It took a long time ...

[Neonsquare]

The mirror is a consequence of both OVF** and PDAF (independently). If you remove just one; OVF or PDAF, you currently still need a mirror. If you get PDAF on the main sensor, then the mirror stops being a consequence of PDAF - it would still be one for an OVF. If this would be a case for EVF, is another story. With OVF being the only thing that needs the mirror, it would certainly a valid design decision to drop out both. Actually this even has been done with the mirrorless designs of today by throwing out mirror + pdaf + ovf altogether. There is no OVF without the mirror; you would at least need a design similar to the old Canon RTs: The OVF equivalent of the PDAF mandated SLT design - an OVF camera with CDAF (or main sensor PDAF) and main sensor liveview. You still would pay the price by having a substandard OVF, which is the best reason to think about next generation EVF systems. As long as good OVFs is a fix point on the design list, you will need a mirror that makes a main sensor livestream impossible; which leads to another thing:

There is a growing demand for cameras which enable a live video stream (think "liveview"). Think of Nikons 3D-Tracking using the 1005 pixel rgb metering sensor, but just using the live stream of the main sensor. Sony is - to some degree - already there. They have main sensor metering and they have a restricted form of object tracking using the main sensor (face autofocus). The SLTs are a chance to develop very progressive camera concepts using this livestream and still offering at least the same AF performance as DSLRs get (PDAF). Think of the performance we have in all those current SoC-designs!

**) With OVF here always meaning the through the lens variant, not the much less interesting variant out of the light path through the lens.

[bfitzgerald]

To me SLT is designed to offer AF (decent AF not the Nikon hmm contrast AF) for video and that was primarily the goal of that. Remember all the talk about "we'll do it right" when they do video for DSLR's?

This was their way of meeting that need. Now the problem for stills shooters is many don't give a flying rats about video or hold the line "no compromises for stills cameras to appease video users" SLT is to a degree (you can argue how much) a compromise for stills shooters.

I also think that if the A77 has at least no option to take shots with the mirror lifted that is a serious design flaw for Sony to have. SLT does at least in some cases cause some loss of IQ you can again argue it's minor and I won't disagree. But be under no illusions that video was the primary reason for this (potential for cost savings later on though) As models such as the A77 are primarily going to be for still shooters and enthusiast ones it will be interesting to see how it fares.

You could argue Sony are taking an innovative route or a very "Sony ish" push on tech..or you could say they're not chasing serious DSLR type users any more. Lots of talk about RIP FF Sony DSLR's it seems the companies shift has moved away from photography to these "all in one wonder convergence" devices. Some places like the Luminous Landscape love convergence but IMO it's a slippery slope that eventually leads to handling compromises too.

[agorabasta]

And what about an SLT with a mirror that reflects only the IR? Then that IR may be used for PDAF, and we also move the IR filter away from the sensor surface thus removing that purple ghosting plaguing most of the recent cams. (That's very similar, yet quite distinctly different from the normal PF.)
Then add a bit of superfast CDAF confirm to keep the AF always perfect. And I mean only the confirm function, without focus readjustment if the focus area contains truly sharp edges.
So no light loss from the mirror, much less purple fringing, super fast and precise AF...

Sounds too good to be true?

But Sony already filed a patent for a design putting the IR filter onto the SLT mirror; and those IR filters do normally reflect the IR rather than absorb it...

So some great stuff may really happen. The question is - when?

[bfitzgerald]

A lot of things sound too good to be true but the problem with the camera industry isn't new technology it's the ability to decide what's actually more practical for the end user. I many cases they do things that are not liked such as removing OVF's from most compacts for one..we already had the pointless power focus and zoom from years past (yes mechanical is better for this)

[Greg Beetham]

Did that purple edge stuff come from just the SLT's? or all of the cameras that have the new type sensor design with the sensels packed much closer together?
I'm not sure having AF work good in the IR section of the spectrum would benefit us poor humans whose receptivity to IR is almost nil, I mean we can see up to around 700nm and IR begins properly at about there and goes on to 1000nm, none of which we can see. What I'm getting at is, what if such an AF system focuses on things emitting strong IR preferentially but those aren't what we wanted it to focus on?
Another thing, I'd be reluctant to mention 'fast' and CDAF in the same sentence...myself, so far anyway, I'd be wondering if CDAF would ever get fast enough to challenge PDAF after reading David's description of how it works.
Greg

[agorabasta]

Greg,

That excessive purple fringing/ghosting is exactly the same between a55/a580. Moreover, it's about the same with 60D and D7000. And it's somewhat weaker with the K-5, though shows the very same geometry. You can check for yourself at the imaging-resource.com looking at the samples taken with fast wide-open macro lenses like this one - http://www.imaging-resource.com/PRODS/A ... 03.ARW.HTM . It all is really about the modern trends in the sensor toppings design that are all quite well shared throughout the industry.

