3-Perf - Depth of Field Query

Hi all CMLâ€™ers,

In a month or so I am about to shoot a short in 3-Perf Super 35mm that will be (for the immediate future) finished on HDCAM SR & and 2K if the money becomes available to go back out to 35mm (i.e. if the short tanks or is "successful").

We really don't have the cash to shoot 4-perf and get the camera package I think is best/appropriate, so 3-perf has become the only practical option (and considering it is going through a DI anyway, it's not such a big deal).

I have only shot Super 16mm before so I hope some people can fill in some stuff for me. The camera is going to be an Aaton 35-III. Aspect ratio hasn't been decided yet. (Would LOVE to shoot 4-perf Academy full gate, but not an option at all really financially/projection-ally & if finishing digitally; Curse HD/16:9 and all it's evil powers. 1.37 is the greatest format ever. If anyone saw the latest Goddard film 'Notre Musique' they will know what I am talking about... looked as lush as 'scope.)

1/. I have been watching a bunch of films on DVD (I know I know not the best way to observe technical aspects) that are shot in Super 35 (I'm basing this from looking up the technical details at the end of AC articles), and many of them seem to have greater depth of field than normal 4-perf 1.85 or Academy - and of course Anamorphic. Looking at the Aaton website and looking through the sizes of the ground glasses, and the Panavision website and the transmitted image areas, it seems like the horizontal image area is roughly the same, but I have been noticing some deeper than 'usual' d.o.f's...

I know it could just be a penchant for wide lenses in movies like 'Seven', but maybe it's more than that? I dunno... all cinematography is sensual as much as it's technical so my "interpretation" could be just that - an "interpretation".

2/. If so, I would need to change the size of my lenses to get the shallow focal depths I am looking for.... Surely I could also just as effectively maintain a small d.o.f. by shooting 1-2 stops wider than normal? Or maybe splitting the difference and increase my lengths by say 15% and being 1 stop more open?

The size of the sets/location dictates that shooting with very long lenses is impossible, and I want shallow d.o.f. to help separate the characters. I like 50mm and 80mm lenses, and the shots they produce - especially the way they were used in the film "Cyclo" shot by Benoit Delholmme in Vietnam in '95.

I guess another other option is to use 'softer' lenses to help smooth fore and backgrounds out, but we really want to see the most maximum detail in the faces of the actors (pores, etc)

I hope these aren't dumb questions, I checked the archives first to see if it was already in there but no joy. It seems to me that this could be a non-issue - but shallow d.o.f. is really important for the types of images I want for this project.

Thanks for whatever help anyone can give. I feel really lucky and grateful to have this forum, the information is very helpful.

Benjamin Rood
Dir/DP

I've been pretty successful at limiting depth of field on S16mm and video by cheating the actors' positions (moving them further away from the background) after moving in for coverage and shooting at a stop wider than I normally would (T2.8 instead of T4).

One scene in particular that I can think of that was shot on a DVX100,(day exterior, two women talking on a porch). We had to cheat the two actresses out about 15-20 ft. from the porch for their close ups in order to throw the background out of focus.

(Fortunately, it was a short scene and they weren't moving around much on the porch and we were shooting on sticks.)

Jessica Gallant
Los Angeles based Director of Photography
West Coast Systems Administrator, Cinematography Mailing List
https://cinematography.net/

Hi Jessica,

Thanks for that advice... but sadly all of this short is going to be shot inside a real hotel room (the biggest we can get, sure - but still a hotel _room_) which unfortunately means no 15-20ft 'fakes' in any direction... 3-5 ft fakes maybe, but nothing more. Fortunately the d.o.f. _will_ be a lot shallower on Super35 than 16mm, but maybe I can just apply a closer to subject/wider aperture and the issue will be solved.

Would love to know from anyone that has tried to do this for 3-perf/Super 35mm. I'm sure there are heaps of music video DP's out there especially that will have done it.

I was thinking about shooting with Ultra speeds also, to shoot at a really open stop, like T1.8 - T1.5. There's not going to be lots of movement in the film or frame really, so I think my focus puller will just manage.

