Depth of Field for S16/U16

[Scott Bryant]

I know CoC is different for 16mm and 35mm but does it also change when shooting in super16 or ultra 16 compared to regular 16? If anything I would think it would get slightly more shallow if noticeable at all. Is this true?

[David Mullen ASC]

I don't know the dimensions of U16, but there's a magnification factor of 1.2X between R16 and S16 if you simply go by the horizontal view and not worry about the vertical dimensions. That's actually somewhat significant in terms of focal length choice to match field of view and thus depth of field -- in practical terms, it probably means about a stop less depth of field with S16 versus R16, maybe? Not sure.

[Bernie O'Doherty]

U16 is still trying to figure itself out now that it might be finally "coming out ".

I offer customers the following 3 options because they asked for them:

11.45mm by 6.35mm which gives a 1.80:1 aspect ratio.

11.45mm by 6.18mm which gives a 1.85: 1 aspect ratio, and

11.66mm by 7.49mm which gives a 1.89 aspect ratio.

Ideally I want to give them the biggest bang for their buck, so I go for 11.66mm by 7.49mm. People want pixels up the wazoo.

Hopefully you can figure the mag factor/depth of field change from this.

Cheers Bernie

[David Mullen ASC]

U16 is still trying to figure itself out now that it might be finally "coming out ".

I offer customers the following 3 options because they asked for them:

11.45mm by 6.35mm which gives a 1.80:1 aspect ratio.

11.45mm by 6.18mm which gives a 1.85: 1 aspect ratio, and

11.66mm by 7.49mm which gives a 1.89 aspect ratio.

Ideally I want to give them the biggest bang for their buck, so I go for 11.66mm by 7.49mm. People want pixels up the wazoo.

Hopefully you can figure the mag factor/depth of field change from this.

Cheers Bernie

 

Well, 16mm is 10.26mm wide, Super-16 is 12.52mm wide, so 11.66mm would be a factor of 1.13X or so.

 

However, I'm not sure of my math in that if doubling the width, a factor of 2X, ends up being a practical 2-stop depth of field difference... then obviously a 1X factor, which is no factor at all... can't be a stop different!

[Bengt Freden]

Re. differences in depth of field in Std.16 vs. Super16 or Ultra16:

 

I should´t think that the difference in depth of field would be that great, if noticeable at all. Even if you do need slightly longer focal lenses for Super16 than for Std.16, the height of the frame/camera gate is still the same. The Super16 format is roughly 2mm wider on one side. For Ultra16, with the image width increase only in between the perforations (roughly 0.7mm on each side), there are many Std.16 lenses that would cover still this format, too. Which means that in Ultra16, you are using a wider image instead of the 'higher' image of Std.16. If you draw an image circle, this is quite obvious. Perhaps depth of field is very slightly shallower in Super16 compared to Std.16 but I doubt it would be noticeable up on the screen.

 

I have checked the diagonal measurements for all three 16mm formats (the uncropped negative or camera gate area), and a 'normal' focal length for each of these formats would be roughly:

 

Std.16 (ratio: 1.37:1): 12,5mm (which is the 'wide angle' setting on a Canon Scoopic zoom lens!)

Ultra16 (ratio: 1.78:1): 13mm

Super16 (ratio: 1.65:1): 15mm

 

As a format comparison, the 'normal' focal length for Super8 is 7mm (vs. SuperDuper8/ MAX-8 where it is 7,5mm).

 

So, going by this, there is a 2.5mm difference in focal lenth between Std.16 and Super16, but only 0.5mm between Std.16 and Ultra16.

 

Going up or down in format size, to 35mm (or 65mm!) or Super8 is a different story, though. You would have noticeably shallower depth of field (selective focus) for 35mm and longer/deeper for Super8mm.

 

Best regards,

Bengt F, photographer

[Bengt Freden]

Hi Bernie,

 

Thanks for the figures on Ultra16. Ultra16, in my opinion, is a smart way to adopt film cameras to 16:9 TV.

 

I have a question re the camera gate modification:

If you machine out the sides of the camera gate to 11.66 mm width, does that still leave the film manufacturer´s edge identifications and edge coding intact? When do they start to interfere with the negative area?

 

I should think that the height of the camera gate, by the way, in all cases would remain the same (about 7.5mm), as there is no machining done here?

 

Another one:

How many Ultra16 conversions have you done to date?

Thanks,

 

Bengt F, photographer

[David Mullen ASC]

I don't think you can use diagonal measurements as a guide -- it would be more accurate if each format were cropped to the same aspect ratio - let's say 16x9 or 1.85 - because if you are going to make comparisons, it would only be relevant for the same project and the aspect ratio would not be arbitrary.

 

So using the horizontal view and size as a basis for comparison and choice of focal length makes more sense to me than measuring the diagonal.

 

But even using your diagonal measurements, it becomes a 1.2X magnification factor between R16 and S16 instead of 1.3X, though personally I still think you'd match fields of view by looking at the horizontal view in each format, not the diagonal view. And then the question remains would it be a practical 1.2-stop or 1.3-stop difference in depth of field? Just glancing at a DOF chart for a 10mm versus a 12mm, it would seem so.

[K Borowski]

I'll have to dig up some calculus, physics, and optics texts to check on this.

 

I know ∏ comes into play somewhere, and the diagonal area of the long axis around which is circumscribed a circle is what is being measured and evaluated when it comes to DOF.

 

I'll see what I can uncover. . .

[Stephen Williams]

Hi,

 

Unless people are using very sharp & expensive new lenses, I would go with standard 16 tables.

 

Stephen

[K Borowski]

True Stephen: Better to have too much DOF than too little!

[Bengt Freden]

David,

 

Thanks for your professional input - it is much appreciated.

I will share my late night pencil scribblings from a week ago, when I tried to figure out the differences in various 16mm camera gate sizes and aspect ratios. The HDTV standard 1.78:1 or 16:9 is the most interesting to me, as I don´t do any work for big screens.

 

These are just very basic and downright simple format comparisons, of ACTUAL camera gate measurements, where I have drawn in the cropped formats for 1.78:1/HDTV - be my guest if you want to make it even more academic or mathematic:

 

 

I still don´t think that there are that big differences btw the Std.16 vs. the Ultra16 or even Super16 formats, with respect to depth of field, with comparable focal lengths (or even the same lenses).

 

Cheers,

Bengt F

[David Mullen ASC]

My calculations suggest that even a switch from a 10mm to a 12mm, for example, by going from R16 to S16, to match field of view, is about a stop loss of depth of field. Now assuming I'm correct (which I may not be) whether that counts as "significant" I don't know.

 

The whole point is to compare lenses with matching fields of view for different sized formats, otherwise what's the point in discussing depth of field differences?

 

If you don't change the CoC figure, look at a 10mm lens versus a 12mm lens, let's say, set to 3' at f/2.8 -- according to a DOF calculator for S16, it's a total depth of 2' 2.25" for a 10mm and 1' 5" for a 12mm. Now you figure that regular 16mm would use a slightly more critical CoC figure than S16, so that difference between the 10mm and 12mm would be minimized a little, but it still seems to be fairly different for the two lenses.

http://www.panavision.co.nz/main/kbase/ref...calcFOVform.asp

[Bernie O'Doherty]

Yes Bengt, you are right about the vertical hight not being cut. I must be up to around 35 to 40 U16 conversions over the past year and a half. S16's would quadruple that figure. Economics is becoming a more prominent factor in the decision making process. People don't like it when you estimate a S16 conversion of a K3 (which cost them $250) for $300. So I offer it hesitatingly.

S16 really happens with SRs, AAtons and Eclairs.

Funny thing is , people( a lot of my customers ) who were in video for years are buying 16 cameras and having us do the S16 mod. They call back and say they think they died and went to Heaven when they view their filmed images. "I don't believe I shot this" is a favorite comment.

Long may they bask in the filmic glow....keeps me off the streets..

Cheers, Back To The Bench Bernie.

[Bengt Freden]

Thanks for the reply, David,

 

One whole stop? WOW!

 

Well, that´s a language I understand. I am sorry if I wandered a bit off-topic with my ranting about different formats and aspect ratios. Knowledge and Experience rules once again!

 

However, the difference in depth of field between two relatively wide lenses, such as a 10 and 12mm, might not be that great, unless you are shooting at near full aperture or very close up. The difference would of course be more noticeable or significant if you switch to longer focal lenses or if you move close to a subject with an environment behind. And of course, if you are working at night with fast T1.3 lenses at full opening, this further exaggerates the difference.

