Mitakon Zhongyi 50mm f/ 0.95 and A7RII sensor max aperture?

[jerome_m]

I would like to know whether the A7RII sensor can actually use the f/0.95 aperture of the Mitakon Zhongyi 50mm. A few years ago, there was a study that pixels internal shading limited the usable aperture of some lenses. The A7RII being a backlight sensor has less pixel internal shading, but a f/0.95 aperture was probably not planned when designing the sensor.

Could someone with the lens and camera try a picture at f/0.95 and f/1.4 and see whether there is a difference in exposure and out of focus highlights?

 

[iiiNelson]

I would like to know whether the A7RII sensor can actually use the f/0.95 aperture of the Mitakon Zhongyi 50mm. A few years ago, there was a study that pixels internal shading limited the usable aperture of some lenses. The A7RII being a backlight sensor has less pixel internal shading, but a f/0.95 aperture was probably not planned when designing the sensor.

Could someone with the lens and camera try a picture at f/0.95 and f/1.4 and see whether there is a difference in exposure and out of focus highlights?

I don't understand the question... Did someone publish a study saying that the sensor can record an image at f/0.95? If so that's crazy as there are 3 native lenses with that aperture and many have adapted other manufacturers lenses as well.

 

[pegelli]

I think Jerome is referring to this article on Luminous Landscape: An Open Letter To The Major Camera Manufacturers. (you have to be LuLa subscriber to read it)

After that there was along thread on the LuLa forum see here where several people (incl. myself) did tests and basically found no evidence of an apparent smaller aperture due to internal pixel shading using then state of the art sensors of 24 MP and 85 mm 1.4 and 1.2 lenses. Also the person from DxO who reported the "flaw" would follow up with another article and more elaborate measurements but as far as I know this never happened.

I have neither a 42 MP camera nor a f0,95 lens (the best I can do is 24 MP and an 1.2 lens) but I would be interested if somebody could do the test and report back here.

 

[Vivek]

AFAIK, the worst case scenario of this was in NEX-7. Beautiful camera, well made and with a terrible sensor. Sony did brighten up things a bit after that.

 

[jerome_m]

I think Jerome is referring to this article on Luminous Landscape: An Open Letter To The Major Camera Manufacturers. (you have to be LuLa subscriber to read it)

Yes, that is the original article. Thank you.

 

[k-hawinkler]

I think Jerome is referring to this article on Luminous Landscape: An Open Letter To The Major Camera Manufacturers. (you have to be LuLa subscriber to read it)

After that there was along thread on the LuLa forum see here where several people (incl. myself) did tests and basically found no evidence of an apparent smaller aperture due to internal pixel shading using then state of the art sensors of 24 MP and 85 mm 1.4 and 1.2 lenses. Also the person from DxO who reported the "flaw" would follow up with another article and more elaborate measurements but as far as I know this never happened.

I have neither a 42 MP camera nor a f0,95 lens (the best I can do is 24 MP and an 1.2 lens) but I would be interested if somebody could do the test and report back here.

What ind of test? :facesmack: :grin:

 

[Makten]

I would like to know whether the A7RII sensor can actually use the f/0.95 aperture of the Mitakon Zhongyi 50mm. A few years ago, there was a study that pixels internal shading limited the usable aperture of some lenses. The A7RII being a backlight sensor has less pixel internal shading, but a f/0.95 aperture was probably not planned when designing the sensor.

Could someone with the lens and camera try a picture at f/0.95 and f/1.4 and see whether there is a difference in exposure and out of focus highlights?

The problem you refer to does exist. However, there is no certain aperture that sets a limit, but a certain angle of infalling light on the sensor. If the exit pupil is very far from the sensor, you could use an f/0.5 lens with no problem. But a fast lens with large exit pupil (which doesn't have to be large just because the entrance pupil is!) and a short distance to the sensor, there will be limitations.
This is the reason some lenses vignette much more on digital than on film, and they don't even have to be fast. Just having an exit pupil "too close", even if it's small. To affect the center though, the exit pupil both has to be large and close.

An 85/1.2 will most certainly have the exit pupil very far from the sensor and will thus not run into these (theoretical) problems. But a very fast, nearly symmetrical rangefinder lens like the Canon 50/0.95, probably gives lesser exposure than it would on film.

I have absolutely no real world evidence of this, but it would be very strange if the angle of infalling light would not affect the pixel light gathering efficiency.

 

[jerome_m]

there is no certain aperture that sets a limit, but a certain angle of infalling light on the sensor.

