[GUB]

Yes but so long as you are using the calculations for comparison purposes it is fine. In the rare situation where I am pre-planning a shot I halve the coc.

[ZombieArmy]

Oh boy, it's this thread again.

[GUB]

Not at all - we are participating in a discussion with no particularly entrenched positions and plenty of courtesy. Dof and Format is really hard to get your head around. I have to say I am currently in the unfortunate position of disagreeing with Photoptimist - unfortunate because he has a strange habit of being right!

[Serkevan]

This is interesting. Depth of field and digital sensors - Artfx
The conclusion is " The depth of field scales engraved on the lenses are handled with care: They are often calculated with the value of a circle of confusion which isn’t adapted to digital cameras. Leica, for example, they are calculated with a value of 25µm, completely valid in film but totally irrelevant if you put your lenses on a M8 or a M9! On a M8 or M9, the size of the photosites are 7µm, multiplied by 1.5 this gives 10.5µm for a realistic circle of confusion! The depth of field scales engraved on the lenses are therefore largely overestimated when using lenses on digital cameras."

Hah, so this explains why I have to set hyperfocal for f5.6 or so when shooting at f8 on my Pentax-M lenses!

This is very interesting indeed, thanks a lot for the info.

---------- Post added 10-06-19 at 11:41 PM ----------

Not at all - we are participating in a discussion with no particularly entrenched positions and plenty of courtesy. Dof and Format is really hard to get your head around. I have to say I am currently in the unfortunate position of disagreeing with Photoptimist - unfortunate because he has a strange habit of being right!

If it makes you feel any better, I'm in the unfortunate position of having no clue anymore

[Dartmoor Dave]

Not at all - we are participating in a discussion with no particularly entrenched positions and plenty of courtesy. Dof and Format is really hard to get your head around. I have to say I am currently in the unfortunate position of disagreeing with Photoptimist - unfortunate because he has a strange habit of being right!

I also find myself in the unfortunate position of not entirely agreeing with @Photoptimist, because I think he's oversimplifying how downsampling works in his example. But having said that, I know from experience that disagreeing with him can end up with me having egg on my face, because he actually knows a heck of a lot more than I do.

I do have to point out to Photoptimist though: The Windows 7 computer that I do my photo editing and viewing on has got no idea what physical screen size I'm using, but only the pixel resolution of the monitor attached. Nor does it have any idea what physical size sensor was used to take any particular photo, but just the pixel resolution. So any relationship between the physical sensor size and the physical viewing size is incidental rather than direct, caused only by the fact that each pixel on the monitor is far bigger than the pixels on the camera sensors (and the pixels in the monitor image are the result of downsampling too). And when we're talking about prints, I think we're better off sticking to non-interpolated prints of the full sensor resolution, because resampling algorithms are a factor better left out of a discussion like this.

Note to Photoptimist: I'll take the eggs on my face fried sunny side up, with nice crispy brown edges please. No doubt you'll be serving them up to me soon, with some math on the side.

[GUB]

But, I reiterate, it needs to know the proportions of the sensor and for that matter the proportions of monitor or print. Otherwise the data would come out a meaningless strip.

[swanlefitte]

DoF, sensor size, and pixel pitch
Gkramer reply #6 seems to have the math behind it. Very confusing because viewing distance vs different display size at 300ppi is compared. (Amongst other confusion)

[Dartmoor Dave]

DoF, sensor size, and pixel pitch
Gkramer reply #6 seems to have the math behind it. Very confusing because viewing distance vs different display size at 300ppi is compared. (Amongst other confusion)

Thanks for that link. I won't get time to read it in detail until later, but from a quick look I think I'm in complete agreement with Gkramer's original post, particularly on the question of "enlargement" and why we can't talk about that in digital photography in the same way that we do with film.

[Serkevan]

DoF, sensor size, and pixel pitch
Gkramer reply #6 seems to have the math behind it. Very confusing because viewing distance vs different display size at 300ppi is compared. (Amongst other confusion)

That's what I was thinking about and the math seems to pass a quick sanity* check. Thanks for another interesting piece of info, this discussion has been lovely.



