So how much light is 1/3rd of a stop? I try and see what difference it really makes.

[KWGeorge]

So I have been thinking about the potential of the new replacement for the A99 and I have been wondering if I am interested in it or not. See the issue for me is that I love shooting just about everything but my absolute favorites and my passion is photographing my grand daughter and wildlife. Now you could say since she just recently turned two that they are one in the same and running after her its hard to argue that but it is what it is. So I own a A850 and a A7II with a LEA4 adapter (Along with a A77, A6000 and NEX6! Yeah I have a problem). The issue with the adapter is it will not support Tele converters but that is another gripe.

 

So as I understand it and I am sure I will be corrected if I am wrong (or corrected even if I am not) with the Sony Translucent Mirror Technology I sacrifice 1/3^rd of a stop of light. Well that begs the question what does that look like in the real world? For me at least it is hard to see how that affects my pictures in the long run so I decided to do some testing, not very scientific but somewhat controlled at least. I took a Tamron 90mm Macro A-Mount lens and shot a picture of my test board with the A850 using the settings 1/25^th sec, f/2.8, ISO 200 with constant photography lights. I then repeated the test with the exact same setting with the A7II with the LEA4 adapter installed and compared the difference. Now these are both full frame cameras and 24 megapixels but there is an age and technology difference however I really don’t think that came into play all that much in this test as both were in manual mode.

 

Here are the side by side results.

 

(The A7II with Adapter is on the left and the A850 is on the right if you have not guessed)

 

Small file

https://flic.kr/p/ubuaKU

 

Bigger Tiff file

https://flic.kr/p/tUh96s

 

So yeah that’s quite a bit of difference and where I am seeing how bad it is interfering with the quality of my pictures is on my A77 which appears to be a “Good Light” only camera as faced with clouds or less than perfect lighting I am finding it to be to noisy to use even at ISO 400! This is much more apparent with longer glass such as the Minolta 300mm f/2.8 with a 2X TC on it or even with the Sony 70-400 GII lens. In any case the point of this is post is to show what that loss of light really looks like or an approximation thereof for people unsure of it like me. In the end I think the A99 will have to be better at low light (by default it will be somewhat due to being a full frame) and that should be a priority for Sony but we shall see.

 

KWGeorge

 

[Golem]

You used different cameras that have different

sensors and processing engines. The difference

in exposure visible in your comparison is much

greater than 1/3 EV, at least according to my

very experienced eyes. The exposure difference

I'm seeing approaches 1 EV. IOW it appears that

the age and technology difference is very much

in play here.

 

You have on hand the items for a very direct

test of the light loss from the mirror. Using

the macro lens that you already used, mounted

on the A850, cobble up a way to shoot straight

thru the adapter mounted to the front of the

lens, as if it were a front mounted accessory.

Then, at the same manual exposure settings,

remove the adapter and shoot again. Now you

have cleared out most of the variables.

 

But you have not cleared out ALL the variables.

It's good that you shot wide open, so any slop

in the auto stop down iris is uninvolved. A 1/25"

shutter speed is not too bad, as that's a fairly

efficient speed, mechanism-wise, but an 1/8" or

1/4" would reduce the margin of error. Finally,

shoot for or five frames with a few seconds idle

between each, for each test exposure. I never

had an A850 but I have encountered variation

across 3 or 4 supposedly identical exposures

from various Sony bodies newer than yours. A

brand new A65 was so sorry on that score that

I returned it.

 

BTW, when you front mount the adapter, the

effect of the mirror on sharpness may not be

represented accurately due to the relocation.

But the effect on exposure will be accurately

represented. And for a more accurate way to

view the exposure difference you can include

a gray scale. A gravure print, such as a dollar,

is quite the opposite of a gray scale !

 

 

`

[KWGeorge]

Thank you for your thoughts I was hoping someone could clarify the difference more for me. I was thinking about your reply and even though I don’t have a 90mm f/2.8 for the A7II (yet) I do have a Rokinon 85mm f/1.4 that I could attach so I grabbed it and removed the adapter and installed it. Set the lens at f/2.8 and all other settings were the same. The results are here;

 

https://flic.kr/p/u9Nqa5

 

The 85mm f/1.4 is the center dollar picture and the 90mm with the LEA4 is on the left leaving the 90mm with the A850 camera on the far right.

 

A bit of the exif from light room on these.

 

https://flic.kr/p/tUxDXy

 

So the color is a bit different and the A850 is still a bit brighter but I have always found that camera when on default to be a bit more saturated and brighter. But still the difference is pretty obvious so what am I doing wrong or what can I do to get them closer?

[Golem]

The left and right dollars now look

like about 1/3 EV difference. Just a

bit closer, just closer enuf, vs the

first pair you posted. Maybe it's a

1/2 EV not a 1/3 ? Hard to tell, so

as a visualization, you got it.

 

I would say that comparing the first

pair of dollars to the second pair

[ignoring the center dollar for now],

the difference shown in the second

pair is only half the difference as of

the first pair.

 

As to the 85/1.4, it's a whole other

optic so don't beat your head against

the wall on that one :-) At minimum,

it's 1.4 stopped down 2 stops vs the

90/2.8 being wide open, so it should

exhibit much better illumination and

contrast vs the 90 wide open, and by

your example, it seems to deliver.

 

 

`

[LeButler]

BTW the LA-AE4 actually includes an SLT mirror within, so any differences seen are not due to mirror Vs no-mirror (it appears to have the Sony A65's assembly).

[Golem]

BTW the LA-AE4 actually includes an SLT mirror

within, so any differences seen are not due to

mirror Vs no-mirror .........

This thread is all about the SLT mirror in the

LA-E4, more specifically the 1/3 EV difference

due to "mirror vs no-mirror". There is no point

in trying to explain that this is not the case.

