Depth of field as numeric range

Given the fact these (full frame)cameras are very smart and know at any point the f-stop and the focal length (even zoom lenses). And since they have very sophisticated auto-focus you would think the cameras could "range find" the focus distance. They would then have all the information to calculate "near" and "far" depth of field as a numeric and be able to display it. What do think?

vss4canon wrote in post #16742847
Given the fact these (full frame)cameras are very smart and know at any point the f-stop and the focal length (even zoom lenses). And since they have very sophisticated auto-focus you would think the cameras could "range find" the focus distance. They would then have all the information to calculate "near" and "far" depth of field as a numeric and be able to display it. What do think?

However, the evidence doesn't support that supposition. Some camera bodies with some lenses will attempt to guess the focus distance and stick that info into the exif data. Sometimes the guess isn't too bad. Sometimes it's unbelievably wrong. Indeed, I have read that some companies have abandoned the idea for now.

Here's my favourite example. The Exif reckons the subject of this image is 41cm away.

There is talk of or is already a function in magic lantern to do this. However, you would have to provide the final piece of info yourself, the distance. It will then tell you what kind of DOF you are looking at. That could be done with Canon's firmware too. Yes it would be possible to get most of the way there, but I am not sure it would be a large enough market differentiator/revenue opportunity for Canon to spend the time developing that function.

The final pieces of information are surely enlargement and viewing distance...

JohnB57 wrote in post #16742908
The final pieces of information are surely enlargement and viewing distance...

Enlargement contains focal length and distance and sensor format, correct? If so the camera knows 2 of those 3.

As to final print results/enlargement needs, that really cannot be an input, because those could vary much later after post processing and final output needs are assessed at that point.

In the field though, the camera knows 3 of the 4 factors for useful DOF numbers, for now a human has to enter the distance, the lens and bodies just aren't good at determining an accurate distance. There are some videos showing Magic Lantern DOF calculations I believe.

http://magiclantern.wi​kia.com/wiki/FileOF_and_Focus_stacking

JohnB57 wrote in post #16742908
The final pieces of information are surely enlargement and viewing distance...

and now we just go in circles of confusion.....

Seriously though if someone is using DOF charts, they aren't concerned about print size and viewing distance, just getting everything sort-of in focus.

Without considering print size or print viewing distance, the camera could assume a Circle of Confusion of .03 (full frame cameras) which would work pretty good, at least as good as most "depth of field" previews. Canon L lenses should be accurate enough to "range find". Internally, meaning for any rotational position the lens would have a corresponding focus distance.


note: Most DOF calculators use CoC=0.03 for full frame. The calculators are useful in combination with a good laser range finders.

vss4canon wrote in post #16743047
Without considering print size or print viewing distance, the camera could assume a Circle of Confusion of .03 (full frame cameras) which would work pretty good, at least as good as most "depth of field" previews. Canon L lenses should be accurate enough to "range find". Internally, meaning for any rotational position the lens would have a corresponding focus distance.


note: Most DOF calculators use CoC=0.03 for full frame. The calculators are useful in combination with a good laser range finders.

Now you're talking...Add a laser range finder to the camera...Problem solved!:p

Distances reported by lenses (even long tele lens) are pathetic in resolution and max distance

https://photography-on-the.net …p?p=16502870&po​stcount=14

I just conducted this test to demonstrate precisely how poorly distance is reported by the lens and stored in the photos...Canon 17-55mm at 51mm FL f/2.8, a series of 20 photos taken at intervals of 1.0' as measured with tape measure, from 4' to 24'. EXIF data for distance recorded for each exposure.

hollis_f wrote in post #16742895
However, the evidence doesn't support that supposition. Some camera bodies with some lenses will attempt to guess the focus distance and stick that info into the exif data. Sometimes the guess isn't too bad. Sometimes it's unbelievably wrong. Indeed, I have read that some companies have abandoned the idea for now.

Here's my favourite example. The Exif reckons the subject of this image is 41cm away.

hollis - maybe "cm" stands for "celestial measurement", where 1 is the distance to the nearest star. LOL.

OP: For me; it would be more useless info I'd ignore - unless they made it into a automated mode. For example; creating a "group shot" mode that, combined with face recognition software, would determine all the faces in the group (ignoring those too distant or close) and picked a setting that has all of them in focus while allowing adequate shutter speed and ISO of 800 or lower.

That said; I have yet to use a "mode" other than P, AV, TV, and M.

TeamSpeed wrote in post #16742956
Enlargement contains focal length and distance and sensor format, correct? If so the camera knows 2 of those 3.

I was referring to enlargement to final image/print size so no, the camera doesn't have this info.

I was being a bit pedantic as actually, I think it could be quite a useful tool, albeit one that requires some interpretation on the part of the user.

JohnB57 wrote in post #16743507
I was referring to enlargement to final image/print size so no, the camera doesn't have this info.

I was being a bit pedantic as actually, I think it could be quite a useful tool, albeit one that requires some interpretation on the part of the user.

Ah gotcha!

Heck, I don't even know that value until I am done with my pics and decide which are going to be printed in what size.

Wilt wrote in post #16743446
Distances reported by lenses (even long tele lens) are pathetic in resolution and max distance

https://photography-on-the.net …p?p=16502870&po​stcount=14

I just conducted this test to demonstrate precisely how poorly distance is reported by the lens and stored in the photos...Canon 17-55mm at 51mm FL f/2.8, a series photos taken at intervals of 1.0' as measured with tape measure, from 4' to 24'. EXIF data for distance recorded for each exposure.

It's hard to respond to your post without knowing exactly how this this EXIF information is determined. My point in regard to "range finding" involves using rotational gear position in relation to the know characteristics of any given lens to determine "focus distance". The rotational position of the ultra sonic motors (USM) gears which drive these lens barrels through their focus, together with a first or second order equation should allow one to calculate the focus distance which can in turn be used with the other know variables (f-stop and focal length) to determine "near" and "far" focus limits.

Did someone have a mode where you could focus first near and then far and the camera would set focus and aperture to capture both in focus?

And, fwiw, DOF scales calibrated to the current standards don't work, IMHO, with current digital imaging. The lenses and sensors are sharper now than when the standards were developed decades ago so we see a significant change from sharp to blurry sooner than we used to. Thus effective DOF is typically considerably less than the scales would have you believe.

Gobeatty wrote in post #16746947
Did someone have a mode where you could focus first near and then far and the camera would set focus and aperture to capture both in focus?

Yes, Canon had this mode in select film cameras, it was called "DEP" mode(Not like A-DEP mode that is or was available on digital Rebels).

In the late 90s to early 2000s there were cameras(such as the EOS 3) which had Eye Controlled Focus Point selection which could be used in conjunction with DEP mode so that all you had to do was look at your nearest and furthest point that you wanted in focus. The camera would then set the focus distance and aperture(if available) so that your desired DOF was achieved.

I sure wish that they would bring back both of those great features.