Choosing how to shoot 16mm (been ages, need help)

[Tomas Brice]

Hello!

Long time lurker, first time posting.

I'm in the middle of prepping for a commercial shoot on 16mm, and I haven't shot film in over 10 years, so I'd like to brush up a bit on technique.

Normally I have been shooting Alexa 35 or Mini LF rated at 200 or 400 ISO since I really don't like noise, and have been enjoying using it that way, as long as I don't clip highlights I find it works wonders. In that sense I think I feel comfortable shooting in the film speeds, but what I'm trying to decide on is how to expose which stock.

If I'd like to shoot clean 16mm should I choose 200t and over expose by one stop and process normally, or pull one stop?

Same thing for 500t, does an overexposed by +1 and normally processed 200t look cleaner than 500t pulled by 1?

What about 250d?

I read on the ASC article for "First Man" and saw that the majority of it was 500t, normally processed and rated at 250, and found it a bit dirty in the grain. Would pulling it help?

Also, I probably won't have access to test rolls, so in general, if rating for box speed, what's the DR on Vision 3? in general, how many under/over if reading on a spot meter?

Thank you all for your time and help

 

 

[Ludwig Hagelstein]

Hey Tomas, 

if you want clean, i'd go for 50D and 200T and pull. Have a look here: 

That is 50D, 200T and 500T, all pulled one stop, super 16, scanned in 4K on Scanity.

[Tomas Brice]

33 minutes ago, Ludwig Hagelstein said:

Hey Tomas, 

if you want clean, i'd go for 50D and 200T and pull. Have a look here: 

That is 50D, 200T and 500T, all pulled one stop, super 16, scanned in 4K on Scanity.

This looks amazing! Thank you for sending it over. Any idea which shots are which speeds?

[Ludwig Hagelstein]

16 minutes ago, Tomas Brice said:

This looks amazing! Thank you for sending it over. Any idea which shots are which speeds?

the night scenes are 500T, the interiors are 50d and i think one dusk scene is 200T if I remember correctly. 

[Tomas Brice]

1 hour ago, Ludwig Hagelstein said:

the night scenes are 500T, the interiors are 50d and i think one dusk scene is 200T if I remember correctly. 

Thank you!
Also, do you have any experience with the difference between Pulling on stop in the development process, or using normal processing and bringing down after the scan?

[Ludwig Hagelstein]

Scanning - if you're choosing a high end scan like ARRISCAN or SCANITY has minimal influence on grain - that happens chemically / physically during processing. Where are you going to send your film to process and scan?

[Tomas Brice]

3 minutes ago, Ludwig Hagelstein said:

Scanning - if you're choosing a high end scan like ARRISCAN or SCANITY has minimal influence on grain - that happens chemically / physically during processing. Where are you going to send your film to process and scan?

Thank you. Cinelab London most likely

[Ludwig Hagelstein]

1 hour ago, Tomas Brice said:

Thank you. Cinelab London most likely

why not a lab in Europe? Much eaiser and no brexit / xray issues...

[Tomas Brice]

2 minutes ago, Ludwig Hagelstein said:

why not a lab in Europe? Much eaiser and no brexit / xray issues...

I think it's who production has a deal with. Who would you recommend instead?

[Ludwig Hagelstein]

1 minute ago, Tomas Brice said:

I think it's who production has a deal with. Who would you recommend instead?

Well there are a few options - DeJonghe, Haghefilm, Transperfect, Silverway, Augustus, and of course Cinegrell in Berlin - the sample I posted above is from there too. You'll definitely get the best scans there - all others do 10 Bit, at Cinegrell you'll always get 16Bit from Scanity, or Arriscan if you choose so,  no extra charge for uploads, and same day delivery if the film arrives until 10 in the morning at the lab. But for transparency, I'm biased towards Berlin, as I work there. All options are good, and if you're in Europe, don't go over the hassle and deal with Brexit uncertainties and possible x-ray hazard in shipping just to get film to the UK. There is plenty infrastructure on the continent, too. 