The AF in IR part of spectrum translates into the visible part of spectrum in a very simple and straightforward way, that's just some quite automatic recalculation/calibration.

The CDAF confirm is as fast as it's possible for the system to evaluate the sample area image. If the critically sharp edges are there, the confirmation is granted without the adjustments. The whole process is as fast as are the necessary calculations.

[agorabasta]

Barry,

The manu's are there to make money and not to please you in particular. In fact, it's up to you if you can take pleasure in their offerings or you can't. Then if no pleasure for you in those offerings, it's just no luck for you. And if it's no pleasure for all the target group, then it's no luck for the manu'.
All's fair, I guess...

[bakubo]

Sony needs an entry-level model without an EVF to make it smaller/lighter and for people who want LCD-only and to attract digicam upgraders. Many users not only don't care about an EVF they dislike them. No EVF, no IBAFM (in-body AF motor), maybe no IBIS, and maybe no tilt/swivel LCD, and no built-in flash. I have seen quite a few posts from people who dislike any kind of VF (EVF and OVF), dislike IBIS (some also dislike ILIS), dislike built-in flash, really like huge lenses attached to small bodies, but want fast PDAF for video and still photos. Sony needs to mop up that group of potential customers. Also, either completely remove the hand grip or make it much smaller since this is an LCD-only camera meant to be held out in front of you. Sony could make it quite a bit smaller, thinner, and lighter than the A33. It could just use all the newer Sony SAM and SSM lenses. Maybe even an A11 (without IBIS and no tilt/swivel LCD) and A22 (with IBIS and tilt-only LCD). I think Sony should go completely to a touchscreen and remove all buttons and other controls for this camera. Might even consider making the shutter button just a touchscreen control.

I did a very quick and dirty Photoshop job to show the new A22 or A11.

[bfitzgerald]

Yuck! Might as well just shove an A mount on a NEX and that at least looks reasonably attractive!
Agorabasta you are right they are not there to "please me" but my views do represent a significant user base out there some folks would agree with my stance at least in general. You can't package all users up in the same box with a neat bow on it and job done. The SLR market exists to please a minority (if taken overall camera sales) of photo users many of whom are very serious about this and have specific needs that many users won't even think about let alone want.

You have to treat each segment for the needs of those users.

Cameras are like computers sometimes the technology might be interesting as a talking point, but on it's own it's not always very relevant to day to day needs. Computers are actually pretty boring and quite straightforward..it is what you can do with them that is a lot more exciting. Same for cameras..live year by year aching for the next big tech change, will ultimately lead to disappointment as human hands have to use them there will be limits to how radical you can make them.

So imagine a world of only EVF cameras no mirrors and no shutters and all makers are doing this. What next? Where does the next "tech fix" come from??

[agorabasta]

Henry,

Your proposal is almost there

You just need to make one more step - remove all controls off that black box. Then add a fast link to a computer, I mean a link as fast as USB3.0. That would be the ultimate tethered solution. Flashes would need to learn Wi-Fi, I guess...

Funnier yet, there is a market for such a thingie...

[Greg Beetham]

I think Henry might be on to something there, as long as those entry levels can do movies they'd be great...you'd also need em to be able to take the odd still of course.
Greg

[bakubo]

I went back and did a bit more work to more clearly and fully show what, I think, is Sony's vision for the entry-level of A-mount bodies. This is even smaller and lighter. It has a touchscreen for all functions. I left the lens release button but, I think, this could be removed and added to the touchscreen too. It could be made even a bit smaller and lighter, but how much unpaid work do I really want to do for Sony?

[Greg Beetham]

I think Sony might be able to make the body a bit thinner than that though Henry, if it doesn't have IBIS that is, and it could use a cellphone type battery so it could be as small as possible. But don't worry about showing the new thinner version I think you've probably done enough free consultancy work already....all we need now is for Sony to pick up the ball and run with it I reckon.
Greg

[bakubo]

I think Sony might be able to make the body a bit thinner than that though Henry, if it doesn't have IBIS that is, and it could use a cellphone type battery so it could be as small as possible. But don't worry about showing the new thinner version I think you've probably done enough free consultancy work already....all we need now is for Sony to pick up the ball and run with it I reckon.

The top view already has about 5mm or so of the thickness trimmed out to make up for no IBIS. Yes, with a smaller battery (good for maybe 100-150 shots) would allow an even smaller body.

[Greg Beetham]

I bet it won't be long before someone demands a FF version too. Also the video will have to be in 3D of course.
Greg