The Sony A7 thread [Merged]

[JimKasson]

Advertisement (gone after registration)

Once I have an RRS L-plate for my A7R I can see whether I also can stabilize the A7R and super teles in Portrait mode.

 

What are your experiences so far? TIA.

 

Sorry, I can't help you. I'm not trying to get the a7R to work with super-teles.All my really long lenses are Nikkors. For them I might as well use the D800E and lock up the mirror. The only thing I give up is the great a7R LV focusing.

 

Jim

[AlanG]

Processing and conversion by the firmware.

 

Yes that can certainly be a factor and may be the explanation. But if done correctly, how would that account for some lenses working better on the A7R and some not?

[AlanG]

AA filters?

 

No AA filter on the A7R version. Besides, as I said about my comment regarding processing, if the cover glass thickness or something else varied enough, then the lack of sharpness should be consistent across all lenses. Measurements of a variety of lenses on the A7R show it to resolve detail very well.

[AlanG]

Right you are. When comparing images from cameras of different resolutions looking for absolute performance, the fair thing to do is res both images to some third resolution that's higher that either. When I compare images from the M240 to images from the a7R, I res both to 10,000 pixels wide.

 

Jim

 

I guess that makes sense as long as the up-scalling is done well. But I can't speak to his methodology, I am trying to understand the logic. In general it makes sense for individuals to test their own equipment, but I don't consider it very scientific and would not project those results universally in any way without an explanation and proof of causation.. Various aspects of the photo can make an image look sharper even if it has lower resolution. E.g. sharpening, un-sharp masking, or turning up local contrast. It isn't as if test charts were photographed and image resolution is being compared.numerically.

[JimKasson]

I guess that makes sense as long as the up-scaling is done well.

 

The word "well" in this context is defined a little differently than for normal photographic upscaling. What you want for this exercise is upscaling with the minimum number of dials, knobs, settings, and proprietary algorithms that may do things you don't understand. I use bilinear interpolation for this, although it's not the algorithm I'd pick for actually making prints.

 

But I can't speak to his methodology, I am trying to understand the logic. In general it makes sense for individuals to test their own equipment, but I don't consider it very scientific and would not project those results universally in any way without an explanation and proof of causation.. Various aspects of the photo can make an image look sharper even if it has lower resolution. E.g. sharpening, un-sharp masking, or turning up local contrast. It isn't as if test charts were photographed and image resolution is being compared.numerically.

 

It would be nice to see images to support the descriptions, and have the details of the methodolgy. I try to do that with all my comparisons.

 

And, by the way, I haven't found any lens that's less sharp on-axis on the a7R than on the M240, although there are many (most?) lenses that have apertures for which the extra resolution of the Sony doesn't buy you anything.

 

Jim

[ModernMan]

No AA filter on the A7R version. Besides, as I said about my comment regarding processing, if the cover glass thickness or something else varied enough, then the lack of sharpness should be consistent across all lenses. Measurements of a variety of lenses on the A7R show it to resolve detail very well.

 

In the LensRantals.com blog (previously linked in this thread) a recent post notes the effect of sensor cover glass thickness. See LensRentals.com - A Bit of A7R Sanity. The relevant passage from the LensRentals blog post is "...People who talk in mathematics I can’t really follow tell me that the further back the exit pupil, the more off-axis resolution (and color shift) will be affected by a thicker cover glass on the sensor. It’s the same reason so many wide-angle M-mount lenses have problems when adapted to NEX cameras."

 

So, as I read this, performance deficiencies with certain lenses may be attributable to the difference in cover glass thickness on the A7r vs. the M240. Importantly however, this effect varies as some inverse function of the distance from the lens's exit pupil to the sensor (or perhaps to the first surface of the cover glass). Further, I speculate that in many cases the proximity of the rear element to the sensor is a rough proxy for the proximity of the exit pupil to the sensor.

[jaapv]

Advertisement (gone after registration)

The last speculation is incorrect; there is no relationship between the position of the rear element and the position of the exit pupil.

