Jim,
My CV15 was shimmed by the previous owner, and it still focuses past infinity. If I shim it to have the correct infinity stop will that improve its performance all around? I'm not familiar with how this would affect field curvature, but it sounds like you're saying that it does (?)
Jacob D wrote:
Jim,
My CV15 was shimmed by the previous owner, and it still focuses past infinity. If I shim it to have the correct infinity stop will that improve its performance all around? I'm not familiar with how this would affect field curvature, but it sounds like you're saying that it does (?)
Jacob, that would imply your M-NEX adapter is too thin. Especially with the CV15, sharpness degrades quickly past infinity.
I gave the thickness of my M-NEX adapter, as an example or guide. I also used lenses that I know are correctly calibrated to Leica M cameras, to shim the adapter. So your CV15 could be focusing past infinity and your adapter could too, for that matter.
If anything, the CV15 should at the most, just barely get to the hard stop infinity focus, to get the corners sharper. Try it on objects several miles away with lens at f/4.5. You will find that the corners sharpen when you back off infinity focus. Then stop down to f/5.6 or f/8 or more to sharpen the center. It's a balancing act.
rscheffler wrote:
...
One thing to keep in mind is that the batch correction applies colour shift as well as vignetting removal. If you want to keep the vignetting but get rid of the colour shift, you'll have to process the files one by one manually in the CornerFix app.
douglasf13 wrote:
I could be wrong, but doesn't batch mode in CornerFix remember your last luminous and color settings, so you can set it to only remove shift?
Jacob D wrote:
I also thought that the batch mode used both the Chroma and Luminance settings. I guess I will have to give it a try to see what happens if I turn the Luminance setting to zero and do a batch conversion.
I could be wrong, but I thought I read somewhere in the CornerFix documentation that it did the full colour and vignetting correction when batched...
JimBuchanan wrote:
Jacob, that would imply your M-NEX adapter is too thin. Especially with the CV15, sharpness degrades quickly past infinity.
I gave the thickness of my M-NEX adapter, as an example or guide. I also used lenses that I know are correctly calibrated to Leica M cameras, to shim the adapter. So your CV15 could be focusing past infinity and your adapter could too, for that matter.
If anything, the CV15 should at the most, just barely get to the hard stop infinity focus, to get the corners sharper. Try it on objects several miles away with lens at f/4.5. You will find that the corners sharpen when you back off infinity focus. Then stop down to f/5.6 or f/8 or more to sharpen the center. It's a balancing act....Show more →
Thanks for this info. Coincidentally, it's how I've been using the CV12 and 15 lately on the GXR and the results seem fine. I'm pretty sure my CV12 is slightly off and the left side smears a bit, which I saw with the NEX 5N and somewhat less on the GXR, when the lens is focused at or near infinity. But if I back it off to around 2-3m, then the results are better...
Thanks for your thoughts. I believe my adapter is ok, or at least a couple of my CV lenses are fine with it. Coincidentally they are both M mount, the CV 35/2.5 and 40/1.4 both seem good @ infinity the hard stop.
On the other hand my CV 90, and CV 15 (both of which are LTM) focus past infinity when at the hard stop. The 90 just ever so slightly, but the 15 goes a little further past and it had already been shimmed (by someone).
My adapter is the Voigtlander brand, .382 inches thick on my dial caliper.
edited to add... also re. your advice on testing the infinity stop; with the 15 most everything miles away will appear fairly small in the frame. Is it necessary to use such distant objects?
Jacob D wrote:
edited to add... also re. your advice on testing the infinity stop; with the 15 most everything miles away will appear fairly small in the frame. Is it necessary to use such distant objects?
With the CV15, I suppose it is not necessary to use such distant objects, as I'm suggesting to focus shorter to sharpen the field. So the closer the reference infinity distance the better, in the case of the CV15.
Just a question for nex 7 owners:
is it worthwhile getting the N7 with expensive leica / zeiss lens considering 1.5 crop and color shift issue?
As for me I wouldn't pay the prime price for leica 35mm lens that will turn into 52.5mm because of 1.5 factor. 50mm is much cheaper than 35mm (just comparing summicron 35 vs 50). I'm not sure how Nex7 performs vs Nex 5N, but at least 5N didn't cause any color shift with Zeiss 21mm...
Sony Zeiss 24mm 1.8 is not sharp at 1.8, not my vote for this lens.
I guess I'm staying with M9 for now.
vovkinson wrote:
Just a question for nex 7 owners:
is it worthwhile getting the N7 with expensive leica / zeiss lens considering 1.5 crop and color shift issue?
As for me I wouldn't pay the prime price for leica 35mm lens that will turn into 52.5mm because of 1.5 factor. 50mm is much cheaper than 35mm (just comparing summicron 35 vs 50). I'm not sure how Nex7 performs vs Nex 5N, but at least 5N didn't cause any color shift with Zeiss 21mm...
Sony Zeiss 24mm 1.8 is not sharp at 1.8, not my vote for this lens.
I guess I'm staying with M9 for now. ...Show more →
I've purchased several ZM, CV and Contax G lenses since I started using NEX cameras. The Leica's haven't been quite worth it for NEX, to me, but I know of several NEX users that use and enjoy Leica lenses on the camera.