I guess also the other option is to shoot with Macros for CU's... but that might be a bit too distorting on the faces... not a big Jean-Paul Jeunet/wide angle CU's kinda guy. At least not for this one, anyway.

But anyway am I right about the 3-perf S35 d.o.f. being greater? from looking at 'specs' I'm not convinced so, I'm just basing it on pictures really.

Benjamin Rood
Dir/DP

Hi Benjamin,

I'm a 50mm man myself and don't mind shooting wide open when it's right for the look. One of my tricks for squashing depth of field is to use diopters/proxars. If you can get achromats they are ideal. They basically pull your focus range in close making background focus impossible. I have used them in situations such as a woman in a car interior in the rain when I wanted the raindrops on the window to have nice fat c.o.c.'s (c's.o.c. ?) and they weren't soft enough with the lens wide open. The achromats will give you much better image quality than single element diopters. Your focus puller will mutter bad things but that is a common result of interesting photography.

As for Super 35 and DoF, that's just a function of the lens. Some people might get squeamish about shooting with a wider aperture due to image degradation in the blow up process causing them to stop down more(?) If anything, you might tend to use a slightly longer lens (a 60 rather than a 50 for instance) on the super format, giving you a shallower depth of field for the same relative frame. Of course you will have more DoF than when shooting anamorphic and that's one of the reasons people shoot S35.

Good luck with your project.

Best Regards,

Anders Uhl
Cinematographer, NY
The DoP Shop
http://www.thedopshop.com

Benjamin Rood writes:

>But anyway am I right about the 3-perf S35 d.o.f. being greater? From >looking at 'specs' I'm not convinced so, I'm just basing it on pictures >really.

Depth of field is not a function of focal length or of format, but rather a function of f-stop and image size. If you make the subject you are filming the same size on the imaging (film) plane -- say 3 perf 35mm and 16mm â€“ and you are shooting at the same stop, the depth of field will be the same, no matter the format size or focal length of the lens.

Put another way, if you frame a subject so that it fills a 3 perf 35mm frame and also frame it the same way on 16mm, the 3 perf will have less DOF because the image of your subject on the film plane will be larger -- provide your shooting at the same stop. 16mm appears to have greater DOF than 35 mm because of the smaller image size.

The only significant determinants of DOF are f-stop and image size.

Brian Heller
IA 600 DP

Thanks Anders and Brian for your info!

Anders, the diopters will be the perfect solution, thank you very much. I will try to see if I can source the anchromats as you have suggested here in Auckland. They sound like they will really solve the small shooting space + desire for small DoF issue. An issue that makes a focus puller (or even better a gaffer) go out of their normal mode is most of the time an encouraging sign for me (Of course, sometimes the requests can be just a silly, wanky idea - hopefully not that often though!).

Brian, of course you are correct. I guess I was just wondering if there was less DoF in 4-perf than 3-perf, because the image area on the negative is slightly different...?

Also I just read your DoF 'discussion' with Dominic Case/Stephen Gladstone/etc in the archives, which help make the technicals of the issue very clear. I don't choose medium-long lenses for the DoF so much as the degree of perspective flattening that I like (i.e. for compositional reasons). But of course that's a bit of a fib because prior to reading the discussion I guess I did kinda just go with the whole 'long lens/short DoF' lazy assumption.

Thanks for your well-wishing for the project. Hopefully it makes it's way to the net or a festival near you at some kind timeframe from now. (Hah!)

Best Regards!

Benjamin Rood
Dir/DP

Benjamin Rood writes:

>I guess I was just wondering if there was less DoF in 4-perf than 3-perf, >because the image area on the negative is slightly different...?

All things being equal, 4-perf would have less DOF than 3-perf. That is, if your subject filled the available negative area for each respective format. Since the subject would be larger on 4-perf than on 3-perf, there would be less DOF, assuming the same f-stop.

