the film, but a while ago. I am going to guess that he shot just before sundown with a clear sky. Interiors seem to have used diffusion over windows (out of frame). In short: soft, natural, medium contrast. If you're shooting 500T unfiltered, maybe overexpose by a third? In digital terms (sorry, I can't quite figure out the term for colour neg), using it unfiltered would be starving the red channel. Overexposing would feed the red channel a bit more, while not overfeeding the blue channel. I'm not sure how to translate that into secondaries. I'd have to check a colour wheel. This looks interesting (I will make this part of my evening reading today): https://theasc.com/ac_magazine/August2011/TheTreeofLife/page1.html

I don't have a technician's answer, but logically, if you underexpose by one stop, you're gaining one stop in the highlights but losing at least one in the shadows. I have heard different people say that for every stop that you push, you lose a stop of DR. If you're pushing, you will lose the highlight headroom you would have got had you just underexposed. So it seems that highlights are the same, but you're losing some shadow detail. I wonder if it would be better to underexpose and correct the negative in the computer. Unless you want optical prints the whole way.

There are web sites for those topics. ;-) But seriously, we all do care about writing and performances (and marketing and producing). We happen to also know and care about lights and cameras. Edit: I finished writing a short film recently. I can't tell you if it's good, but I can tell you that I enjoyed writing it.

When I was 18, I wore white denim jeans. I hope to never make that sort of mistake ever again.

I am really enjoying this discussion. There have been some very interesting comments. Perhaps peripheral to the OP's question, but interesting nonetheless. And who doesn't love hardware porn? If image capture was as simple as audio (it can't be, sadly), we wouldn't need these discussions. The most we would argue about is whether we should want an analog intermediate (some recording technicians really like the 'thicker' sound of tape) between our digital master and our digital release. But to respond to the original question directly, I have two answers: 1. Digital has killed film for IMAX, IMO. RED has at least two flavours of 8K sensor, and this means that you have movie cameras the size of a Hasselblad 500 that can shoot images that fill a 100' wide screen. So in one specific instance, film has arguably lost. I would counter that by suggesting that 8-perf 65mm could replace 15-perf for IMAX, thanks to modern stocks. 2. When digital sensors can render light sources (like traffic lights) properly, only then can we begin discussing the OP's question. It has not begun yet. We still have aesthetic considerations, of course - which is why I said 'we can begin the discussion'. That is not correct. DR is not affected by the addition of a global shutter. You do lose a stop, that's true, but not DR. Unless you're talking about a specific camera? BTW the notion that we prefer what we grew up with is nonsense (no offence, David!). In any case, I grew up with both interlace TV (both NTSC and PAL) as well as projected film in a variety of situations. I've seen video displayed on many types of device, from big, triple-lamp front-projection TVs to plasma TVs to different types of CRT to LCDs to... you name it. Oh and for my fellow Twin Peaks fans, season 3 did not look as good as it could have if it was shot on film. Don't get me wrong - it didn't ever look bad. It just didn't look great. I loved Deming's work on Mulholland Drive, so I don't blame him. I blame the camera.

Well I see no point in DSLRs for video. They're almost completely redundant for photography (not quite yet, but almost). But for video, what are you going to use the mirror box for? The A7, or any mirrorless camera, is compatible with many more lenses than the D500 is. IMHO, FWIW, I'd prefer using APS-C cameras like the Fuji XT series - preferably the recent ones. I don't have heaps of experience in film & video, but I do like using smaller sensors with lenses from larger formats if they're available (in this case, VistaVision capable lenses). For S35, you can get really good cinema lenses - proper ones - for very little money. Worth thinking about.

Perry, I agree with 100% of everything you wrote!

Good question. I forgot! I remember calculating from a nominal resolution for 16mm. I think ARRI was one source. And also a few people saying that they got more out of Super 16mm with a 3K scan than a 2K scan. I know there's another source out there but I don't recall. So now I'm going to try some theory - bear in mind that I don't know how to properly interpret MTF charts. I have a photocopy of a data sheet for Eastman EXR 200T 5293. It says that the maximum resolving power, at high contrast, while averaging the R G and B layers, is about 175 cycles/mm. This seems a little high to me so I'm going to assume 100 cycles/mm to be conservative. If I am correct, 100 cycles/mm translates to 200 lp/mm, which translates to 200 lines/mm. A Super 8 frame has a specified height of 4.01mm, which I will round to 4.00. So 4.00 x 200 = 800 = 800 lines. I remember doing this sort of calculation for Super 8 Kodachrome 40 back in the '90s. At the time, I recall that K40 had a resolving power of 125 lines/mm, which gives a maximum of 500 lines of resolution, which was pretty good. But PAL was still a bit higher. But all this assumes that the numbers are meaningful and interpreted correctly. Obviously we're not saying that we're going to get 720 lines out of pushed 7219. And you need a sharp lens - consumer Super 8 cameras do not have the best lenses.