[Tomas Brice]

25 minutes ago, Ludwig Hagelstein said:

Well there are a few options - DeJonghe, Haghefilm, Transperfect, Silverway, Augustus, and of course Cinegrell in Berlin - the sample I posted above is from there too. You'll definitely get the best scans there - all others do 10 Bit, at Cinegrell you'll always get 16Bit from Scanity, or Arriscan if you choose so,  no extra charge for uploads, and same day delivery if the film arrives until 10 in the morning at the lab. But for transparency, I'm biased towards Berlin, as I work there. All options are good, and if you're in Europe, don't go over the hassle and deal with Brexit uncertainties and possible x-ray hazard in shipping just to get film to the UK. There is plenty infrastructure on the continent, too. 

Thank you so much for the suggestions, I'll definitely pass them forward to production.

Also, for colour temperature, do you ask for the lab to correct tungsten to day light? Or all is done after the grade? And if you were to shoot something digitally at 4300k, what would you ask the lab to do?

 

Also, when pulling, would you overexpose even more? So lets say 200t, pulling one stop. Do you think you might rate it at ISO80 or 64?

Edited November 9, 2024 by Tomas Brice

[David Mullen ASC]

Overexposing doesn't reduce the size of the grains, which determines the speed of the stock - it just exposes the smaller, slower grains in between the larger faster grains, giving the impression of a smoother, tighter grain structure. I don't think there are any advantages of overexposing more than 1-stop to reduce the impression of graininess.  And I don't think it matters whether you pull it to compensate versus develop normally and darken it in post to look correct but I don't know for sure, it all depends on whether the scanner is OK with a denser negative without creating noise in the highlights from pushing more light through it.

Also, a decade ago, some labs did not offer pull-processing for 16mm, just push-processing -- though today with film processing being rare in general, maybe it is offered now.

You shoot a grey card or chart at the head of the roll in the color and light level you want to be neutral.  So if you leave out the 85 filter in daylight and overexpose the stock by one-stop, you'd shoot the grey card or chart (and a face holding them helps too) so that whoever color-corrects the transfer or makes a print knows that you want that to be neutral. The lab itself doing the development can only go by what you order -- normal, pull, or push, etc.  They don't know what's on the negative until after it has been processed.

If you want only half the color cast to be corrected in outdoor photography, perhaps you pulled the 85 filter but want some of the blue tone left in, then you'd shoot the grey card with a half-correction (81EF for example) in and then shoot the scene with no correction. Indoors you can do with with gels on whatever light you use for the grey card, or set the LED light on the grey card for somewhere halfway between daylight and tungsten, then shoot the scene in daylight lighting and get a half-corrected color from the colorist.

[Ludwig Hagelstein]

a good log Scan will have a neutral black at around 10% on the Scopes, but color temperature is usually not corrected, as during the scan, no creative decisions are being made - that is up to the colorist. Once you receive a scan that neither crushes the blacks, nor clips the whites, you can work in grading to achieve whatever look you desire. 

Re Pull: I wouldnt go down to 64, but between EI 100 and EI 80 there is not so much difference I think. In brightly lit scenes, go for 100, and in scenes with less light, expose a bit more I'd say

[Ludwig Hagelstein]

9 minutes ago, David Mullen ASC said:

Overexposing doesn't reduce the size of the grains, which determines the speed of the stock - it just exposes the smaller, slower grains in between the larger faster grains, giving the impression of a smoother, tighter grain structure. I don't think there are any advantages of overexposing more than 1-stop to reduce the impression of graininess.  And I don't think it matters whether you pull it to compensate versus develop normally and darken it in post to look correct but I don't know for sure, it all depends on whether the scanner is OK with a denser negative without creating noise in the highlights from pushing more light through it.

Also, a decade ago, some labs did not offer pull-processing for 16mm, just push-processing -- though today with film processing being rare in general, maybe it is offered now.

You shoot a grey card or chart at the head of the roll in the color and light level you want to be neutral.  So if you leave out the 85 filter in daylight and overexpose the stock by one-stop, you'd shoot the grey card or chart (and a face holding them helps too) so that whoever color-corrects the transfer or makes a print knows that you want that to be neutral. The lab itself doing the development can only go by what you order -- normal, pull, or push, etc.  They don't know what's on the negative until after it has been processed.

