An informal test: J3 and Canon 50/1.8

[mjflory]

I've been curious about the Jupiter 3 for a long time, and when Brad Bireley posted an ad here for a nice J3 checked over by Brian Sweeney I figured it was time to have a look. The lens arrived over the weekend and my first test roll is back. Here are some comparison shots taken with the J3 and with a Canon 50/1.8, on the same roll of Kodak 400UC in a Leica IIIf. Since I've enjoyed the lens tests posted on RFF I thought I'd take a chance that these might interest someone. I'll apologize in advance for the poor quality of the scans and the washed-out quality of the shots. (I intentionally overexposed by a stop or two per a good meter and the negs were still a bit thin; could the drugstore's chemicals be a little old?) I wanted chiefly to check the focusing and the sharpness, with a passing interest in flare-resistance, so I didn't spend much time adjusting for scan quality. First, the J3: from left, wide open (f1.5), almost wide open (f2.0, for a better comparison with the Canon), and at a more optimal aperture (f8.0). All shots were taken with the lens set to focus at infinity.

 

[mjflory]

Details of J3 test shots

Details of J3 test shots

Here are details taken from the center of each of the shots above.

 

[payasam]

This is of special interest to me as I'm expecting a J-3 to arrive: but I'm afraid these pictures do not say a great deal.

 

[mjflory]

Canon 50/1.8, for comparison

Canon 50/1.8, for comparison

Here are comparable shots taken with the Canon 50/1.8: from left, at f1.8 and at f8.0.

 

[mjflory]

Details of Canon test shots

Details of Canon test shots

Here are details from the Canon shots, same order.

 

[mjflory]

Payasam, I'm afraid you may be right. I realize it's not a very rigorous testing, though I've tried to stick with the same picture in testing quite a few lenses and picked the convenient view down the sidewalk out front for its mix of distant and close features. (The shift in perspective after the very first shot is because I didn't want to get in the way of our landlady, who was kindly sweeping away the piles of bark that the sycamore on the right of that picture sheds at this time of year!)

Do you think I should go back to the resolution charts on the wall?

 

[mjflory]

A couple of flare tests

A couple of flare tests

Finally, here are a couple of flare torture-tests. These pictures are taken straight into the beam of a very bright LED-flashlight. (My usual use for it is to search for haze and dust in lenses!) Left, J3; right, Canon.

I'm not sure how much these poor tests say about the lenses, but I was impressed by the relatively good showing of the J3 in comparison with the Canon. The Canon f1.8 and f1.4 didn't show themselves to be the sharpest lenses in Raid's tests, so perhaps it's not the most stringent comparison, but I don't think I'll hesitate to use the J3 to allow myself another half-stop or so of speed. (Curiously, the J3 stops down to f22 vs. f16 for the slower Canon lens.) I can't pretend that the flare tests will apply to any samples of the lenses other than the ones I have, but those lenses both look clean to me save for some cleaning marks on the Jupiter. The tests reassured me that those marks didn't have a disastrous effect on flare-resistance.

 

[payasam]

There's a place for resolution charts, Michael, but this is not that place. If lens performance is to be judged, processing and scanning should not introduce any errors or variation.

 

[mjflory]

Payasam, thanks for the stern but sage advice. I've gone back and fought with the scanner parameters for a while and have changed the parameters rather radically. I realized that I was using parameters that gave a good overall range to the image, but what I was really interested in -- and what I had exposed for -- was the small area of detail far down the block. While the areas off to the side interested me for what they could tell me about out-of-focus character and overall contrast, they were incidental to the main point of the test. While the negatives still look thin to me, I don't think it's worth it to reshoot them; I've gotten much more contrast in the detail areas now and they do tell me, at least, what I wanted to know about the sharpness I could expect from the lenses.

Here, for what it's worth, are the scanner parameters I used for the rescans. The principal differences from the original parameters are that I raised the "white level" from its default (which was zero, I think) and lowered the brightness from .8 to .5. The result looked very dark indeed at first glance, but the central area seems much more informative. Color balance is taken by setting a brightly illuminated area of the sidewalk to neutral. The film type became irrelevant once the color balance was set manually. ("Kodak Professional 400UC" is not listed among the many film types listed in the version I'm using.)

 

[mjflory]

Rescanned tests

Rescanned tests

Here, then, are rescanned images of the J3 tests at f1.5, f2.0, and f8.0.

