Kodaks spec for the E1 sensor mentions microlenses, however we also need to take into account the efficiency of older microlens designs an less dense sensors, Clearly this is one area which has advanced.
as does the deeper sensor crop, 2x will be more resistant than 1.6x
--
Oblique angle = angle of light cone. Sorry that is the right interpretation of this little bit of web text translated from the French. To get to it you need to work through it, not just rely on one particular interpretation of a single (translated) source. I got onto it when I used to go around saying there were no downsides to smaller pixels. I was challenged on the basis of this f-number drop-off and was referred to this paper:
'Stacked microlenses' are one solution. It's like making a retrofocal microlens, it can have a longer back focus for a short FL, and the inner lens can actually be made to fit inside the light tube formed by the metallisation.
That has nothing to do with this issue. Telecentricity does not change the angle of the light cone, it is constant for a give f-number.
Lots of the tricks used for phonecam and P&S sensors can be applied, light pipes, dual element microlenses, and even BSI.
That is true. The factor that matters is the silicon fill factor relative to the microlens so the olf 5D fares well despite having a really inefficient sensor because its huge pixels allow a high effective fill factor under a less than 100% microlens.
from JW
--
since no anti-equivalence clowns have bothered to respond to you I shall proffer an answer on their behalf. On the matter of framing or perspective when I shoot photos, I can say absolutely that I care, but likely care so much I NEVER waste a moment on relating what im doing to 135 or any other format.
i believe it, but Im not going there to check
Just stop and think about the geometry of the situation. Suppose we have an f/2 lens. The rays reaching a point on the sensor from the exit pupil fill a cone from that point to the exit pupil. The angle of that cone is 28 degrees. It is the same for all f/2 lenses, and does not depend on the position of the exit pupil. Take an f/1.4 lens, the angle of the cone is 39 degrees, again, independent of the position of the exit pupil. Take an f/1.0 lens and that angle is 53 degrees. It is thin increasing angle that causes the rays at the edges of the light cone to fail to be refracted onto the photoreceptor. This is an effect that occurs even in the centre of the field, where the light cone is perpendicular to the sensor, telecentric or not.
i believe some bloke did it in his spare time.....i am not joking btw.Where'd you get this chart:
It's not hosted on DxOMark. Seems you'd want to link the source, right?
The question really is the source of data for the EM-5, and whether it is comparable with the source for everything else (DxOmark), and credible.i believe some bloke did it in his spare time.....i am not joking btw.Where'd you get this chart:
It's not hosted on DxOMark. Seems you'd want to link the source, right?
Even though those two requirements might make sense for Nikon/Canon they don't make any sense when you are talking about the m43 System since it's a much more open format and there are amazing gems that compete with the best of the best m43 glass. The two F/0.95 Voigts or the Samyang 7.5mm (which exceeds the Panasonic fisheye solution in almost every IQ aspect) are notable examples.
Your recollection is at fault, then. Microlenses have been an integral part of image sensor design since the 1990's. The Kodak sensors around which the Four Thirds system was designed all had microlenses.
The need for 'telecentrity' arises because of the use of microlenses, which have a limited acceptance angle. The FillFactory CMOS chip used in the Kodak DSLR's did away with microlenses (note, it was not pre microlens, it was designed specifically as a supposedly superior alternative) as an approach to the same problem. And of course, 'telecentricity' was a sensible design course and still is, which is why all the manufacturers used it in their digital lenses.
The reduced sensor to mount distance does not make telecentricity more difficult at all, in fact it makes some interesting telecentric optical configurations possible. The provision of a distant exit pupil does not mean that there must not be any elements close to the sensor, in fact there can be with advantage, typically field flatteners. The poor thinking of Olympus, IMO, was imposing a long register which excluded such solutions, and also more compact non-telecentric lenses when offset microlenses became available.
Bob & Joe, see here:The question really is the source of data for the EM-5, and whether it is comparable with the source for everything else (DxOmark), and credible.i believe some bloke did it in his spare time.....i am not joking btw.Where'd you get this chart:
It's not hosted on DxOMark. Seems you'd want to link the source, right?
Part of the reason for the register is the mirror. Olympus DSLRs don't have a lot of clear space. Then when you get the sideways-swinging mirror in the E-3xx you have no clear space.
I'm pretty sure they said that a long register assists telecentric design. You seem to be arguing that it doesn't, but also seem to be adding extra lens elements. Given that lens designs are a compromise, and must be competitive, are Olympus more correct on this point than you?
Olympus never claimed "unique", to be fair. They did, indeed, push the designed for digital from scratch message. At the time, a good idea.
I haven't seen that claim by Olympus.
Thanks, and since he posts the raws, he has allowed others to validate it if they want to.Bob & Joe, see here:The question really is the source of data for the EM-5, and whether it is comparable with the source for everything else (DxOmark), and credible.i believe some bloke did it in his spare time.....i am not joking btw.Where'd you get this chart:
It's not hosted on DxOMark. Seems you'd want to link the source, right?
http://forums.dpreview.com/forums/read.asp?forum=1041&message=41615417
I find the sensor diagonal to register ratio is basically no different from anything else, and certainly not an outlier.
--
All DSLR's have a mirror. Canons have a register 1.83 times the frame height, the FT register is 3.0 times the frame height.
If they engineered it as Canon did, they could have a 24mm register. Kind of interesting, if they'd made the register just 5mm more on mFT, they could have used the mount also for DSLRs.
How would it do that? Take the same design, put an extension tube behind it, put it on a short regisetr. hoe has the long register aided that design?
I am not adding them, I say you can add them, with advantage. A short register give that opportunity.
Well, we're only going on your memory of your interpretation of what Olympus marketing material said. That is a long way from what their optical designers said and thought, which you simply don't know.
Yes, an excellent marketing idea. However, whatever they really said, a fair few of the Olympus faithful seem to have the idea that it only applies to Olympus.
Again, it is something some people seem to have got into their minds. Otherwise we would not have these interminable propositions that FT is better because of telecentricity, when it is something all systems have and Olympus does not have universally.
so you are saying this table by JW
--