[AlanG]

The cover glass thickness shouldn't be a factor with on-axis sharpness. And I am not sure it is much of a factor on any lens designed for a reflex camera. This should be pretty easy to measure in test shots of resolution charts.

[jdlaing]

Yes that can certainly be a factor and may be the explanation. But if done correctly, how would that account for some lenses working better on the A7R and some not?

 

No explanation for that.

Each programmer that sets the firmware to render the sensor data will do it differently.

[ModernMan]

The cover glass thickness shouldn't be a factor with on-axis sharpness.

 

Correct for rays incident on the cover glass at exactly 90 (perfectly on-axis). But even for a point in the middle of the sensor, that point may be illuminated by a cone of rays if the lens is not a telecentric design.

[ModernMan]

The last speculation is incorrect; there is no relationship between the position of the rear element and the position of the exit pupil.

 

Yes, I agree that the speculation about exit pupil location and rear element location is not correct. In the case of telecentric lenses the exit pupil is infinitely far from the image plane.

[AlanG]

Correct for rays incident on the cover glass at exactly 90 (perfectly on-axis). But even for a point in the middle of the sensor, that point may be illuminated by a cone of rays if the lens is not a telecentric design.

 

If this makes a significant difference with typical reflex lenses, then why would Sony and others use such a thick cover glass?

[k-hawinkler]

If this makes a significant difference with typical reflex lenses, then why would Sony and others use such a thick cover glass?

 

Great point.

Sony can afford to do that as they basically start from scratch.

Leica can't, so they are willing to accept some levels of infrared contamination in their digital Ms.

[AlanG]

Great point.

Sony can afford to do that as they basically start from scratch.

Leica can't, so they are willing to accept some levels of infrared contamination in their digital Ms.

 

Sony is only starting from scratch on whatever lenses are designed for the A7. I don't think this somewhat thicker than Leica's cover glass is a factor in limiting sharpness of other legacy or newer SLR lenses. And other manufacturers use similar cover glass I presume.

 

Considering that woorob reported that telephoto R lenses worked better on the M than the A7R then cover glass thickness would seem to me to be a non-factor with such lenses since they are telecentric.

[k-hawinkler]

Not necessarily true. At this stage I would take everything with a grain of salt.

[ModernMan]

Considering that woorob reported that telephoto R lenses worked better on the M than the A7R then cover glass thickness would seem to me to be a non-factor with such lenses since they are telecentric.

 

Telephoto is not the same as telecentric. Telecentric refers to a property where all rays incident on the image plane are on axis (and consequently all parallel), rather than from an incidence cone. More on Wikipedia, and most interesting, if you like: Telecentric lens - Wikipedia, the free encyclopedia. As I understand it, a simple telephoto lens may be much closer to telecentric than a normal lens (narrower incidence cone), but not actually telecentric, strictly speaking.

 

Presumably the A7r's offset microlenses and cover glass are a design compromise that favors the lenses Sony thought would bring the most business benefit to them, which is not a set composed principally of Leica M ad R lenses.

[AlanG]

Telecentric refers to a property where all rays incident on the image plane are on axis (and consequently all parallel), rather than from an incidence cone. More on Wikipedia, and most interesting, if you like: Telecentric lens - Wikipedia, the free encyclopedia. As I understand it, a simple telephoto lens may be much closer to telecentric than a normal lens (narrower incidence cone), but not actually telecentric, strictly speaking.

 

Presumably the A7r's offset microlenses and cover glass are a design compromise that favors the lenses Sony thought would bring the most business benefit to them, which is not a set composed principally of Leica M ad R lenses.

 

Of course the rays will be somewhat closer to parallel with some lens designs and this is easier to achieve with longer focal lengths. But they are not going to be parallel in regular camera lenses. I probably shouldn't have used the term telecentric because it really doesn't apply to these kinds of lenses despite being used (incorrectly?) in some lens descriptions and marketing as well as in general discussions. In any case lenses that are 280mm long will not have an issue with cover glass thickness.