 

[mjflory]

Here are the details of the rescans for the J3: f1.5, f2.0, f8.0.

 

[mjflory]

These are the Canon shots: f1.8, f8.0.

 

[mjflory]

Finally, here are the Canon details.

As I look over the rescanned images, I see how the light was shifting more than I realized during shooting, and how the exposure was too low for the J3 at f1.8... Misery!

 

[payasam]

Michael, you could one day do another series of five pictures somewhere else and then scan all the negatives with exactly the same settings. Possibly you could do a flare test nearer to real life. [edit] You might also like to use a tripod.

 

[mjflory]

Thanks, Payasam, they are all good points. I did try some more realistic flare tests, but on looking at the results I saw the light was shifting too quickly for useful comparisons. This is probably the peril of trying to squeeze this sort of test into a brief window of available time; better just to wait for a cloudless day. Of course the attempt to do this all handheld was just pure laziness.

Nonetheless, I think I did get some sense of the behavior of the J3, and I think it will be useful. (I suspect you'll like yours when it arrives, too. Enjoy it!)

 

[raid]

Michael,

The Canon 50/1.8 is a sharp lens and so is the J-3 when adjusted. Take a look at the following images taken on a tripod and scanned the same way on the same film and taken with the same camera:


Canon 50mm/1.8 @ 4.0:
http://photo.net/photodb/photo?photo_id=5288230




Canon 50mm/1.2 @ 4.0:
http://photo.net/photodb/photo?photo_id=5288268


Canon 50mm/1.5 @ 4.0:
http://photo.net/photodb/photo?photo_id=5288258



Also, the Nikon 50/2 is a good lens. Here. @ 4.0:
http://photo.net/photodb/photo?photo_id=5288252

FSU lenses:
J-8 50mm/2.0 @ 4.0:
http://photo.net/photodb/photo?photo_id=5288213

Industrar 50/3.5 @ 4.0:
http://photo.net/photodb/photo?photo_id=5288163

Jupiter 3 50mm/1.5 @ 4.0:
http://photo.net/photodb/photo?photo_id=5288274

 

[mjflory]

Greetings Raid! I had your lens tests in mind all the time as I was stumbling through my attempts to achieve some consistency.

My impression that the Canons were not the very sharpest lenses was from the series of comparison shots you did of the glass cabinet, the candle, etc., but the other contenders included many extraordinary lenses, including Summicrons, so I suppose I should have mentioned the context.

I'm going to continue to play at testing lenses as I find time. I do have trouble putting all the observations I've seen, and those few I've made, in context, establishing a grounds for comparison. It would be good to be able to quantify further... I suppose I'm with Kelvin on that one: "When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind."

 

[payasam]

Good luck, Michael. Yes, I have high hopes for the J-3.

 

[alternatve]

Just some thoughts of mine.

Comparing both the detailed shots of the J3 and the Canon 50/1.8, I would say that the Canon is a tiny bit sharper then the J3 and produces a cooler image. The J3 produces a tiny bit softer image then the J3 and a warmer image. Anyone else feels the same?

I'm thinking of getting a J3 soon, so I'm eagerly gathering any information about it now. 🙂

Regards,
Samuel

 

[raid]

Hi Michael,

When I test 10-26 lenses at the same time, it is very possible that I make errors. On the other side, I usually have replicate photos that can affirm the optical qualities of a lens under certain conditions. The other factor is that vintage lenses can differ a lot from lens to lens due to age.

Overall, the Canon 50/1.8 is more refined than the J-3, in my humble opinion. Still, the J-3 can be an excellent performer when you get it adjusted.

I don't agree with the statement that you have to get something numerical before you can compare it. I am a Statistics Professor, so I work each day with numbers, but there are cases where you need more than numbers to see "the big picture", as they say [whoever "they" are!].


Raid

Greetings Raid! I had your lens tests in mind all the time as I was stumbling through my attempts to achieve some consistency.

My impression that the Canons were not the very sharpest lenses was from the series of comparison shots you did of the glass cabinet, the candle, etc., but the other contenders included many extraordinary lenses, including Summicrons, so I suppose I should have mentioned the context.

I'm going to continue to play at testing lenses as I find time. I do have trouble putting all the observations I've seen, and those few I've made, in context, establishing a grounds for comparison. It would be good to be able to quantify further... I suppose I'm with Kelvin on that one: "When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind."