Help judging performance of my FE55z, FE35f2.8z horizon/infinity

Started Jul 5, 2016 | Questions
l_d_allan
l_d_allan Veteran Member • Posts: 5,091
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

slstr wrote:

All my fe35 shots posted were indicated by camera as infinity during manual focus. Focused on big TV radio tower 5km away, even the FE55 has trouble on 24mp A7 sensor to show much detail at 5km.. Amazing how you can see the lines that stabilize the structure with both 35 and 55 leases.. You need shots focused further away than 5km?!

My speculation is you (and others) are encountering atmospheric issues ... heat, haze, smog, heat, humidity, heat, pollution, wind, heat ...

Did I mention heat? Variants of haze?

My impression is that Jim K. can get away with near-infinity tilted horizons because his situation in California has occasional times of sufficient dryness, low haze, low wind, and other factors ... and even then usually in the morning only before it heats up.

My situation is probably even better/clearer than JK's here on the Front Range of Colorado at 6500' altitude, but I am still leery of near-horizon targets for lens testing.

MTYEWTK ?
Early on clear mornings maybe 30 times a year, we have what the aviation weather service describes as "unlimited visibility". That corresponds to clearly visible avalanche chutes on the Spanish Peaks 100+ miles away (if I'm at my sunrise vantage point with 360 deg FOV).

I think the assumption that near-infinity horizons are acceptable for lens evaluation is an obsolete carry-over of the [bad] old days of film.

IIRC, A.A. relied on testing with trees on the horizon, and what I think he described as "infinite detail the closer you looked".

Maybe that was fine with the much lower demands placed on lenses back then. Now? Not so much?

I'm am hoping that much simpler de-centering tests turn out to be adequate to filter out almost all "bad copies" so the odious and error prone brick wall testing can be greatly reduced, if not relegated to irrelevance.

 l_d_allan's gear list:l_d_allan's gear list
Panasonic Lumix DMC-TZ3 Panasonic Lumix DMC-FZ47 Canon PowerShot S110 Canon EOS 5D Mark II Canon EOS 600D +27 more
JimKasson
JimKasson Forum Pro • Posts: 26,986
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

l_d_allan wrote:

I think the assumption that near-infinity horizons are acceptable for lens evaluation is an obsolete carry-over of the [bad] old days of film.

IIRC, A.A. relied on testing with trees on the horizon, and what I think he described as "infinite detail the closer you looked".

In terms of cycles per picture height, an 8x10 view camera with film in it is nothing to sneeze at, even by today's standards. Such negatives are sufficiently sharp that there is a clear difference between an 8x10 silver-gelatin contact print on F paper and an identically sized print made in an enlarger.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +4 more
Nephi Contributing Member • Posts: 531
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

slstr wrote:

All my fe35 shots posted were indicated by camera as infinity during manual focus. Focused on big TV radio tower 5km away, even the FE55 has trouble on 24mp A7 sensor to show much detail at 5km.. Amazing how you can see the lines that stabilize the structure with both 35 and 55 leases.. You need shots focused further away than 5km?!

Ahhh... you have a A7.  That explains why infinity looks different.  I though you were using an A7R II.  My mistake.

Here's the scoop on the Sony 35mm.  ALL of them have a tilt that will manifest itself in a noticeably less detailed right or left side.  The only difference between the lens is how blurry the side is and wither its left or right.  Stopping down only helps marginally.

Your 35mm looks well centered however that's because you are using an A7.  An A7R II would show a blurry side.  I would keep the copy you have because on the A7 you are using it looks fine.  I see no issues.
Chalk the decentering up to Sony's awesome quality control.  The 35mm 1.4 is even worse as are several other Sony lenses.

Someone post a shot with the following criteria:

1.  Sony 35mm 2.8
2.  2.8 - 5.6 aperture
3.  A7R II
4.  Level shot.
5.  Infinity with detail on each side or corner. 
6.  Unedited (RAW or JPEG).

