Macro APS-C, FF confusions

Susurs · Oct 17, 2017

Hi!

I am trying to figure out some things with regards to macro photography and comparison between FF and APS-C sensor while using the exact same FF lens (Sony SEL 90M28G).

I'd be happy if someone would confirm my assumptions, or explain tings a bit for me...

If the sensor in both cameras is 24.3MP, and magnification of the lens is 1:1:

1) @ min focus distance 1:1 magnification, an object which is ~23x15mm will fill the output image for APS-C (6000x4000 or whatever, same res for both cameras).

2) If yes, same object @ same min focus distance 1:1 on FF will look smaller on the image, and include some surroundings.

3) Sensor areas are ~366mm2 (APS-C) and 855.62mm2 (FF) but pixel density is lower on the FF. Will both images in question 1, 2 have the same IQ? Smaller pixel can take in less light, but density is higher....

4) When FF macro image @ min focus distance 1:1 magnification with an object which is ~23x15mm will be manually cropped so that the 23x15mm object fills the image - will the IQ be the same as that which is taken by the APS-C?

5) In macro photography with objects larger than 23x15 @ 1:1 I would have to step back with APS-C to fill the frame? What is the impact on IQ?

6) In practice - taking pictures with the same SEL90M28G on FF (or take some another lens) will I get “worse” macros for smaller objects than 23x15mm (even after cropping images manually)? Are there other aspects which would make FF better/worse in this case? For example, sensor areas, pixel density, DOF?

DxO Mark tests with A7R say that sharpness of the SEL90M28G is a lot higher than on APS-C, but does it depend only on MP (A7R is something like 36MP)?
https://www.dxomark.com/Lenses/Sony/Sony-FE-90mm-F28-Macro-G-OSS-mounted-on-Sony-A7R__917

Susurs · Oct 17, 2017

P.S. In practice I am comparing A5000 level APS-C sensor (20.1MP) vs A7II FF (24.3MP) with HAF, IBIS etc. Just not to make things/numbers more complicated I decided to simplify my questions and assume there is a comparison between APS-C 24.3MP with FF 24.3MP.

admwright · Oct 17, 2017

When you are working at 1:1 the size of the object is the size on the sensor. So if the object fills a 'full frame' sensor then a smaller sensor will only show part of the object - it is cropped with respect to the 'full frame'.
If the pixels are the same on each sensor then this implies that the image will display larger from the smaller sensor but it will not be showing all of the original object.
If the object is smaller so that it only completely covers the smaller sensor then the 'full frame' will be 'wasting' some of its pixels as these are showing the surroundings not the object.
As for IQ at 1:1 the resolution (pixels on object) will be higher from the smaller sensor where the object is smaller than the 'full frame'. Whether this results in a better image will depend on the specific characteristics of the two sensors.

dwig · Oct 17, 2017

Susurs said:
Hi!
...

1) @ min focus distance 1:1 magnification, an object which is ~23x15mm will fill the output image (6000x4000 or whatever, same for both cameras).

Only on the APS-c camera. It would not fill the frame on a FF camera.

2) If yes, same object @ same min focus distance 1:1 on FF will look smaller on the image, and include some surroundings.

There is no yes/no answer to #1. The 23x15 subject will look smaller on the FF camera since at 1:1 the FF camera's field of view would be 36x24mm.

3) Sensor areas are ~366mm2 (APS-C) and 855.62mm2 (FF) but pixel density is lower on the FF. Will both images in question 1, 2 have the same IQ? Smaller pixel can take in less light, but density is higher....

There are VASTLY more issues that influence IQ than pixel density. There is absolutely no way to ever, now or at any time in the future, to make any such simplistic conclusions based solely on the different formats.

4) When FF macro image @ min focus distance 1:1 magnification with an object which is ~23x15mm will be manually cropped so that the 23x15mm object fills the image - will the IQ be the same as that which is taken by the APS-C?

Again, there are VASTLY too many issues for any such comparison to be meaningful in the real world. If, and only if, two sensors were made using EXACTLY the same technology with the FF sensor being 24mp and the APS-c sensor being 12mp then the cropped FF image would match the uncropped APS-c image. Two such sensors, of course, have never existed and are totally unlikely to exist. Even if they did you would have to specify the exact camera models being compared and not merely the formats.

5) In macro photography with objects larger than 23x15 @ 1:1 I would have to step back with APS-C to fill the frame? What is the impact on IQ?

Of course not, duh!!. With a larger subject you can't shoot at 1:1. "Stepping back" would mean you were not shooting at 1:1.

6) In practice - taking pictures with the same SEL90M28G on FF (or take some another lens) will I get “worse” macros for smaller objects than 23x15mm (even after cropping images manually)? Are there other aspects which would make FF better/worse in this case? For example, sensor areas, pixel density, DOF?

There are too many issues for any specific answer other than to say that the less you crop the less the IQ loss. With any given lens setup, the smaller sensor will require less cropping.

Susurs · Oct 17, 2017

dwig said:
Only on the APS-c camera. It would not fill the frame on a FF camera.

There is no yes/no answer to #1. The 23x15 subject will look smaller on the FF camera since at 1:1 the FF camera's field of view would be 36x24mm.

There are VASTLY more issues that influence IQ than pixel density. There is absolutely no way to ever, now or at any time in the future, to make any such simplistic conclusions based solely on the different formats.

Again, there are VASTLY too many issues for any such comparison to be meaningful in the real world. If, and only if, two sensors were made using EXACTLY the same technology with the FF sensor being 24mp and the APS-c sensor being 12mp then the cropped FF image would match the uncropped APS-c image. Two such sensors, of course, have never existed and are totally unlikely to exist. Even if they did you would have to specify the exact camera models being compared and not merely the formats.

Of course not, duh!!. With a larger subject you can't shoot at 1:1. "Stepping back" would mean you were not shooting at 1:1.

There are too many issues for any specific answer other than to say that the less you crop the less the IQ loss. With any given lens setup, the smaller sensor will require less cropping.

Thank’s! For sure, I ment APS-C in the No.1. Corrected it now.

[doublepost=1508276168][/doublepost]I tried to take pictures (and cropped some to the max) of a 1,5cm flower (indoors and did not had a lot of time) - Thanks to my local shop for allowing me to test the equipment (My lens is in service for cleaning).

To tell the truth I do not see so much differences...Seems good in a case where FF images are manually cropped a lot more for my test purposes (maybe some differences, but it is also very hard to tell - I held the cameras in hands, indoors, angles were probably a bit different).

In any case IBIS and fast AF, HAF made it so much easier for taking macro images...

Crop, APS-C, f/5.0