p.17 #1 · A thinner sensor stack may be possible after all!
hiepphotog wrote:
I thought RF calibration deals with cam action to make sure that the RF moves accordingly. I don't think any of the first party can guarantee a highly precise distance from the sensor to mount surface, let alone these adapter makers to get a precise thickness across their production.
So that is why adapter like Hawk's and Phigment would be of great value (infinity adjustment). Unfortunately, they tend to loosen over time and would cause mount misalignment, especially if you do a lot of helicoid extensions.
The problem I'm having right now is that Michael said his Novoflex adapter is spot-on. But Novoflex officially state that they produce shorter adapter on purpose to accommodate camera mount and lens mount variations. For the first time, I hope my Rayqual is not too accurate . ...Show more →
You are right, but there is a limitation on how much you can accommodate for the cam action, so some lenses do need to be shimmed to be spot on. This does not affect the A7 series other than there may be a variation in the infinity stop position. I am not concerned about the infinity stop going over but with the Metabones I had a number of lenses that did stop short.
p.17 #2 · A thinner sensor stack may be possible after all!
charles.K wrote:
I would just add with respect to the infinity stop calibration which affects both M9/M240 and A7 series. Generally the M lenses are calibrated with compromise to fit within the envelope of tolerances. The mount will vary and with a number of my M lenses, I sent back to Camera Clinic when I had the M9/M240/MM to match the infinity stop. Some cases it stopped short, and some were over. Every time the near, medium and far distances were calibrated at different f stops, there was a compromise to suit the best style of shooting. Sometimes the lens was well outside the tolerances and the cam iteself needed to be modified or replaced in Solms.
I do think once the "thin filter" mod evolves, then having the adapter matched to a group of lens/lenses would be a great to eliminate to question of de-centering and finding the optimum flange width.
There is a "correct" way to to calibrate Leica lenses at infinity (on Leica M bodies). The one popular exception is the ZM 50/1.5 due to the focus shift when stopping down. Yes, it does make the adapter situation a little tricky when using the M lens on both RF and E mount cameras.
M mount lenses are designed around a reference hard infinity stop and that hard infinity stop is used to register the M rangefinder cam mechanism to accurately focus through the distance range down to a typical .7 meter close focus. The lens is designed to have the least focus shift when stopping down to improve focus accuracy at any f-stop, to have a helical that moves a fixed distance between infinity and .7 meters to allow the camera rangefinder to be accurate, to be small and not obscure the viewfinder, and numerous attributes. So, there is a reason M lenses don't focus closer than .7 meters. That is the reasonable close range distance that a rangefinder can handle. As a side note, I would not count on premium optical performance on M lenses below the .7 meter close focus distance.
With the above as a premise, I would point to the M-to-E adapter as the key to using several M mount lenses on a Sony E mount camera. Otherwise, the solution is to have one adapter for each M mount lens and infinity adjust by shimming each adapter individually. I received a Leica M lens from Leica service that I used as a reference for the infinity hard stop. I then shimmed a good quality adapter that gave exact infinity on a factory Leica camera. I have referenced the dimensions on this site previously. With a proper adapter, it will be evident if the M lens itself, needs shimming for infinity on the E mount camera. This maybe the case for older M mount lenses, more so than recent lenses. And back to the ZM 50/1.5. By "optimizing" that lens to focus accurately at f/1.5 at close focusing, infinity focus is compromised. Not nessecarily a bad thing, as infinity shots are usually stopped down and of course the ZM 50/1.5 is not known as a lanscape lens anyway.
I would also find confidence that the distance between the Leica M camera flange mount and sensor has a high degree of accuracy with the robust magnesium body, after all the system's focus accuracy depends on this deminsion. With a set of M mount lenses that adhere to the rangefinder design parameters, the camera body focuses them accurately, but if not then the camera has the cam and cam arm that is adjustable. However, if there is a good hard stop infinity on a Sony E camera, this is immaterial.
As long as the Sony E to Leica M adapter has uniform thickness, it will not contribute to any decentering of the M lens. I do acknowledge that if a "thin filter" mod of a Sony E camera doesn't place the sensor in a parrallel plane to the camera mount, then there could be lens decentering issues.
p.17 #3 · A thinner sensor stack may be possible after all!
JaKo wrote:
Not sure what is the end percentage needed for Leica to badge their lenses 'Made in Germany' but aren’t their lenses mostly made in Portugal?
M lenses are made in Germany (somebody correct me if they know better) Alot of the M240 is made in Portugal. My newer M builds are far superior to the Cosina ZM, though I love those lenses too.
