CFN PlanApo 100x/1.4 vs CFN PlanFluor 40x/1.3

[JohnyM]

As 40x, 60x and 100x 1.3 / 1.4 lenses offer similar resolution, but two later are never sharp and produce dim fluorescent images, i've tried to compare the results between them (Nikon doesnt Offer PA40x so it's PF40x vs PA100x). I wanted to check if the theory supporting that 40x/1.3 is more than enough on mine 3,9 micrometer A6000 camera is correct. Also if Apo glass deliver "crispier, more contrasty images" as it's commonly belived.

According to nyquist criterium 40x/1.3 would need pixel as small as 4,2 while 100x1.4 only 10 micrometers.

First lets start with light gathering power or maybe we can call it T-Stop.
In fluorescence, when signal is always weak and light is more... monochromatic, PF is wiping the flor with PA:
1/5 sec exposure for 40x

0,5 sec exposure for 100x


This can also be very helpfull with DIC imaging of fast moving subjects as on below photos i've used 1/400 sec for 40x, but only 1/50 for 100x:



So lets take a look at those red circles i marked on above pictures. I've cropped 300x300 pixel box from 100x and ~120x120 pixel box from 40x then upscaled it to match 100x (should actually do it otherway, as 100x is ovesampled and i get some artifacts from upsampling). Left row for PA, right for PF. BF on top DIC on bottom:

I actually see more lateral CA on PA than PF, which is suprizing. But with DIC images we begin to see longitudal CA which PA is almost free of and PF get a slight amount. Resolution wise, lenses are indistinguishable unless sample is perfectly prepared.

Finally i wanted to look at loca, where PF finally gets behind PA:


PA produces much better results here, but PF is eazy to correct in post processing:


All images SOOC unless stated otherwize. In the last serier we can see that PF is clearly more contrasty, it's because DIC prism have higher shearing value, and have nothing to do with lens correction. Is PA 60x & 100x worth it? I decided to sell 60x and keep 40x for photography, and 100xPA for 2k movie recording. 60x would be better for 4k.

[Pau]

Interesting test.

Finally i wanted to look at loca, where PF finally gets behind PA:
...
PA produces much better results here, but PF is eazy to correct in post processing:

Is the last image of the set the first one with the LoCA corrected?
If so you did a good work! Could you explain your method?

[JohnyM]

Yes. The last one is corrected. I use CameraRaw plug-in for Photoshop CC. There is automatic correction checkbox for LaCA and two sliders for LoCA. You can do it once, and then save profile for your lenses (including vignette, distortion, sharpening etc).

That 40x is quite a beast, dwarfing other CFN lenses with eaze:

[Ichthyophthirius]

Is PA 60x & 100x worth it?

Hi Johnny,

Thanks for the test. The 60/1.40 has a working distance of 0.17 mm; 100/1.40 only 0.10 mm. This can lead to problems (makes the 60/1.40 more susceptible to RI mismatches from Temperature diviations etc.) but for most users, a larger working distance would be an advantage. Just one more thing to keep in mind!

Regards, Ichty

[Pau]

Johny, thanks for the tip. Unfortunately Camera Raw 7 (CS6) doesn't allow for LoCA correction

I once tested the Nikon Plan Apo 60/1.4 for a friend and it is outstanding, the same quality of my Zeiss Planapo 63/1.4 but with much more comfortable working distance. It wouldn't be an objective I would sell!. I much prefer 60X over 100X, you can crop (like you do with the 40/1.3) as both are outresolved by modern cameras. And at a second examination of your images, despite lower contrast, the PA seems to resolve finer details than the 40 PF

[JohnyM]

Im sure that i was using this with CS6... camera raw 8.X (the one that supports A6000 from sony), check if there is an update for your CR plugin.

Well, in theory, there is about as much difference between 40x and 60x as between 60x and 100x. And that's what i feel when comparing 40x with olympus Splan Apo 60x (not a fair comparision for photography, as olympus is eyepiece corrected).

Bear in mind that presented images need to be resized and compressed to 300kb, so dont judge resolving power on images that werent supposed to serve this purpose (diatom slide). Apparent resolving power is bigger, because of much higher spatial resolution or 100x, althrough resolving power is very similar for those lenses. I've scanned that diatom slide in search of something that 100x could resolve, but 40x couldnt and failed to do so.
While i've found some frustules that were seen "more clearly" in PA, but after photographing and doing post-process (best i can afford for both images separately) i failed to keep this difference. Also this difference would vanish immidiately after something was moved out of perfect alignment in microscope.

Big plus (only real advantage over PF except being LoCA free )of 100x lens that i didnt mentioned (but i should, as it might not be immidiately obvious) is that when images are cropped to display same FOV, 100x will present much less noise (as can be observed on diatom photos).

I've decided to keep 100x for 2 reasons. For 2k video recording without any post-processing and for my future confocal plans.

[JohnyM]

So i keep exploring those lenses.
FOV 40x VS 100x:


When i said i've failed to find any differences in resolution, with those lenses, i was right then but this need clarification. I've realised that they've been tested on aquaeous mounted specimen, so NA1.4 lens was really limited to 1.33

This time i've mounted diatoms with immersion oil and looked for something that 100x was already struggling with, that's about the biggest difference i've found (keep in mind that it was cropped to ridiculously small size of ~200x200 pixels, then upscaled 2x, images were processed this time):


I think im just gonna keep 40x/1.3 and get magnification changer for sake of eyeballing through micro.
Anyone interested in buying PA100x with dic slider?