By using Diopters, you are enabling any given lens to focus closer, just like increasing the strength of reading glasses. Just keep in mind that by focusing closer, you are essentially increasing the subject size which is one of the two prime determinants of DOF. Also keep in mind that as you approach closer and closer to your subject, other issues start to intrude.

Namely camera shadow and film plane to subject distance, where slight distance changes can result in significant changes in subject size. Remember, it's still the focal length that determines the angle of view.

If you are considering using diopters, you might want to test simple diopters vs. Achromats vs. compound diopters. The sharper the image, the more visible focus errors become. There may be a happy middle ground between seeing the pores on an actor's face, and being distracted as individual pores slide in and out of focus.

If reducing DOF is your goal, then using the widest stop possible is the first step -- think SuperSpeeds. There's little net gain in diminishing DOF by closer focus, if you have to stop down. Use the fastest lens you can, wide open, and control the stop with ND's or slower film.

>Hopefully it makes it's way to the net or a festival near you at >some kind timeframe from now. (Hah!)

Please keep us informed.

Brian Heller
IA 600 DP

>,,,,,,I have used them in situations such as a woman in a car interior in >the rain when I wanted the raindrops on the window to have nice fat >c.o.c.'s (c's.o.c. ?),,,,,

Bokeh's ?

>,,,,,Fortunately the d.o.f. _will_ be a lot shallower on Super35 than >16mm, but maybe I can just apply a closer to subject/wider aperture >and the issue will be solved.,,,,,

Have you considered swing-and-tilts?

Probably not much more trouble than shooting full-time with diopters. If you're using 35mm kit there are some nice sets out there.

David Perrault, CSC

Brian Heller wrote:

>All things being equal, 4-perf would have less DOF than 3-perf.

Only if you're framing for 1.33:1, which is unlikely given the tone of this thread. The gate width of 3 perf and 4 perf is the same (24mm, basically S35) provided you're using a common centre and that it's the optical centre. Only the height has any significant difference. If you're shooting in just about any wide screen format (1.77, 1.85, even 2.35 flat) the image area is, for all practical purposes, identical, as the extra area in 4 perf is basically a waste area that essentially serves as a wide frame line. Even if you're shooting Academy, the image area is basically identical unless you're shooting 1.33.

Mike Most
VFX Supervisor
IATSE Local 600
Los Angeles

Mike Most writes:

>Only if you're framing for 1.33:1, which is unlikely given the tone of this >thread. The gate width of 3 perf and 4 perf is the same (24mm, basically >S35) provided you're using a common centre and that it's the optical >centre. Only the height has any significant difference.

In didn't want to get into a format discussion, but Mike is, of course, absolutely right. My post was purely from a DOF/image size perspective.

However, from personal experience I have shot 3-perf and was seriously asked to protect for 1.33 and TV.

I just said : "Sure, good idea." Further discussion seemed pointless.

Brian Heller
IA 600 DP

Mike:

Thanks for explaining the 4-perf/3-perf DoF - it's pretty much what I thought. In 'Notre Musique' / 'Our Music' by Godard last year, they shot in Academy/full-gate 4-perf and it looked _amazing_.

Brian :

What about PV Ultraspeeds? Don't they go even wider than SuperSpeeds? - and I don't really like the 'colour' of Zeiss images, especially not for this one. Slightly more luscious and removed from reality while also being crisp and picking up detail is my desired set of glass (it's probably many people's fantasy lens). Would Ultraspeeds fit this bill (I know, I know - TEST them. I will, I just want to know if I'm on the right track), or are they too old & soft? Sadly Primo's and Cooke's aren't going to be fast enough to work at an effective stop to reduce the DoF like we're talking about.

Ultimately the wider-stop method is easier, but the diopter method is also viable -and may let me use the lenses I would prefer. But more importantly it will make a different type of image... another piece of visual vocabulary so to speak. To discreetly use both methods could be quite interesting and expressive.

Especially in a short when you are looking for discreet visual effects that can help with characterisation for a short time span. I prefer the idea of an optical difference in facial rendering to 'overt' differences in lighting between two characters in the same small room. That said, appropriately using both techniques is of course the best way.