If you want only half the color cast to be corrected in outdoor photography, perhaps you pulled the 85 filter but want some of the blue tone left in, then you'd shoot the grey card with a half-correction (81EF for example) in and then shoot the scene with no correction. Indoors you can do with with gels on whatever light you use for the grey card, or set the LED light on the grey card for somewhere halfway between daylight and tungsten, then shoot the scene in daylight lighting and get a half-corrected color from the colorist.

the advice regarding the color cast is more valid for a classic telecine approach, or for analog rushes printing. Not so much for finishing scanning or a scan once workflow, where you directly scan at maximum resolution / bit depth and then the colorist can make whatever decision he / she intends to.

[Robin Phillips]

I would just note that pull processing is not going to give you more detail, if anything it slightly softens the image. if you are doing a 4k scan, you really want to over expose. personally I usually over expose by at least 1/3rd of a stop and thats enough to make a difference. IMO over exposing a full stop is very much overkill unless you are shooting 500T and want that extra detail that comes from the over exposure. TBH though I dont think I've ever shot 500T at anything more than 2/3 over (320), so Im not sure if going a full stop over will not actually get you much.

 

[David Mullen ASC]

1 hour ago, Ludwig Hagelstein said:

the advice regarding the color cast is more valid for a classic telecine approach, or for analog rushes printing. Not so much for finishing scanning or a scan once workflow, where you directly scan at maximum resolution / bit depth and then the colorist can make whatever decision he / she intends to.

Sure — except that grey cards / scales are a way of communicating to that colorist what the intent was if you’re not going to be the one color-correcting the scan and any log files delivered to create a display image.

[Ludwig Hagelstein]

3 minutes ago, David Mullen ASC said:

Sure — except that grey cards / scales are a way of communicating to that colorist what the intent was if you’re not going to be the one color-correcting the scan and any log files delivered to create a display image.

yes that is absolutely true - a color chart is always a useful reference. I was just making the point that during the actual ingest of the scan, the color will not be corrected as that is a creative decision while working with the log scans. The scan itself is supposed to acquire the best possible data, and at that stage, no intervention regarding color bias is supposed to happen. What happens in dailies grading or finishing is an entirely different story, of course. 

[Tomas Brice]

Thank you all for hopping in on the discussion. Your comments are well appreciated and I feel more confident going forward. I'm still waiting on details from production but I'm feeling inclined with these decisions as of now:

-If the director wants it contrasty I'll develop normally but overexpose by one stop, if more contrasty I'll over expose by one stop and pull

-I'll do the quick reference shot with the filter, it's a great idea as a reference base of colours for the grade, even if just for the first balancing step.

-Probably going to shoot 200t, and only really do 500t if in dire need.

-Also will look into the new labs that I didn't know about!

 

Camera wise, any particular recommendation between Aaton, 416 or SR3?

 

[David Mullen ASC]

17 hours ago, Tomas Brice said:

-If the director wants it contrasty I'll develop normally but overexpose by one stop, if more contrasty I'll over expose by one stop and pull

You mean if the director wants it less contrasty, you'd overexpose and pull... though I think you'll find that it is easy to make it look less contrasty in color-correction with an overexposed and normally developed negative.

[Tomas Brice]

On 11/10/2024 at 2:41 PM, David Mullen ASC said:

You mean if the director wants it less contrasty, you'd overexpose and pull... though I think you'll find that it is easy to make it look less contrasty in color-correction with an overexposed and normally developed negative.

Exactly, thanks for correcting, that's what I meant!
Is there any real benefit in Pulling then? Ludwig vouched for it, and I really liked his example

[David Mullen ASC]

If you are going to print, then definitely overexposing combined with pull-processing will give you a smoother grain structure with less contrast.

But if you are going to scan the negative, it's somewhat of a controversy whether to compensate in development. Some colorists say yes, others say no.  

You push or pull process to change the final density, so if you overexpose by 1-stop and pull-process by 1-stop, the resulting negative should be of normal density. If you had just processed normally, then the negative would be 1-stop denser than normal. (I wouldn't factor in the change in contrast too much -- a log scan is much flatter than any low-contrast look that most people would aim for so color-correcting for a softer contrast is fairly easy.)

