Minolta DiMAGE Scan Elite 5400 Lens Test at 1.2x

[RobertOToole]

Robert,
how would a "normal" macro lens stopped down to 5.6-8 look in comparison to the Dimage 5400 lens at 1:1 magnification?
regards
Lothar

Hi Lothar,

In my limited shooting with the Minolta lens, due to work, I would guess that it would do very well, probably better.

In the past I have tested all kind of lenses, enlarger lenses, bellows lenses and you know what, the Sigma 50 or 70mm beats them all. But with the scanner lenses I think the APO correction and the faster aperture would beat a normal macro lens.

I plan to do lots more testing with the lens later in May when I am back from a trip to Africa. I will post the results here.

Thanks.

All the best,

Robert

[Lou Jost]

Adding a bit to what Robert said, good scanner lenses like these are working at a nominal aperture of about 2.8 (maybe even 2.5 for the Dimage). A regular macro lens at a nominal aperture of 5.6 or 8 will lose much resolution due to diffraction relative to the scanner lens. If the lenses had no other aberrations at their optimal apertures, the macro lens at 5.6 would have only half the resolution (in the aerial image) of an f/2.8 scanner lens. However, if your sensor has big pixels you may not be resolving the full resolution in the aerial image, so in practice the difference might not be noticeable.

[Smokedaddy]

However, if your sensor has big pixels you may not be resolving the full resolution in the aerial image, so in practice the difference might not be noticeable.

Would you mind giving me an example of a camera with a big pixel and one with a small pixel? Also, maybe what is considered to be a small pixel for a astronomy video camera.

Thanks,
-JW:

[Lou Jost]

An example of a pixel size that would resolve the differences between these lenses is that of my 20Mp MFT sensor, which starts to show the effects of diffraction at an effective aperture of 5.6-8. At 1:1 the scanner lenses have an effective aperture of around 5.6 or a little less, while a macro lens at nominal f/5.6 will have effective aperture of 11.

If you look at published lens tests on newer MFT bodies, the resolution curve generally starts decreasing around f/5.6. Pixels are about 3 microns wide on that sensor. In contrast many older FF cameras have pixels twice as wide.

[lothman]

If you look at published lens tests on newer MFT bodies, the resolution curve generally starts decreasing around f/5.6. Pixels are about 3 microns wide on that sensor. In contrast many older FF cameras have pixels twice as wide.

Hmh but don't we speak of 1:1 on sensor, so why would a macro lens "react" other on a small MFT or Fullframe semsor. We have a smaller field of view on the smaller sensor but everything else is related to the Pixel pitch. Or am I wrong?

On full frame even on a 42MP A7rii my 70mm Sigma macro lenses are better at 5.8 or 8 than on 2.8 so I don't see diffraction.

[Lou Jost]

I am not sure I understand your point. In this context we were talking about 1:1 (or close to it). With magnification (rather than field of view) the same between lenses.

Your A7RII is not an older FF camera with big pixels. With a good lens on an A7RII you are already seeing diffraction lower your resolution beginning at f/2.8. Example:

https://www.lenstip.com/517.4-Lens_revi ... ution.html

So on this camera, if your lens isn't at its best until f5.6, your image is suffering from diffraction loss. If aberrations are well-corrected in the scanner lenses (and they are, based on Robert's tests, and my tests of the Scanner Nikkor) then, since they are working at f/2.8 or less, they could be sharper on your camera.

[Smokedaddy]

I thought this was informative.

https://www.cambridgeincolour.com/tutor ... graphy.htm



-JW:

[ray_parkhurst]

I completed my first test of the Minolta 5400 lens. As per my usual method I used a Lincoln Cent to do the comparison. My reference was the 85mm Macro Varon at f4.5. I used my HRT2i. Lighting was identical, but due to the differences in focal length, there are noticeable differences.

