Which system should I pursue?

[Justwalking]

I have marked in pink one set of important things to see. These things are the same in both setups. Both setups have the same number of pixels on the sensor, the same number of pixels in the COC, and the same number of pixels across each diffraction blur. If you displayed both of these images at the same size, they would look identical.

Must be identical if diffraction of the lenses is not e?ist. But it depends from F-number and F-number must be large for another focal range to cover big sensor so system with crop 5 sensor wins 5X in optical resolution (150lp/mm vs 30 lp/mm). All others on both pics would be the same.

What I am trying to help you see is that when you keep the same FOV and the same NA, both optical images look identical except that one is bigger than the other. The larger sensor gets a larger image, but it has exactly the same information content.

Difference only is that lens diffraction kills information contens on FF.

And only 3.5MP on FF will be enough to show this limit of optics.
Again I agree. But I also point out that this is exactly the same for the smaller sensor. Have you done the calculation?

Yes, exactly 3.5MP with f/50, but 3.5MP with f/10 still can resolve 149lp/mm on 5 crop and for FF f/10 still unavailable and optics can't resolve more than 30lp/mm on f/50.

[Justwalking]

There is nothing arbitrary about the f/50. By laws of optics, it is the only value that will give same FOV and NA on both these sensors.
I have marked in pink one set of important things to see. These things are the same in both setups. Both setups have the same number of pixels on the sensor, the same number of pixels in the COC, and the same number of pixels across each diffraction blur. If you displayed both of these images at the same size, they would look identical.

Would if diffraction is not exist, but can't due to lens difraction difference. That is main thing. (Green)

Again I agree. But I also point out that this is exactly the same for the smaller sensor. Have you done the calculation?

Yes. It would be the same 3.5M. The difference that on the 5X crop 3.5 still can resolve 149lp/mm for f/10 and for FF f/10 is unavalable and f/50 can resolve no more than 30. All others are same.

If you need experimental confirmation of that fact, please consider the following animation. I made it by starting with a crop from your actual-pixels image.

I have said that it is not actual dump of the camera Pic. This pic have no
to say about sensor resolution and it's ability. it's very simple camera very limited. I use it for live video most
If i want to take good pics i need something like this with corresponded soft
https://www.microscopeworld.com/p-2703- ... tware.aspx

[Lou Jost]

Hint to justwalking-- calculate the resolution in line pairs per picture width for both.

[rjlittlefield]

Other hint to Justwalking -- translate that resolution on sensor, to resolution on subject.

149 lp/mm on sensor at 0.8X is 119.2 lp/mm on subject. (149*0.8=119.2)

29.81 lp/mm on sensor at 4X is 119.24 lp/mm on subject. (29.81*4 = 119.28 )

The difference is only rounding.

Both systems have the same resolution on subject.

I am appalled that you cannot seem to keep straight what these numbers mean. It's like you've learned the formula to calculate diffraction, but you have no understanding how to apply it. Very frustrating.

--Rik

[JohnyM]

Try to divide LP/MM by crop factor.
Another brilliant plot twist incoming...

[Justwalking]

Hint to justwalking-- calculate the resolution in line pairs per picture width for both.

Resolution of what? Thel maximum theoretical resolution is above on calc.

[Lou Jost]

Those figures are the resolutions per mm on the sensors. But the sensors are different sizes. How many line pairs total across the sensor width for each sensor?

[Justwalking]

Other hint to Justwalking -- translate that resolution on sensor, to resolution on subject.

149 lp/mm on sensor at 0.8X is 119.2 lp/mm on subject. (149*0.8=119.2)

29.81 lp/mm on sensor at 4X is 119.24 lp/mm on subject. (29.81*4 = 119.28 )

--Rik

Wait a minute. That magnification already was calculated by inserting F#

No matter that resolution of the subjct is same
It is pretty easy thing to understand that FF can't resolve such resiolution on f/50.

[rjlittlefield]

Wait a minute. That magnification already was calculated by inserting F#

No. The f# was calculated based on magnification.

FOV and sensor size determines magnification.

Magnification and object-side NA determines effective f#.

--Rik

[Lou Jost]

"It is pretty easy thing to understand that FF can't resolve such resiolution on f/50"

Now you don't believe your own calculator?

[rjlittlefield]

This seems like the sort of confusion that could be resolved in a couple of minutes with the student at a whiteboard.

Draw picture of subject, small sensor, big sensor.

Point to subject, say "119.2 lp/mm here at the subject..."

Point to small sensor, say "... is equivalent to 149 lp/mm here (because this image is only 0.8 as big as the subject)..."

Point to large sensor, say "... and is equivalent to 29.8 lp/mm here (because this image is 4 times as big as the subject)".

If that doesn't work, then???

--Rik

[Lou Jost]

Or even simpler, note that the large sensor is 5x bigger than the small sensor, and the resolution per mm of the large sensor is 1/5 the resolution per mm of the small sensor. So they both have exactly the same total resolution across the senor (in line pairs per pixel width).

Or just multiply the resolutions per mm of each sensor times the width of the sensor in mm. You get the same number for both sensors. What more could you need to convince yourself?

[rjlittlefield]

If I had this problem while teaching in a classroom, I would stop, put a puzzled expression on my face, then say "OK, I'm stuck. I cannot figure out what to say that will help Andrew look at this problem the right way. Can anybody else in the class help?"

That works because often another student happens to have been confused about the same thing at some point, and can tell what got them un-confused.

So I'm going to try that approach here:

If anybody else reading this thread has ever had the same confusion, and can help get Andrew un-confused, please feel free to chime in!

--Rik

[Justwalking]

"It is pretty easy thing to understand that FF can't resolve such resiolution on f/50"

Now you don't believe your own calculator?

Why? I do, but others won't.

Let's start a count of real achievable resolution of object by math not by manipulation of diffraction limited resolution on sensors.

Pixel pitch = 0.0015mm
Object space resolution: 1.5um/0.8 = 1.875um
Object space resolution lp/mm = 1000/1.875/2 = 266.7 lp/mm
(not 119 after manipulation)

From this resolution Crop5 sensor on f/10 can resolve 149 lp/mm

_________________________________________________________

Let's see about 1X 16 MP sensor

Pixel pitch = 0.00749mm
Object space resolution: 7.49/4 = 1.875um (same as on small sensor)
(not the silly 119)
Object space resolution lp/mm = 1000/1.875/2 = 266.7 lp/mm

From this FF can resolve 30 lp/mm on f/50

How to explain easy?

[Lou Jost]

I can't think of anything more to say that hasn't been said....if he calculates the number of line pairs resolved across each sensor, and finds them to be the same, what more can he want?