But when using a diopter of some form, what focal lengths work well in combination with a diopter and it's macro-izing effect to avoid heavy facial distortion in a CU? Or even an MCU? My gut for a CU says around the 50-65mm area, and for an MCU around the 30-45mm range. Because the Close Focus effect is only going to take effect at a close distance, otherwise the Diopter factor is going to be negligible, correct?

This list forum is amazing. I'm kind of a question-whore, so I hope this isn't going over too much that you've dissected many times before. Thanks heaps everyone for the help.

Benjamin Rood
Dir/DP

David wrote:

> Have you considered swing-and-tilts?

Sorry David, I hadn't. I honestly don't know anything... about... them...?

If you don't want to go into detail you could just refer me to a previous thread in the archives.

Cheers,

Benjamin Rood
Dir/DP

Brian Heller wrote:

>However, from personal experience I have shot 3-perf and was >seriously asked to protect for 1.33 and TV. I just said: "Sure, good idea."

Nearly all television programs that are shot on 35mm are shot on 3 perf. They are all shot 16:9, protecting for 1.33 using a common centre. This is standard operating procedure today and will continue to be so until 16:9 screens are in a hell of a lot more homes than they are right now.

Mike Most
VFX Supervisor
IATSE Local 600
Los Angeles

One other variable that depth of field is also dependent is the amount the image is magnified when displayed. The apparent depth of field on you TV will be much greater than the apparent depth of field on a large cinema screen. Yet, one more thing to take into consideration. And Panavision does have a 50mm T1.0 Lens that Caleb Deschanel used on the film "It Could Happen To You" (1994) to achieve some super soft back grounds.

Eric Jones
Los Angeles, CA

Mike Most writes:

>Nearly all television programs that are shot on 35mm are shot on 3 >perf. They are all shot 16:9, protecting for 1.33 using a common centre.

Do these shows furnish any tips on simultaneous composition for 16:9 and 1.33?

Benjamin Rood writes:

> What about PV Ultraspeeds?

I don't think the speed difference between UltraSpeeds and SuperSpeeds will be very apparent. Panavision does have a couple of faster lenses, but they're not very available. By all means you should use whatever lenses you prefer.

>Ultimately the wider-stop method is easier, but the diopter method is >also viable -and may let me use the lenses I would prefer.

Using a diopter will not reduce the DOF; the diopter will only allow you to focus closer, thus obtaining a larger image.

>But when using a diopter of some form, what focal lengths work well in >combination with a diopter and it's macro-izing effect to avoid heavy >facial distortion in a CU? Or even an MCU?

There's no real answer to this question; it depends on your preferences and the effect you are trying to achieve.

My gut for a CU says around the 50-65mm area, and for an MCU around the 30-45mm range. Because the Close Focus effect is only going to take effect at a close distance, otherwise the Diopter factor is going to be negligible, correct?

You only need a diopter if the lens you are using cannot focus as close as you would like to get. Normally diopters are used on zooms and longer lenses with longer minimum focus distances. The primes you are talking about may focus as close as you need without a diopter.

Brian Heller
IA 600 DP

Brian Heller wrote:

>Do these shows furnish any tips on simultaneous composition for 16:9 >and 1.33?

There isn't any way to do two different aspect ratios in a way that avoids compromise for one, the other, or both. It's something that we're stuck with and have been for a number of years now. Having said that, it sometimes surprises me how well the 16:9 version often turns out in spite of the primary framing being for 1.33. In general, overs work reasonably well, as do most singles, even with the added "air.". It's the wider shots, where the entire cast is crammed into the centre of the frame, that don't work at all. This is most prevalent on sitcoms, where group shots are more often used.

Mike Most
VFX Supervisor
IATSE Local 600
Los Angeles

Thanks Brian.

Guess it's time to shut up and do some tests to before embarrassing myself any further. Thanks for the suggestions and advice, I will try to put it to good use I promise.

Benjamin Rood
Dir/DP

>Have you considered swing-and-tilts?
>Sorry David, I hadn't. I honestly don't know anything... about... them...?