As to lenses "falling apart" this what happens when a wobble, whether in a ZM 35 or SEM 21 or 28 Cron is not addressed. I've had DAG fix several lenses which had this "problem". It's just interior screws which need a tighten, but to get to them you need a special tool to un screw the ring around the rear element. " target="_blank" rel="nofollow">Here you can see how it's done
I like Cosina, but I do find it bizarre that Zeiss does own a camera lens facility. Or maybe they do? Isn't the 85/2 and 15/2.8 made in Germany?
@toothwalker Yes it seems obvious Zeiss would be better making a design from scratch for the A7 series, but obviously that would have been more expensive.
A good part of my resentment on this issue is that the A7 was specifically marketed as a camera good with many lenses. "Where are the native lenses?" we asked in the beginining. "Oh, you can use all sorts of lenses, don't worry" This was coming from all sorts of sources in Sony marketing and dealerships. To be fair, they warned about RF wides. But Brian Smith http://briansmith.com/blog/
Who had a ZM18 and A7R before the rest of us, assured me on twitter the lens was good edge to edge on the A7r. Of course he is a sponsored Sony shooter.
How easy would it have been to just make a camera friendly to film glass, or at least one version? Very Very easy.
Now they are stuck on a one way street, though with many fans, like Huff, who simply ignore the issue. What the Loxia proves is this system needs special formulas to work at full potential. What a pain for third party lenses makers. What a pain for those invested in lenses they like. And even worse, way way worse, is the need to make big lenses, kept SLR distance from the sensor. So the whole coup of a compact full frame is compromised.
Why? To save us from dust? To save us from moire? (have yet to see it with A7.mod) Ridiculous.
On a more positive note, imagine the A7s with this sensor mod. These little CVs might really come to life
p.17 #4 · A thinner sensor stack may be possible after all!
uhoh7 wrote:
M lenses are made in Germany (somebody correct me if they know better) Alot of the M240 is made in Portugal. My newer M builds are far superior to the Cosina ZM, though I love those lenses too.
At least some of the lenses are also actually made in Portugal, which you can see in the below video. Perhaps final QA or adjustment is performed in Germany (however little is required to be able to mark the lens "Made in Germany").
As a side note, some of the best lenses I have used were those from Canada. I do wonder if this whole "Made in Germany" marketing scheme where bodies from Portugal are simply re-assembled and such in order to use the "Made in Germany" tag might be backfiring on Leica a bit. I would rather them just be more honest and upfront. If something is actually Made in Portugal, state such. Dont' try to get around it using silly procedural/ legal definitions.
p.17 #5 · A thinner sensor stack may be possible after all!
uhoh7 wrote:
On a more positive note, imagine the A7s with this sensor mod. These little CVs might really come to life
I have wondered a lot about this. I don't own an A7s, but it seems that, assuming it has a stack with the same optical properties (necessary for the designed-for lenses to all work optimally), the only reason the A7s works 'better' (from a sharpness perspective, not considering color fringing, which we know is the sensor design) is that it just has lower resolution so the smearing is less obvious (i.e. smeared over fewer, larger pixels).
Can anyone here confirm or deny that with actual experience and images?
p.17 #6 · A thinner sensor stack may be possible after all!
IMO, the a7S is equally bad. I have tested all three with a whack of RF lenses, and compared to either an M9 or M240 'reference' body, all of the unmodded Sonys underperform with most 35mm and wider lenses, and some 50s, too.
But this is in reference to infinity distance. For example, the 28 Cron looks particularly bad here, yet I used it a lot on the a7S for nearer distance work and thought it looked OK...
That link is for the a7S test and each lens links to 12MP full rez files of a cityscape near infinity scene (M240 files downsized to a7S dimensions).
Charles, interesting points. When I did some of these a7 tests, I was also given a Metabones adapter to try, along with the Novoflex I own, which focuses a touch past infinity with most lenses I tried on the various a7 flavors (though my impression is it was better with the a7R and a7S than the earlier a7, which was a preproduction camera)... Anyway, the Metabones did not allow lenses to reach infinity. It was especially problematic the wider the lens. Oh, and the fit was so tight to the lenses that I was afraid I was going to break them (some of the Canon LTMs are rather delicate feeling), so I gave up after a couple lenses and stuck with the Novoflex.
p.17 #7 · A thinner sensor stack may be possible after all!
Thanks Ron. That's what I would have expected. I was just surprised by all the talk about how 'friendly' it was. Could not figure out why that would be true.
p.17 #9 · A thinner sensor stack may be possible after all!
Now that there are some modified A7 out there, could any of you check for internal reflections? I'd like to know whether there has been an improvement or it's even worse. Similarly, could you check for field curvature? It's been said that it's induced or worsened by the glass cover. Is it really true? Is it flatter?
Thanks so much in advance,
p.17 #10 · A thinner sensor stack may be possible after all!
nicoimages wrote:
bump - I am waiting for my A7R to come back from Kolarivision - do any of the early adopters have any more test images to share? - Thanks Nicholas
No, but I'm very interested in hearing your feedback
p.17 #12 · A thinner sensor stack may be possible after all!