 

And any lenses designed for the angled microlenses of the A7R also have to work on the A7 that has regular microlenses.

 

I'd just like someone to explain how having a cover glass that is .8mm thick vs.one that is 1mm thick will produce a significantly sharper image with reflex lenses. And do we know how thick the cover glass is on these two cameras? Is the displacement of the projected image greater than the depth of focus? Are the subjects a flat plane that are being photographed wide open? Or is the tiny additional thickness of the glass causing some overall decrease in sharpness when light has to go a tiny additional distance through it when at an angle than when going straight? Why would it do that?

[pop]

Is there any data on the precision with which those cameras and adapters are built? In particular, are we quite sure that the (Leica) lenses are mounted with their axis perpendicular to and centered on the sensor?

[k-hawinkler]

Is there any data on the precision with which those cameras and adapters are built? In particular, are we quite sure that the (Leica) lenses are mounted with their axis perpendicular to and centered on the sensor?

 

 

You should read the LensRentals article on the subject.

I believe it was referenced earlier in this thread.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[k-hawinkler]

K-H,

 

Thanks for your comments. I would certainly never take offense to your question. I’ve learned a lot on this forum and, like most people on this earth, can still learn a thing or two. I’m enjoying my M immensely and find the long lenses really outstanding on the M.

 

As you note, though, focusing at anything over about 300mm is tough. My comment about focus peaking was made in the context of trying to figure out how to make it work best. The Sony has three settings – low, medium and high – but so far as I can tell the manual doesn’t even mention these settings. Experimenting with them, it seems that the “high” setting covers quite a bit, so that nothing would be in critical focus, and the “low” setting often will not show any focus peaking at all. On the M there are no settings and it seems to be more critical in terms of what it picks up as a focus point. On both cameras there also seem to be differences in focus peaking dependent on the lens used and the f/stop of the lens as well. So I quite agree with you that focus peaking is a tool but not a guarantee of critical, definitive focus.

 

When I took these comparison shots I would start by trying to use focus peaking and then switch to the highest focus magnification setting and carefully try to get a critical focus that way. I’m guessing that you use the same approach, but if there’s a better way, please let me know!

 

What I discovered, and have gathered from your previous posts on the subject, is that it is extremely difficult to obtain accurate focus and stability at high levels of magnification.

 

To make matters worse, we have a major storm brewing and yesterday afternoon it was rather windy in advance of the storm (a “dangerous and historic cold air outbreak will continue through Tuesday morning. The combination of extremely cold air temperatures and gusty west winds will bring wind chill values of 45 to 65 below zero”). There were episodes of considerable wind yesterday afternoon when I took the pictures, which may have been a factor in the outcomes. I always use the best tripod I have when using long lenses and a 10 or 12 second delay on the camera’s self timer, but it is certainly possible that wind was a factor. If time permits, I might reshoot some of the long-lens images and post my results later. In the meanwhile, feel free to download these full size RAW files if you’d like to examine them more carefully:

 

 

16-18-21 Tri-Elmar

 

M (L2000791)

https://www.sugarsync.com/pf/D243783_74498898_300619

 

a7R (_DSC0070.ARW)

https://www.sugarsync.com/pf/D243783_74498898_300762

 

19mm Elmarit-R II 1:2.8

 

M (L2000792)

https://www.sugarsync.com/pf/D243783_74498898_300612

 

a7R (_DSC0071.ARW)

https://www.sugarsync.com/pf/D243783_74498898_300775

 

75mm APO-Summicron-M f/2.0 ASPH

 

M (L2000793)

https://www.sugarsync.com/pf/D243783_74498898_300626

 

a7R (_DSC0073.ARW)

https://www.sugarsync.com/pf/D243783_74498898_300779

 

100mm APO-Macro-Elmarit-R 1:2.8 with 2x-APO extender

 

