Question about DOF

[rut]

Hi All,

a doubt: a len mounted into a S2 multiply its focal from 1.5x.

DOF change from original ?

Thx.
R.

[amazingthailand]

No Change.

Declan

[rut]

Originally posted by amazingthailand
No Change.

Declan

me happy :)
Thanks,
R.

[Stormfield]

If you are asking if the DOF will be different for the same focal length, distance, and aperture using a 35 mm film SLR vs the S2, it will be (the hyperfocal distance will be greater). I posted this discussion to a similar question in the dpreview S2 forum:

Tho DOF is determined by aperture, distance to subject, and lens focal length, it is dependent on the recording media's size (e.g. 35mm film or CCD size). The actual calculation of the DOF (near/far) is dependent on the value of the "circle of confusion" that is used, which relates to the recording media size (for the most part, to keep it simple). The COC is how 'in focus' is defined mathematically. In any case, if a digital camera's CCD (or CMOS) is smaller or larger than 35mm film, then the tables which were calculated based on 35mm film will not be valid. Thus a new value for the COC must be derived to recalculate the DOF in the same tables for the sames lenses, etc. The new COC is calculated: COC for 35mm/ digital FL multiplier. Thus, for the S2, the COC calculates to be 0.022 (using 0.033 for 35mm film, and an S2 multiplier of 1.5). It is interesting that the various lens manufacturers don't seem to publicize the exact value for COC that they are using in their lens DOF calculations, but convention suggests using 0.033 for 35mm film. Thus, the Hyperfocal Distance for a digital camera with a CCD/CMOS smaller than 35mm film, given the same subject distance, aperture, and lens focal length, will actually be GREATER than for 35mm film (and thus as calculated in most DOF tables). The issue of COC values for digital cameras is actually a bit more complicated when you have to take in to account CCD pixel size, spacing, and array interpolation, but the above values are probably sufficient to make practical use of in-field DOF calculations for the S2. Thom Hogan in his new S2 eBook will apparently be providing an Excell spreadsheet formula to use to calculate DOF for various lenses used on the S2. You can see and calculate the numbers yourself with a DOF/HFD calculator here: http://dfleming.ameranet.com/dofjs.html

[rut]

Thanks for exaustive response!

So, I can't use printed grid into my 50/1.8 lens for hyperfocal, right ?

R.

[Stormfield]

Rut: That is correct, the DOF scale on your lens will overestimate the DOF for the S2. For example, if you are taking a picture with the S2 with your 50mm @ 7' with f/2.8, your DOF will be 6.7 to 7.4, or only 8 inches; for 35mm film, it would be 6.5 to 7.5 (a total of 12 inches). You can compensate by either stopping down (@ f/4 the DOF on the S2 is 6.5 to 7.5, or 12 inches--the same as the 35mm film DOF at f/2.8), or by increasing your distance: increasing to 9' distance will give you 13 inches DOF on the S2 @ f/2.8. When DOF is crucial- such as tight portraits or objects, it is probably best to consult tables for the DOF--as noted, Thom Hogan will be providing an Excel formula for the S2, or you can calculate each DOF situation by using the formula for DOF, or use the calculator that you can find on some sites (such as the site noted in my previous post). I've noticed that some discussions about 'problems' with focus with the S2 are actually issues with a compressed DOF with the S2--particularly noticeable with close-up macro shots and close subjects using wide-angle lenses. Hope this helps (and I hope I haven't increased the Circle of Confusion about this topic!).

[Marvo]

Hi John, can you help me out here?
If I were to shoot onto a 35mm piece of film in a regular N80 with 50mm f1.8 and the subjuct is rendered "sharp" from 8ft to 10ft from the lens and then was to shoot the same scene at the same settings with same lens on the S2, surely the subject will still be rendered sharp to the same degree in the same 8ft to 10ft range?
OK, the the image will have effectively been cropped by the fact that the CCD is smaller than the 35mm film area, but the physics of what's sharp in the area of the image in an area equivalent to the CCD will be the same - for a given lens and apperture. I always thought that DOF was a function of the "ratio of reproduction" of the subject-to-image in combination with shooting apperture.
If you can tell me where I'm wrong, I'll reprogram the grey matter!
All the best,
Marvo.

