5x Test Coming Soon but I Need Some Feedback First Please

[RobertOToole]

Finalizing a 10 lens 5x test now but I need some help with some feedback if you have a couple of seconds.

Each center and corner image is 1250px across.

Whats the best or easiest way to view these in your opinion?

Center/Corner in one 2500px image.

Center/Center then corner/corner in one 2500px image.

Centers in 1250px images than corners in 1250px images posted separately.

Anything animated like a GIF or compare-view plugin is going to crush the quality since they are 1250px crops.

Any ideas or comments appreciated.




This is the list of lenses in case anyone is interested:

Canon 20mm f/3.5 Macrophoto Lens

Canon MP-E 65mm f/2.8 1-5x Macro Photo lens

Mitutoyo M Plan Apo 5X/0.14 Objective

Mitutoyo M Plan Apo 7.5X/0.21 Objective @ 4.7x

Nikon 5x-A MM Measurescope Objective Lens

Nikon CFI LU Plan Fluor 5X/0.15 Objective

Olympus MPLFLN 5X/0.15 Semi-Apochromat Objective

Qioptiq Mag.x LD Plan Apochromat 5x / 0.20 Objective @ 4.7x

Sigma 150mm OS + Schneider-Kreuznach Xenon 28mm f/2 Lens Combination

Tominon 17mm f4 Lens

BTW there is a big surprise in this test

[Adalbert]

Hi Robert,
I would prefer the first option: “Center/Corner in one 2500px image.“
and I'm looking forward to seeing the results of your test!
BTW, my favorite is Lu Plan (I have only :-)
BR, ADi

[RobertOToole]

Hi Robert,
I would prefer the first option: “Center/Corner in one 2500px image.“
and I'm looking forward to seeing the results of your test!
BTW, my favorite is Lu Plan (I have only :-)
BR, ADi

Thank you ADi.

The Lu Plan is a nice lens and they aren't too expensive and the NA is good. I have the 5x and 10x versions.

Robert

[typestar]

Hi Robert,

I would like to see Center/Center then corner/corner in one 2500px image.

And I am very exited about the results - especially the sharpness results of the 7,5x Mitu in this "grail" ;-)
Thankyou again for your great efforts - for all of us...!

Christian

[enricosavazzi]

[...]
I would like to see Center/Center then corner/corner in one 2500px image.
[...]

This would work best for me too. When comparing two images, nothing beats seeing them side-by-side and as close as possible to each other.

Top-to-bottom (vertically stacked) and as close as possible may also work for my eyes, but the brightness, contrast and/or color rendering of many LCD desktop screens changes with vertical view angle, which makes it impossible to compare accurately vertically stacked images.

Incidentally, the latest generation of HP laptops that we were given at work a few months ago (at a large telecom company) have a very narrow horizontal viewing angle, which may make unbiased comparisons of side-by-side images difficult. Perhaps the idea with these narrow-angle screens is preventing snooping by strangers sitting on a nearby seat on a plane or train, or perhaps it is a way to save battery charge, but it does make it difficult for two people to share the laptop screen.

[rjlittlefield]

When comparing two images, nothing beats seeing them side-by-side and as close as possible to each other.

I disagree. For my eyes, "flash to compare" is way better. (Alternate display in exactly the same place.)

I've been doing a lot of stack retouching the last few days, and despite years of doing that task I'm still frequently surprised at how different the PMax and DMap results look when flashed, even though in side-by-side display I may be inclined to think "OK, nothing to do here."

Unfortunately (as Robert has pointed out), flash-to-compare in a browser either loses a lot of quality or requires some special plugin.

So if I get down to flash-to-compare, I'll probably be pulling the images into Photoshop or StereoPhoto Maker anyway. (Swap Left/Right in SPM is quick and effective if the image pairs are formatted for it.)

--Rik

[RobertOToole]

Thanks for the input everyone. I appreciate it.

Good points Rik has made. I will mention that in the post, it would be best to download and pull the images into an app to compare on your computer.

Will work on the test results over the next couple of days when I get time.


Robert

[Chris S.]

I was planning to write exactly what Rik said. For my eyes, flash-to-compare is the only method I trust. I test a fair number of lenses, and it is the only method I employ for comparison. (This is why I almost never publish lens-test posts, though I do, whenever asked, share layered-up comparisons in a Photoshop file.)

