carstenw wrote:
There is a tiny little dot next to the Leica logo dot which can measure the light independently, and can thus be used to estimate the aperture. In the Leica DNGs, there is a field for this estimated aperture, although it is not part of the public tags.
That's so clever. This inspired me for the following method to recover the aperture:
Shoot with F1.4 first, then the desired aperture.
Then in Lightroom use match exposure for these two files.
The EV deducted from the 2nd file would be the required EV to add to the aperture.
One question, when you say it is not in the public tag, can Lightroom sees that? I am thinking of the Jeffrey's plugin to extract that information and updated them via LensTagger. Though not really matters to me since I probably won't spend that much money on M9. (To me A7(R) are the poor man's Leica, apparently)
Lightroom reads it, I imagine, but it isn't shown. TIFF/DNG has the information in tags, and there is something called private maker's tags or something like that. Programs aren't meant to show it.
Come to think of it, I think that Leica made it harder to read at some point, and I am not sure where they might have put it now. A couple of people over on l-camera-forum used to dig around in this kind of information. I am sure someone over there would know. I wrote a TIFF reader once (apart from the image data, I was just curious about all the meta data), but I don't even know where the code is any longer.
sebboh wrote:
there's plenty of direct comparisons showing less color shift on the a7 vs the a7r. there's more confounds with comparisons of smearing between the two (in the comparisons i've seen) so i'm not so certain of that.
I just didn't keep score well enough to recall seeing the same lens on BOTH the A7 & A7R to be able to make such a comparative assessment for myself as to which one had more color shift (same lens / same lighting). I just saw enough to know that we still have no real overall solution ... i.e. relegated to case by case @ degree by degree.
That being said, I realized I missed something along the way that would have others saying the A7 performed better than the A7R. Just thought I should fill in on what I missed in that regard.
michaelwatkins wrote:
Except that both cameras have ~20mm more depth between flange and sensor than the A7 / A7r and isn't that the key difference here?
I wasn't comparing the A7/A7R to the D800E/D700 ... yes those are different flange/sensor distances that have different outcomes +1 @ the trig. That wasn't what I was trying to say when asking about why folks are suggesting the A7 is performing better than the A7R. I meant that if someone said that a Distagon performed better on the D700 than the D800E ... it would also raise the question of why someone would say the D700 would possibly outperform the D800E?
I was inferring that their is pixel size difference between the A7 vs. A7R, similar (not exact) to the pixel size difference between the D800E and the D700 (or more closely the D600). Is the pixels size difference in in any way contributory to the smearing or color shift issues?
Rusty, thanks, now understood. Maybe it's the glass of wine I just finished but I still don't feel any more informed as to "why" the difference between A7 and A7r when with the D600 and D800 you would not expect to see such profound differences in performance.
Is it trig? Is there a point where moving the lens X distance farther out makes everything better/easier/generally workable?
Notes by shooter, Leo Roos:
"Raw converted jpeg; F13, focus @ 1 mtr
Standard processing in CS 6
No purple edges; slight blur at the extreme edges
But in all in all excellent performance for this UW WA RF lens!"
p.114 #10 · A7/A7r - performance with WA RF lenses
wfrank wrote:
Sorry Joakim to pick a sentence from your normally enlightening posts but that is pure bollocks.
This thread is about comparing RF WAs on the A7 and the A7r. There's absolutely nothing showing that the presence of offset microlenses in the A7r makes it better than the A7. On the contrary.
So you've seen what the A7R would do if it had microlenses with less offset? Because that's what's it's all about... Not that the smaller pixel camera would in any way be better than the larger pixel camera (in this case).
There's a huge difference in pixel structure height vs active pixel aperture width between the A7 and the A7R. Unless they added more offset, the difference in color shift would be even bigger.
So I do think you're missing the point here, totally.
p.114 #11 · A7/A7r - performance with WA RF lenses
michaelwatkins wrote:
Rusty, thanks, now understood. Maybe it's the glass of wine I just finished but I still don't feel any more informed as to "why" the difference between A7 and A7r when with the D600 and D800 you would not expect to see such profound differences in performance.