Basically, a bellows with a lens in front that can be articulated to manipulate the focal plane and image geometry.

These are often used in large format architecture photography. Trendy uses over the years include fashion and music videos. There are packages that can be rented that include the bellows and lens sets that work with motion picture gear.

Anyhow, your rental house is the place to see these. Clairemont has them for sure. Canon and ARRI make shift/tilt lenses that are self-contained but I find these (the Canon's) don't offer enough excursion to be very useful - just found out about the ARRI so cannot comment. There are three pictured at the ARRI website.

David Perrault, CSC

Hi David,

I have used 5x4 and 10x8 Camera's a little before for a project when I was in art school, so I understand the premise of this suggestion now.

It sounds very interesting... but obviously you would want to experiment with them a little. Panavision NZ in Auckland has a shift & tilt lens set (or so their website says). I will make enquiries!

Thank you for the suggestion.

Benjamin Rood
Dir/DP

Brian Heller writes:

>Do these shows furnish any tips on simultaneous composition for 16:9 >and 1.33?

Centre all your subjects.

And make sure the sun is behind you. f/8 in bright sun. 5.6 in the shade.

Dan Drasin
Producer/DP
Marin County, CA

Greetings

Depth of field tables are calculated using the entrance pupil of a lens and not the film plane. Entrance pupil of some lenses can be well behind the film plane.

Does that mean that for a given aperture, a zoom lens set on 50mm magnification 6 feet away from a subject might have greater depth of field than a 50mm prime 6 feet away from the subject (all other variables being equal)?

Behzad Olia-Rosenkranz
European based Cameraman / Assistant
Luzern
0041 78 644 89 13

Behzad Olia-Rosenkranz writes:

>Depth of field tables are calculated using the entrance pupil of a lens >and not the film plane. Entrance pupil of some lenses can be well >behind the film plane.

In a word, yes. What's more, with modern zoom lenses, the entrance pupil is not stationary, but moves rearward as the nominal focal length of the lens is increased. This can place the entrance pupil as much as 3 feet behind the film plane. For a more complete explanation of why this is so, for methods of compensation, and for an excellent overview of lens terms and cardinal points, consult David Samuelson's 'Hands-On' Manual for Cinematographers.

In any case, Mr. Samuel sonâ€™s opening statement on DOF says it best: "The calculation of DOF is a well-defined mathematical procedure, whereas the assessment of what is acceptably sharp and what is not, is a matter of personal judgement."

Brian Heller
IA 600 DP

Behzard wrote :

>Does that mean that for a given aperture, a zoom lens set on 50mm >magnification 6 feet away from a subject might have greater depth of >field than a 50mm prime 6 feet away from the subject (all other >variables being equal)?

Hi Behzard,

All Zoom lenses are unique. If you look at pages 223/4 of David Samuelson's Hands-On Manuel, you can see a chart of the Panavision Primo 24-275 zoom lens. At 150mm and above the entrance pupis is behind the film plane. Therefore IMHO with a zoom lens there are times that there is more DOF than a Prime lens. With this lens at 100mm or wider there would be less DoF than a prime lens..

If I have this wrong I am sure Brian Heller will correct me!

Cheers,

Stephen Williams DoP
Zurich
www.stephenw.com

Behzard wrote :

Brian Heller wrote : In a word, yes

Brian,

You answered while I was writing my message! If you look at David Samuelson's Hands-On manual the answer should IMHO be No? assuming 50mm and 6 feet!

Stephen Williams DP
Zurich

Stephen Williams writes:

>If you look at look at David Samuelson's Hands-On manual the answer >should IMHO be No? assuming 50mm and 6 feet!

Only if you're assuming a number of other things. What zoom lens? What prime? What format --16 or 35mm?

Behzad 's question was :

The answer is still yes, it might and probably does. Remember focus is still set from the film plane. But given a different set of lenses, it might not, or the difference may be purely academic, like this discussion.