Robert, any chance you could make such a Photoshop file available for download? While not everyone here uses Photoshop, many of us do.

Cheers,

--Chris S.

[ChrisR]

I agree, flash to compare is clearest for me too.

Whether it's healthy or not is another matter:? .

I very often scrutinize tests.
Very often I reflect that if I'm looking at a 1000 pixel actual crop which fills my screen, I'm looking at part of an image which would fill my wall.

Then I think "Hey, that's pretty good," and pause to wonder on how some bee's knee or gnat's appendage would look, so magnified.

Then I click the image and the horror of my naïvety crashes down, as the new image is so superior.
"How could anyone accept that first image? It's not even a big wall!!"

[RobertOToole]

I was planning to write exactly what Rik said. For my eyes, flash-to-compare is the only method I trust. I test a fair number of lenses, and it is the only method I employ for comparison. (This is why I almost never publish lens-test posts, though I do, whenever asked, share layered-up comparisons in a Photoshop file.)

Robert, any chance you could make such a Photoshop file available for download? While not everyone here uses Photoshop, many of us do.

Cheers,

--Chris S.

Hi Chris,

In regards to the crops.I actually recommend as a best practice to download the full size 100% crop files to compare on their own computer, at least on the last couple of tests I posted.

The file you see on the forum is linked to a larger full size file from my site, so you can click and then right click or two finger press and save image as....

I will make sure I that is clear on the new 5x test.

In regards to the full size un-cropped files. If people are interested I can share the full test files from the 5x test no problem. I have a couple of online storage options that I can use.

Thanks for comments, I appreciate it.

Robert

[RobertOToole]

I agree, flash to compare is clearest for me too.

Whether it's healthy or not is another matter:? .

I very often scrutinize tests.
Very often I reflect that if I'm looking at a 1000 pixel actual crop which fills my screen, I'm looking at part of an image which would fill my wall.

Then I think "Hey, that's pretty good," and pause to wonder on how some bee's knee or gnat's appendage would look, so magnified.

Then I click the image and the horror of my naïvety crashes down, as the new image is so superior.
"How could anyone accept that first image? It's not even a big wall!!"

Some lens results look fine on their own, like an MP-E file, only when you compare it to the Minolta Elite 5400, for example, does it look weak.

For the 5x test I am working on now I planned to make un-cropped and sharpened 2500 pixel files available since it turns out all of the lenses in next test look perfectly fine on their own at this size. At 100% its a different story of course.

BTW I think the 5x test results will raise some eyebrows.

Robert

[Chris S.]

In regards to the crops.I actually recommend as a best practice to download the full size 100% crop files to compare on their own computer, at least on the last couple of tests I posted.

The file you see on the forum is linked to a larger full size file from my site, so you can click and then right click or two finger press and save image as....

I've certainly done that.

In regards to the full size un-cropped files. If people are interested I can share the full test files from the 5x test no problem. I have a couple of online storage options that I can use.

These full, uncropped files are the ones I think folks should layer up. Then it's easy to compare center, sides, corners, or features they think particularly reveling--all on a single file.

Robert, I may be lazy in suggesting this, but I think it would be most useful if you created a Photoshop file containing all the full-images, layered & labeled for clarity, and offered this PSD file for download. This takes time, I know. But done this way, the time is spent only once--and over and over by each person interested in interpreting your data. I think this might result in more people fully appreciating the distinctions you've discovered in your experiments.

But sorry to be suggesting you do even more work, Robert!

Cheers,

--Chris S.

[Smokedaddy]

... how does monitor calibration come into play when comparing images?

[RobertOToole]

... how does monitor calibration come into play when comparing images?

Brightness might be a little off if it's not a calibrated screen but I purposely shot the last two tests a little dark so they wont hurt your eyes when you look at the images on a non-calibrated monitor or phone.

I plan to upgrade to a pixel shift body in the future and these have better dynamic range I believe so the files and crop images will look even better and easier on the eyes also.

Robert

[typestar]

...
I plan to upgrade to a pixel shift body in the future and these have better dynamic range I believe ...

Robert -- which body, how near the future for this change... ?

typestar