Is it a problem to have a negative that is 1-stop denser on average?  Probably not though some older scanners and telecines do pick up some noise in bright areas of the image (like hot skies) because that portion is so dark on the negative, but I'm not sure how much of an issue that is anymore.  In fact, that was one of the main changes made to create Vision-3 film stock -- Kodak added "micro grains" that are extremely slow in order to add another stop of detail in overexposed areas. So it's harder to clip information in the highlights with today's Kodak Vision-3 stock.

But I've been meaning to test this notion of whether to push or pull 1-stop versus adjust for the density difference in the color-correction. It's more of an issue with underexposure and push-processing; some colorists would rather get a negative with more density so prefer the push-processing; some don't like the shift in color/contrast and would prefer to fix a thinner negative.

[Tomas Brice]

15 hours ago, David Mullen ASC said:

If you are going to print, then definitely overexposing combined with pull-processing will give you a smoother grain structure with less contrast.

But if you are going to scan the negative, it's somewhat of a controversy whether to compensate in development. Some colorists say yes, others say no.  

You push or pull process to change the final density, so if you overexpose by 1-stop and pull-process by 1-stop, the resulting negative should be of normal density. If you had just processed normally, then the negative would be 1-stop denser than normal. (I wouldn't factor in the change in contrast too much -- a log scan is much flatter than any low-contrast look that most people would aim for so color-correcting for a softer contrast is fairly easy.)

Is it a problem to have a negative that is 1-stop denser on average?  Probably not though some older scanners and telecines do pick up some noise in bright areas of the image (like hot skies) because that portion is so dark on the negative, but I'm not sure how much of an issue that is anymore.  In fact, that was one of the main changes made to create Vision-3 film stock -- Kodak added "micro grains" that are extremely slow in order to add another stop of detail in overexposed areas. So it's harder to clip information in the highlights with today's Kodak Vision-3 stock.

But I've been meaning to test this notion of whether to push or pull 1-stop versus adjust for the density difference in the color-correction. It's more of an issue with underexposure and push-processing; some colorists would rather get a negative with more density so prefer the push-processing; some don't like the shift in color/contrast and would prefer to fix a thinner negative.

Thank you for the reply. I like having a dense negative, I even do it digitally , shooting on the Alexa mainly between 200 and 500, as I really don't like noise. I think grain looks better than noise, but not necessarily on 16mm. 

If as you mentioned, over by 1 and pull by 1, and resulting in a normal density negative, is that density different due to it's processing if it's an image with the same exposure, does it look cleaner? Or it only looks cleaner by having a denser negative and pulling back in the DI? 

Quoting from Linus in ASC Babylon Issue: “Something fascinating happens with film in the development process,” he continues. “I find that the mids come out first, and then the highlights and blacks are developed. If you develop in the time that Kodak says you should, then the mids have developed and the blacks and highlights come along. But if you keep developing, the mids stay where they are and the blacks and highlights keep developing, pushing more contrast as they build on the negative and the highlights start to bleed a bit into the mids. If you pull one stop, the mids develop, but the blacks and highlights underdevelop, leading to less contrast."

So I guess it doesn't generally mean it's going to be a cleaner image, but an image that is even more "loggy"?

[Mark Dunn]

2 hours ago, Tomas Brice said:

Quoting from Linus in ASC Babylon Issue: “Something fascinating happens with film in the development process,” he continues. “I find that the mids come out first, and then the highlights and blacks are developed. If you develop in the time that Kodak says you should, then the mids have developed and the blacks and highlights come along. But if you keep developing, the mids stay where they are and the blacks and highlights keep developing, pushing more contrast as they build on the negative and the highlights start to bleed a bit into the mids. If you pull one stop, the mids develop, but the blacks and highlights underdevelop, leading to less contrast."

This is how development works- the longer, the contrastier. It's not peculiar to MP film.

[Dirk DeJonghe]

It is called gamma, more processing time, more gamma, the whole B&W process is based on gamma measurement via sensitometric strips.