Example of shooting out of date stock.

Adam Frisch FSF · June 17, 2007

I asked for some advice in the Professionals forum on the 6th of May on shooting on failed stock and I though the answer might be interesting to post in an open thread. This was my question:

"I'm doing a music video early next week and we're doing a take on Boogie Nights, basically. The film-within-the-film is an early 80's type of porn film that we'll be cutting back to.

In trying achieve that "Debbie Does Dallas"-look in camera we first explored using reversal. But the Kodak 100D is just too contrasty and has too snappy colors - what we want is more washed out and soft. So we decided to do a print and TK that. So far so good. But I also asked if the production could get hold of some older EXR 7298 to get the right grain structure and make it look more correct for that period, which they did. But not from Kodak, it was from the rental house and they got a good price on the whole 4-year old batch. Only catch was they had to buy it to clip test it.

Anyway, clip test failed, not surprisingly. This is fine since I've shot with failed stock before. But what's happening now is that production is freaking out a bit and I basically have to more or less give them a guesstimation of the grain structure. Obviously, I'm confident it will be usable, but they want to know how the grain will look. They've been very good at not putting my head on the chopping block - we all agreed this was a risk we were willing to take, so that's fine, but still it does end up in my court. And to compund the matter further, there has not been enough time to shoot a test (it's a bank holiday here), which would have been the best thing to do.

Here are the density values:

Reference : R 0.19, G 0.59, B 0.96

Actual: R 0.38, G 0.91, B 1.31

So doubly raised values in all color layers. Nothing earth shattering in my 35mm experience, but might be out of my comfort zone on 16mm. Has anybody got any real world experience shooting on 16mm with similar values?"

And here's the follow-up:

I did end up using the film.

But all up until the very morning of the shoot the producers were to-ing and fro-ing (quite understandably) if we should use the old stock or not. Finally it was a no - we were not going to take the risk. But the producer said that he'd go with it IF the production insurance would cover a re-shoot based on faulty stock (bit of license there). And lo and behold it did, so the old stock was back on.

I rated the 7298 at 250ASA and off we went. All went fine, that is, until after about 14hrs of shooting when I notice that it says 7289 on the mag tape. I ask my clapper/loader if he's just put it down wrong and he says "no, it's correct - 7289".

I get a lump in my throat.

"So it's not 7298, it's 7289 we're shooting?". He shows me the can. "Yeah - didn't you know? I did ask you this morning if it was the correct one and you said yes". He did, but I must have misunderstood or heard him wrong - I had been under the steadfast notion all through pre-production that this was 7298, the old 500T stock. Now, the 800T 7289 was a notoriously grainy stock in 16mm even when it was brand new - 4 year old 800T wasn't going to be any less grainy - especially when it had been this unknowingly and heavily overexposed.....

I inform the producer and he informs the director and we all freak out a bit. They're sh***ing themseleves, concerned there won't be an image at all on there, whereas I'm perhaps a little more sure there will. Still, I did not feel comfortable with it all and all I could guarantee at this point was that it "probably will look quite close to super-8, if anything" - hoping I was right...

To add insult to injury, we wouldn't be in TK until the evening the day after, making the whole next day an agonising and stressful limbo-ish wait. The producer finally couldn't take it any longer and begged the lab and asked if I could pop down and just have a quick look at the sequences. I did and....

....it was all fine.

Just goes to show you how extremely sturdy film is. Even when it's almost 2 stops overexposed, 4-year old, grainy 16mm 800ASA that has failed it's density value test, it STILL manages to keep not only an image, but does it surprisingly grain-free. In fact, in the end we almost added some extra grain in post.

Anyway, the video can be viewed on my website at this link - shot all the film-within-the-film stuff on the old stock with an ancient Cooke zoom, some Fogal ladies nylon stockings bluetacked to the back of the lens and with a Varicon flashed with some magenta. The "normal" scenes were shot on 7217 straight:

http://www.adamfrisch.com/music/dragonettes/indexmovie.html

Lesson learned? I've never been very good with keeping stock numbers in my head, so I normally just tell production "get me Kodak Vision2 200T stock - can't remember the stock number, so check it out". This is something I'm going to continue doing, because those numbers are all to easy to screw up as was the case here.

Dominic Case · June 17, 2007

Lesson seems to be that if you plan to get a "different" look, you won't be disappointed.