Is it trig? Is there a point where moving the lens X distance farther out makes everything better/easier/generally workable?
As you move the exit pupil in towards the sensor (shorter flange distance, semi-symmetrical WA lenses), the diagonal angle from "left" exit pupil edge to "right" sensor edge grows dramatically.
Since the D800/600 both have a realistic absolute minimum exit pupil distance of about 40mm, and this necessitates distagon-type WA lenses, that distance grows even more.
You could basically say that the minimum focal length of the rear-end part (main) of the lens is about 50-55mm for a DSLR type camera. To get a WA you then have to make a 50mm lens with a wide converter built into the front. And for image quality, flatness of field, aberration correction and so on this is often better - though it necessitates a more complex construction. A cheap (but complex!) retrofocal is often worse than the same build-price symmetrical if the image recorder isn't a problem.
RF or MILC cameras don't have that rear end distance restriction, so those constructions tend to push the angles harder - the lenses get smaller, and also cheaper to make. But then they put a higher angle load on the sensor part of the camera system.
p.114 #12 · A7/A7r - performance with WA RF lenses
sculptormic wrote:
I have seen examples with the 28 cron on the M9 wide open and it does not look good!
You could call it smearing.
It gets better @ 5.6.
I think we're getting to critical around here.
Who shoots landscapes or other well chosen subjects wide open anyway?
OMG the voice of reason!
I shoot a million scapes with the 28 cron, 90% at 5.6 on the nex 5/n.
When I shoot wide open it's indoors usually (not always) and I never think to check a corner.
I cannot believe--well yes i can---how people look at the chinese smog shots I found with that lens showing minute detail at the edge @ 5.6 through the smog, midday, and say "Oh, that's not right!"
No smog here, usually, so I'll know soon enough.
as to the seemingly obvious "the A7 is better!" we aren't going to know really untill we see some really good tries, with really good PP.
People see the colorshift and cry OMG look!
I can tell you, the original nex 5 shifted colors like mad on the CV21/4, while the 5n did not. Were the edges better on the 5n? No. Not in my tests anyway.
PP is a pain, so that might be reason enough to pick the A7, and I'm sure there must be some lenses which smear less.
To early to confirm which.
The CV 12 would be a good test. I'd like to see it @ f/13 on the A7r and check the edges-- Lee ROos shows they are decent on the A7 above.
I've looked at that lens for years, but I'm too spooked to get it yet for the A7r, and I suspect there may be something native within a year.
I love the shot wfrank picked in his recent post, which is in Chile I believe. A7r does seem to love the 35/1.4 FLE, but I don't think it's near so friendly with the plain asph, forgive my not making this distinction earlier.
I would not have guessed this since floating elements not always great on the emount.
Nov 22, 2013 at 11:50 PM
Steve Spencer Online • • • • • • • Upload & Sell: On
p.114 #13 · A7/A7r - performance with WA RF lenses
uhoh7 wrote:
OMG the voice of reason!
I shoot a million scapes with the 28 cron, 90% at 5.6 on the nex 5/n.
When I shoot wide open it's indoors usually (not always) and I never think to check a corner.
I cannot believe--well yes i can---how people look at the chinese smog shots I found with that lens showing minute detail at the edge @ 5.6 through the smog, midday, and say "Oh, that's not right!"
No smog here, usually, so I'll know soon enough.
as to the seemingly obvious "the A7 is better!" we aren't going to know really untill we see some really good tries, with really good PP.
People see the colorshift and cry OMG look!
I can tell you, the original nex 5 shifted colors like mad on the CV21/4, while the 5n did not. Were the edges better on the 5n? No. Not in my tests anyway.
PP is a pain, so that might be reason enough to pick the A7, and I'm sure there must be some lenses which smear less.
To early to confirm which.
The CV 12 would be a good test. I'd like to see it @ f/13 on the A7r and check the edges-- Lee ROos shows they are decent on the A7 above.
I've looked at that lens for years, but I'm too spooked to get it yet for the A7r, and I suspect there may be something native within a year.