First problem I ran into was one I expected...the short FL makes it difficult to mount and achieve low magnification. Robert uses a Sony camera with short register, but I use a Canon, and in order to get mags where I wanted them (~0.75) to take pics of a Lincoln Cent, the lens needed to be mounted essentially directly to the camera. This would normally be impossible while still giving some level of magnification (extension) adjustment, but the excellent adapter from Raf allows infinite positioning flexibility since it is just an 18mm diameter tube with outside threaded for RMS. I was able to dial-in just the right magnification for this compo.

Of course I have the lens reversed, with short side / paint mark toward the camera instead of the normal orientation toward the lens, as it is used for m=1.2 in the 5400 scanner.

Here are the full-coin pics, downsized 4x and cropped square:

Minolta 5400 Scanner Lens:


Schneider 85mm Macro-Varon:


Center Crop Animated Comparison:


edited to add:

Here is the comparison animation of the WE in IGWT, basically the short edge APS-C.

[rjlittlefield]

In the animations, how can I tell which is which?

--Rik

[ray_parkhurst]

In the animations, how can I tell which is which?

--Rik

For the center/chin, the more reddish image is the 85MV, while the one with the extra small reflection on end of chin is the 5400M.

For the edge/WE, again the more reddish (and slightly darker) image is the 85MV, while the one that appears to expand is the 5400M.

[Smokedaddy]

I haven't got my flashes yet but they're on the way so I was using a couple of diffused Jansjo lights. This is a 100% crop of a section of the overall image. The lower bottom left to be exact. It's a small microrontroller die, about 2.93mm x 2.66mm, that I have. It was shot at about 2.5x on my MM-11 with a Canon 50D. That's the least mag I can do on the MM-11. No relay lens, with the shortest Kenko adapter and the dovetail adapter.







-JW:

[RobertOToole]

Whoa!

Great job with the die in the hand for scale!

Robert

[zzffnn]

It's a small microrontroller die, about 2.93mm x 2.66mm, that I have. It was shot at about 2.5x on my MM-11 with a Canon 50D.

Isn't Canon APSC sensor 22.2mm in length? That is more likely 7.6x (22.2/2.93), rather than 2.5x? I apologize, if I missed something.

[Smokedaddy]

It's a small microrontroller die, about 2.93mm x 2.66mm, that I have. It was shot at about 2.5x on my MM-11 with a Canon 50D.

Isn't Canon APSC sensor 22.2mm in length? That is more likely 7.6x (22.2/2.93), rather than 2.5x? I apologize, if I missed something.

I trust absolutely 'nothing I do" in microscopy, so yes I could be wrong. When I looked it up on the Internet it said the Canon 50D sensor was APS-C (22.3 x 14.9 mm). I thought the procedure was 22.3/10 or about 2.2x (via the mm scale I photographed). Looks like I guessed wrong on the 2.5x (if this is all correct to begin with).



-JW:

[rjlittlefield]

It's a small microrontroller die, about 2.93mm x 2.66mm, that I have. It was shot at about 2.5x on my MM-11 with a Canon 50D.

Isn't Canon APSC sensor 22.2mm in length? That is more likely 7.6x (22.2/2.93), rather than 2.5x? I apologize, if I missed something.

I think the disconnect is that you're assuming the image shown was the full frame.

But JW wrote that:

This is a 100% crop of a section of the overall image. The lower bottom left to be exact. It's a small microrontroller die, about 2.93mm x 2.66mm, that I have.

The way I read this, we're seeing the lower left 1024 x 772 pixels of an original larger image, probably native camera resolution 4752 x 3168. In the whole frame, there must have been quite a bit of spare space to the right and above the chip.

I thought the procedure was 22.3/10 or about 2.2x (via the mm scale I photographed).

Right, magnification = sensorWidth / subjectWidth. By counting pixels to the centers of ruled lines, I measure your subject width a little more precisely as 9.92 mm, so 22.3/9.92 = 2.25X.

--Rik