The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Leica 24 3.8 on sony a7 testing

karlfoto

New member
Hi forum

I did these tests about mid april with my a7 and leica 24 3.8 to see whether i could live with the soft corners for my landscape work which is predominantly shot while multi day backpacking (hence the desire for a small kit).

The leica i still have, but the a7 is up for sale as i am happy with the sony platform for this application, but want to get the a7rii when the price becomes a bit more realistic.

Just some info on the testing for the 5 street scene images. The camera was on a tripod, using a cable release. The lens adaptor is a voigtlander VMII. I just set the focus as per the indicators on the lens at f11. There are other images that i took where i manually focused using the focus magnification through the efv at the bottom centre and upper left of the image, however these 5 images worked out to be more accurate for the test. I processed them through C1 using an LCC and them sent them out as tiff files with the std C1 sharpening. I then pulled them into CS enlarged them to 100cm and gave them a final USM of 150;1.2;0. These are the 100% views. I often print to 100cm so the viewing distance would be about a meter.

I eventually settled on a point about 1.8-2m as being the ideal focus point to get the best corners and good sharpness throughout. 3m is better for the central point, however with the USM on the 2m image i don't think it would be to noticeable on a print. (i have not printed these, just pixel peeped on the screen so am not sure if the sharpening has been to aggressive)

I also tested using helicon focus and took three of the different focus points and stacked them that result was also quit good.

The setup for the image of the tent and mountains was exactly the same as above and the lens was taped at 2m and f11 to stop accidental shifting. Process was the same, however in C1, i added sharpness falloff of 150% to one of the images and not to the other to see the difference.

I know that some might see this is being a bit ocd, and that soft corners dont really matter. However, I have always shot my scenic work with the intent of getting the whole image sharp. The images that i got from my pentax 67ii where always exceptional, and this is what i would like to see from a small camera package like this.

So are these acceptable corners at f11 without the kolari mod? Ultimately, I want to get the A7Rii and i wonder if using a similar set up i would get a similar result. I dont usually shoot below f11 for the scenic work but sometimes with my canon and wa lenses I do shoot open and try to throw the focus.
 

Attachments

spence

New member
Wow...if these are 'soft corners', sign me up! I've never shot Leica, but these look damn good to me.
 

uhoh7

New member
Sorry can't tell a thing from that test shot.

Any kind of close foreground is going to disguise the FC induced by the sensors.

Try a proper infinity landscape where the corner details are at least several hundred feet away, for example:


a7M_CV21_11 by unoh7, on Flickr

This is why there is so much confusion about lens performance with the Sonys. Many highly misleading test shots which seem to show a "good" corner but that's because the "smearing" is a matter of focus, not "softness". You can't get both a good center and a good corner UNLESS the corner is closer. Which it is in your shot.

Many wanted the SEM 24 to work on the A7, I never heard of anyone satisfied with it's general use.

However the 400USD Kolari mod does restore 90% of the lost SEM 24 performance. Here is a similar lens, the SEM 21 on the kolari, at f/4 no less. Near WO!


a7m_sem21_f/4 by unoh7, on Flickr

I doubt any UWA in the world on a stock A7, any flavor, can beat this performance.

But the point is, when testing any sony A7 WA or UWA, your details must be at least this distant in the foreground and at least one corner. I marvel this seems to not be clear after two years of back and forth about what lens can do what on the camera. People who know the cameras pretty well otherwise (better than me), besides this well-meaning OP, still present test shots which tell nothing about the true problems. They mean well, and go to trouble to show them, but in the end it's even worse than naught. People, unintentionally, are misled.
 
Last edited:

karlfoto

New member
Hi Uhoh

Thanx for your input and the image examples. Please remind me what you mean when you say 'FC induced by the sensors'.

So what i take from your reply is that if I want near and far sharpness i.e. rocks in the immediate foreground and mountains at infinity, then my present setup should be ok, but if the foreground is several hundred feet away i.e. a huge chasm in front of me and then mountains in the distance, the rocks on the other side are going to be soft or smeared, or is that to simplistic an understanding.

I suppose that 98% of my wa scenic images are shot with the foreground details right up close, not at a great distance and this was the motivation for the test that i did. The examples you mentioned, I have seen before however perhaps did not register that the f/grnd objects were closer to infinity.

I did mention though that i had to shoot it at f11 and that any work that was shot near open would not really be possible. That is somewhat of a limitation to creativity that is true.

So it looks like its either the kolari or the batis.



Sorry can't tell a thing from that test shot.

Any kind of close foreground is going to disguise the FC induced by the sensors.

Try a proper infinity landscape where the corner details are at least several hundred feet away, for example:


a7M_CV21_11 by unoh7, on Flickr

This is why there is so much confusion about lens performance with the Sonys. Many highly misleading test shots which seem to show a "good" corner but that's because the "smearing" is a matter of focus, not "softness". You can't get both a good center and a good corner UNLESS the corner is closer. Which it is in your shot.

Many wanted the SEM 24 to work on the A7, I never heard of anyone satisfied with it's general use.

However the 400USD Kolari mod does restore 90% of the lost SEM 24 performance. Here is a similar lens, the SEM 21 on the kolari, at f/4 no less. Near WO!


a7m_sem21_f/4 by unoh7, on Flickr

I doubt any UWA in the world on a stock A7, any flavor, can beat this performance.

