SLT Contrast Loss...

[Greg Beetham]

OK, let's see...

1. The EVF power consumption is supposed to be less than that of the back LCD; all modern cams have some, all tiny P&S use the LCD (some have EVFs also) and their tiny batteries are no issue.

Ok, I take that under advisement, I saw somewhere that the power consumption on this EVF unit was even higher than the rear LCD, and extrapolating that up to FF standards....

2. The a900/a850 have the SSS and don't suffer too much power drain. So that's no issue.

Yep granted, also have the AF motor, and the RAMDAK GPU processing unit and memory buffer/bus etc. situation normal...so far, BUT..we are suddenly talking a FF movie camera at probably 30MP or more with the main sensor turned on just to see through the EVF and it has to stay on for as long as you want to see....that is going to add up to lots of heat my friend....LOTS.

3. Memory cards capacities grow faster than the sensor resolution. This as well is no issue.

Yeah I sortof chucked that one in just for good measure...but you still need to consider what you are going to do with the camera stills or movies or both and stock up accordingly.

Well, Sony used some pretty laughable batteries in the SLT/Nex cams. That's easily correctable - use the older a200/300/350/700/900/850 batteries; so no issue as well..

Yeah I guess the A99 would have a power grip available so at least the battery/s would probably last a reasonable amount of time, one would still have to carry quite a few spares though if one was doing extensive coverage of some event, just to be safe.

So what exactly were you so joyously struggling with, Greg?

That about covers it as far as struggling goes, but there is always the issue with ghosting to be solved, I don't see too many 9 category owners accepting that, but I could be surprised.
Greg

[bfitzgerald]

I'm not a Kurt fan but I came across this..
http://www.kurtmunger.com/sony_transluc ... id181.html

Something seems a bit odd.

DPR did miss those issues which is kinda funny as I would have thought they would see it. They've missed a few things before on the G1 no mention of banding yet some pretty massive examples in their gallery.

[agorabasta]

Greg,

The LCD/EVF power consumption is no more than a few single-digit percent points of total power consumption of modern cams, and it will get only better. All you need for EVF is a few mW of emitted light power to make it immensely bright. So the total EVF power is more about the efficiency of driving circuits, and that's very 'optimisable'.

There already are FF cams with the main sensor LV, and their sensors really may overheat pretty fast. But once again, the heat of current gen of sensors is far lower than it used to be just a couple of years ago when those old FF sensors were devised. So that's doable even for the full res readout which you don't even need for LV; and some low-power binned operation may reduce the power down to the P&S levels if only one in every four sensor columns are driven for the LV readout.
But I still think that extensive video operation would need larger battery packs, about the same as those used with decent camcorders.

The issue with ghosting could be possibly improved by a few more stops down, but it is there to stay anyway. That is likely to avert me from using an SLT myself; but the general reception in the market should be very good...

P.S. I have some personal experience of vacation shooting with the Nex5. And the worst offender at power consumption is not even the Nex body itself, it's the lens' SSS and AF drive. I could easily take about 1000 single shots on one battery in a day with the 16mm (no SSS), the 18-55 was good for about 400-700 single shots depending on the zoom. And that with the LCD often at 'Sunny Weather' and 5min powersave. Later I got an 18-200, and that thing eats 25% of charge in 10min with 20 shots at 200mm!

[agorabasta]

Barry, Kurt's results are not confirmed anywhere else. You could try and look into the available raw samples around the web and see for yourself.
One of the possible reasons for his results could be different focusing with some LoCA showing with lens wide open.

[bfitzgerald]

I don't know why he has that problem but it looks like blooming if the mirror is a factor in that I can't say myself. Something is going on though and there must be a reason for it.

[David Kilpatrick]

Just to clarify the power consumption, the EVF uses more power than the rear LCD. The difference is officially 380 shots with the EVF, versus 450 shots using the LCD. The GPS, if switched on, reduces these figures to 430 and 370 approximately (handbook, page 33, figures for A55 and A55V without flash).

David

[Greg Beetham]

At a guess, if I had too guess, those fringes could be caused by focus error as well as anything, including the interference by the inclined mirror to some extent.
Greg

[David Kilpatrick]

The A55 does not suffer from chromatic fringes any more than any other camera. In fact, I'd say that CA was less obvious with the 16-80mm CZ than it was on the Alpha 100 I first used this lens on. CA with the 16-80mm was pretty strong on the A100, it's not very strong at all on the A55.

David

[Greg Beetham]

I'd like too see the same test done again with MF and with and without a polarizer.
Greg

ps. Thinking about it for a sec. what's the point of the example purple fringing test shots anyway even if the focus was correct, I'd say not many would shoot landscapes at f2.8?

[bfitzgerald]

So far we don't have an explanation as to why this happened. I won't doubt the points raised but something is going on here..so what exactly is it?

[David Kilpatrick]

It looks like a filter reflection not the mirror. The mirror ghost images appear below the light source, not to one side, and are much closer to the source - just three or four pixels distance, not like that at all.

David

[agorabasta]

The true nature of purple fringing seen with most digital sensors is rather simple and straightforward - it's caused by the metal wiring on the surface of the sensor chip. Thin layers of metals like gold, copper, aluminium are semitransparent and appear green/blue/violet in transmitted light. The other parts of the incident light spectrum are hence more effectively reflected off the sensor chip surface. Then that reflected coloured light is also partly reflected back by the sensor microlenses and other sensor toppings. That's it.

Obviously, the purple fringing depends greatly on the properties of the sensor toppings, so even the cams sharing the same basic sensor may exhibit quite different fringing strength. The effect of lens output flux collinearity is also quite understandable, as smaller apertures and longer exit pupil distances both act to make incident flux more collinear and hence the reflections from the sensor wiring spread around less than it would be otherwise at larger apertures and shorter exit pupil distances.

Removal of the wiring from the sensor active surface effectively removes all the purple fringing as it is readily demonstrated with back-illuminated sensors.
(Just check an Exmor-R output here - http://www.imaging-resource.com/PRODS/H ... C00107.HTM )

[agorabasta]

Here's a typical coloured reflection off the sensor wiring. Mind also that it's basic purple colour is attenuated by the primarily green Bayer array, and still the purple reflection is very saturated and strong.

[Greg Beetham]

There could be a certain amount of ambient influence in that reflection, have you got a white light, like a led panel, led ring light, to see if it's still the same colour with all other lights off?
Greg

[agorabasta]

Greg, it's always the same, no matter what lighting. You get a strong purple reflection at oblique angles, and you get a weaker green (Bayer coloured) diffuse reflection at all angles. It's easier to notice with the Nex because the sensor is not sitting as deep inside as it is with DSLRs.
That particular example was shot under a single tungsten source with a yellowish shade.