M (L2000796)

https://www.sugarsync.com/pf/D243783_74498898_300620

 

a7R (_DSC0075.ARW)

https://www.sugarsync.com/pf/D243783_74498898_300758

 

280mm APO-Elmarit-R 1:2.8 with 1.4x-APO extender

 

M (L2000801)

https://www.sugarsync.com/pf/D243783_74498898_300623

 

a7R (_DSC0083.ARW)

https://www.sugarsync.com/pf/D243783_74498898_300752

 

500mm MR-Telyt-R 1:8.0

 

M (L2000803)

https://www.sugarsync.com/pf/D243783_74498898_300636

 

a7R (_DSC0085.ARW)

https://www.sugarsync.com/pf/D243783_74498898_300765

 

500mm MR-Telyt-R 1:8.0 with 2x-APO extender

 

M (L2000804)

https://www.sugarsync.com/pf/D243783_74498898_300630

 

a7R (_DSC0086.ARW)

https://www.sugarsync.com/pf/D243783_74498898_300769

 

I’d also be interested in anyone’s insights into tripod use with longer lenses. I have a nice Velbon carbon tripod (model “El Carmagne 540”) and a Manfrotto ballhead (model 486RC2) that have served me well, but I may be reaching (or have exceeded) the limits of this combination for anything over, say, 300 mm. Right now I can’t really afford to spend a whole lot on a new heavy-duty tripod/ballhead combo, but, then again, if a good deal showed up on that certain auction site or elsewhere I might be tempted.

 

The biggest challenge I face with this tripod/ballhead is that at magnifications of 500 mm or greater, I get a slight shake and, more significantly to me, it becomes difficult to frame the subject properly. You’ll notice that the last image linked above, at 1000 mm, is not centered because of this. My Manfrotto ballhead does not have any fine-tuning capabilities and I wonder if Linhof or other manufacturer makes a ballhead where tiny adjustments, both vertically and horizontally, can be dialed in using a toothed gear or other similar mechanism.

 

All in all, I’m enjoying my foray into long-lens photography. The M with EVF works fine for landscape or tripod use, but would be pretty inadequate for action sports photography (where Nikon and Canon truly excel).

 

 

 

Hi Bob,

 

Many thanks for your post.

You did an incredibly terrific job under very difficult weather conditions.

That looks like it's really nasty out there.

I would last maybe 5 minutes before I would have no feeling left in my fingers.

WOW!

 

Now, please let me give you some feedback.

 

16-18-21 Tri-Elmar

The WATE images differ in that the M240 image has a little bit of shading and discoloration whereas the A7R image has more intense shading but to my eyes no discoloration. I can not detect any corner smearing so both images can be put in excellent shape with Lens Cast Corrections LCC. When you shoot the correction image please tape up the hood corner with the hole with some black electricians tape so that you getter a proper correction image. While you shoot regular images the tape is not necessary. However, if you shoot alternatingly images and correction images in the same light I would leave the black tape on. BTW, I have no focus problems with the WATE on either my M9 or A7R in nice weather. In nasty weather all bets are off!

 

My conclusion: You should be able to get excellent images from both cameras with the A7R giving you more detail.

 

 

19mm Elmarit-R II 1:2.8

Interestingly here the A7R has a more benign shading whereas the M240 has more pronounced discoloration and shading. Again I don't see corner smearing.

 

My conclusion: Same as WATE, needs LLC to give you great images.

 

 

75mm APO-Summicron-M f/2.0 ASPH

Both images look fine to me if properly focused. LLC doesn't seem necessary but I would prepare corrections images to have them and to see what they look like. I am looking forward to use my copy of this lens.

 

 

100mm APO-Macro-Elmarit-R 1:2.8 with 2x-APO extender

I would shoot both cameras with and without extender and carefully focus.

In downtown Minneapolis there are enough indoor walkways so that you don't have to venture outside at all to practice focusing. I would take advantage of that. Inconclusive based on your images for the adapter use.

 

 

280mm APO-Elmarit-R 1:2.8 with 1.4x-APO extender

Same feedback as above.