The d-min figures (+0.19, +0.32; +0.35 from reference) suggest that if you overexposed by almost 2 stops, you would push the density of the negative enough to get clear of the fog level (which is where the extra gain is mostly lurking).

Also, that you could expect some colour twists, probably with the shadows picking up a cyan cast (since the red-sensitive layer is least affected by the age fog).

As you shot it, I guess the stocking on the lens was there to get some diffusion, and to flatten the blacks a bit, and the Varicon to warm up the shadows.

What you've ended up with is probably ideal: you (partly unwittingly) overexposed by exactly the right amount, and the cyan colour twist in the shadows was offset by the magenta Varicon flash - giving the blue shadows in the final result.

Of course (note for others) if you had wanted the 4-year-old, age-fogged 800T stock to look normal, and cut in with other, new, stock, there is no way you would have got away with that amount of fog.

One more comment: there never was much rhyme or reason in Kodak stock numbers, though they shouldn't be any harder to remember than phone numbers. The trouble with telling production to "get the V2 100T" is that they have a fighting chance of ordering 16mm instead of 35mm or vice versa.

David Mullen ASC · June 17, 2007

What drives me nut are handwritten labels where you can't tell if it says "5298" or "5248". I had that happen once on a shoot -- the mag was color-taped for 5298 but the handwritten label said 5248 and the loader couldn't remember which he actually loaded. Luckily the problem has gone away more or less -- I don't think any of the current color neg stocks could be mistaken for the other if badly handwritten. (01, 05, 12, 17, 18, 29) -- I guess 4's and 9's are the only two numbers that can look similar. So if you plan on buying any recanned 5274 and 5279, look carefully at the label...

The video looks great, Adam! You almost made the 1970's recreation stuff look a little too good... It's very clever.

Jon Kukla · June 18, 2007

How much magenta did you add with the Varicon? I'm just curious bc I've been sitting in on some telecine work on very old faded prints, and it's really amazing just how pink/magenta the prints are. Luckily the colorist was very good and was able to bring back most of the yellow/green/blue spectrum, but of course it's not quite what the original was.

Alessandro Machi · June 18, 2007

If that had been the exact same old film, but 500 ASA instead of 800 ASA, do you think the result would have been as good, as in looking as good as it did?

Your internet codec played really well, no dropped frames for me, what's your method?

Phil Rhodes · June 18, 2007

Hi,

It's 480x270 h.264 with at about 750Kbps overall, with 128Kbps AAC audio.

"Web High 2-pass" preset in Compressor, Adam?

Phil

Rupe Whiteman · June 18, 2007

'Nice looking work as always Adam...

Adam Frisch FSF · June 21, 2007

How much magenta did you add with the Varicon? I'm just curious bc I've been sitting in on some telecine work on very old faded prints, and it's really amazing just how pink/magenta the prints are. Luckily the colorist was very good and was able to bring back most of the yellow/green/blue spectrum, but of course it's not quite what the original was.

Can't really remember, but I think it was full minus green or something. Those old Varicons are a bit rickety and not out on jobs that much anymore, so they're not always in the

best of shape. Think I turned it up in intensity to about 8 (on a scale of 12), but it wasn't very precise.

Adam Thompson · June 21, 2007

Wow, looks great man. But what did you do after you knew the real stock? Did you change you stops fully or meet it in the middle somewhere? Also, why would going over 2 stops be that big a concern?

Jamie Metzger · June 21, 2007

Hey adam,

great looking footage. I actually enjoyed the video as well. I haven't been impressed by a lot of the concepts of videos these days.

I think the film came out great, and I only noticed one shot where there was an intense amount of grain, and it was a little past midway, and you can see it in the main actresses hair. Nothing huge.

I did something similar with thinking I was shooting 50D and I was really shooting 250D. We pulled one stop at the lab, and fixed the rest in telecine. No grain at all.

