Ever wondered what will a Zeiss 100MP do with a Canon 2xIII?

[fegari]

Out of boredom this afternoon I decided to clean a bit my lenses. At some point all were lying on a table and theZeiss 100 Makro-Planar was laying besides the Canon 2XIII extender so....thought of matching them. And match they do! Well, kinda because the extender will not let focusing beyong 10m but's that more than enough for macro or portraits (and up to 5m if you put the 1.4x II though I did not bother to take pics with it)

Below a copuple of comparisons of the same pic taken with the Zeiss alone and with the extender. Both pics are taken at minimun focus distance, this means the Zeiss "only" subjct is magnified 1:2 (as per the lens design, obviously) and the extender subject fills 1:1. This is important, the normal zeiss pic is shown zoomed 2:1 to match the extender pic

Pics taken on tripod, aperture priority, liveview mode, cable release. As I had a crappy light in my house I decided to shoot ISO 800 with a Canon 5D III. This proved to be an interesting choice when looking at the effect of noise. I'm no lens test expert so there may be flaws in this test, don't hesitate to pint them out.


(you may want to click to go to the flicker pic at 100% otherwise the differences are lost at this sizes)

The Original Field of view for each pic
The "A" in PANAM is about 2mm wide. The coin is a 20 cents of €


Original shots II par Fegarix, sur Flickr

To the left the Zeiss only at 200%, to the right the Extender pic at 100%


Zeiss 100MP and 2x comparison f4 (cropped) II par Fegarix, sur Flickr

The Original Field of view for each pic
(The "I" in Colombia is about 1mm wide)



Original shots par Fegarix, sur Flickr

To the left the Zeiss only at 200%, to the right the Extender pic at 100%


Zeiss 100MP and 2x comparison f4 (cropped) par Fegarix, sur Flickr

My personal conclusion
a) IF your subject does not FILL your frame with teh Zeiss alone, by all means use the Canon Extender!!!! Even for the Zeiss 100MP which is THAT freaking good at any aperture and holds well zooming in however blowing up the Zeiss "only" will cause artifacts and dramatically increase the apparent noise so to me the extender pic gives you more margin to work under those conditions.

b) keeping the Zeiss alone pic at 1:1 and reducing the extender pic to 1:2: the Zeiss alone clearly wins, but your subject does not fill the frame....

c) I'm impressed, did not expect those results.

d) Really happy I got myself a Zeiss 200mm Makro-Planar f/4

A couple of Notes
- When using the extender the camera reads the natural Zeiss aperture range, starting at f/2. In reality you are shooting two stops above it so a extender pic at f/2 = f/4. This is easily seen in the shutter speeds. This also means when shooting at f/22 you get ridiculous small aperture of f/44. Did not bother to try it, probably crappy IQ

-The selections you saw were matched to trying and get the (roughly) same DoF (remember the to calculate double the aperture number in the extender pic to have an idea of the real one)

-have a bunch of other apertures tests in case you're courious!

 

[Harry Muff]

If only I could fit my 2x to my 100L... :-[

 

[WillThompson]

Use an ext tube.

 

[noisejammer]

Sorry - I really don't see any benefit here.

It's pretty clear from your examples that the 200x images from the Zeiss lens are dramatically better than the 100x images from the Zeiss + 2x converter. In all honesty, I am not surprised.

In addition to wrecking the image sharpness, the tc also costs you two stops of light. It drives the lens closer to diffraction (if not completely off the cliff) and adds copious dollops of CA to a lens that's almost entirely free of it.

In what way can this possibly be better?

 

[fegari]

Sorry - I really don't see any benefit here.

It's pretty clear from your examples that the 200x images from the Zeiss lens are dramatically better than the 100x images from the Zeiss + 2x converter. In all honesty, I am not surprised.

In addition to wrecking the image sharpness, the tc also costs you two stops of light. It drives the lens closer to diffraction (if not completely off the cliff) and adds copious dollops of CA to a lens that's almost entirely free of it.

In what way can this possibly be better?

To you, which one is the one with the TC and did you see the original files in flickr? cause I really see the opposite in my screen. The loss of 2 stops is not a terrible penalty for a lens that starts at f/2 and is sSO good as the Zeiss, and diffraction cliffs cannot be seen in the posted TC examples. Remember the purpose here is get into 1:1 teritory not to compete below 1:2 magnifications. You loose a bit of contrast but to me the Zeiss+2xIII at 100% resolves better than the normal zeiss blown to 200% with a lot less apparent noise. That TC is a handy feature to me.

And for what is worth, with the TC you get twice the working distance for the same 1:2 magnification

A tighter crop of the same pics, different area. Left normal at 200%, right with the 2xIII at 100%.
In all, I think adding the TC is a very nice compromise.


Test par Fegarix, sur Flickr


And below, even more convincing when zomming the pic without TC to 400% and the pic with the 2xIII to 200%. Again, same pic as before. Here I see a significant difference in favor of the TC. Click to see the original full screen in flickr:


test 400x par Fegarix, sur Flickr

 

[noisejammer]

Hi Fegarix

I've looked at your images again and came to the same conclusion. Please take a look at your image of the bus. I have no doubt that the left hand image (labeled as a bare ZE 2/100 MP) is much sharper.