Brian Heller
IA 600 DP

Brian Heller wrote :

Only if you're assuming a number of other things. What zoom lens? What prime? What format --16 or 35mm?

Brian,

Do you know of any 35mm zoom lens that at the wide end (50mm or less) when focused at the close end (with diopter if necessary) have any 'extended dof' due to entrance pupil being very close or behind the film plane? I shoot a lot of table top work, so such a lens would be most helpful.

Many thanks

Stephen Williams DoP
Zurich

Stephen Williams writes :

>Do you know of any 35mm zoom lens that at the wide end (50mm or >less) when focused at the close end (with diopter if necessary) have >any 'extended dof' due to entrance pupil being very close or behind the >film plane?

I think there's a bit of a misunderstanding here. By definition, if the image size and the f-stop are the same, the DOF should be the same. The entrance pupil location is used to calculate any given lens' DOF. I don't think your going to get more actual DOF by using a zoom lens with an entrance pupil behind the film plane, but that sound's like an interesting idea to check out.

Off hand I donâ€™t know of any zooms that have an entrance pupil behind the film plane at 50mm. You might want to check with the manufacturers.

Brian Heller
IA 600 DP

Brian Heller wrote :

>I think there's a bit of a misunderstanding here. By definition, if the >image size and the f-stop are the same, the DOF should be the same. >The entrance pupil location is used to calculate any given lens' DOF

Brian,

In theory I totally agree that the DOF should be the same between 2 lenses of the same focal length. However if DOF is measured from the entrance pupil, IMHO it would seem possible that when the entrance pupil is further away there should be more depth of field.

Last week I did a shoot and tested a Cooke 20-100 v Zeiss SuperSpeeds at T4 I realize that DOF is related to F stops and not T stops. The Cooke zoom, being a F2.8 & T3.1 (about 1/3 stop). I used an 85 & 135 Zeiss. Looking at the rushes on a Spirit there seems to be a far greater DOF on both the Zeiss lenses (IMHO much more than 1/3 stop). The Cooke is from 1978 and the Zeiss From about 1981 (Leonetti/Ultracam). I would conclude that the entrance pupil is further forward with the Cooke Zoom.

On page 187 of Blain Brown's book 'Cinematography & Practice' I quote "Thus if you are shooting a close-up at the wide end of a zoom it's as if you were 10" closer to your subject matter. Being closer you of course have less depth-of-field. This is one of the reasons that zooms are seldom used in macro, table-top and other critical focus applications."

Stephen Williams DoP
Zurich

Stephen Williams writes:

>In theory I totally agree that the DOF should be the same between 2 >lenses of the same focal length. However if DOF is measured from the >entrance pupil, IMHO it would seem possible that when the entrance >pupil is further away there should be more depth of field.

DOF is not measured from the entrance pupil. It's only computed using the entrance pupil. DOF is measured in the "focus space" in front of the lens.

>Last week I did a shoot and tested a Cooke 20-100 v Zeiss >SuperSpeeds at T4 I realize that DOF is related to F stops and not T >stops. The Cooke zoom, being a F2.8 & T3.1 (about 1/3 stop).

OK.

>I used an 85 & 135 Zeiss. Looking at the rushes on a Spirit there >seems to be a far greater DOF on both the Zeiss lenses >(IMHO much >more than 1/3 stop).

That doesn't necessarily follow. And I'm not sure your test is a valid indication of what you claim. You have to check with the manufacturers to determine where the entrance pupil is located. Deducing the entrance pupil location by a visual test is problematic. So is setting the f-stop on a t:stopped lens or setting the exact focal length on a zoom. Remember DOF is all about image size and f-stop. Focal length is not a factor.

>On page 187 of Blain Brown's book 'Cinematography & Practice' I quote >"Thus if you are shooting a close-up at the wide end of a zoom it's as if >you were 10" closer to your subject matter

I can't speak for Blain, but I believe that he is only referring to the wide end of the zoom. By closer, I'm assuming he means that the subject image will be larger and therefore have less DOF. As you know from table top work, switching to a wider lens to get more DOF and then having to move the camera closer to maintain the same image size is an exercise in futility, even though many directors will still want to try it.