That is true, the angle on in falling light depends on the position of the exit pupil and of its size, which depends on the aperture. Still: I have no idea where the exit pupil of the 50mm f/0.95 is situated, the best way to find out would still be to test what happens on the A7 RII.

 

[Makten]

That is true, but the angle on in falling light depends on the position of the exit pupil and of its size, which depends on the aperture.

Wasn't that exactly what I wrote? :grin:

 

[thompsonkirk]

There's really no practical problem here. I've just finished a project with A7rII and wide-aperture lenses and am starting to print it as a book:

DARKSCAPES - Kirk Thompson Photography

The images are made with f0.95 Noctilux, f1 Nocti, and f0.95 Mitakon (often from 2 or more stitched files). The images with the sharpest centers (but the most CA) are from the newer Nocti; those with a bit of flare around the highlights (for example the last in the series) are from the Mitakon. The f1.0 Nocti vignettes more than the other two and can barely be corrected with LR/ACR Lens Correction.

If you're going to use this sort of lens wide open, you're aiming for effects that won't be undone by what you're worrying about, so relax and enjoy! Mitakon is fun, and at the end of the project I realized I could have shot the whole of it with Mitakon and not put out >$10K to try the other lenses. The only advantage of the Noctis was for other uses – they autofocus very nicely with TAP.

(BTW, Anyone wanna buy the 0.95 Nocti? Now that the project is finished, I'm keeping only the Mitakon.)

Kirk

 

[Vivek]

Mitakon is fun, and at the end of the project I realized I could have shot the whole of it with Mitakon and not put out >$10K to try the other lenses. The only advantage of the Noctis was for other uses – they autofocus very nicely with TAP.

(BTW, Anyone wanna buy the 0.95 Nocti? Now that the project is finished, I'm keeping only the Mitakon.)

Kirk

You are as bad as I am, I think. Very bad at selling the excess lenses that I own.

 

[jerome_m]

There's really no practical problem here. I've just finished a project with A7rII and wide-aperture lenses and am starting to print it as a book:

DARKSCAPES - Kirk Thompson Photography

The images are made with f0.95 Noctilux, f1 Nocti, and f0.95 Mitakon (often from 2 or more stitched files). The images with the sharpest centers (but the most CA) are from the newer Nocti; those with a bit of flare around the highlights (for example the last in the series) are from the Mitakon. The f1.0 Nocti vignettes more than the other two and can barely be corrected with LR/ACR Lens Correction.

If you're going to use this sort of lens wide open, you're aiming for effects that won't be undone by what you're worrying about, so relax and enjoy! Mitakon is fun, and at the end of the project I realized I could have shot the whole of it with Mitakon and not put out >$10K to try the other lenses. The only advantage of the Noctis was for other uses – they autofocus very nicely with TAP.

(BTW, Anyone wanna buy the 0.95 Nocti? Now that the project is finished, I'm keeping only the Mitakon.)

Kirk

I like the pictures you presented, but what I am looking for is quite different. I want to take pictures in low light and wonder whether I should get the Mitakon or simply use a 50mm f/1.4, one more step of ISO and accept a higher noise level.

What is "TAP" which allows autofocus with the Noctis?

 

[thompsonkirk]

Well, you were asking how f0.95 lenses might work, wide open, with the sensor. You should be able to draw a proper inference regardless of the subject matter. All that matters is that the lens is at largest aperture.

IMO you'd be better off with a 1.4 lens and a higher ISO, bcz A7rII will yield lots of shadow detail without much noise.

TAP is TechArt Pro Adapter, with extensive info on this forum and on FM forum; just do a search. It lets you autofocus wide open with Nocti or Summilux, with occasional glitches.

Kirk

 

[jerome_m]

Well, you were asking how f0.95 lenses might work, wide open, with the sensor. You should be able to draw a proper inference regardless of the subject matter. All that matters is that the lens is at largest aperture.

Since you have the camera and lens, can you please take a picture full open and at f/1.4 and see if the exposure and depth of field changes?

TAP is TechArt Pro Adapter, with extensive info on this forum and on FM forum; just do a search. It lets you autofocus wide open with Nocti or Summilux, with occasional glitches.

Thank you. With the full name of the adapter, I was able to find the discussions.

 

[thompsonkirk]

Sorry. My A7rII has the Kolari mod, so it won't be exactly what you're worrying about.

 

[Vivek]

Since you have the camera and lens, can you please take a picture full open and at f/1.4 and see if the exposure and depth of field changes?

I can tell you that with the Canon 50/0.95 on the A7r, etc and the Leica MM, the exposure and DOF do change.