MFT Lens Turbo II Flare

Hoods with a rectangle mask reduce the unneeded light even more, a square mask already has a 37% smaller opening compared to a round mask. The percentage increases when the sensor sizes ratio gets to 2:3 or beyond. It is however not easy to calculate the right shape of the mask as the extension from the lens front element, lens design, focal length and aperture opening play a role then. I converted an old Canon SLR (opal light source in the film plane, mirror + shutter for ever open) to get some idea about the needed size but even then it becomes a trial and error system to get the best mask size. After making the mask the FF digital camera is set at infinity on a copy table, widest stop an image is made of the white evenly lit table. Then the vignetting is examined compared to an image without the hood. Mask size changed accordingly. I tolerate some extra vignetting in the corners but not on the edges.

Before I printed 3D lens hoods, I made some designs of masks that fit into lens filter rings and had them laser cut out of polycarbonate. Using small stamped magnetic vinyl discs glued into the masks I could attach even smaller masks for APS sensor size on the masks for FF. While the Sony FFs allow making APS size images too I will not use that much so the concept is there but may not be used on the 3D hoods.

Faint image of the projected illuminated film gate, that image normally is copied with a digital camera on the copy table in a dark room. Grid is in mm. Considering a widest aperture projection through a transparent mm grid to the ceiling now for a better representation of what is needed.

On the left the converted SLR for mask estimation, must be one of my ugliest creations ever.

Met vriendelijke groet, Ernst
750+ inkjet paper white spectral plots: OBA content etc.
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

E Dinkla wrote:

Hoods with a rectangle mask reduce the unneeded light even more, a square mask already has a 37% smaller opening compared to a round mask. The percentage increases when the sensor sizes ratio gets to 2:3 or beyond. It is however not easy to calculate the right shape of the mask as the extension from the lens front element, lens design, focal length and aperture opening play a role then. I converted an old Canon SLR (opal light source in the film plane, mirror + shutter for ever open) to get some idea about the needed size but even then it becomes a trial and error system to get the best mask size. After making the mask the FF digital camera is set at infinity on a copy table, widest stop an image is made of the white evenly lit table. Then the vignetting is examined compared to an image without the hood. Mask size changed accordingly. I tolerate some extra vignetting in the corners but not on the edges.

Before I printed 3D lens hoods, I made some designs of masks that fit into lens filter rings and had them laser cut out of polycarbonate. Using small stamped magnetic vinyl discs glued into the masks I could attach even smaller masks for APS sensor size on the masks for FF. While the Sony FFs allow making APS size images too I will not use that much so the concept is there but may not be used on the 3D hoods.

Faint image of the projected illuminated film gate, that image normally is copied with a digital camera on the copy table in a dark room. Grid is in mm. Considering a widest aperture projection through a transparent mm grid to the ceiling now for a better representation of what is needed.

On the left the converted SLR for mask estimation, must be one of my ugliest creations ever.

Met vriendelijke groet, Ernst
750+ inkjet paper white spectral plots: OBA content etc.
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

Fabulous idea! Makes perfect sense. You could design a set for popular lenses and market this accessory. Do a Kickstarter campaign.

Alan WF wrote:

Tom Caldwell wrote:

My own little comment is that I may be too casual to notice issues. If something doesn't work out I tend to just assume that my skills are to blame and my good shots were just a lucky streak.

I think you're being too modest.

If anyone wants to look for this, try taking a photo of a patch of interior wall right next to window, with the sensor edge at the boundary between the wall and the window.

Regards,

Alan

Thanks for the tip Alan.

Just tried it with an unlikely suspect: GX85 + Metabones Ultra + Canon EF 70-200mm f2.8 L.

Seems to work perfectly as it should - spot on AF as well.

Have to cut grass (lots of it*) so it will curtail my immediate efforts.  I will try and find time to drag my "original" lens turbos out of the archives.  I have a few focal reduction adapters I can try but I never made it to the LTII version.

* yes the grass stays green and keep growing hereabouts in winter but not very fast - I just have not cut it for a while

Alan WF wrote:

Alan WF wrote:

I might look at that with my 50/1.4 and 100/2.8.

Yeah, they flare too.

I pointed the camera at an interior wall just below a brightly illuminated window. The window was just outside the top of the frame. I tested the New FD 100/2.8, 50/1.4, and 28/2.8 on an unbaffled LT2. I shot each lens wide open focused on the wall, once with no shade and once with a 15 x 20 cm piece of cardboard above and in front of the lens to shade it from the window. This is pretty much a torture test for this sort of flare.

Here is the BTS:

And here are the results. The top row is the 100/2.8, the middle row the 50/1.4, and the bottom row the 28/2.8. On the left, without the shade. On the right, with the shade.

So, yes, they all suffer from this flare.

Interesting, the flare changes character with focal length. With the 100/2.8, it seems to be a fairly uniform veiling. With the 50/1.4 and more so with the 28/2.8, the flare acquires more structure and is worse at the top of the frame.