I love the shot wfrank picked in his recent post, which is in Chile I believe. A7r does seem to love the 35/1.4 FLE, but I don't think it's near so friendly with the plain asph, forgive my not making this distinction earlier.
I would not have guessed this since floating elements not always great on the emount.
I didn't say it before, but part of the reason I am not impressed by the Summicron 28 "smog" photo is that if you look at his photo stream, you can compare those 28 cron edges @ f/5.6:
That comparison makes it pretty clear that the 35 can still produce a very sharp image despite the atmospheric conditions and the 28 cron just looks pretty weak in comparison. I know it isn't fair to compare the centre from the 35 cron with the edges from the 28 cron, but he didn't frame the shots with the same centre focus point so it just turned out the centre of the 35 cron is at pretty much the same spot as the edge of the 28 cron shot. It would have been nicer if the centre focus point would have been the same, but it still suggests that the edges of the 28 cron shot could be quite a bit sharper and probably are much weaker than its centre just as they are weaker than the 35 cron's centre.
Oh, and if you look at the edge crops for the 35 cron they are unfortunately even weaker than the 28 cron:
So, unfortunately although the centre is excellent on the 35 cron it seems to have issues as well. These sorts of edges may be fine for some, but I am left less than impressed and I would still like to see some better comparisons between lenses of the same scene. Unfortunately, these shots leave a lot of question unanswered for me and are far from definitively showing that either the 28 cron or the 35 cron ASPH won't show any serious problems.
By the way his shots also show excellent performance from the 50 lux Asph, the 90 AA, and the 135 ASPH, so the conditions were good enough to show strong results from these lenses at both the centre and the edges.
p.114 #14 · A7/A7r - performance with WA RF lenses
sculptormic wrote:
Thank you very much!
I was very curious about this for a long time!
This looks quiet good. Is there an easy trick to get rid of the colour cast?
If so, I am going for it. I used it once and I like it a lot. Very little distortion as well.
I think all things considered it is pretty good too and quite useable with additional pp.
I tried the DNG flat field plugin in Lightroom to fix the colour shift but I keep getting green corners instead which is weird since it works fine with my Nex-7. Has anyone else had this problem? I think they may need to update the plugin. Cornerfix didn't work also. I need to investigate further perhaps.
Makten wrote:
Edit: OK, it was the new Flickr UI that screwed everything up. Now I know how to view the image at 100%, and it actually looks really good! Even better than I'd expect from a fast 21 mm SLR lens.
Overall it appears to outperform the OM24mm f/2.8 I tested it against. It's good to know that some rangerfinder WA lenses may work fine on the A7R.
sebboh wrote:
yes, you have to go to flickr to see full sized crops. they look bad but not really smeared at 100% till f/5.6 where they look decent to good for a $1k 21mm.
I like the fact that it does double duty on the wide open end too. The bokeh rendering is surprisingly smooth and aesthetically pleasing to view with nice colours.
naturephoto1 wrote:
The problems with the hood relate to the usage of filters including rectangular Grad filters, square or rectangular ND filters, rotating variable ND filters, etc. Additionally, if a round 58mm filter jams or is over tightened, it could become exceedingly difficult to remove.
Rich
I don't find it hard to untighten, there is ample grip for twisting on or off. With a circular filter, ND and polarisers aren't a problem. I use a graduated filter lightroom preset for those times I need that effect. I would rather not have the built in hood as it's fairly angular and doesn't feel that nice.
p.114 #15 · A7/A7r - performance with WA RF lenses
theSuede wrote:
Since the D800/600 both have a realistic absolute minimum exit pupil distance of about 40mm, and this necessitates distagon-type WA lenses, that distance grows even more. [...]
I think I just heard something click. Ok, I get how slapping mirrors (or not) and history have shaped choices in lens design.
Is it ironic that a slapping mirror forces a particular design style and but now that the slapping mirror is gone sensor size (and ? density or resolution ?) is forcing the same design decisions, only sans mirror?
p.114 #16 · A7/A7r - performance with WA RF lenses
I've been amazed at the performance of the A7r stopped down to f/16 and beyond. At RFF they laughed at me for this, saying "oh, the A7r needs a nobel for breaking the laws of physics"
"Diffraction blurring reduction function is whether the effect of how much?