[Stormfield]

The answer to your question would be, no, the DOF for the same settings IN YOUR RECORDED IMAGE would be different. Since it is the recorded image we are interested in, we have to take into account the recording media size. Here in lies the issue with Circle of Confusion (COC). There are basic mathematical relationships which ALWAYS apply in DOF calculations, but the COC number (used in those calculations) is empirically derived FOR THE SPECIFIC recorded imaging requirements. However, to reduce the amount of chaos in dealing with this one aspect of picture-making (DOF), convention (generally) has it that the COC is derived for viewing an 8 X 10 (inch) photo at normal viewing length, with the average human eye resolution of 3 or 5 mm per line (works also for 5X7 at normal viewing, etc). The various lens manufacturers don't seem to disclose what they base their determination of COC on for the DOF scales on their lenses, but most seem to be using a COC of about 0.03 for 35mm film.
Back to your original issue. In your example, the DOF for your 50mm lens at f/4 (for example) would be from 8.7 to 11.7 feet for 35mm film. Since the S2's CCD is smaller than 35mm film, the COC is smaller [COC = 1/ Resolution factor, where the Resolution factor is defined as 1625/ diameter of the media. the 1625 is derived from 5mm per line resolution X diameter of an 8 X 10 diagonal in mms]. The DOF for the S2 with same parameters would be 9.1 to 11.1 feet, a smaller DOF range, and a bit farther away.
The important thing about the different Hyperfocal Distance and different DOF on the S2 over 35mm film, is that it directly impacts your picture making in the real world, and that the numbers on your lenses and 35 mm film DOF tables do not apply. As I noted previously, some of the "sharp focus" problems that some have talked about in various forums in regard to the S2 are actually issues relating to the different DOF for the same lens, F-stop, and focal distance for the S2, where the S2 will usually have a smaller (contracted) DOF compared to the 35mm film tables.
I have not been able to find out if the DOF preview has been 'adjusted' by Fuji for the S2's CCD--I suspect it has not. The DOF preview, therefore, will not be accurate for the final recorded image! The up-side is that the DOF differences make the most difference in close focal distances and Macro, the landscape photographers won't likely notice the difference except for very wide angle lenses with close focal distances, but portrait photographers need take heed.
In summary, the reason DOF changes with S2 is because the smaller (than 35mm film) CCD is magnified more by final enlargement than is 35mm film, and that factor alone is sufficient to change the DOF in the final recorded image. You might want to explore the site I gave in the earlier post, fill in numbers into the calculator, and explore the formulas yourself.

[amazingthailand]

The image is NOT MAGNIFIED AT ALL.

The lens produces the exact same image on the S2 as it would on an F5. The difference is the S2 CCD is smaller than the 35mm film aperature and so we only record a smaller portion of the image and that gives us an APPARANT 1.5x magnification. BUT, the image at the focal plane is IDENTICAL and therefor the DOF MUST be identical.

The rest of the discusion above is simply smoke and mirrors.

Declan

[Seachicken]

I love this thread... :)

Until yesterday I too was a firm believer in "no change" in the depth of field. The lens is the same, so why should anything change just because the sensor is silicon rather than film?!

Unfortunately, now I see the problem... what a swine it is too...
My view of the problem is that to make a full frame S2 print the same size as a full frame 35mm film print, the S2 image must be enlarged more... and when you do this, things that seemed sharp at the smaller degree of enlargement, now seem no longer acceptably sharp... and this changes the depth of field... the range of distance in the real world for which the image is acceptably sharp.

Anyway, thanks for taking the time to explain all this in such detail... it's something I wasn't aware of.

Andy.

[Stormfield]

Andy hit the nail on the head. No smoke and mirrors. Just real world physics (of course, that's what real magic is all about too!).

[Marvo]

Nice explanation John, we've all now got something to think about. :rolleyes:
What a revelation!:eek:

Thanks for your time on this one.

Marvo.

[amazingthailand]

It's still smoke and mirrors.

The subject matter of the discussion has changed from DOF to duplicating the image size in an enlargement. Of course, to produce the same image on a 35mm film and on an S2 sensor would require that the subject to camera distance be different and THAT would result in a difference in the depth of field.

However, that was not really the initial question, or at least not how I interpreted it.

If you take 2 cameras (an S2 and an F5) and take the same photos from the same position with the same lens, then the DOF is identical.

Declan

[memobug]

Originally posted by amazingthailand

If you take 2 cameras (an S2 and an F5) and take the same photos from the same position with the same lens, then the DOF is identical.