If the shot shows equal or near equal sharpness on each side I'd be stunned.

Nephi Contributing Member • Posts: 531
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

osv wrote:

slstr wrote:

Test protocol FE55 , manual focus max magnification+low peaking on every f-stop (focused on tall tv/radio broadcast tower far back over horizon, focus using center of lens, but in pictures it is located on the left side of picture).

useless test because the photos are downrezzed... you don't need raw for this, ooc jpeg is generally good enough.

don't do focus/recompose for lens testing... it's probably not an issue at those distances, but it's not a good habit to be in.

you have the right idea with your setup, but you need to be up higher, with more objects to make comparisons with.

see this korean 35/1.4 junk lens test that i did, for example, at f/2, it's perfectly focused at the sign area this side of the freeway, the problem areas should be obvious:

I wish I had that scene in my area.  That is my definition of a near perfect place to test for decentering.  You don't have a 35mm 2.8 you can test do you?

Nephi Contributing Member • Posts: 531
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

JimKasson wrote:

holin wrote:

JimKasson wrote:

Most people's difficulty in getting good MTF Mapper numbers is, I believe, the result of several issues:

  • Alignment
  • Getting a high-resolution target
  • Alignment
  • Getting the target big enough.
  • Alignment
  • Getting the target far enough away
  • Alignment
  • Lighting the target properly
  • Alignment
  • Vibration
  • Alignment
  • Field curvature

Oh, and alignment is important, too.

I would like to add flat target the above list. Folded and then straightened newspaper taped to a wall from two corners is not quite there. And alignment.

You are absolutely correct. I have found from experience that just dry mounting a target to a piece of foam core won't do it.

Jim

I printed 3 large ISF charts and posted them on foam core.  Despite all the effort a miniscule bend in a corner can give blurry false results.  Half the time you won't know if its the chart, your camera, or your lens.

I wish I could fine the link for a lens test I found earlier online.  Basically it is a $500 light box on a stand that has a test chart, perfectly flat, and perfectly lit.  While knowing if you are perfectly parallel could be a challenge at least you could rule out the target!

I'm still looking for a DYI centering test that is reliable and accurate enough to detect decentering.  I tried the flashlight test taking a shot of a test chart in each corner however I found the results be less than consistent and somewhat hard to interpret.

I've heard taking a shot at infinity and then flipping it upside down works however I'm not quite sure what you need to flip a camera and how in the world you would line it up upside down.

-Brian

l_d_allan
l_d_allan Veteran Member • Posts: 5,091
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

JimKasson wrote:

l_d_allan wrote:

I think the assumption that near-infinity horizons are acceptable for lens evaluation is an obsolete carry-over of the [bad] old days of film.

IIRC, A.A. relied on testing with trees on the horizon, and what I think he described as "infinite detail the closer you looked".

In terms of cycles per picture height, an 8x10 view camera with film in it is nothing to sneeze at, even by today's standards. Such negatives are sufficiently sharp that there is a clear difference between an 8x10 silver-gelatin contact print on F paper and an identically sized print made in an enlarger.

Not meaning to be argumentative (moi ?), but surely you aren't claiming/extrapolating that what was sufficient for the 8x10" view camera you may still own is therefore fine and dandy for a 42 mpx Super-Exmor?

And I do realize that your climate situation may very well allow you to "get away with" near-horizon tilted captures ... especially since they are right outside your door and thus much handier than a brick wall you'd have to drive to.

I'd suggest that a reasonable guideline for the rest of us might be that:

  • IFF you have a suitable climate (which we both do)
  • IFF you can walk to a suitable scene with near-infinity flatness of field targets (see below for my situation)  ... then
  • Knock yourself out with tilted horizons
  • Otherwise if you have to drive, consider using a brick wall (of which I have 2+ within walking distance so I don't have to drive to do that testing)

180 deg FOV pano from FE28 within easy walk from my front door. Not intended as test scene since there was no intent to have a flat field with all trees at near-infinity ... but that I am fortunate to have a usable near-infinity scene close by. Even better, it would work with a wide variety of focal lengths.

