Sounds great if you could get iso 400 quality at iso 1600 sensitivity...or iso 800 quality at iso 3200 sensitivity...

At some point you have to ask yourself, is it possible? Shooting in complete darkness at 1/50th second?

I think its possible, and might even be a reality at some point in my years. Not that I'm implying that this technology is capable of that, but it seems a direction that is trying to be obtained.

To me I'd rather get 16 stops or more out of a sensor, but maybe that is being worked on too?

Carl

I doubt it. This will be on the next camera that they come out with after that. It's an ever improving thing. Impossable to stay on top of the curve.

Dave

Although Sony does tend to shock and surprise. But I lean your direction I think overall.

I did daydream and think "wow, all those comments about how 24mp is too much on a FF sensor, what about low light performance yadda yadda.... this could totally fix any possible issue with that".

:)

If Sony did include this surprise backlit array with higher ISO and low noise in the Flagship, it would get the attention of the world and blow the competition out of the water. We don't know how well the back lit array can be scaled. Can it be made in a full frame size? We will soon know. Sony is great for thinking outside the box. Let's move the anti jitter from the lens to the array... Lets flip the chip and let the light hit the bottom side and avoid the land wiring patterns.... Obvious now that we see it done. "Why didn't I think of that" says I, and the engineers at Canon and Nikon. Maybe because Sony has been making your Image arrays (the heart of the camera) for you all along. I'm sure Sony has patent protection on both of these great ideas and will have a legal monopoly for the next 17 years. Are you listening Sony?

I think they know.

Dave

...

Where is Sony going to be 2-5 years from now in DSLR technology.

Great comments Dave :) Excellent discussion.

Carl

A very quick (5 minute) patent search shows that Sony received a US patent on how manufacture a back-illuminated solid state imaging device, on Sep 5, 2006. Way before they said anything about a FF flagship. Not sure what this means but it does show that Sony knew how to do this for many years. It is a certainty that Sony had this patented in Japan way prior to applying for a US patent. That's the way it's done. For your information the US patent number is 7,101,726

This is only one patent. I'm sure Sony has others. Before I retired and became a full time photographer/hiker, I was a patent searcher. Before that I worked at IBM where I designed and patented and published a couple imaging arrays and how to manufacture them.

I just wondered how long Sony has been sitting on this back lit thing. The answer is years!

Please don't ask me details or questions about this array. I don't know or care and it's too much like work to find out. I'm retired now.

A brief abstract of the patent:
A solid state imaging device having a back-illuminated type structure in which a lens is formed on the back side of a silicon layer with a light-receiving sensor portion being formed thereon. Insulating layers are buried into the silicon layer around an image pickup region, with the insulating layer being buried around a contact layer that connects an electrode layer of a pad portion and an interconnection layer of the surface side. A method of manufacturing such a solid-state imaging device is also provided.

Dave (the hiker)

I see two advantages to this:

Larger light receptor with the same pixel count

And looking at the diagrams, a sensor that sees the lens image more like a film surface.

Could this be the reason the Flagship is taking longer? could this be a 24MP FF sensor with the low light performance of an 8MP CMOS?

Pardon me for droolling....

if you mean that the surface will be less reflective. This could be good news for folks with old Minolta lenses. They will be less likely to suffer the spot due to reflections between the lens and imaging array. Ken, I don't see how back illumination will make a big improvement in light sensitivity. The drawing provided above were for conceptual purposes and are not to scale. I think the advantage of a back lit array will be that you can now put more circuitry on the front side and not mess up the light path, less fringing from adjacent cells. I would expect to see better and faster on chip A/D conversion. This could result in less noise.

Like you, I can't wait to see what the flagship brings to us.

Dave

That I lack the technical backgrounds of both of you. However, I can still keep up and understand where this is all going.

If this technology isn't in the FF sensor, the good news is that the photosites are still bigger than the A700's and should perform just a bit better overall in noise etc.

I figure thats the least we can expect from the FF sensor.

Very interesting background you have Dave, quite different than my day job ;).

Carl

The diagrams may not be scale, but..

