Edgar Nyari

There are plenty of brand new scans done of original negatives of 60s films. The entire early James Bond films collection for example is sourced from OCN scans on a high end Imagica scanner. I'm not saying this makes for any sort of "proof", just mentioning it as an example of "how a negative looks like", give or take some color tweaks.

As far accuracy of color reproduction goes; well Kodak, Fuji, Agfa and other manufacturers have been trying to improve that for decades. The orange mask itself was invented because the dyes are not "pure" enough, and it has been 50 or more years of Kodak trying to minimize unwanted color absorbtions even with an orange mask in place. They must have had SOME progress, enough for us to assume that Vision 3 has "cleaner" color reproduction even without a test. That and the fact that we have eyes and can see this.

 

I'm don't how you can be sure how 5251, which was discontinued in 1968, compares with modern stocks like vision 3, except in terms of grain. There are simply too many variables.

I didn't say how exactly 5251 would perform. All I'm saying is they are different products, separated by decades of research dedicated to getting more realistic color rendition. Those early materials, like the old Kodacolor negatives, and cinema products like 5251, had a pretty distorted way of seeing color compared to modern products. Sometimes it's difficult to articulate what one is seeing. One particular thing I do see in almost all telecined or printed materials from that time is a lot of development inhibitor artefacts sometimes (though rarely) to the point of dancing halos on the edges of objects, usually directional; in the direction opposite to the transport of the film in the processing machine. Why that happened I have no idea as agitation was always the best in motion picture processing machines compared to still image processing. This sometimes gives the objects that "outlined" feel. I don't see such things to that extend in modern materials.

Yes, Love Witch was amazing! I wish I could see a print of it; it would only accentuate the effect. But even the bluray looks amazing. Seems you had everything perfect, except of course the filmstock, which gave a slightly more realistic/modern color rendition than the old 5251 would have.

and did optical printing tricks, but those are the exceptions.

I have always been currious whether anyone ever attempted a crazy thing like contrast masking or somethging of that sort in an optical printer, like was normally done in a darkroom. Are you talking about crazy effects like seen in "2001" where they'd invert one color layer, or something more akin to what I just said?

In terms of the feeling or intensity of pure black or pure white, that's affected by the contrast range of the display method, and it's intensity.

Yes, it's a bit difficult to even discuss this because he is really talking about video/digital transfers of old vs. new films and not prints.

What I was aiming at is white clipping (in some DI work...), not the actual brightness of the projected "white".

And the "whiter whites" that Kodak advertises in their emulsion brochures has nothing to do with maximum density or brightness on screen, but

probably with lack of color distortion and grain in highlights compared to older emulsions.

This is a really interesting thread.

I'm no expert, and don't have much experience with film but I'm going to mention a couple of factors which I observed that might contribute to the difference in "look":

1. when seeing video transfers of older films, you are often (still) seeing a telecine transfer from an original IP , or even duplicate negative. That amplifies any kind of color-crossover issues and inter-layer effects that give that old-school film color palette. A scanned negative of an old film will look a bit more "modern" though, as seen on a lot of bluray "remasters" of old films, scanned in 2k or 2k from original negatives.

2. I know you might not want to hear this, but lighting plays a major role too. Hard light was used in old movies, and it brings out that classic "film look" even in more modern emulsions.

3. Old emulsions are just more "dirty" in their color rendition. Which is a point similar to my number 1, but relates to the negative itself.

4. There is rarely any WHITE in old films. That's for two reasons I think. Older emulsions don't really have clean highlights, and not only do modern emulsions have clean transparent and undistorted highlights, the stylistic choices of modern transfers, and DI , allows for clipped whites. Same thing goes for shadows.

"more real" is a matter of perception. I don't think older emulsions made a more realistic image. It's quite the contrary actually. But it does FEEL somehow more organic to me also, yes, which probably has a lot more to do with psychology than with image science.

UPDATE: Further thinking leads me to this conclusions. A lot of the scratches and lines are GREEN in color, which would suggest they are emulsion side scratches on the top magenta layer, which would suggest it's a print. AFAIK the top layer on the negative/intermediate film is yellow. Can someone comment on this line of thought?