I'd also acknowledge that my non-aversion to brick walls may at least be partly due to having several good ones within walking distance.

 l_d_allan's gear list:l_d_allan's gear list
Panasonic Lumix DMC-TZ3 Panasonic Lumix DMC-FZ47 Canon PowerShot S110 Canon EOS 5D Mark II Canon EOS 600D +27 more
Nephi Contributing Member • Posts: 531
Re: Heat waves

osv wrote:

these are both 600mm f/6.3 shots from my tamron 150-600, with fast shutter speeds, same scene that i posted earlier, shot across a busy freeway... the "hospital san diego" text is reasonably sharp? you can still almost read the license plate numbers, but this looks like a possible heat wave problem to me... it's ooc jpeg, a nightmare compression scenario.

the alignment on the brick wall shot looks fairly close, because the bottom corners look equivalent to each other; the outer center/top corner region on the left side looks weaker, however.

the first thing you are looking for with any test shot is that the sides match each other, followed by how do the sides compare to the center.

Dan,

As others have noted the brick wall test is very hard to do correctly.  Even the slightest lack of a being  parallel and you will get skewed results.  Not to mention a slight difference in the way the light hits the wall can cause sharpness differences.  Personally, I would only use the brick wall test to check for gross decentering

osv Veteran Member • Posts: 9,970
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

Nephi wrote:

I wish I had that scene in my area. That is my definition of a near perfect place to test for decentering. You don't have a 35mm 2.8 you can test do you?

no, sorry i don't, i wish that i had that lens.

-- hide signature --

dan

osv Veteran Member • Posts: 9,970
Re: Heat waves

Nephi wrote:

Dan,

As others have noted the brick wall test is very hard to do correctly. Even the slightest lack of a being parallel and you will get skewed results. Not to mention a slight difference in the way the light hits the wall can cause sharpness differences. Personally, I would only use the brick wall test to check for gross decentering

agreed, the blocks aren't in a perfectly parallel line, and even tho i leveled out the camera, i didn't check to see how close the wall was to being parallel.

this rough surface is really good for using with focus peaking tho, as someone mentioned earlier, it's better than bricks or wood.

i'm sticking to the big scenes tho.

-- hide signature --

dan

JimKasson
JimKasson Forum Pro • Posts: 26,986
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

Nephi wrote:

JimKasson wrote:

holin wrote:

JimKasson wrote:

Most people's difficulty in getting good MTF Mapper numbers is, I believe, the result of several issues:

  • Alignment
  • Getting a high-resolution target
  • Alignment
  • Getting the target big enough.
  • Alignment
  • Getting the target far enough away
  • Alignment
  • Lighting the target properly
  • Alignment
  • Vibration
  • Alignment
  • Field curvature

Oh, and alignment is important, too.

I would like to add flat target the above list. Folded and then straightened newspaper taped to a wall from two corners is not quite there. And alignment.

You are absolutely correct. I have found from experience that just dry mounting a target to a piece of foam core won't do it.

Jim

I printed 3 large ISF charts and posted them on foam core. Despite all the effort a miniscule bend in a corner can give blurry false results. Half the time you won't know if its the chart, your camera, or your lens.

Yep.

I wish I could fine the link for a lens test I found earlier online. Basically it is a $500 light box on a stand that has a test chart, perfectly flat, and perfectly lit. While knowing if you are perfectly parallel could be a challenge at least you could rule out the target!

Imatest has a bunch of boxes like that. They're not cheap, and neither are the special targets.

But the big problem is getting them big enough, especially for short focal length lenses. You don't want to have to focus the lens to a super close distance where you'll never use it. That's looking for the keys under the lamppost. (If you haven't heard the joke, let me know)

I'm still looking for a DYI centering test that is reliable and accurate enough to detect decentering. I tried the flashlight test taking a shot of a test chart in each corner however I found the results be less than consistent and somewhat hard to interpret.