Scott, thanks very much for that example. I agree with your rationale - the higher resolution you scan with, the larger you can display the footage, regardless of its actual resolving power. If you are willing to share a couple of raw frames I'd love to have a look at them. I especially like the mountain shots. Very Twin Peaks like. ;-) In this case the lens was the weak point of the footage, so imagine what a good lens will show. The grain was very well controlled - I assume that you did not use grain removal? In any case, proper cine scanners never exaggerate grain, unlike photographic scanners such as the Pakon F135 or the Noritsu or the Coolscan. So it seems that Super 8 is the new 16mm. I did a rough calculation once, IIRC, and based on the fact that Super 8 can resolve 720 lines, 5-perf 65mm is equivalent to 10K across - 50% more than a RED 8K sensor once you take into account debayering (though keep in mind that the RED sensors & cameras are unparalleled achievements).

Well deserved!

Adam, thank you for showing us that. Super 8 is somewhat underestimated. However, in that example, the lab did apply grain removal and a small amount of sharpening. But in order to do that well you need a high resolution scan. 4K is perhaps overkill but I can see the logic of it.

Thanks for doing this. I look forward to seeing the 16mm stuff. BTW I don't suppose you could post a few scans of individual frames?

That print would not have been easy to produce. In fact I wouldn't want to try! But you got me thinking: do we grade shots or scenes? IMO, if I'm converting a movie to b&w, I'd do it a scene at a time, then do a second run to reduce any stark differences from scene to scene, if needed. Doing each shot separately would not be smart. I'm happy to lose some shadow detail, but I certainly would not want to clip important highlights, such as on an actor's face. I wonder if a DVD rip would give me a file that I can work on properly. I guess there's only one way to find out. ;-)

I'm all for it! In fact I've been thinking of converting a certain movie into b&w... ;-) But I need a good version of it. I think the free version of DaVinci Resolve should be able to handle that kind of project? I have never used it, BTW. I am surprised that nobody has made a TV show - of any budget - in b&w. Why not? It's a medium begging for a production to use it. Of course some works truly need to be shot in colour. The Wizard of Oz is the obvious example. Blue Velvet is another. The only criticism I have of your example is that there is not enough contrast. Contrast, contrast, contrast! I don't like looking at grey mush, if you get what I mean. B&w has to be crisp. Edit: Maybe the problem with your example is softness as well as contrast.

At least some scenes in War Horse were pull processed. This is a good start but it isn't very specific: https://britishcinematographer.co.uk/janusz-kaminski-asc-war-horse/ Lala Land was entirely pull processed, according to this: http://www.kodak.com/motion/blog/blog_post/?contentid=4295000679 I haven't seen either, but I'm looking forward to watching Lala Land eventually.

I am just a spectator in these sorts of discussions, but if I understand your scouting data correctly, it seems that you're going to want 200T. That allows you to get your exposure at night but you will have to use probably 2 stops of ND in total for the daylight shots. Stopping down or using NDs won't make a difference to the operator of course, but the focus puller would probably appreciate the narrower aperture. Maybe compromise so that one stop is taken care of by filters, the other by the aperture. I'm sure that 200T can easily handle massive overexposure, but it seems that this is not what you want, as it will introduce inconsistencies between the day and night scenes. Just my two cents. :-)

Josh, that was quite impressive, given the limitations. However, the one thing that I have noticed for quite some time now is that hand-held 8mm and 16mm cameras need compatibility with stabilized lenses (Micro 4/3 lenses would work really well with these formats).

I do recall that Kodak had the intention of producing small batches of old stocks. That was a few years ago and they may have abandoned that plan. I think what David said in his second comment makes sense. IMHO I would try and find (or make, as sometimes happens) filters that subtly affect the image in the way that I want. I have no experience with filters, but imagine experimenting with Ultra-con, diffusion, saturation and correction filters etc. If you combine that with multiple internegative and interpositive options, lenses, teleconverters, pre-development, chemistry etc, you have a very diverse set of tools. In theory. :-)

Robert, that would be very helpful information. The photographer I cited uses 5222 in his film cameras.

A photographer explains it briefly: https://iftimestoodstill.net/2017/04/11/before-after-2/ So, has anyone done this with movie film? If so, was it worth it?