Which leads further to the possibility that the 2K DI part was scanning those scratched prints. Or maybe an internegative was struck from those prints in order to get the density levels down for scanning.

Sorry to resurrect this thread, but I've been looking into this issue again after quite some time and have found new information about it.

In one interview the film editor, the now late, Sally Menke, explained the process of creating film scratches and other film damage for the first part of the movie. I quote:

"SM: We’d take a pen, a needle or some other implement, and scratch the film. Nina Kawasaki, my assistant, would go out and thrash it against the bushes on the driveway. We have video of it; it’s actually kind of funny. We kept asking the lab to make this section dirtier. We never even got it; we were too careful. We should have gotten it dirtier in some places. The lab had a lot of fun, though, not being careful. Want to smoke a cigarette over that? No problem."

This is from the volume 30 of the EDITORS GUILD MAGAZINE.

It seems that there is really nothing fake in Death Proof's "look". It's actual film damage, and I assume real splices, even though she doesn't mentioned the splices.

Another interesting thing is that the end credits feature a negative cutter, titles typographer, and color timer at Deluxe, and no mention of any DI work in any shape or form. It seems like this film didn't go through a regular DI at all . IMDB mentions a 2k DI process though and it could be at the stage of combining it with Planet Terror, using a photochemically finished IP maybe? The fact that print dailies (as seen in DVD/Blluray bonus materials exist of the entire film (even the second half) doesn't suggest a full DI was done.

But DI or no DI , what I wonder is, what sort of elements were they thrashing against the bushes. Would it have been a color timed IP, or prints?

Is there any way to tell from the damage if it's an IP or a print? And how likely would it be that all those tape splices were done on an IP?

Either way it must have been quite a strange workflow.

Thank you Tyler. That sounds as a convincing explanation.

No. The only transportation that the film has seen was via mail/courier. And after production this roll was in the same package as several other rolls which didn't have this problem at all. So whatever happened must have happened before it came into my posession. And I'm also sure it didn't happen on the way to me when I purchased it, because it was in the box with many other rolls that turned out fine. That being said, it's a short end, so it could have happened while it was still in posession of the previous owner (who has purcahsed it and resold it by the way...)

The important thing for me is that it's not a magazine or camera fault. Working with short ends of old film (this is 5218 so It could be pretty old) is a gamble, I know, but having faulty equipment is much worse.

More info:

looking at the rhythm of the flicker as it changes from the start of the roll to the end, I find something strange. At the start it goes in a different pattern:

flash flash flash flash (4X).....pause.....flash flash (2X).....pause (longer)........then 4X...2X.......4X...2X,

very strange. Towards the end of the roll (in the clip you see) it doesn't come in bursts of 4 and 2, but has regular intervals as you can see in the clip

Greetings,

I have this strange case of flicker on a single roll of film we shot. The entire roll seems to have the flicker but it's much more visible in dark scenes. At first I thought that one of my mags was faulty so the motor couldn't establish the film speed correctly so it kept speeding up and down. But I have a cinematography-electronics crystal control for Arri 35-III, and I doubt that it would behave like that and I don't remember seeing the light yellow/orange (which happens when it's not in full speed). Then I thought maybe the batter was too low.

But later I noticed that the brighter parts are not really overexposed (as it would be in the case of speed fluctuations), but they seem to be bluer and hazier/granier. Also when I stop the brighter frames and move frame by frame, I see a "wave" of brightness somehow pass upward on the image, instead of the entire image being brighter equally. And then towards the end of the roll it suddenly stops and the exposure seems normal and colors a bit warmer. On the frame where it switches to normal, it seems the frame is split and the upper part is a bit brighter and lower darker.

All this led me to conclude that it's not the motion of the camera, and that the roll was "flashed" somehow only on one side of it or something to that effect. If that were the case the rhythm of the flicker would accelerate or decelerate over time, depending on whether that happed prior or post exposure in camera (because the head and tail would switch their places from the core to the outside part of the roll). But I didn't measure flicker intervals to determine that. This is a short end, so it could be a loading error either on my part or the person that recanned it prior to me.