What the big boys use is a dedicated setup that is aligned by measurement. Not practical for somebody like me.

I've heard taking a shot at infinity and then flipping it upside down works however I'm not quite sure what you need to flip a camera and how in the world you would line it up upside down.

Here's what I use:

http://www.reallyrightstuff.com/CRD-87-Package-Camera-Rotation-Device

Also not cheap, but a lovely, precise tool that you can use in real photography.

If you're using adapted lenses, you can get an adapter with a rotating feature.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +4 more
JimKasson
JimKasson Forum Pro • Posts: 26,986
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

l_d_allan wrote:

JimKasson wrote:

l_d_allan wrote:

I think the assumption that near-infinity horizons are acceptable for lens evaluation is an obsolete carry-over of the [bad] old days of film.

IIRC, A.A. relied on testing with trees on the horizon, and what I think he described as "infinite detail the closer you looked".

In terms of cycles per picture height, an 8x10 view camera with film in it is nothing to sneeze at, even by today's standards. Such negatives are sufficiently sharp that there is a clear difference between an 8x10 silver-gelatin contact print on F paper and an identically sized print made in an enlarger.

Not meaning to be argumentative (moi ?), but surely you aren't claiming/extrapolating that what was sufficient for the 8x10" view camera you may still own is therefore fine and dandy for a 42 mpx Super-Exmor?

I sold my Arca Swiss 8x10 about 8 years ago. Actually, I traded it for part of a new Nikon 400/2.8. I still have 'chromes made with that camera. There is nothing in digital photography like putting one of those on a light box. With panos, you can beat the resolution, but you can't print that sharp. I also own 8x10 and 8x20 silver contact prints. They are amazing, too.

Consider this: with 20 cy/mm on an 8x10, you'd get 4000 cy/ph. I did that in my head; I hope I got it right.

And I do realize that your climate situation may very well allow you to "get away with" near-horizon tilted captures ... especially since they are right outside your door and thus much handier than a brick wall you'd have to drive to.

I'd suggest that a reasonable guideline for the rest of us might be that:

  • IFF you have a suitable climate (which we both do)
  • IFF you can walk to a suitable scene with near-infinity flatness of field targets (see below for my situation) ... then
  • Knock yourself out with tilted horizons

But remember that it's qualitative. And then there's focus. And field curvature.

  • Otherwise if you have to drive, consider using a brick wall (of which I have 2+ within walking distance so I don't have to drive to do that testing)

Alignment? How are ya gonna do that? not saying it's impossible, but what's your proposal?

180 deg FOV pano from FE28 within easy walk from my front door. Not intended as test scene since there was no intent to have a flat field with all trees at near-infinity ... but that I am fortunate to have a usable near-infinity scene close by. Even better, it would work with a wide variety of focal lengths.

I'd also acknowledge that my non-aversion to brick walls may at least be partly due to having several good ones within walking distance.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +4 more
JimKasson
JimKasson Forum Pro • Posts: 26,986
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

Nephi wrote:

slstr wrote:

All my fe35 shots posted were indicated by camera as infinity during manual focus. Focused on big TV radio tower 5km away, even the FE55 has trouble on 24mp A7 sensor to show much detail at 5km.. Amazing how you can see the lines that stabilize the structure with both 35 and 55 leases.. You need shots focused further away than 5km?!

Ahhh... you have a A7. That explains why infinity looks different. I though you were using an A7R II. My mistake.

Here's the scoop on the Sony 35mm. ALL of them have a tilt that will manifest itself in a noticeably less detailed right or left side.

How do you know that?

The only difference between the lens is how blurry the side is and wither its left or right. Stopping down only helps marginally.

Your 35mm looks well centered however that's because you are using an A7. An A7R II would show a blurry side. I would keep the copy you have because on the A7 you are using it looks fine. I see no issues.
Chalk the decentering up to Sony's awesome quality control. The 35mm 1.4 is even worse as are several other Sony lenses.