I do see some noise present on the LHS of your most recent comparison. (Look at the edge of the monitor bezel.) This looks very like the signature of a slightly over-sharpened image. I can't tell which has more detail.

That speckle will certainly be reduced when the extender is applied - but this does not automatically mean that there is more information present in the resulting image. For example - if an image was 2x larger but the spot size was 4x larger, it would appear far smoother but actually contain far less information.

I understand that the 2x would give you twice the effective distance at the same image scale. The problem with this is that image scale is largely irrelevant. Once again, the important thing is to maximise the amount of information present in the image after capturing and processing it to the same scale. Since the extender clearly blurs the image (quite a lot by my observation), it really can't be helping you.

(For what it's worth, do I own a ZE 2/100 MP and a 1.4 III tc. In the interests of science I'll do an experiment or two tomorrow. I'm not expecting magic though... some matches are really not made in heaven.)

Regards, NJ

 

[Mt Spokane Photography]

I find it almost impossible to manually focus a lens without tethering it, so doing a really good and valid comparison would be difficult, since the plane of focus can be just a hair different and give the impression that something was wrong. A flat target might overcome the issue, merely because its so difficult to use manual focus on a 3 dimensional target and focus at the exact same sopt.

 

[fegari]

Hi Fegarix

I've looked at your images again and came to the same conclusion. Please take a look at your image of the bus. I have no doubt that the left hand image (labeled as a bare ZE 2/100 MP) is much sharper.

I do see some noise present on the LHS of your most recent comparison. (Look at the edge of the monitor bezel.) This looks very like the signature of a slightly over-sharpened image. I can't tell which has more detail.

That speckle will certainly be reduced when the extender is applied - but this does not automatically mean that there is more information present in the resulting image. For example - if an image was 2x larger but the spot size was 4x larger, it would appear far smoother but actually contain far less information.

I understand that the 2x would give you twice the effective distance at the same image scale. The problem with this is that image scale is largely irrelevant. Once again, the important thing is to maximise the amount of information present in the image after capturing and processing it to the same scale. Since the extender clearly blurs the image (quite a lot by my observation), it really can't be helping you.

(For what it's worth, do I own a ZE 2/100 MP and a 1.4 III tc. In the interests of science I'll do an experiment or two tomorrow. I'm not expecting magic though... some matches are really not made in heaven.)

Regards, NJ

Well I still don't see that, we'll need a third opinion

All images have exactly the same sharpening in lightroom (55, 0.8, 35, 5 masking). Zero noise reduction.

Something to keep in mind though: the bus image was taken from slightly different angle as I had to reframe to fit the 2x pic. This means the DoF area does not fall exactly in the same place. In that same bus pic, take a look at the bottom left corner, I see clearly that the in focus zone is sharper than the corresponding in focus zone of the bare Zeiss. Additionally I think the TC actually reduced the magenta cast of the bare Zeiss and the blacks are blacker. Slight, but for some reason clear to me.

The 2x certianly reduces contrast but compensates in resolution vs pulling the non TC pic to the same magnification. I still believe in overall the TC wins if you want to reach 1:1 magnification, that's all.

I have the 1.4x II as well but being only 40% more my feeling was the bare zeiss will be better when upsampling to the same level so went directly to the 2x IIIl.

I'll hope you can post your findings with the 1.4x, for the time being I think I need a few hours of sleep :=)

 

[fegari]

I find it almost impossible to manually focus a lens without tethering it, so doing a really good and valid comparison would be difficult, since the plane of focus can be just a hair different and give the impression that something was wrong. A flat target might overcome the issue, merely because its so difficult to use manual focus on a 3 dimensional target and focus at the exact same sopt.

That's true and that's what happened with the bus pic when had to reframe and change a bit the angle in order to fit the 2x pic. Comparing that pic is tricky casue the in focus zones do not fall exactly in the same spots and I've realized now they do not have teh same ISO.

However I was more careful with the second set (the little airport trucks) so the same ange, ISO and focus point was maintained very very closelly.

 

[fegari]

So I though to put my money were my mouth is

I field shot of the Zeiss 100MP + 2xIII. I would have never gotten it with the bear zeiss, the bugs would fly away as soon as you approach just that much. Here the added working distance saved the day. Very acceptable details and sharpness. Handheld.

 

[hamada]

Sorry - I really don't see any benefit here.

It's pretty clear from your examples that the 200x images from the Zeiss lens are dramatically better than the 100x images from the Zeiss + 2x converter. In all honesty, I am not surprised.

In addition to wrecking the image sharpness, the tc also costs you two stops of light. It drives the lens closer to diffraction (if not completely off the cliff) and adds copious dollops of CA to a lens that's almost entirely free of it.

In what way can this possibly be better?

grain is often perceivced as "sharper".

but it´s just grain.... not usefull information.

 

[LetTheRightLensIn]

I agree with Fegari that the 2x TC III definitely gives it more true macro reach. I have no clue what those who claim the bare lens zoomed 2x shows more detail are seeing since the macro+2x TC clearly is bringing in more true and real fine detail.