But I could be very wrong about this. Maybe it's even a lab mistake? Or is it the camera?

Can someone please take a look at this clip and comment? This is the very last roll of our shoot and a some of the crew staid for champagne to celebrate the wrap so we shot that.

Sorry for the YT compression and softness, but you can see what I'm talking about.

I've gotten a really good offer from a lab in Italy (which I will not name because the offer was far from their standard scanning price), for an HD 24p Spirit transfer. I did have to wait a bit for the Spirit machine to be available when most other scanning jobs were finished. If you think 16mm for 45 euro is cheap, we'll let's just say this was even cheaper. Ridiculously cheap. I'm so fortunate that they were willing to hook me up like that. I had about 22 minutes of footage to transfer. The HDD should arrive any day now. Can't wait to see the materials. This is not the first lab that showed interest in "negotiations", but this one turned out to be really ready to help out a small production like mine was.

I just found this:

This clip features some alternate takes where actors seem to give their regards to the editor. The thing is, these clips seem to have really been lifted from a print, because there are pen writings near the slate and that same print look (only more warmer in tone) even for the second part of the film (which on bluray looks a lot different and more like a regular film transfer). It seems that for some reason they made film dailies even though the film obviously went through DI. Maybe they used them for reference in the grading process to recreate a print look, or perhaps they actually used them in scanning of the first part of the movie.

Hi,

Does anyone have any information regarding the DI process for the film "Death Proof"? The first part of the film looks like it has been scanned from a print, but I'm not refering to the, probably fake, scratches and damage, I mean the way the image looks. Either that or it's a very good approximation of a film print done in grading. It's hard to put my finger on it though, what exactly gives that look. It's not just the contrast. There's something "in your face" about film prints which seems to be hard to emulate in grading.

Hi there,

To save me hours of googling, can someone recommend a relatively cheap scan service within EU in 2K. I noticed some labs use a Spirit Datacine (which AFAIK is not true 2K)....I'd prefer some true 2K scan like Arriscan.

thanks

Edgar

Yeah, I am indeed guilty of being unashamedly sentimental about film. I'm not even going to go into the whole issue of how insane it is to shoot 35mm on our budget. We'll have a Black Magic for backup, but I personally hope we never end up needing it. :)

Yeah. I was going by the orientation of the road. By the way, I am the clapper/loader in this frame, because I switched with my actual clapper/loader, just to have a souvenir appearance on film for a second or two :) You can even see the light-meter hanging around my neck. :)

Yes I'll do that for sure eventually. We are preparing a small production this month, which I hope to scan in 2K by the end of this year, probably Arriscan so I'm sure there won't be any such problems. Thanks everyone for your contributions. I'm relieved that I don't have a leaking mag or something that drastic.

David I think I know what you are seeing. I see a slight darkening also, but the problem is, it's on the wrong side of the frame. The side of the frame where the yellowish "leak" is in the teleciine is in the upper part of the neg on this photo. And over there I really see nothing. And as for car lights, and panning flashlights, as I mentioned in one of the previous posts, the entire roll of film has the same consistent yellowish line. It seems to be equal in width and intensity in every frame of the entire roll.

Yeah, thanks guys.

I think I'll make another photo then invert it and increase the contrast to the point where it would have to reveal any kind of density changes....stay tuned.

Finally I got the negative back in mail today. I see absolutely nothing out of the ordinary. But I guess I'd have to create an even-light background to see subtle changes in density. It's a very thin part of the negative. This shot was one or two stops under I think.

and the other side

I managed to make new photos of the crucial part of the gate. Here it is.

Here is the eyepiece "problem" I was talking about. Notice a mirror image left to the black framed area that contains the groundglass image. This mirrored image happens in the door, and not around the actual groundglass. Even when I open the doors and look away from the groundglass window and the camera body, into empty space, I can still see this. My other doors don't show this problem and have everything black around. Is this normal, or did something maybe fall off inside the door?