As long as they are enlarged the same amount, this is true, but in the real world, why would you ever do that?

One would probably want comparable photos; relative subject size AND perspective, and that subject size would require a different focal length on each camera.

Regards,

Matt

[tony field]

If I remember the arithmetic correctly, for all images (within the limits of defraction, resolution contrast, etc) taken with a camera that are shot at the same f-stop and the final images are enlarged to have the same image content have the same depth of field.

For example, a portrait with an F5 and an S2, both using a 85mm lens at F5.6, will have identical depth of fields if the printed portrait images are identical in size and image content.



The depth of field preview on the S2 is "perfectly accurate" in the following sense: take a picture of a scene with your S1 at a certain f-stop. Immediately print the full frame on your handy portable printer :-) Using your right eye, look through the viewfinder of the S2 and enable the depth of field preview on the scene that you just shot. At the same time, view the print you just made of the scene with our left eye. Adjust the viewing distance of the picture so that it looks identical in size (in your left eye) to the viewfinder image (in your right eye). The visual depth of field will be identical - within the limits of the ground glass roughness limits.

One additional, and possibly unexpected, observation is that perspective is purely determined by the location of the optical centre of the lens when the shutter is pressed. It does not depend on the focal length of the lens - only the angular view (i.e. NOT perspective) is determined by the lens / film size combination.

tony

[Seachicken]

a portrait with an F5 and an S2, both using a 85mm lens at F5.6, will have identical depth of fields if the printed portrait images are identical in size and image content.

... so the DOF is the same :)

BUT...
The answer to the question "Does the DOF scale marked on my lens give me the DOF I will get in the final S2 image?" is NO.

... so the DOF is not the same :mad:

Both of the above are true...

... "So what! I'll just use the DOF preview and forget all this confusing nonsense!" .. but what about the DOF preview? Is that accurate? What you see is what you get after all... or NOT, since the viewfinder image is smaller than a 35mm SLR's so the same "smoke and mirrors" trick can be used here to deprive you of that precious DOF.

Andy.

[memobug]

But it isn't possible to do what you suggest below and get "identical size and image content."

The subject would be a different size (occupy a different % of the frame) unless you move closer with the film camera. Moving closer to the subject would change both the perspective and the DOF.

To get an "equivalent' digital image with your S2, you need to keep the same subject distance, use a shorter focal length lens and open the aperture nearly 2 stops to compensate for the difference in DOF.

Regards,

Matt

> For example, a portrait with an F5 and an S2, both using a
> 85mm lens at F5.6, will have identical depth of fields if the
> printed portrait images are identical in size and image content.

[tony field]

The depth of field preview is accurate in the following sense.

1) Take a picture of a scene and immediately print it on your portable printer full frame :-)

2) View the scene again (same fstop) with the S2 in your right eye and press the preview.

3) View the printed image with your left eye and adjust its viewing distance so the image "appears" identical in your left eye (the print) and your right eye (the S2).

4) The visually perceived depth of field will be identical (within the limits of the poor ground glass).

5) If you change magnification, viewing distance, etc, the preview will be relatively inaccurate.

The above is ALL you can say about DOF preview - S2, F5, Hasselblad, 4x5 etc. It is only a GUIDE to the nature of your picture . Even if you have a magnifying glass to view the ground glass (I can use a 3x Emoscop for detail examination), the actual accuracy is in question.


tony

[tony field]

The depth of field markings on the lens are designed by circle of confusion calculations for a 35mm full frame. This is inappropriate for the reduced size frame of the S1. For a quick and dirty solution to this, read the DOF scale for one stop larger than you are shooting. If you shoot at f5.6, look at the DOF for F4, etc. (maybe 2 fstops - i have not done the arithmetic - but probably 1)

Remember that the DOF scale for any lens of any format (35mm, 6x6, 6x9, etc) is for (approximately) an 8x10 image viewed by normal people with normal eyesight at arms length (i.e. a "normal" viewing distance). If you make 16x20 prints, your images will be fuzzier than expected, if you make 4x6 prints your images will be sharper than expected when held at arms length.

tony

[amazingthailand]

Depth of Field has nothing to do with print size and certainly was never built around any one print size.

Depth of field is simply defined as the zone of sharp focus, in front of and behind the object upon which you are focusing.