But the point is, when testing any sony A7 WA or UWA, your details must be at least this distant in the foreground and at least one corner. I marvel this seems to not be clear after two years of back and forth about what lens can do what on the camera. People who know the cameras pretty well otherwise (better than me), besides this well-meaning OP, still present test shots which tell nothing about the true problems. They mean well, and go to trouble to show them, but in the end it's even worse than naught. People, unintentionally, are misled.
 

uhoh7

New member
Hi Karl:

The thick cover glass over the sensor induces "FC" i.e. a field curvature, in short register wider angle lenses especially, but can be seen also in SLR lenses 35 and wider to a lesser extent.

I also like a nice close foreground with details in many of my UWA shots, but as a test for lens performance on the Sonys, the FC will wrap around and it's hard/impossible to tell what is what in such a shot.

Shoot a distant flat field, with focus alternately on the center and edge, and one in between. That WO, f/4ish, F5.6, 8 and 11. Then you can really evaluate the lens.

Critical focus without an infinity stop is not easy and can also mislead. :)
 

Amin

Active member
Hi Karl:

The thick cover glass over the sensor induces "FC" i.e. a field curvature, in short register wider angle lenses especially, but can be seen also in SLR lenses 35 and wider to a lesser extent.

I also like a nice close foreground with details in many of my UWA shots, but as a test for lens performance on the Sonys, the FC will wrap around and it's hard/impossible to tell what is what in such a shot.

Shoot a distant flat field, with focus alternately on the center and edge, and one in between. That WO, f/4ish, F5.6, 8 and 11. Then you can really evaluate the lens.

Critical focus without an infinity stop is not easy and can also mislead. :)

I don't agree with this. Distant flat field tells you one thing, close foreground tells you another. Neither tells you everything. I find a close foreground test to be the more useful of the two since most images I would take with such a lens will involve a close foreground.

That said, one has to try multiple focus points and f-numbers to know which one is best for a given lens and scene given the real world effect of field curvature.
 

uhoh7

New member
I don't agree with this. Distant flat field tells you one thing, close foreground tells you another. Neither tells you everything. I find a close foreground test to be the more useful of the two since most images I would take with such a lens will involve a close foreground.
You miss the point, sorry. We are not doing a generic "lens test", in which all sorts of factors are in play and you might even shoot a chart.

We are testing a particular issue with the Sony sensors, which is heavily documented. There is only one way to do that: a distant, detailed flat field.

I've learned this the hard way.
 

karlfoto

New member
Of course, field curvature, doh. It is not something that i have really paid much attention to before i arrived at the sony a7/leica lens discussions.

Perhaps you could explain what you mean by this comment. 'Critical focus without an infinity stop is not easy and can also mislead'. I also shot the lens at its infinity stop, but noticed that the sharper images were closer to the 5m engraving mark, and that at infinity point, it seemed softer. Would this mean that the lens with this adaptor is focusing past infinity?

Being in Australia, the kolari mod is a bit more difficult to organise and if there are any issues with the mod, the backwards and forwards would be painful. I suppose that i was looking for a way out of that option. I also shot this test against my canon 24 IS USM (not the sharpest canon lens i know), and while the canon was sharper in the corners at f11, it was not much, and I wonder if it would be noticeable on a print at the correct viewing distance.

It would be better to have a lens that can shoot from almost open all the way to f11, but as i usually only use f11 this might work in the interim and wait for the supposed wide angle loxia that is rumoured. I should print some images and see if i am happy with the results.
 

uhoh7

New member
Of course, field curvature, doh. It is not something that i have really paid much attention to before i arrived at the sony a7/leica lens discussions.

Perhaps you could explain what you mean by this comment. 'Critical focus without an infinity stop is not easy and can also mislead'. I also shot the lens at its infinity stop, but noticed that the sharper images were closer to the 5m engraving mark, and that at infinity point, it seemed softer. Would this mean that the lens with this adaptor is focusing past infinity?

Being in Australia, the kolari mod is a bit more difficult to organise and if there are any issues with the mod, the backwards and forwards would be painful. I suppose that i was looking for a way out of that option. I also shot this test against my canon 24 IS USM (not the sharpest canon lens i know), and while the canon was sharper in the corners at f11, it was not much, and I wonder if it would be noticeable on a print at the correct viewing distance.

It would be better to have a lens that can shoot from almost open all the way to f11, but as i usually only use f11 this might work in the interim and wait for the supposed wide angle loxia that is rumoured. I should print some images and see if i am happy with the results.
Very few M adapters have an accurate infinity stop. If any. The Hawks CF adapter has a way to adjust the stop, that's most desirable. The lens is a shadow of itself on the A7 cameras unmodified. You love the lens, and you have every reason to do so, as it is probably the best 24mm lens ever made. With an M9 you would be in heaven. You know they can be had now for as little as 2200? And Leica will replace the sensors for life?

Now there is the BATIS 25/2, which I don't think is the equal of this lens, but on the Sonys it's far and away the best option, and very fast too. It's easy to find a 24 SEM and they are getting cheaper. If you can handle the BATIS footprint and want to keep the A7 as is, I would trade the SEM 24 for a BATIS. Then later you can get a M body, which is really the way to shoot these lenses, and is getting more affordable every day. M240s now for 3800.

The fact that you want to get this lens going tells me that is your future ;)
 
Top