I am not too sure about test shots, but images with the A7Mod and A7rMod are appearing in the normal A7's thread. Both Jack and Charlie have illustrated with numerous shots so far
p.17 #13 · A thinner sensor stack may be possible after all!
My A7 is back. I'll be posting some photos in the A7 thread shortly. I did a few quick tests compared to the A7ii yesterday, and the edges and corners are definitely better with the mod, using the CV15, LM21/3.4 and LM28/2. I'll try to post examples here, though they weren't scientific tests (they were going through a window, in fact).
(I had a brief heart-attack when unboxing because the diopter correction on the EVF was out of whack, so I thought they screwed up the focus adjustment! ).
p.17 #14 · A thinner sensor stack may be possible after all!
So, one thing I'm wondering about: it seems that the white balance leans pretty heavily towards the green now. I also have noticed this in Charlie's shots posted with his modified A7. I guess the new cover glass has a different cast than the old?
p.17 #15 · A thinner sensor stack may be possible after all!
Taylor Sherman wrote:
So, one thing I'm wondering about: it seems that the white balance leans pretty heavily towards the green now. I also have noticed this in Charlie's shots posted with his modified A7. I guess the new cover glass has a different cast than the old?
Just in case, please check all your camera setting and make sure to adjust back to what you had prior conversion. Kolari makes changes to setting for their conversion/testing so make sure you have RAW, instead of JPG, correct colour balance, mode, etc.
p.17 #16 · A thinner sensor stack may be possible after all!
Taylor,
Are you seeing that with AWB or (for example) daylight shots with Daylight setting? I generally use AWB since I process from RAW, and have not seen anything that strikes me as a problem. I've been meaning to pull out my strobes that I know are 5500K and check it, but have not gotten round to that yet.
Certainly it is possible that the WB changes slightly since the stack shapes the response in the visual range to compensate for the non-flat sensor response. Sony would fine-tune the firmware for the combined response of the factory combination in the preset WB modes.
Michael
Post Script...
I did a quick test using a tungsten light source, shooting in my bathroom where I have a white sink. Set white balance on Tungsten, took the RAW file into light room and used the WB eyedropper on the sink. Your observation seems correct...
Lightroom shows the 'As Shot' at 2750K and +5 toward Magenta on the Green-Magenta slider.
Balancing on the sink moves to 2650K and +16, slightly more Blue and more Magenta (less green).
The eyedropper adjustment looks right on my calibrated monitor. I don't have an un-modified camera handy to compare, but I think it's likely there is a slight shift.
You can adjust the presets with the A-B and G-M fine tuning adjustments in the Sony. At least on this quick test, shifting +2 Magenta and +2 Blue in the Sony bring the 'As Shot' and the eyedropper settings into line.
p.17 #17 · A thinner sensor stack may be possible after all!
I noticed it using the Daylight WB outside. Though when I picked the LR Daylight WB I also had to adjust quite a bit towards the magenta. It's possible there was just something about the light today?
Hmm, also, I have been using the Huelight A7 Standard color profile. Maybe that is related. It is possible I have just forgotten what this profile is like, since I've been using the Mk2 for the past month and there isn't a Huelight for that so I've been using Adobe Standard.
This picture is set at 5450K, +33 Magenta in LR5, but with the Huelight profile. When I switch back to Adobe Standard it is much less green.
p.17 #19 · A thinner sensor stack may be possible after all!
That looks nice and sharp Taylor
Check all your settings, as many of mine had been changed---but no worries, that just shows they were testing it. I needed to re learn the camera anyway.
I also normally shoot AWB.
Here is ZM 35/2 WO: DSC01842 by elleliv, on Flickr
Derek.....you still think the Kaaragon has bokeh issues? It's a matter of taste but I like that above
I'm no longer used to the stock A7 WB, so am a poor judge. I think it is certainly possible the thinner cover has an effect, and nice to see it can be adjusted for
I also note that electronic front curtain shutter does not like some lenses, might possiblly be more sensitive than before, or not. Anyway with 50 and up I recommend you compare at least or just turn it off.
Anyway, I love the mod. For me it seems a different camera. Still CMOS, still with various annoyances of course, but what a relief to slap the ZM 35/2 on there and not worry How they managed to bury items like the EFCS, and even the format command so deep in the menus, with no shortcuts allowed, I will never know!
I'm also so relieved I did it because my M9 needs a new shutter That would have been especially brutal without the A7.mod
One last with the biogon f/4 or faster: DSC02073 by elleliv, on Flickr
Note edges Not a definitive test or anything, but there they are, and pretty fast too.
.
It's a matter of taste but I like that above 
That would have been especially brutal without the A7.mod 
