Wouldn't it make sense for Sony to acquire Sigma's

>>3060621
>backlit
That's not how stacked sensors work, retard.

>>3060686
The new Sony A9 sensor is a stacked backlit, so please explain.

>>3060691
Not him but by "stacked" he clearly meant 3 colors on top of each other.

Sony's "stacked" sensors "stack" the electronics on the photosites, they don't stack 3 photosites on top of each other like Sigma does.

>>3060621
>126MP (42x3)
smegidiots' math is even worse than fourturders math.

>>3060686
You're saying BSI isn't compatible with the Foveon colour filtering?

I call bullshit on that.

>>3060736
I suppose you mean relatively worse, and on an unrelated calculation.

There is light hitting 126k pixels in his model...ergo?

>>3060621
It might be a good idea. But I don't actually know how Foveon works.

All I know is it's a ton a data the sensor outputs. It's so much data that Sigma is unable to process it fast enough to get good video out of the sensor.

But Sony is developing a lot of new ways to increase the speed of the sensor, so these recent breakthroughs they have might actually be very useful to the Foveon sensor.

>>3060691
>>3060785
Foveon is dependent on the thickness of the silicon layer for absorbing certain wavelengths of the spectrum. Blue on the upper parts, red goes the deepest and green is absorbed in the middle. There is literally no difference of a backlit design because you still need certain thickness for the photosites to properly record color information.
The design of foveon is very different from the regular single layer bayer array sensors where there is a benefit of doing a backlit design since there are no layer thickness limitations.

File: foveon-sensor-diagrams.jpg (46KB, 475x474px) Image search: [Google]

46KB, 475x474px

>>3060809
That's kind of fascinating. So Foveon essentially already had "Backside Illumination" build into it from the start?

Either this is extremely smart solution, or this is extremely clever marketing that's good at hiding its caveats.

But I can see this benefit hugely from the faster sensor readout technology in the Sony sensors.

>>3060621
foveon is garbage
high MP is bad for low light

bayer can easily be replaced with CMY, RGBW, etc. which would increase low light performance

>>3060621
Sense or no, Sigma is a family company and making cameras with the Foveon gig seems to be a thing they do because they like/think it's cool, not necessarily because it is profitable.

I'm sure they'd be open to partnerships or licensing for foveon, but unless the company goes in the deep red or something I doubt it has an interest to actually sell its camera business to anyone.

all i want is a monochrome. Is it really that hard to skip adding a bayer filter to a batch of image sensors?

>>3060854
Monochrome images are hard to correct if the lens is bad with CA.

Why not just convert them to black/white in post?

>>3060856
the whole point of wanting a native monochrome sensor is to gain a couple stops of light. you cant do that in post

>>3060858
I see. I guess that could be pretty cool.

But it would be pretty necessary to use lenses that don't suffer from CA in any way.

>>3060859
Thats pretty interesting. You could use color filters; this would slightly negate the extra stops of light, but not by much. You could also use mirror lenses, which do not suffer from CA. Also, many lenses have CA less than the size of the resolution of the image sensor, so it wouldnt get picked up

>>3060621
Foveon should always stay inaccessible, because plebs don't deserve it, and because i like when my images look a bit different. I would hate if it became widely used.

I would love to see improvements in the long exposure IQ department though.

File: spectral response Sigma-dpQuattro-sensor-spectral,sm.jpg (37KB, 424x312px) Image search: [Google]

37KB, 424x312px

>>3060621
>color characteristics of the gods
The colour reptoduction of Foveon X3 is godawful. Stop implying that if you like metamerism of Foveon X3 everyone else should consider it superior.

The RGB layers don't just capture RGB. Red layer captures all colors. Green layer captures all colors. Blue layer captures all colors.

This is the reason they switched to the Quattro design, to make efficient use of this phenomena while improving low-light performance.

>>3060854
Leica does just that.

>>3060863
>You could also use mirror lenses

And for astro photography is makes a lot of sense.
I think most astro CCD's are in fact monochrome.

>>3060825
>having to get through two layers of sensors before hitting the red layer.
I guess that explains the image quality, lol

I just nabbed a mint DP2 Merrill on eBay after selling my Sony e mount 28mm f/2, which paid for it. I'm so hype to shoot with it.

The Merrills cost as much as a good landscape lens on emount so if this works out I'll sell my wides and just keep the A7ii for portraits/50mm lowlight.

After comparing landscapes from the dp Merrills to A7riis I'm sold. Sigma looks like it made an amazing product for the price.

>>3061434
Supposedly the rationale is the long red wavelength has greater penetration power than the short blue wavelength.

This all works very well as smart marketing to fool school children who knows just that from their physics class.
We don't know how efficient it is, or what other caveats it has, as per this Anons claim>>3061394
apparently it isn't as all a dance on roses, apparently some of the red is indeed lost to the blue and green layers.