 

Are you familiar with the A7-II and the A850 ?

When you mount the LA-E4 on the A7-II you

create an SLT camera. An A850 is not an SLT

camera. It's an SLR.

 

Further, the thread is NOT about what camera

uses which technology. It's about visualizing a

1/3 EV exposure difference, regardless of how

that difference came about. The visualization

could just as well have been facilitated using

an ND 0.1 filter, but that would have been way

too easy :-)

 

FWIW, when I earlier suggested just putting the

LA-E4 on the front of the lens "like a filter", that

method *IS* the "way too easy" one. The mirror

in the LA-E4 is essentially an ND 0.1 filter.

 

`

[KWGeorge]

“The visualization

could just as well have been facilitated using

an ND 0.1 filter”

 

 

Dang! I had not thought about it like that for some reason. Why is it not explained like that in the first place! This is a very simplified and relatable way to understand the difference at least to me. Sometimes you can't see the forest through the trees J 

[olemartinorg]

...or you could just use the dial on the A7II to underexpose by 1/3 stop.

[Golem]

...or you could just use the dial on the A7II to underexpose by 1/3 stop.

Except that you can't count on the gear to

actually deliver an accurate 1/3 EV shift

in any of the 3 parameters [iSO-iris-time].

 

`

[Golem]

“The visualization

could just as well have been facilitated using

an ND 0.1 filter”

 

 

 

Dang! I had not thought about it like that for some

reason. Why is it not explained like that in the first

place! This is a very simplified and relatable way to

understand the difference at least to me.

 

Sometimes you can't see the forest through the trees J 

Oh. I thought that the whole point was to have fun with trees ....

 

`

[olemartinorg]

Sure, Golem - but i didn't take this thread as a quest to find the exact scientifically corrct measure of 1/3 stop of light, but more "huh, how.much light is that, roughly?". Also, an exposure reduction in RAW processing should yield an exact (scientifically accurate) 1/3 stop underexposure, should it not?

 

Depends on the photographer, but personally i don't really care about 1/3rd stop exposure compensation at all. It doesn't matter that much for me, as i'll do the fine-tuning in post anyway (and i never need to push the exposure _that_ much anyway). For me, the 1/3 underexposure is absolutely worth it in the LA-EA4, as long as it's not affecting my depth of field or anything like that (which an aperture value change to decrease exposure would do). Modern (full frame especially) cameras have so much room to play with that i never feel limited by 1/3 stop handicap.

[weinhous]

So I have been thinking about the potential of the new replacement for the A99 and I have been wondering if I am interested in it or not. See the issue for me is that I love shooting just about everything but my absolute favorites and my passion is photographing my grand daughter and wildlife. Now you could say since she just recently turned two that they are one in the same and running after her its hard to argue that but it is what it is. So I own a A850 and a A7II with a LEA4 adapter (Along with a A77, A6000 and NEX6! Yeah I have a problem). The issue with the adapter is it will not support Tele converters but that is another gripe.

 

So as I understand it and I am sure I will be corrected if I am wrong (or corrected even if I am not) with the Sony Translucent Mirror Technology I sacrifice 1/3^rd of a stop of light. Well that begs the question what does that look like in the real world? For me at least it is hard to see how that affects my pictures in the long run so I decided to do some testing, not very scientific but somewhat controlled at least. I took a Tamron 90mm Macro A-Mount lens and shot a picture of my test board with the A850 using the settings 1/25^th sec, f/2.8, ISO 200 with constant photography lights. I then repeated the test with the exact same setting with the A7II with the LEA4 adapter installed and compared the difference. Now these are both full frame cameras and 24 megapixels but there is an age and technology difference however I really don’t think that came into play all that much in this test as both were in manual mode.

 

Here are the side by side results.

 

(The A7II with Adapter is on the left and the A850 is on the right if you have not guessed)

 

Small file

https://flic.kr/p/ubuaKU

 

Bigger Tiff file

https://flic.kr/p/tUh96s

 

So yeah that’s quite a bit of difference and where I am seeing how bad it is interfering with the quality of my pictures is on my A77 which appears to be a “Good Light” only camera as faced with clouds or less than perfect lighting I am finding it to be to noisy to use even at ISO 400! This is much more apparent with longer glass such as the Minolta 300mm f/2.8 with a 2X TC on it or even with the Sony 70-400 GII lens. In any case the point of this is post is to show what that loss of light really looks like or an approximation thereof for people unsure of it like me. In the end I think the A99 will have to be better at low light (by default it will be somewhat due to being a full frame) and that should be a priority for Sony but we shall see.

 

KWGeorge

 

 

The experiment is "clouded" by the difference between f-stops and t-stops.

 

See http://petapixel.com/2014/09/30/your-lens-aperture-might-be-lying-to-you-or-the-difference-between-f-stops-and-t-stops/

[olemartinorg]

I don't really think the exact aperture has anything to do with this.. A 1/3 stop of light less is the same reduction of light regardless of what aperture you're shooting with (be it 2.8, 5.6, 1.4 or whatever). Remember that cameras measure light and try to find the best settings for an evenly exposed image (yes, you can of course change this default setting). A 1/3 stop less light than the optimal exposure is always that, a 1/3 stop less. Wanna find out how that looks in an image? Leave all the variables up to the camera, use exposure compensation to underexpose by a 1/3 stop, and compare the exposure with a non-corrected exposure taken in the same lighting conditions. This is pretty easy, really.

[Golem]

Yup. Back to basics of the OP. He just wants

to visualize how that looks. That idea about

going 1/3 EV in RAW is perfect. It avoids the

problem that the cameras often can't even do

a burst of jpegs on manual exposure without

showing a visually obviously shift in contrast,

color, and brightness among the images.

 

`