Jamie

K Borowski · June 25, 2007

Hey Adam, GREAT video. I agree, most of the poop out there these days doesn't get half as good as this, and that's saying something when a lot of them have multimillion-dollar budgets. I agree that hitting the 800 with an extra stop boosted the exposure past base fog. That stock wasn't a true 800 anyway, so even if it weren't 4 years old it would've benefited from an extra 2/3 stop. I did notice a little bit of possible fog (unless the flash was added intentionally or it was lens flare or something; there's a corresponding flash in the fresh stock too, so either you did a very good job of camouflaging it that way or it is indeed deliberate) in the part where the *Star* is whispering "You gotta take it like a man" into her "leading man's" ear. Frankly, for footage of that type, I think you should have gone for more grain rather than worrying about it. We all know the production values that those shoots had. I don't think fogged footage would have bothered them enough to reshoot anyway. Also, I really enjoyed the subtle touches: the rounded corners in the "production footage", the outfits, the surly characters behind the camera, the cue cards, the vintage camera, and the varicon affect made it look better, I'd say. Too bad no one else seems interested in getting it right in camera these days. I certainly think there's a real improvement when you get something just right on the negative.

Again, excellent job, great lighting, good cinematography, and great song too!

Regards,

~KB

Edited June 25, 2007 by Karl Borowski

thomas-english · July 6, 2007

great looking promo adam.

I have had a couple of miscommunication s recently. I overexposed a load of vision1 500t by 3 stops. I regretted requesting a pull of 1 stop but still the client was happy (focus puller thought there was an ND6 in but had forgotten to put it in). Bought back in Tk it looked ... ok

We underexposed a load of fuji 500t by 2 1/2 stops on a run through (s16 handheld 50mm) for a track to cut with the wide. I re did the run through anyway at the correct stop. When we bought back the underexposed run through in TK it looked ... OK ... but they still used it. I notice the cuts but they seem happy.

Basically if you ve made the mistake, relax ... just make sure you shoot everything that s gonna cut at the same mistake.

Adam Frisch FSF · November 29, 2007

Lesson seems to be that if you plan to get a "different" look, you won't be disappointed.

The d-min figures (+0.19, +0.32; +0.35 from reference) suggest that if you overexposed by almost 2 stops, you would push the density of the negative enough to get clear of the fog level (which is where the extra gain is mostly lurking).

Also, that you could expect some colour twists, probably with the shadows picking up a cyan cast (since the red-sensitive layer is least affected by the age fog).

As you shot it, I guess the stocking on the lens was there to get some diffusion, and to flatten the blacks a bit, and the Varicon to warm up the shadows.

What you've ended up with is probably ideal: you (partly unwittingly) overexposed by exactly the right amount, and the cyan colour twist in the shadows was offset by the magenta Varicon flash - giving the blue shadows in the final result.

Thanks, Dominic.

Since I'm facing the same thing again with new stock that has surprisingly failed although it's only a month old (it was bought for the Amy Winehouse video she never turned up to), what

amount of over-exposure would you recommend for these d-min values (Fuji 400T)?

Actual: 0.32, 0.67, 0.90

Reference: 0.24, 0.53, 0.71

Also, is there a list of all the reference numbers for all film stock I could get a hold off? I understand they vary from stock to stock.

Mark Dunn · November 29, 2007

What drives me nut are handwritten labels where you can't tell if it says "5298" or "5248". I had that happen once on a shoot -- the mag was color-taped for 5298 but the handwritten label said 5248 and the loader couldn't remember which he actually loaded. Luckily the problem has gone away more or less -- I don't think any of the current color neg stocks could be mistaken for the other if badly handwritten. (01, 05, 12, 17, 18, 29) -- I guess 4's and 9's are the only two numbers that can look similar. So if you plan on buying any recanned 5274 and 5279, look carefully at the label...

Don't you use the can tape or the sticky can label any more?

David Mullen ASC · November 29, 2007

Don't you use the can tape or the sticky can label any more?

This was back when I was shooting features on recan stock, but even when using new stock, once you get down to using up your leftovers, you end up with hand-written labels usually.

Dominic Case · November 30, 2007

The manufacturers tend not to publish "aim" figures for d-mins, as they would then be asked for tolerances, and then they are in the position of having unsaleable film stock - on the one hand - or on the other hand, stock that they have approved which subsequently affects the results in critical area.

Your lab might have done the work of testing new samples of every available film stock - they'd be the best source to get information from.

On the figures you've shown, you might need to increase your exposure about a stop and an half to get clear of the raised d-min, and expect to see the usual shift to bluish shadows. But as I've always said, it's a mysterious art, and depends a lot on the look that you are after, and the nature of the scenes you shoot.

Is that non-commital enough - or what?