Someone post a shot with the following criteria:

1. Sony 35mm 2.8
2. 2.8 - 5.6 aperture
3. A7R II
4. Level shot.
5. Infinity with detail on each side or corner.
6. Unedited (RAW or JPEG).

If the shot shows equal or near equal sharpness on each side I'd be stunned.

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +4 more
(unknown member) Contributing Member • Posts: 854
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

Here's an unverified idea for cheap aligment. Aligning to a wall is difficult, even with lasers it takes some work and there are multiple sources of uncertainty. However, floors (and ceilings) are level in modern buildings, usually done with a self leveling compound (think how water surface levels itself). My arca clamp has a bubble level that can be used to point the camera straight up or down, which should make it parallel to a level floor or ceiling. Maybe this would work, at least better than trying to align to a wall, if no specialist setup is available.

Nephi Contributing Member • Posts: 531
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

JimKasson wrote:

Nephi wrote:

slstr wrote:

All my fe35 shots posted were indicated by camera as infinity during manual focus. Focused on big TV radio tower 5km away, even the FE55 has trouble on 24mp A7 sensor to show much detail at 5km.. Amazing how you can see the lines that stabilize the structure with both 35 and 55 leases.. You need shots focused further away than 5km?!

Ahhh... you have a A7. That explains why infinity looks different. I though you were using an A7R II. My mistake.

Here's the scoop on the Sony 35mm. ALL of them have a tilt that will manifest itself in a noticeably less detailed right or left side.

How do you know that?

The only difference between the lens is how blurry the side is and wither its left or right. Stopping down only helps marginally.

Your 35mm looks well centered however that's because you are using an A7. An A7R II would show a blurry side. I would keep the copy you have because on the A7 you are using it looks fine. I see no issues.
Chalk the decentering up to Sony's awesome quality control. The 35mm 1.4 is even worse as are several other Sony lenses.

Someone post a shot with the following criteria:

1. Sony 35mm 2.8
2. 2.8 - 5.6 aperture
3. A7R II
4. Level shot.
5. Infinity with detail on each side or corner.
6. Unedited (RAW or JPEG).

If the shot shows equal or near equal sharpness on each side I'd be stunned.

Good question. I've tested 7 copies myself.  I also spoke to Roger C. and echoed a similar sentiment.

Same problem applies to the 35mm 1.4 as well. Though I've never used that lens because I find the 2.8 to be, ironically, sharper at some apertures.

I don't mind posting a few example if you are interested. So while there may be a centered copy I've yet to see one!  Maybe someone is holding a centered copy and can chime in?

I

JimKasson
JimKasson Forum Pro • Posts: 26,986
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

Nephi wrote:

JimKasson wrote:

Nephi wrote:

slstr wrote:

All my fe35 shots posted were indicated by camera as infinity during manual focus. Focused on big TV radio tower 5km away, even the FE55 has trouble on 24mp A7 sensor to show much detail at 5km.. Amazing how you can see the lines that stabilize the structure with both 35 and 55 leases.. You need shots focused further away than 5km?!

Ahhh... you have a A7. That explains why infinity looks different. I though you were using an A7R II. My mistake.

Here's the scoop on the Sony 35mm. ALL of them have a tilt that will manifest itself in a noticeably less detailed right or left side.

How do you know that?

The only difference between the lens is how blurry the side is and wither its left or right. Stopping down only helps marginally.

Your 35mm looks well centered however that's because you are using an A7. An A7R II would show a blurry side. I would keep the copy you have because on the A7 you are using it looks fine. I see no issues.
Chalk the decentering up to Sony's awesome quality control. The 35mm 1.4 is even worse as are several other Sony lenses.

Someone post a shot with the following criteria:

1. Sony 35mm 2.8
2. 2.8 - 5.6 aperture
3. A7R II
4. Level shot.
5. Infinity with detail on each side or corner.
6. Unedited (RAW or JPEG).

