Which shot which? Spot the indefinables! — Two at 42.5mm

[JamesD172]

I really like @alex g 's idea of posting side-by-side images to see if we can actually see any noticeable difference between lenses with purported rendering differences. I can contribute a comparison between the Lumix 42.5 & Nocticron.

Since it's hard to take near-identical portraits with two lenses, I have two shots of some tulips sitting our living room.

Subscribe to see EXIF info for this image (if available)

A

Subscribe to see EXIF info for this image (if available)

B

The above images were shot with:

1) Panasonic Lumix 42.5mm f/1.7
2) Panasonic Leica 42.5mm f/1.2

I adjusted the WB between the two images to match one another. (one required +200K and -4 tint). Other settings for both photos were as follows:

ISO 800, 1/80, f/1.8

Which lens rendered which image?

 

[TNcasual]

Call it a noob question: Why should the WB be different? Same camera, same focal length, same ISO same SS same Fstop. That doesn't compute for me, unless the lighting changed between shots.

 

[gnarlydog australia]

I don't know those two lenses (apart from reading and seeing samples) but I have no preference over one or the other image that you presented.
The only possible difference to me could be if the images were both taken at their respective maximum aperture, but that would be an easy give away, right?
I think we need to create a whole library of these samples so prospective buyers of theses lenses can see for themselves the difference (if any) side by side.

 

[JamesD172]

My understanding is that it has something to do with lens coatings. The WB has always been different between these two lenses in every side-by-side shot I've taken.

 

[gnarlydog australia]

My understanding is that it has something to do with lens coatings. The WB has always been different between these two lenses in every side-by-side shot I've taken.

OK. I often hear that some people love the colors of certain lenses and I don't deny that I can see differences between color temperature in my lenses.
I think the tones that a lens would give was critical in film days but, as you have proven, a few seconds in digital post production and any lens can be adjusted or brought to a totally new level, and if I want, not necessarily reflecting the original scene as seen by my eyes.
If one loves to take OOC Jpegs, then the story might be a bit different

 

[TNcasual]

Coatings didn't even occur to me. That makes sense.

 

[retiredfromlife]

Voted.
Hope these challenges prove popular, thanks for hosting this one.

 

[alex g]

Hah! Excellent! I'm in the process of posting another set of images and panicked a moment ago when I saw a second one had appeared on the forum's thread list... I thought I'd accidentally submitted an unfinished post! Good on ya!

These two images are teasers too — the differences are far from obvious, to my eye at least! I think I'll have to try to base my decision on which one looks most expensive...
Edit: I appear to have voted for the least popular choice, so perhaps I don't have expensive tastes?

Re: variation in colour — I've noticed sometimes quite dramatic differences, especially between different brands and/or ages of lens. That would tie in with the coatings theory. A lens has a transmissivity curve which indicates how much light of a given frequency is passed through, and although presumably a straight line is the designer's goal, those I've seen published have been far from straight.

My thought was to leave the WB the same across all the lenses in a set, so that the colour "signature" remains visible, but matching the WB as you have is interesting — it makes it even harder to match the lenses!

 

[AussiePhil]

My understanding is that it has something to do with lens coatings. The WB has always been different between these two lenses in every side-by-side shot I've taken.

The coloring is part of the rendering difference and in my opinion should not have been adjusted out, whilst with a couple simple tweaks it becomes very similar from the raw file on a SOOC JPG it becomes part of the final rendered image

 

[JamesD172]

The thing is, though, that the built-in auto WB will generate different WB settings for the two lenses. So there are three ways to approach WB when comparing lenses with different light transmission profiles:

1) Leave the WB settings at whatever was auto-detected by auto WB in-body.
2) Manually select identical WB settings for photos from both lenses in RAW
3) Adjust WB in post to achieve closest visual similarity between photos

For this comparison, I went with option #3. I'm also happy to post results from #1 and/or #2, as well. At the very least, it will be informative. Manually selecting identical WB settings in RAW results in very different colors.

 

[JamesD172]

For the curious, here are the same two images, but with identical WB settings, in order to demonstrate the different color casts:

Subscribe to see EXIF info for this image (if available)

A

Subscribe to see EXIF info for this image (if available)

B

 

[alex g]

The thing is, though, that the built-in auto WB will generate different WB settings for the two lenses. So there are three ways to approach WB when comparing lenses with different light transmission profiles:

1) Leave the WB settings at whatever was auto-detected by auto WB in-body.
2) Manually select identical WB settings for photos from both lenses in RAW
3) Adjust WB in post to achieve closest visual similarity between photos

For this comparison, I went with option #3. I'm also happy to post results from #1 and/or #2, as well, perhaps when the contest is over. At the very least, it will be informative. Manually selecting identical WB settings in RAW results in very different images.

I actually discovered something in the process of shooting another set of images — I shot a white card and saved a custom WB in the E-M1 mk II and then shot the subject with that lens. I then changed the lens and shot the subject again using the same custom WB I'd saved previously. When opened in Lightroom, the RAW files had different WB values, which came as a surprise to me. I chose to adjust them both to the same value, but as you say, there are a few different options

 

[John M Flores]

There's practically no difference between the two that can't be erased with a modicum of post processing.