Manoeuvring a big zoom to get a WA close-up on a table top, or keeping the camera and lens shadows out of the shot or when using an overhead rig are other reasons to avoid zooms, but many people prefer them for ease of framing.

The newer ones are plenty sharp.

Brian Heller

In close work, tabletop, macro, what _counts_ - is nonetheless is the dof at the subject position, right ?

Really it seems to me what one is concerned with is subject, film plane, and what comes between the two.

I think the principal reason to avoid a zoom would be so you don't knock the twinkies over when the director says "Closer ! Closer ! I can almost taste them ! Zoom wider, we want to experience their voluptuous roundness"

With all due respect I think Blain's statement is a bit misleading, although I sort of know what he's trying to say and am admittedly reading it somewhat out of context.

Maybe I'm not getting where you're going with this...

Sam Wells

>I think the principal reason to avoid a zoom would be so you don't >knock the twinkies over when the director says "Closer ! Closer ! I can >almost taste them ! Zoom wider, we want to experience their voluptuous >roundness"

Absolutely!

The zoom is simply too big and doesn't focus close enough. It's difficult to use on camera rigs, like the zero gravity, that get you close to your subject. The only time I've used a zoom on tabletop is for the wide lock off.

BTW - You never want to knock the twinkies over on a tabletop shot. The client goes nuts and will have a legitimately good defence for any irrational actions.

Jim Sofranko
NY/DP

Sam Wells :

>With all due respect I think Blain's statement is a bit misleading, >although I sort of know what he's trying to say and am admittedly >reading it somewhat out of context.

Sam the Heading in Blain's Book is "Zooms & Depth of Field" I Quote "Zooms have some special characteristics when it comes to depth of field. As we discussed previously, depth-of-field is not measured from the film or video receptor. In fact it is measured from the nodal point of the lens. Depth-of-field charts compensate for this in a general way by adding a fixed amount in front of the focal plane. This is why you may see different DOF charts for zooms and primes at equivalent focal lengths.

I only just read Blain's book yesterday for the first time, and in view of my tests thought it an interesting thread!

Brian Heller wrote :

>Remember DOF is all about image size and f-stop. Focal length is not a >factor.

Brian,

All I am saying is that there could be exceptions to this 'rule of thumb'

For example

400mm with 200 foot distance @5.6, 200mm with 100 foot distance @5.6, 100mm with 50 foot distance @ 5.6, 50mm with 25 foot distance @ 5.6 all follow the rule.

However a 25mm lens with 12.5 foot distance @5.6 has more than twice the DOF than the other examples. With the last example IMHO it's because the focus is set near to the hyper focal distance of 14'5".

Stephen Williams DP
Zurich

Stephen Williams writes:

but it's probably more correct to say Blain Brown wrote :

>"Zooms have some special characteristics when it comes to depth of >field. As we discussed previously, depth-of-field is not measured from >the film or video receptor. In fact it is measured from the nodal point of >the lens."

It might have been less confusing had Blain written calculated, instead of "measured."

"Depth-of-field charts compensate for this in a general way by adding a fixed amount in front of the focal plane. This is why you may see different DOF charts for zooms and primes at equivalent focal lengths."

This is very true, and has led to a lot of confusion.

However, when measuring DOF for focus splits, the essence of what is done in this business and why DOF tables were created, the footage marks on the lens are used, which are set from the image plane, and not the nodal point or entrance pupil. So you would say that DOF is measured from the film plane, but calculated using the nodal point. (The nodal point really only comes into play in process and miniature work.)

Brian "and that's enough about this" Heller
IA 600 DP

Stephen Williams writes:

>All I am saying is that there could be exceptions to this 'rule of thumb'

It's not a rule of thumb. It's not an observation. It's not a pet theory. It's a basic principle of optics.

Depth of field tables on the other hand, are generic and are intended as a guide, not as the final word.

A zoom lens that has it's entrance pupil a full yard behind the film plane is not going to fit perfectly into any generic DOF table designed for standard primes.