If the shot shows equal or near equal sharpness on each side I'd be stunned.

Good question. I've tested 7 copies myself. I also spoke to Roger C. and echoed a similar sentiment.

Same problem applies to the 35mm 1.4 as well. Though I've never used that lens because I find the 2.8 to be, ironically, sharper at some apertures.

I don't mind posting a few example if you are interested. So while there may be a centered copy I've yet to see one! Maybe someone is holding a centered copy and can chime in?

I've got a copy that has been fine. I can take a hard look at it.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +4 more
Nephi Contributing Member • Posts: 531
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

JimKasson wrote:

I've got a copy that has been fine. I can take a hard look at it.

Jim

-- hide signature --

Perfect.  This will be a good test because you've tested it and did not see decentering.  In my experience the best way to see the decentering is to take a level infinity test with somewhat equal detail on each far side (i.e. same trees, or similar structures).  Obviously a test shot at 2.8 is the best way to see the issue.  Though to be clear, stopping down does not seem to help much.

You have Imatest.  Is there not a way to put up a test chart and measure sharpness left to right?  I would think that would be the ideal approach.  You'd be the guy I'd trust to get the test right.

I'm anxious to see what you find.

l_d_allan
l_d_allan Veteran Member • Posts: 5,091
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

Nephi wrote:

I printed 3 large ISF charts and posted them on foam core. Despite all the effort a miniscule bend in a corner can give blurry false results. Half the time you won't know if its the chart, your camera, or your lens.

If feasible, I like to be 100x Focal Length or so from the target. That can present a problem with some/many focal lengths.

I wish I could fine the link for a lens test I found earlier online. Basically it is a $500 light box on a stand that has a test chart, perfectly flat, and perfectly lit. While knowing if you are perfectly parallel could be a challenge at least you could rule out the target!

I'd also be concerned about being able to get far enough away with many focal lengths.

Hmmmmm .... wonder if address'ing this niche issue could be a valid crowd-funding'ish project? IIRC, I had such related correspondence to Jim K's razor blade based protocol plus slanted wall or door.

I'm still looking for a DYI centering test that is reliable and accurate enough to detect decentering. I tried the flashlight test taking a shot of a test chart in each corner however I found the results be less than consistent and somewhat hard to interpret.

Drat. I've had my fingers crossed that such an approach could weed out the vast majority of "bad copies". However, that always seemed like wishful thinking on my part to make tedious, error-prone brick wall testing be obsolete.

I haven't tried either the OOF LED approach, or the more involved test from LensRentals several years ago.
https://www.lensrentals.com/blog/2012/05/testing-for-a-decentered-lens-an-old-technique-gets-a-makeover/

I've heard taking a shot at infinity and then flipping it upside down works however I'm not quite sure what you need to flip a camera and how in the world you would line it up upside down.

I wondered the same thing ... and "not quite sure" ... but baffled. Seems like you'd need a rather advanced / specialized tripod head or a "rigid 3 sided L bracket".

I suppose a lens with a tripod collar would work, but those lenses are few and far between. My pano-head might work, especially if I could come to terms with using near-infinity horizons as a target.

 l_d_allan's gear list:l_d_allan's gear list
Panasonic Lumix DMC-TZ3 Panasonic Lumix DMC-FZ47 Canon PowerShot S110 Canon EOS 5D Mark II Canon EOS 600D +27 more
l_d_allan
l_d_allan Veteran Member • Posts: 5,091
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

JimKasson wrote:

Here's what I use:

http://www.reallyrightstuff.com/CRD-87-Package-Camera-Rotation-Device

Also not cheap, but a lovely, precise tool that you can use in real photography.

Link didn't work for me with embedded RRS items ... this may be a better alternative:
http://www.reallyrightstuff.com/Heads/Camera-Rotation

If you're using adapted lenses, you can get an adapter with a rotating feature.

Hadn't thought of that. Just checked ... my mid-level FotodioX has a removable "foot" but it is in a fixed location.