Normal Sensors lose light in two ways from what I can see
A) Some of the circuits HAVE to take up space that blocks light.
B ) putting the light sensitive surface at the back means that light the hits at an angle doesn't always make it to the sensor. This has been an issue with smaller apertures on digital as the aperture gets smaller the angle of the light hitting the edge of the frame is greater so more loss. Its almost like a the sensor looks down a small tunnel.

This array puts the all of the light sensing area at the top..so no circuit shadows and no tunnel effect. It should perform better in low light and might perform better in light that needs a smaller aperture.

I understand your point. I was not aware of the issue of tunnel effect with small apertures, thank you. I read someplace that this back lit array has a 2X improvement in light sensitivity. The more I think about it, the more I think that this is the array that will be used in the flagship. A larger cell with an unobstructed light path has got to result in better sensitivity. I have read about all the on-chip/image array data processing that they will doing on the flagship, and wondered how they were going to squeeze all that stuff between the cells. Now we know.

I think you made a typo on your point "B" above. Did you mean to point out that some light coming from an angle is lost when the sensor is on the front, but not lost when it's on the back? I think that was your intent.

I'll break open my piggy bank when this comes to market. First the camera, then a new flash to trigger my old flash, then an expensive lens to complement the potential of the array, a new case to carry my new bigger equipment, spare battery, more memory, etc.

Sony is going to make some money from me and many others. I look forward to it! I'm leaving for a cruse to Alaska in early September that will include a hike. I hope I get a chance to have the Flagship with me.

Dave

the naming of the technology is confusing...
The name is about how the engineers see the chip. but in reality back lighitng really means the light senative media is on the front of the sensor by putting what is traditionally the bottom/back of the chip up front. It means that they have to build the chip differently with the first layer of material goes on last.

So the light hits the sensative media first since it is in "front"

I think that if Sony does start stuffing this technology into the Alpha line they should withhold selling it to competitors for a while. In my opinion, Sony needs their own D3 and this technology could pave the way. They need a flagship camera that will tempt pros to jump ship with its status quo busting abilities.

Hi Brian,

This is interesting technology, and like you I think Sony should keep the best bits for themselves. However as was pointed out on DPR, Sony sells way way more sensors than it does DSLRs (at the moment at least) and with the rate at which Samsung is progressing- see the GX20 and Pentax K20d- the last thing they are gonna want to do is push Nikon away.

Perhaps they could have some form of exclusivity of the new sensor for say 6-12 months, and then sell it to Nikon et al? An if it went into the flagship then yes it could really be something. Imagine if the flagship had say 1-2 stops better noise than the EOS 1Ds MKIII?

Andrew

Better yet, Imagine if the flagship had the ISO performance of the D3 and a higher pixel count than the 1DsIII...that would be something to write home about.

I thought Nikon was trying to wean themselves off of Sony sensors a bit? As long as Sony is building their image sensors, Sony will have some advanced knowledge of their competitors camera. That alone stands a good chance of driving Nikon away IMHO.

and good point about the competitive advantage. I'd imagine their reliance on Sony sensors must be quite a prickly subject for them. I wonder if there's ever been any kind of official stance by them on that subject, or if Nikon just prefers to avoid any public discourse on the matter.

I do not think that flagship's chip will use this - the press release states they "announced the development" of 5mpix chip. I understand it to be far away from 24mpix FF chip.
sorry... I think maybe in Flagship II

If one holds off on buying until the next wonderful gizzmo that's just around the corner, you will never buy anything because it's a moving target.

Dave

The Sony A900 has by far the most noise of any FF sensor on the market (check out dpreview.com or any other pro review). This technology--IF it's everything Sony advertises--will only bring the A900's noise level down to Canon 5D Mark II levels if they're lucky (and by then who knows what the rest of the competition will have). Sony's marketing department is getting you too excited. I bet you the real-world improvement will be very small and definitely not a doubling of the sensitivity.

Is only one of the holy trinity of image quality factors.

The A900 is a bit noisy, but it also has other highly excellent qualities that make up for that. It is only about a half stop to a full stop behind the competition on noise which really isn't as much as some dubious sports tend to think, but, the A900 is at least a full stop ahead on color retention on higher ISO's. You don't get something for nothing with other brands either.

Some people read too much into too many reviews that are slighted by the hand that feeds them, along with the forums on those sites that occupy them.

Its so easy to be a follower.

Carl