Similarly -- and depending on conditions -- a focus split may appear soft even though it is within the parameters of the DOF table, or vice versa. A poorer quality lens may very well appear to have more DOF than a high quality lens, but it doesn't. It simply appears that way because the overall image is not as sharp. Likewise a lens may be razor sharp at f:5.6, but much more forgiving for focus splits at f: 4.5 even though according to the tables there should be less DOF.

The only way to determine the DOF of any given lens is to test it.

Brian Heller
IA 600 DP

> "Zooms have some special characteristics when it comes to depth of >field.

Nonetheless a zoom lens at say 50mm will behave like a 50mm lens; at 25mm like a 25mm lens, no ?

Maybe it would be better to state it as "Zoom lenses have to be given special consideration as to how depth of field is calculated" but even then, you might just as well say the same thing if you're talking about a retro focus wide angle lens as compared to a "normal" lens it would seem to me.

Sam Wells
films/nj/usa

The problem of relating the DOF of zooms to that of fixed focal length lenses has always been a problem.

I first became aware of it when I produced the first Samcine calculator which was designed for me by Bill Pollard, the originator of the Kelly calculator and was based on the same arithmetic. It was while I was using it that I noticed that it aligned very well with Cook's tables for the Speed Panchro lenses but not for the Cooke and Angenieux 10:1 zooms. I went back to Bill and said to him 'we have a problem'.

Six months later he had come up with new arithmetic and designed the Samcine Mk II calculator which had different scales for fixed and for zoom lenses. The difference was because DOF works from the front entrance pupil of a lens and which on a zoom lens can be as much as 16 inches in front of the tape hook so when you think you are focusing on, say 3 ft, the lens is actually focusing on a plane that is only 1 ft 8 inches away so if anyone gets something out of focus you can blame the calculator if it doesn't say which type of lens it is designed for!

(From where to measure from with a diopter supplementary lens is another thing again!)

Trouble is that the Samcine type of calculator is very difficult and comparatively more expensive to manufacture ... as I have found out when my son and I fairly recently updated and redesigned it and called our's the Mk III type.

The other thing that a knowledge of where the entrance pupil is useful for is for nodal panning (I am told that this is an important factor for Movement Control and Movement Capture work and would be grateful for feed back on this as I am thinking of expanding my listings of nodal positions in the new edition of my manual that I am currently working on).

Sincerely,

David Samuelson

David Samuelson writes:

>Trouble is that the Samcine type of calculator is very difficult and >comparatively more expensive to manufacture ... as I have found out >when my son and I fairly recently updated and redesigned it and called >our's the Mk III type.

The money was well spent â€¦ The Mk I Samcine Calculator I bought at Criklewood/Broadway is still in use today. I gave it to a friend when I bought the Mk II, OK maybe I sold it to him. He in turn passed it down the line when he got his Mk III, and so on. The loader who has the Mk I doesn't really get to use it much, but it's great to practice with

>The other thing that a knowledge of where the entrance pupil is useful >for is for nodal panning (I am told that this is an important factor for >Movement Control and Movement Capture work and would be grateful >for feed back on this as I am thinking of expanding my listings of nodal >positions in the new edition of my manual that I am currently working >on).

It is indeed extremely useful to know even the approximate position of the nodal point. Trial and works, but having a fairly accurate starting point is a great time saver.

Maybe you could also include some type of GPS locator in the binding of the Manual. Mine seem to keep wandering away

Brian Heller
IA 600 DP

With regards to the discussion of zoom vs primes DOF, there might be other factors at play that contribute to the slight differences in DOF.

Zooms lose more light and have a higher discrepancy between t stops and f/stops, and its the latter that determines DOF.

Newer primes tend to be sharper, more contrasty, higher MTF glass, and therefore may exhibit a more critical DOF at the same stop. In essence, some insanely high-end primes might warrant a smaller CoC than an old 10:1 zoom (on sharp Vision2 and excellent projection, for example).

Mark Doering-Powell
LA based DP