Perhaps a relatively cheapo, but still well made fully manual adapter might suffice, if available? Seems less than likely it would be rotatable, or have mounting holes for a removable "foot" that were 180 deg opposite.

YAUNCFP? (yet another unlikely niche crowd funding project)

 l_d_allan's gear list:l_d_allan's gear list
Panasonic Lumix DMC-TZ3 Panasonic Lumix DMC-FZ47 Canon PowerShot S110 Canon EOS 5D Mark II Canon EOS 600D +27 more
l_d_allan
l_d_allan Veteran Member • Posts: 5,091
Easy to do invalid brick wall testing, but not beyond many motivated Zony owners.

Nephi wrote:

As others have noted the brick wall test is very hard to do correctly.

As is all valid "good copy vs bad copy" testing for those not owning an optical bench or commercial ImaTest setup. (but I'm hoping to be yet again in error and that the OOF LED approach works out)

Even the slightest lack of a being parallel and you will get skewed results.

My experience has been that:

  • with sufficient attention to detail,
  • after a learning curve,
  • a motivated purchaser of a pricey Sony/Zony lens ... especially those with known crap-shot q/c issues ... (term from LensRentals that applies to FE90, several zooms, and especially the notious FE35 f1.4 ... ten of ten which were rejects)
  • should be able to accomplish adequate alignment with a brick wall
  • at a rigor level appropriate for "good copy vs bad copy" testing.
  • But NOT for valid, repeat-able MtfMapper numbers (bummer)
  • ...
  • Obviously ... GIGO ... but you don't need 100% alignment down to to the micrometer to get adequate alignment. You can be the slightest bit off, but obviously past a certain threshold, evaluation will be invalid. 
  • I don't happen to know precisely what that percentage is, but I don't think you need to be a rocket scientist at JPL to achieve satisfactory alignment. YMMV.

Not to mention a slight difference in the way the light hits the wall can cause sharpness differences.

Along with multiple other issues ... especially sloppy focus. I'll speculate that testers who have trouble with meticulous alignment might also have trouble with meticulous focus.

Personally, I would only use the brick wall test to check for gross decentering

Sorry to be blunt, but perhaps that says more about your testing proficiency than the adequacy of brick wall testing ... done correctly?

I've got my fingers crossed that the relatively simple, but still unproven OOF LED test is satisfactory to weed out the great majority of out-of-spec lenses.

One of the "known unknowns" is how much of these out-of-spec issues are from de-centering. I've got an email to Roger C. at LR to check whether he might have an estimate he could share.

In the meantime, I'd be reluctant to purchase ANY FE lens if you aren't confident in your experience and expertise to do "good copy vs bad copy" testing.

 l_d_allan's gear list:l_d_allan's gear list
Panasonic Lumix DMC-TZ3 Panasonic Lumix DMC-FZ47 Canon PowerShot S110 Canon EOS 5D Mark II Canon EOS 600D +27 more
JimKasson
JimKasson Forum Pro • Posts: 26,986
Re: Help judging performance of my FE55z, FE35f2.8z horizon/infinity

Nephi wrote:

JimKasson wrote:

I've got a copy that has been fine. I can take a hard look at it.

Perfect. This will be a good test because you've tested it and did not see decentering. In my experience the best way to see the decentering is to take a level infinity test with somewhat equal detail on each far side (i.e. same trees, or similar structures). Obviously a test shot at 2.8 is the best way to see the issue. Though to be clear, stopping down does not seem to help much.

You have Imatest. Is there not a way to put up a test chart and measure sharpness left to right? I would think that would be the ideal approach. You'd be the guy I'd trust to get the test right.

I'm anxious to see what you find.

I'll get to more exhaustive, more time consuming testing, but here's the OOF LED:

I've seen better. Looking at the lens when it's wide open, I can see the problem; the blades on one side are not fully retracting. That shouldn't affect normal photography.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +4 more
Keyboard shortcuts:
FForum MMy threads