Photoclubalpha

I've had a chance to really aim the A77 into difficult light conditions, repeatedly. I think this example, which I have put up on pBase at full size (a rather amazing 17MB of Level 12 JPEG data) will tell you all you need to know about whether Sony's new coatings on the SLT , and changes to internal design of the darkchamber, have cured ghosting problems.

Basically I've not been able to produce any ghosts or flare patches. This is taken, also, with Sigma's DG UV filter on the CZ 16-80mm. I have established that this filter, out of all those I've been able to try from my collection, produces no visible relfections and also has the highest quality glass. It's also slim enough to use with the 16-80mm and fits the lenscap unlike my previous B+W slim UV. The combination of this lens, filter and the A77 is what I'm normally using right now and it appears to be rock solid with into the light situations.

For info - ISO 100, no NR (of any kind), strong contrast curve, ACR 6.5 black level set to zero (so I am not crushing any hard to see flare or ghosts). I think I've front focused a touch but that is me not the camera, I did focus in front of the boat to secure a better sharpness on the water, but I thought f/11 would give sufficient depth of field. Wrong; at 24 megapixels, you really can't rely on depth of field at all if you want a sharp pixel view at screen size. In a print, it looks perfect.



This link opens the original size: http://www.pbase.com/davidkilpatrick/im ... 9/original in pBase and you can also see the EXIF and the GPS location.

David

I wonder if the mirror frames are physically compatible between a77 and in the former SLTs.
Then if they are, I would love to see the results of swapping them around between a77 and the a55

David, it was already quite clear from the earliest raw samples floating around that the ghosting is at least deeply suppressed in the 77/65 SLTs. But your example is really extreme, no astro shots can compete here as they have much more air in the path to blur out the star images.

But I would love to learn if the excessive purple fringing present in the a55/33 is also as seriously suppressed, at least to the levels present in a700.

I've noticed that in the dpreview studio raws, that the ghosting noticed with the A55 (for example under the white cross bar of the cross) is not visible in the A77, which is very good news.

tom

Agorabasta - there is still some purple bleed from extreme highlights. The effect in a vertical version of this shot, at f/8 with less depth of field and more blur especially in the foreground, is to create a kind of green-purple shift from focus points behind and in front of the subject.
Enlarged to 100X and taking the extreme bottom, the purple fringe blooming is very clear indeed:
It's certainly no better than any other camera system or sensor and reminds me of some of the results I had from the Alpha 100 with the old 70-210mm beercan f/4 lens (very sharp but with similar subtle bokeh-related colour shifts). I have not experienced this from the CZ 16-80mm before. It can be reduced of course by turning colour NR on, as with the standard ACR default settings.

David

Isn’t this our omnipresent friend, garden-variety lateral CA caused by the lens? If so, this has nothing to do with the camera itself, I would think.

Are some sensors more prone to this type of CA? Or do some sensors or SLT designs exacerbate longitudinal CA or the mysterious purple fringing?

Bruce

No, this is not lens CA. This lens does have CA, but it looks quite different and a profile has been applied to correct it. This is an effect which varies from sensor to sensor.

David

Why is that violet flare only on one side of a blown highlight and not all around it?
Greg

That's because of the way the filter toppings, cover glass, structure of the CMOS interact with the image-forming light and also with the residual aberrations. The halo effect is absent in the centre of the field. Remember this is a small section of the worst example I could find, and the image is also out of focus at this point, which emphasises the effect. In more normal situations this won't happen but there are some conditions, like twigs or branches against very bright sky, where I expect a very fine purple fringe will happen on the 24 megapixel sensor.

David

Thanks, David.

As long as there's no PF in the frame centre, that's OK for the most part. That's at least not worse than average like the a55 is.

I was hoping though, they could have moved the hot mirror onto the SLT mirror. They filed for a patent regarding exactly that. That would kill most of the sensor toppings related PF.

In the guts of the A77
http://www.youtube.com/watch?feature=pl ... vkCMbEO8bQ#!