These calculations are for a perfect optical system. Of course we could add calculations. The interest is to verify the theory by practice. I have already checked my resolution curves in practice .... The experience is useful but the interpretation of the experience can only be done correctly if we ...

So the MTF of my lens does not vary with the FOV as long as the sensor is sampled correctly. In this case the MTF of the sensor decreases when the a FOV increases and the MTF of the assembly decreases. On the other hand, the resolution of the system increases when the FOV increases. When the sensor ...

The problem was a bit complex. But I have finished functional sheets. This ultimately does not change my initial curves, but I can put an MTF with it. Indeed the MTF of the sensor normally tends to decrease when FOV increase. When the sensor reaches its limits for sampling then the effective frequen...

At first I had a doubt with what you were saying Rick. But I have no more doubts. I am sure of my thinking at the moment. You calculate a partial MTF or you have to calculate an MTF for the objective and for the sensor. My curves do this with a minimum criterion of 2.5 pixels for a pair of lines. Th...

For a FF 24 MPIX. I think about it a bit (I'll take a bit more time as soon as possible). Initially I was ok for your calculations but I am no longer ok at all. I have to clear it all up ... I have improved my spreadsheets but I still have to complete my thoughts. for the 5x 0.14 sensor MTF is indee...

[ But perhaps I missed something. It is hard to tell because you show only graphs and do not reveal the math behind them. --Rik Yes I did not show the math behind. Besides, it would be interesting to see if we can modulate the resolution more subtly by the contrast on the same graph? To calculate t...

Yes i agree. But i must see if find something else. This shows despite everything that it is the whole system whose MTF is measured. Because it is the interpretation of the sensor and its sampling that produces the contrast. After that, it would be necessary to see if the corrections of the contrast...

MTF = MTF objective x MTF sensor

I must seen that when i have more time. You study just MTF sensor.

Yes i agree. But you need a 61 Mpix for that.... with a FF 24 Mpix or APS-C 24 Mpix MTF should be not very different.

And 7x 0.21 must be better .

and I find it hard to think of any reason other then the resolution s higher, and that with a 61MP FF bayer sensor we can see the difference. You don't even need 61 MP. The diffraction-only MTF curve for NA 0.21 is higher than for NA 0.14 at all spatial frequencies. That higher MTF translates into ...

Example for FF 61 Mpix 3.5 mm to 5 mm => 10x 0.28 with 3600 LP/mm 5 mm to 7.5 mm => 7.5x 0.21 with 3600 LP/mm 7.5 mm to 10 mm => 5x 0.14 with 3600 LP/mm Yes it's better with 5x 0.21..... you have 5 mm to 10 mm with one objective but circle 7.5 x is acceptable until 9 mm... It's more economic to take...

The image circles of the mitutoyos are large. Also rather than buying a 5x 0.21 I prefer with a 61 Mpix, it will be more economical to use a shorter tube focal length to get the same sampling (and the same resolution) with the 7.5x 0.21. It is therefore necessary to place for a given FOV on a range ...

Yes sure 0.21, do a better job until 7- 8 mm fov with à 61 Mpix => Mitu 7x 0.21 > Mitu 7x 0.14 Resolution is better ( 3600 LP/mm for most of fov). This resolution is inferior until 7-8 mm fov for Mitu 7x 0.14 but after it's same except sampling. After sensor can't see the difference with 180 mm or l...

Having compared the 10x HR and normal 10x head to head on an 18mp aps-c sensor the difference was easily visible. The theory in this case doesn't hold up for whatever reason. Or experience is not valid for an other reason.... It's not normal. Then it is normal that the objective less open resolves ...

As I said only the test with a 208mm focal length tube lens with pixel shift is valid for testing the lens. If we do the test without pixel shift at 61 Mpix, it is normal to observe due to the downsampling with the 5x 0.21 the absence of colorimetric imprecision. We will see more clearly but not bec...