Thanks to Jim for saving me from returning my new m-mount CV15. I really disliked the Sony 16 (which also had some purple issues in addition to fairly weak optics...), so I bought a CV15 figuring I's someday own a NEX-7n or similar that treats WA RF's better.
Anyway, I snapped it on my trusty Fotodiox M adapter, which focuses exactly at infinity with my ZM 50. I focused at infinity, got lots of focus confirmation, and shot samples at f/4.5-f/8. After cornerfixing, they were all pretty blurry in the corners, as well as the outer third of the image circle. Overall, I'd say on a par or worse than the Sony 16. Ugghh. Then I recalled Jim's post about curvature, and reshot the sequence focused at infinity, 3m (on the barrel marks), 2m, and even 1m. All the shorter focused shots were drastically better in the corners, and gave up a little at the center of the frame. I decided f/8 looked a touch better than f/5.6, and the optimum compomise seemed to be at a 2m focal point. Here are the results:
First, the f/8 & 2m shot out SOOC -- very ugly!
http://sheltons.net/temp/CV15/DSC01487.jpg
Here's the final result after cornerfix and some LR4 work:
http://sheltons.net/temp/CV15/DSC01487_CF.jpg
Here's a 100% view of the center focused at infinity:
It's not perfect, but it's pretty sharp IMO. The 1m focused shots are even sharper in the edges at the expense of some center sharpness. I suspect focusing approach explains the wildly different opinions of tis lens on the NEX-7.
Shelt, keep in mind that if you adapter nails focus exactly on one lens, it doesn't mean that it will on others. My various M lenses very quite a bit in regards to infinity focus. My ZM 35/2 seems to nail focus with my new Hawk's adapter, but everything else focuses past infinity to some degree.
Either way, it seems that astigmatism is exaggerating field curvature with the NEX-7 more so than with the 5N.
You say you weren't satisfied with the 16/2.8, but I'm not sure I see anything in that shot that is any better than the 16mm @ f/8. The corners certainly aren't better.
What really bothered me with my copy of the Sony 16 was the lack of resolution in the center 2/3 of the frame. I found a small circle near the center that was sharp, and that was it. The outer third-half of the frame was terrible, including the corners. I tried to live with it, but gave up. To be fair, I hadn't thought of MF on the corners to see if that helped.
The CV15 is better across most of the frame. I agree the extreme corners aren't the best, but if you look at the full res photo, the overall frame is very nice. I find myself pixel peeping a bit too much with 24 megapixels as well -- the corner crops are pretty small portions of the frame.
Here's a Sony 16 crop, shot at f/8, just away from the frame center. It looks as bad as the extreme corners on the CV15:
http://sheltons.net/temp/CV15/DSC00225.jpg
The CV15 may be the best of a bad situation, but it will do until the NEX-6 or NEX-7n arrives... I'm guessing this lens is really sharp on the 5n, whereas the Sony 16 is probably still pretty rough.
Weird. I've found the 16/2.8 to be sharpest at the center, too, but by f/5.6 it's acceptable to the edges. It's really hard to compare unlike images, but by f/8, the corner problems don't seem to extend as far into the frame as what you've shown here with the CV15.
The shot you added from your 16 doesn't look like what I'm getting. It may still not be as sharp as the CV15, but it's not that bad. I can see why you were frustrated.
I'm in the middle of a LR upgrade, but I can follow up later with some images, if you like. Somewhere in this forum are a series of sample shots I posted from the 5n, and I haven't seen any major differences with the 7.
Pixel peeping on the 24mp sensor can be maddening, that's for sure!
I'm glad my experience with the Heliar 15 & NEX7 has helped in getting the maximum image quality out of the lens. It is a very unique lens. I can't say enough about M lens adapter thickness and focusing /exposing for the corners, especially in the case of the Heliar 15.
I may not have much else to contribute on this thread because in the end, an effective 22.5mm FOV is a bit wide for my style, and I have sold my lens. There is a bit of seller's remorse, because with optimized use and Cornerfix, it is a workable NEX7 solution. But the high pixel count demands and max aperture of f/4.5 suggests a large fraction of a second shutter at f/8 or so, and this requires a tripod.
Actually, almost all 15-18mm lens solutions with max aperture f/3.5 (RF or SLR), on the NEX 7 camera will require stopping down to get the corners and cost ISO without a tripod. So, this may be the wall a crop sensor NEX7 user will run into when better quality wide angles are required, full format or not.
NEX-7 and CV-15 is an interesting combination regarding performance. The reasons why the edges and corners are suboptimal is in the NEX-7 sensor pixel pitch (increased electrical crosstalk (electrical and optical) -> smearing, reduced edge QE, increased difference in QE of different wavelenghs of light), quality of microlenses (the same QE-issues), thickness of the IR&AA filters (again, the same QE issues and some optical crosstalk), general performance level of CV-15 (field curvature) and the position of the exit pupil.
In practise this lens is on the very edge of either being a total disaster on NEX-7 and being a stellal performer. For optimal performance one should try to keep the exit pupil as far as possible from the sensor (focusing closer moves the exit pupil away from the sensor) and close the aperture as much as possible (f/8 seems fine at infinity, wide open at 1m or 2m) - doing these two forces all the light rays to hit the sensor as straight as possible. However, the magenta fringing won't disappear, so CornerFix is still needed, but resolution near the edges should be nice with care.
(Also one should make sure that the lens isn't focused beyond infinity when at infinity setting.)