- So, please tell me the features of picture making.
Hotokesaki: image processing engine will change to "BIONZ X", there was the evolution of the three. The first is the "detail reproduction technology". who ever felt image is look unnatural by contour is emphasized in digital photography might also ever been, but to raise the texture in the natural depiction while maintaining the sense of resolution this time .
The second is the reduction of diffraction by the aperture. I think you refine, such as landscape photography, had been concerned about the fact that sharpness is lost also come. This time, we have to reduce the diffraction blur by multiplying the optimal filter in accordance with the aperture value.
The third, coupled with detail reproduction technology, noise reduction is divided into areas that traditionally has improved both in terms of noise and detail.
Hotokesaki Jian, who was in charge of quality design.
- The detail reproduction technology do you have what kind of processing in particular?
Hotokesaki: I had that line, such as the border is attached to the edge of a simple edge enhancement of the past, the process has put a mechanism in the shallow, to restore the detail in a process different this time.
- That's called "process was different" Would whatever they?
Hotokesaki: I can not answer this for more information, but I give contrast local processing sharpness and adaptive processing in accordance with the subject.
- The diffraction reduction process followed, but the process of claiming the same effect in each company has begun to be adopted today. Do you have realized in any processing Sony?
Hotokesaki: calculates the filter in reverse to return the diffraction blur from the data of the lens, it has been applied to the image.
- The data of the lens, or would you use the data of design time or do you keep measuring individual lens in advance?
Hotokesaki: Based on the design data, we have optimized to confirm the influence.
- Of that data for correction do you are in the better of the lens.
: Hotokesaki please refrain more about it too.
- Does this also supports any lens?
: Hotokesaki you have any effect in any lens. I also supports A-mount lenses.
- If you want to get a deep depth of field, what all right in understanding that all right, narrow down any number of do not care about the diffraction blurring?
: Hotokesaki does not mean that any number. For example, in the case of lens best, at the same image quality as much the F5.6 to F8, F5.6 has kept a decent quality even F11. However, you will see the influence of diffraction slightly and squeeze until F16.
- If you do so, what all right even squeeze in extra is about two-stage from the F value of the diffraction blur acceptable yourself?
: Hotokesaki Yes. I think because there is a place by the pixel size of the sensor and the characteristics of the lens of what stage improvement numbers, it is good when I have to understand the degree to "image quality degradation is a concern squeeze" as there was prior art to be improved."
A7r may be new benchmark for f/8-f/22.
Which is highly convenient, since we may need it for the edges on some lenses
But on close reading I think this may not apply to MF glass.
p.114 #17 · A7/A7r - performance with WA RF lenses
sebboh wrote:
i put a kim wipe (kleenex would be fine i'm sure) in front of the lens using a rubber band to secure it uniformly and flat then point the camera towards a clear blue sky and shoot through the aperture range.
yes, i'm using the adobe flat field plug-in.
there's lots of ways to record the aperture in the field, but i don't do any of them.
the easiest way i can think of is to use hand signals and just take a picture of your hand before or after the shot making the appropriate signal.
if you don't remember you can always just guess and then change profile pics if it doesn't look right.
I ordered a piece of Medium White (#2447) Plexiglas from Amazon. It's similar to what's used in the professional LCC filters. Just hold it in front of the lens. $7 buys enough to make 9 of them!
PS. There are two focus points in infinity test (1.At Center Building 2.At Lower-Left Building) in order to present clearer concept of "corner smearing".
Notes by shooter, Leo Roos:
"Raw converted jpeg; F13, focus @ 1 mtr
Standard processing in CS 6
No purple edges; slight blur at the extreme edges
But all in all excellent performance for this UW WA RF lens!"
Be sure to check the full size at my Flickr account.
I tested the 12/5.6 on the pre-production a7R too and apart from the magenta edges, it was not bad at all; so I wasn't surprised by how clean and good this shot turned out on the 7.