Regards,

Alan

This is the sort of torture test that usually produces the "blue spot" issue.

Hello!

I did a few more test snaps with the Metabones Speedbooster and Minolta Tele Auto Rokkor 100/2.0, without lens hood, wide open:

window below

Nothing came out as bad as your examples. I did find a flare pattern that tends to show itself without hood though:

A lens hood (even a regular one) does work wonders on my lens combo.

Best,

Alex

As before, your window scenes don't appear to have much brightness range compared to Alan's tests (and mine), where instead, inside the window there's almost no light and outside it's very bright. They seem to have diffuse lighting that doesn't challenge the lens system.

Hello Alan!

You are most welcome - if you are interested in anything specific, I will try to test it for you.

Best,

Alex

Hello!

Thanks for the input - I don't know, is there any specific minimum exposure difference you would suggest? I will gladly retest.

I did not see any examples of your's btw, but I tried to get close to the images in the original post, which to my eye don't sport a much more extreme exposure difference. I could be wrong though and am open to suggestions.

I only have this mobile phone sample snap of our holiday home which does show the "test range" and the conditions the snaps were done in though (except for the last sample):

Best,

Alex

SiFu wrote:

Hello!

Thanks for the input - I don't know, is there any specific minimum exposure difference you would suggest? I will gladly retest.

I did not see any examples of your's btw, but I tried to get close to the images in the original post, which to my eye don't sport a much more extreme exposure difference. I could be wrong though and am open to suggestions.

I only have this mobile phone sample snap of our holiday home which does show the "test range" and the conditions the snaps were done in though (except for the last sample):

Best,

Alex

Hi Alex,

I didn't shoot any tests, but just visually tried four LTii adapted Nikon lenses at a window with no inside lighting, looking out on a sunlit scene -- I mean direct sunlight on urban buildings across the way and concrete pavement outside the window. My results were similar to Alan's, so I didn't bother to record them.

Your scene (shown here, at least) has only scant oblique sunlight on the wood deck and most of the distant scene has a low brightness range, which as I thought, doesn't challenge the lens system.

Looks like a very pleasant place to be, though! I'll have a Mai Tai, please. I don't know what Alan's view out his window looks like.

Hello!

Thanks for the pointers! "Unfortunately" this is pretty much the scenery for me atm, I will try again when we return home in August.

Do you think that reflections off your buildings caused the flare?

Best,

Alex

SiFu wrote:

Do you think that reflections off your buildings caused the flare?

Absolutely. Back lighting is the killer.

Below is a photo that demonstrates the veiling flare. Shot with Olympus E-M5 MkII, Nikon 105mm f/2.8 lens and Lens Turbo II, with a deeper lens hood added, and an old Hoya U-V filter. 1/160th sec., I think it was f/11 (f/8 actual with FR). The filter might have added to the flare. Shot on tripod as shown.

Downsized from originals to 1080h. The first image is the OOC JPEG. Next is my processed RAW. As you can see, I successfully removed the flare in PS using a Levels Adjustment Layer and moving the bottom slider up, then applying a mask and locally erasing it. Of course, it would be better not to have it in the first place, but it doesn't spoil the sharpness, contrast, or color quality if you're working from a Raw file.

OOC JPEG

Processed from RAW file

I know there's degradation, but I just don't find it to be that much of a problem with most scenes, considering the ability to correct veiling flare in PP.

Here's a before-after shot with Micro-Nikkor 55mm f/3.5 lens and LTii (downsized to 1200h). First, the OOC JPEG; second, the corrected processed Raw. Below that are 100% crops, so you can see the improvement in detail & contrast (unfortunately a bit more noise due to sharpening). Corrections were made on separate layers, so further adjustments could be made (i.e. African-American man's face).

View the last two at 100% Details to see a good comparison.

before-after 1200h

100% crop, OOC JPEG

100% crop, retouched processed Raw

Alan WF wrote:

I could well agree that there's a loss of contrast here, but I don't see a strongly structured flare like the one over my face in my original example with the 28 mm, coming down from the window in my later examples with the 28, 50, and 100 mm, or coming in from the upper left in your earlier image of the woman walking.

I think the reason is that the scene is so bright. I find the flare ruins the image when I have a dark scene with bright lighting around it. Here, there's not enough contrast between the scene and the sky for the flare to completely spoil the image.

I posted the marina scene to show how veiling flare happens due to bright surfaces causing it within a scene. As with the previous image, it can be corrected. Besides your test shots, have you encountered this severe "strongly structured flare" in your normal shooting? I haven't done much with the LTii lately, but looking through my files, I'm hard pressed to find any examples, except some night shots where street lights have extra glow around them. I mean, were you seeing defects in your normal work, which prompted you to do the tests and the mod?

Here's the Nikon 105 w/ LTii again --

I actually think the flare from a stoplight here makes the shot more interesting.