*Debatable and probably overestimated as it's Monday morning.

[swanlefitte]

Rereading that I think the math says scaling pixel pitch to sensor size will be very similar to film and not scaling pitch will exaggerate this at the same viewing distance so probably similar to film.
Is that posts math correct? Is my understanding of it correct? So many conversions and variables that I can't keep them all in order.
I think with uncertainty I am going to the Photoptimist side.

[slartibartfast01]

Just as a thought experiment. What if the sensor measured 10" x 8" and print size was the same. The CoC would be exactly the same on the print and the sensor.

[GUB]

Thanks for that link. I won't get time to read it in detail until later, but from a quick look I think I'm in complete agreement with Gkramer's original post, particularly on the question of "enlargement" and why we can't talk about that in digital photography in the same way that we do with film.

I think Gkramer is outright wrong here and the correct response is BJL as far as the dof calculator goes. But to reap rewards from forever increasing megapixels the images have to be viewed in a pixel peeping way (massive enlargements) and then his comments come true.
EDIT Isn't it funny to look at the same argument from 12 years ago.

[Dartmoor Dave]

I've only had time to have another quick read through it over lunch, but I strongly agree with Gkramer that CoC makes more sense when stated in terms of the number of pixels it covers on the sensor than it does when stated in terms of physical size. Physical size is such an analog concept.

There's an awful lot of people in that LuLa thread refusing to acknowledge any difference between digital and film though. One guy even refers to Ansel Adams as an authority. About film, yes, an incontrovertible authority. About digital, not so much.

[BrianR]

Photoptimist and BrianR, can you think of an imperative where the computer has to know the size of the sensor?

The computer would not need to know the physical size to display the image. Aspect ratio would be needed though (I assume there's provisions for pixels that aren't square).

The key point behind trying to measure DoF is figuring out how a blur disc of a given size that's being projected on your film/sensor gets mapped to a disc in the final image you're viewing. Figure out how to do that and you're good. It doesn't matter if the digital file contains no size information, I know what camera it came from, I know how large the sensor is, I can figure out how many pixels across a blur disc was and I can figure out how this translates to a print. Of course it's an approximation to skip from the projected image straight to the print by a straight up multiplication by (print size)/(sensor size), but at the sizes of blur discs we're usually concerned with, I'd wager you'll be just fine. If it weren't, digital photography would have such low fidelity that our images would all be unrecognizable mush.

Be very careful with changing assumptions like different viewing distances, print sizes (or 100% view on a monitor), subject distances, etc. Defining DoF in terms of pixel pitch is fine if you want to run with the base assumption that you're viewing everything at 100%. That's perfectly fine, but in practice that's not how I work when switching from one camera to another or cropping. e.g. if I'm printing a bunch of 8x12's for a little display out of my k5iis and I have a nice photo from my k100d I'm going to include, I don't print the k100d image smaller.

---------- Post added 10-07-19 at 08:51 AM ----------

Just as a thought experiment. What if the sensor measured 10" x 8" and print size was the same. The CoC would be exactly the same on the print and the sensor.

This is the fun situation of a pinhole camera with large paper negatives. I use 5x7" in mine, but it's a good way to experiment with large 'film' sizes on the cheap.

---------- Post added 10-07-19 at 09:23 AM ----------

I've only had time to have another quick read through it over lunch, but I strongly agree with Gkramer that CoC makes more sense when stated in terms of the number of pixels it covers on the sensor than it does when stated in terms of physical size. Physical size is such an analog concept..

I just had a quick skim, but note the reliance on pixel pitch, which is a physical property of the sensor. Accepting a definition that uses pixel pitch but rejecting one involving the sensor size (or vice versa) would be inconsistent as they're saying pretty much the same thing. The real difference is what assumptions you're making about viewing size and either formulation should be flexible enough to include any viewing condition.

Deciding one starting point is more in line with your use case would be a subjective decision.

[clickclick]

Oh boy, it's this thread again.

Yeah. These things happen....