 

I just read in David Samuelson´s "Hands-On Manual for Cinematographers" (p. 217, 2nd ed.) that there are a number of factors that determine the depth of field, the most important being (in this case for LESS depth of field):

 

• Longer focal lens

• Larger lens aperture

• Smaller acceptable Circle of Confusion (one of the posts in this thread were dealing with high-resolution lenses - my comment!)

• Close plane of critical focus

• LARGER FILM FORMAT

 

- other factors are for example; the resolving power of the lens in use, the quality and luminance of the lighting (hard/softer contrast, brightly lit objects), the addition of diffusion or fog filters, and the resolution of the presentation system.

 

In summary, I am inclined to accept now that there may be a slight depth of field difference between Std.16 and the WIDER Super16 format. However, I do not think that there is a difference with regard to Ultra16, because if the conversion is done to 1.78:1 standard, both formats could be inscribed in the same image circle and you would use the same lens - for example in the Canon Scoopic 16 MS camera, where the zoom lens is permanently mounted to the camera, and where you may switch from Std.16 to Ultra16 (in a converted or modified camera) and back on the same roll, depending on how you choose to scan or crop the native camera image.

 

Best regards,

Bengt

[Bengt Freden]

David,

 

Thanks for including the link to the Panavision Depth of Field calculator. I went to the calculator, for a lens of 10mm focal length:

 

I set the subject distance to 3 meters (say a person in a Mid Shot), the aperture to a pretty 'normal' value of f=2.8, and the format of choice to Super16 (HDTV-transmitted 1.78:1. Circle of Confusion 0.0005mm)

I got a depth of field for this relatively wide lens (at 2.8) of 1.4 meters to infinity: http://www.panavision.co.nz/main/kbase/ref...lcFOVresult.asp

 

That is a pretty good depth of field, even if the best resolving power is of course at 3 meters and slightly beyond.

 

Best,

Bengt

[Bengt Freden]

This link to film formats might be relevant to the ongoing discussion, although all measurements are in inches or tiny fractions thereof:

http://en.wikipedia.org/wiki/List_of_film_...ts#Film_formats

 

David,

I am beginning to be a bit uneasy about the width of the camera gate for Super16, in relation to what we have talked about. I have seen many different figures for this width. Do you know what the SMPTE standard is for the native camera area, in Metric mm please?

 

Thanks.

Bengt :-)

[Bengt Freden]

Well, David,

 

I found this: http://en.wikipedia.org/wiki/16_mm_film

It seems that there is some kind of consensus on the width of the Super16 camera gate being 12.52 mm, just as you mentioned in one of your posts. That´s a bit wider than the one I found. There may, of course be slight variations in the width of the camera gate from different manufacturers (Aaton, Eclair, ARRI, Ikonoskop A-CAM ...) but I believe that SMPTE has set up some kind of standard for this format, too?

 

Could you clarify?

Thanks.

 

Bengt :-)

[David Mullen ASC]

There may be SMPTE / ANSI standards... but it is less critical for a camera gate as long as it isn't smaller than any standards for extraction or projection (not so much an issue for S16 since it is not a projection format.) It is more critical for projector gates to follow common standards than the (generally) larger camera gates.

 

It's always a bit confusing though, for example, for film-out work -- do they record digital files out to sizes slightly larger than the projector gate standards? You'd think so. So when I set-up 2.39 framelines in an HD camera for extraction and then recording out to 35mm anamorphic, do they film-out the area I framed for, and then the projector gate slightly crops this further, or do they film-out slightly more than what I composed for, or is the filmed-out image exactly the same dimensions as the projector gate, thus risking seeing some black around the edges on some theater screens?

[Scott Bryant]

Thanks everyone for putting your answers (and physics) out there. This helps shed some light although, it is making my brain hurt a bit (4 years since freshman physics). I will have to read and reread these posts. I can't thank everyone enough for their help.

 

Scott

[Chris Keth]

However, the difference in depth of field between two relatively wide lenses, such as a 10 and 12mm, might not be that great,

 

Remember that at the wide end of the spectrum, 2mm of difference is a much greater visual difference than 2mm difference is between longer lenses. Consider it in percentages. 10mm to 12mm is an increase in focal length of 20%. 50mm to 52mm is an increase of 4%. 100mm to 102mm is an increase of 2%.