So if you focus on an object at 10 feet, then use the DOF lines on the lens (if present) to determine the DOF. This will be the same for the same lens at the same aperature if focused on the same object at a distance of 10 feet, be it attached to an F5 or an S2.

Check out the following: http://www.kodak.com/global/en/consumer/pictureTaking/cameraCare/cameCar6.shtml

Jeeze, hasn't anyone heard of KISS (Keep It Simple Stupid). You guys are just going to too much trouble to complicate a very simple issue.

Declan

[tony field]

> Depth of Field has nothing to do
> with print size and certainly was
>never built around any one print size.

Re-read Stormfields previous description of circle of confusion. His comments are accurate. The circle of confusion has everything to do with print size, viewing distance, etc. DOF scales are entirely based upon circle of confusion calculations.

> Depth of field is simply defined as
> the zone of sharp focus, in front
> of and behind the object upon which
> you are focusing.

Create two prints of the same scene with good foreground and background detail shot with a wide open aperture. Make one 16x20 and one 4x5. View both of them from 1 meter. Each print will have different perceived "zones of sharpness"

> So if you focus on an object at 10 feet,
> then use the DOF lines on the lens
> (if present) to determine the DOF.
> This will be the same for the same
> lens at the same aperature if focused
> on the same object at a distance of
> 10 feet, be it attached to an F5 or an S2.

Only if you create prints that are of identical size and identical image content.

Explain why a 90mm is considered to be a telephoto on 35mm, a normal lens for 6x9 and a wide angle for 4x5. In each case, the DOF lines for the lens are dramatically different.

By implication, since the F5 and S2 have different image (sensor / film) sizes, they should have different DOF scales for any given lens - just as the 90mm lens DOF scale depends upon size of the negative.

> Check out the following:
http://www.kodak.com/global/en/consumer/pictureTaking/cameraCare/cameCar6.shtml

Unfortunately, Kodak left out the important IMAGE SIZE and VIEWING DISTANCE parameters.

> Jeeze, hasn't anyone heard of
> KISS (Keep It Simple Stupid).
> You guys are just going to too
> much trouble to complicate a very simple issue.

Do not confuse simplicity with completely missing the point.

Examine any serious photography book (something like the ancient Ilford Manual of Photography would be good).

tony

[amazingthailand]

You are confusing apparant sharpness in a print of a given magnification, which has nothing to do with the depth of field produced by a lens. Two different subjects.

Anyway, I give up, you guys can blow all the smoke and mirrors you want.

Bye
Declan

[Stormfield]

Physics is useful for me (nowdays) only so far as it can show me how to do something better (or correctly). I just have to ask the right question, and I have to recognize that the laws of physics will continue unabashed whether I like what it tells me or not. My favorite professor used to harp to me: Simplify! Simplify! So:

Question #1. Can the DOF scales and tables I've been using for years on my F5 still be used, with the same lenses, on my S2? Short answer: no.

Question #2. Why not? Short answer: Though my Nikkor lenses are the same, my cameras are different. Of all the differences between my F5 and my S2, only the different sizes of recording media between the 2 cameras matters (F5=35mm film; S2=smaller CCD). Depth of Field depends only on:
[1] focal length of lens- same on F5 and S2
[2] Aperture- same on F5 and S2
[3] focal distance- same for F5 and S2
[4] Circle of Confusion- DIFFERENT for F5 and S2-- the COC for the F5 is GREATER than S2, no matter how you determine it, because the S2's CCD is smaller than 35mm film. period.

Simplified: The COC is a CONSTANT for any given camera (regardless of the lens used), and is different only between different cameras that also have DIFFERENT SIZED recording media (35mm film, or APS, or CCD, etc). If the CCD is the SAME size as 35mm film, then I can use the len's manufacturers scale on my Nikkor 35mm lens for DOF- Kodak's new DCS 14n is such a camera (no more confusing DOF dilemmas if you buy this digital camera).

Last question. So why does all this really matter? Short answer: Because I like to get the DOF in my photos right the first time. Physics is better than trial and error.

[tony field]

> Simplified: The COC is a CONSTANT for any given camera
> (regardless of the lens used), and is different only between
> different cameras that also have DIFFERENT SIZED recording
> media (35mm film, or APS, or CCD, etc).

COC is a CHOSEN value for a specific degree "unsharpness" in the final image size that is tollerable for a specific viewing distance and resolving power of the eye. And out-of focus area that results in a point image dispersion that is the same size or smaller than the COC is considered to be "sharp".
tony

[Marvo]

Can I chip in with my UK's 2 penny worth of simplification for anyone who's interested and not fallen asleep yet.
With reference to Depth of Field, what a lot of people don't realize is that when shooting at ANY apperture, only one minute point/wafer thin plane in the subject is rendered truely "sharp" on the film/CCD plane. All other points/planes in the subject are rendered to the film/CCD with variing levels of "unsharpness". Close to the principal point of focus, these levels of unsharpness are inperceivable to the eye and so are deemed "sharp". At some point in front of and behind the principal point of focus, some of these levels of unsharpness ARE perceivable to the eye and so these areas are deemed "unsharp".
This is all happening at the film/CCD plane, so as this information is enlarged (e.g. when printing the image), progrssively, more of the once perceived "sharp" information appears to be "unsharp". This is why the amount of our subject that appears "in-focus" at a given apperture is directly related to the size of reproduction of our images for a given viewing distance.

Marvo.

[tony field]

Originally posted by Marvo
With reference to Depth of Field, what a lot of people don't realize is that when shooting at ANY apperture, only one minute point/wafer thin plane in the subject is rendered truely "sharp" on the film/CCD plane. All other points/planes in the subject are rendered to the film/CCD with variing levels of "unsharpness". Close to the principal point of focus, these levels of unsharpness are inperceivable to the eye and so are deemed "sharp". At some point in front of and behind the principal point of focus, some of these levels of unsharpness ARE perceivable to the eye and so these areas are deemed "unsharp".

Marvo.

Exactly!!!! the COC is the assignement of a numerical value to the degree of unsharpness - and is one of the primary numerical basis for all estimates of things like depth of field tables, markings on lenses, etc.

tony

[mrrjm]

First I have to say I'm an new at all this. I read all the posts and seem to understand most of it. I think I know the answer. When Fugi releases the S3 I assume it will have a 35mm size CCD. So we all need to trade in out S2 for and S3. I apologize if this sounds stupid. But if the CCD was 35mm size problem solved. And we would probably have results that aproach med format.


Ray

[steve bingham]

Circle of Confusion is what determines depth of field. This is more than simply focal length, f stop, and distance focused at. See http://www.northnet.org/jimbullard/CoC.htm for a very simple explanation.

However,
http://www.nikonlinks.com/unklbil/dof.htm will give you a better understanding AND formulas to determine for YOURSELF the correct answer. None of which is close in all the above explanations. Sorry.

[steve bingham]

If the LENS SIZE (focal length) is the same, the larger format always has greater depth of field AT THE SAME DISTANCE. Take it to the bank. (35mm film is a larger format than a S2 CCD)

On the other hand, if the angle of view is the same, the smaller format always has greater depth of field AT THE SAME DISTANCE.

No smoke and mirrors. Simply high school physics. So in using my 50mm lens, I realize the actual depth of field is actually less - kinda like using a 75mm. Pretty simple. Of course when you back away from your subject all bets are off. You have changed the parameters.

Don't you just love it!

[memobug]

Well, in most cases I don't think I would take a picture of the same subject with an F5 and an S2 at the same focal length, aperture and shooting distance. Why? because the subject would be nearly twice as big on the S2. The temptation would be to back up to decrease the subject size, but that itself changes the perspective of the shot and completely changes the DOF calculation.

Instead, if I wanted to take an "identical" picture to the one I took with film, I would do the following.

1) Maintain the exact same subject distance for identical perspective. See Note 1
2) Use a wider focal length to get a proportionally identical subject size. See Note 2
3) Open the aperture more (smaller numeric Fstop) to compensate for the DOF change which resulted from the wider lens in #2. See Note 3

Notes:
1. Perspective is determined by shooting distance, not focal length
2. For S2, the required focal length can be found by taking the film camera focal length and dividing by 1.5
3. For S2, the aperture setting required is F=Fn/1.5 where Fn is the 35mm film aperture. e.g. for Fn=5.6, F=3.73 so use about the f/4 setting for similar DOF.

Summary: If you want to take an S2 photograph that is as close in content to your film image (subject size, perspective, DOF) you will need to decrease the focal length and decrease the Fstop number (open the aperture).

Regards,

Matt

PS - glad I caught my error earlier. I shouldn't type and try to hold the infant!