 

[0000]

I think we need to create a whole library of these samples so prospective buyers of theses lenses can see for themselves the difference (if any) side by side.

Yeah, I don't want to miss any of these, let's keep 'em all together! If not a separate subcategory, then there should at least be a "which shot which" tag.

The thing is, though, that the built-in auto WB will generate different WB settings for the two lenses. So there are three ways to approach WB when comparing lenses with different light transmission profiles:

It's actually even more complicated than that - m43 being the "smart" system that it is, my Olympus bodies use a different set RGB multipliers for "standard" (daylight, shade, etc) WB settings depending on the lens. These aren't strictly accurate, in my experience, but are in the right direction... for instance, the Oly 60/2.8 macro has a greenish cast compared to most other glass I own, and the built-in WB settings reflect that, but not quite to the extent of reality. All of which to say, I think normalizing the white balance of the images is the Right Thing for this purpose... if we were all still shooting film, that would be a different story.

One little nitpick about these two lovely images: the only real difference that jumps out at me, personally, is the slightly different focus points... I only mention it because when the OOF rendering is what's being evaluated, this can be significant, in my opinion, since it can affect the rendering of large parts of the scene. Other than that, I have almost no basis for distinguishing them - maybe there should be a "not different enough for me to guess" option in the poll? That might provide more meaningful information than people who don't feel strongly either way just picking one or the other at random...

 

[JamesD172]

One little nitpick about these two lovely images: the only real difference that jumps out at me, personally, is the slightly different focus points... I only mention it because when the OOF rendering is what's being evaluated, this can be significant, in my opinion, since it can affect the rendering of large parts of the scene.

They're actually focused on the same petal. I took multiple shots with both lenses, using magnification to aid in focusing, and chose to compare shots with the closest framing and focus. The differences in apparent focus may represent differences in sharpness away from the center, perhaps? Both were focused at this point:

Subscribe to see EXIF info for this image (if available)

I'm not sure if this would have any relevance to the point of focus, but the Lumix is actually a little longer than the Nocticron (my guess is about a 1mm focal length difference), so I had to be a little further away with the Lumix to keep the same framing.

... maybe there should be a "not different enough for me to guess" option in the poll? That might provide more meaningful information than people who don't feel strongly either way just picking one or the other at random...

I agree, for future polls, this should be an option.

 

[alex g]

One little nitpick about these two lovely images: the only real difference that jumps out at me, personally, is the slightly different focus points... I only mention it because when the OOF rendering is what's being evaluated, this can be significant, in my opinion, since it can affect the rendering of large parts of the scene. Other than that, I have almost no basis for distinguishing them - maybe there should be a "not different enough for me to guess" option in the poll? That might provide more meaningful information than people who don't feel strongly either way just picking one or the other at random...

I find that AF is often prone to front/back focusing at close distances, even when using CDAF, so manual focus is probably a good plan if the goal is identical focus points. While I agree that would be helpful in terms of assessing blur characteristics, I feel that the spirit of the exercise doesn't really require it — slightly different focal distances just forces us to be more cunning in our analysis. Personally, I'd rather see loads of people posting slightly imperfect comparisons that have a smaller number of painstakingly-shot ones. It's the cumulative effect that counts.

Regarding the idea of a "dunno" option in future polls, I see what you're getting at — it might potentially be a source of useful information. E.g. "In comparison Expensive_X vs Cheap_Y, 30 out of 35 people couldn't tell the difference, so Cheap_Y is clearly a steal!". Buuuut, in view of the law of averages, people guessing randomly between X and Y would give a similar result — so we would be able to draw a similar conclusion if we so desired. Plus, it's the process of trying to puzzle it out which I think is the worthwhile thing, not so much the actual outcome of the poll!

Let's field some more opinions first!

 

[0000]

The differences in apparent focus may represent differences in sharpness away from the center, perhaps?

Spoiler

I guess it's possible, but it looks like a Z-axis thing to me: if you move your gaze to the upper-right from the area you marked, you can see that A has sharper focus in areas closer to the lens, and B in areas further from it. I suppose this could also be related to different, asymmetrical "shoulders" to the DOF curves for each lens? I dunno.

Edit: added spoiler alert, in case it applies.

 

[0000]

While I agree that would be helpful in terms of assessing blur characteristics, I feel that the spirit of the exercise doesn't really require it — slightly different focal distances just forces us to be more cunning in our analysis.

Yeah, I feel that way on the one hand, but on the other, I can't deny that the difference, in a case like this of a very layered subject, can create a very different overall impression, which, if one happens to not notice the reason for, could be a misleading one. But as you say, that's all part of the fun, I suppose.

 

[retiredfromlife]

There's practically no difference between the two that can't be erased with a modicum of post processing.

Unless it's me doing the PP

 

[Gerard]

Buuuut, in view of the law of averages, people guessing randomly between X and Y would give a similar result — so we would be able to draw a similar conclusion if we so desired. Plus, it's the process of trying to puzzle it out which I think is the worthwhile thing, not so much the actual outcome of the poll!

Let's field some more opinions first!

I tend to agree with the - in view of the law of average - argument, but then we need lots of voters.