Reading mode: Light Dark

Are 14mp the limit of a 24 x 36 mm chips area ?.

Go to last post
Mikael Risedal

Mikael Risedal

Senior Member
Messages
4,646
Reaction score
2,053
Location
skane, SE
Have we now reached the limit of pixels on a 24 x 36 mm chips area ???.

A 14 mp chip have a pixel size of about 6 micron. Because of lens difraction problems the maximal f -stop will be F6. This means also that all f-stops above ( example F-stop 11 ) will not gain a higher resolution or better results from the CCD or CMOS today . Because of the smaller iris ( F-stop 8. 11. 16.) the light beam will bend more than 6 micron and not only reach one pixel at the time.. This phenomenon is caused by the lens difraction . F-stop 8 are similar to 8 micron difraction and so on.

Shall all new lenses have the maximum resolution and contrast up to about F-stop 6 on a Kodak 14 mp camera body ???.

And what about the future , let us speculate in a 22mp 24 x 36 mm chip , now the pixel size is about 3.8 micron and a lens at F-stop 4 produce a maximum result ???? What kind of DOF does this camera have?

Difraction problem or not, does the physical law of light put an end of more smaller pixels on a 24 x 36 mm area.??

Can some psysical lens and ccd experts help me out here
Best regards
Mikael Risedal
Photographer
Sweden
 
Mikael Risedal said:
Have we now reached the limit of pixels on a 24 x 36 mm chips
area ???.
Click to expand...
I don't see any problem about moving to 30MP, if you see some of the existing consumer sensonrs in 3/4 MP cameras they are already denser that that. The big question is why do it?

Smaller sensors will not have the same dynamic, diffraction will create other problems as well. I don't think that we are goign to have much more quality out of a 30MP sensor. Eventually a similar sensor will have better quality than today's Kodak 14MP but when this will happen will be also possible to have a much beter 14/15 MP.

Instead of going to 30MP I would prefer to stay in the 14/15 MP on 3 layers. That would be a kick a$$ sensor.

--
Regards
Gabriele Sartori
 
Already now on a 6MP Camera and F-stop 8 and above that, causes diffraction problems. If the light bending so much inside the

lens when its passing the iris ( f-stop8 and above) and the light area will be bigger and hits more than one pixel at the time, then the other ones seems to me unnecessary. ????

Now it is more than one pixel who take cares of the same information.

Maybe Im out and bicycling here (Swedish expression)
Best regards
Mikael Risedal
Gabriele Sartori said:
Mikael Risedal said:
Have we now reached the limit of pixels on a 24 x 36 mm chips
area ???.
Click to expand...
I don't see any problem about moving to 30MP, if you see some of
the existing consumer sensonrs in 3/4 MP cameras they are already
denser that that. The big question is why do it?
Smaller sensors will not have the same dynamic, diffraction will
create other problems as well. I don't think that we are goign to
have much more quality out of a 30MP sensor. Eventually a similar
sensor will have better quality than today's Kodak 14MP but when
this will happen will be also possible to have a much beter 14/15
MP.

Instead of going to 30MP I would prefer to stay in the 14/15 MP on
3 layers. That would be a kick a$$ sensor.

--
Regards
Gabriele Sartori
Click to expand...
 
Mikael Risedal said:
Already now on a 6MP Camera and F-stop 8 and above that,
causes diffraction problems. If the light bending so much inside
the
lens when its passing the iris ( f-stop8 and above) and the light
area will be bigger and hits more than one pixel at the time,
then the other ones seems to me unnecessary. ????

Now it is more than one pixel who take cares of the same information.

Maybe Im out and bicycling here (Swedish expression)
Best regards
Mikael Risedal
Click to expand...
If I can join in this small discussion (so far) I am not sure smaller pixel areas yield better results. I agree with designers that increasing active area to 24x26 is good (yields a "no multiplier" factor) but decreasing pixel size doesn't make an image better per se. Color depth, etc. are strong factors here as well. That's why the original D1 was such a strong performer with a small magapixel rating (the pixel size is so large). We need to constantly be thinking about the final output of the image. Many folks want really high megapixel files but they will output to printers with low pixel requirements (because of ink bleed). I don't have much to contribute on the diffraction issue other than even with film cameras f22 yields very high diffraction. There are other means to get "depth of field than to 'stop down'.
--
Ross Frid
http://users.netonecom.net/~rsfrid/
 
36x36mm would go further into the fringe area of the lens, beyond its design tolerance. If you want a square imager with the same diagonal, you'd be limited to 30.6mm on a side. That's only 8% bigger area.

No idea what would happen to the mirror or other internals. Just doing the math on the lens circle.
Joev said:
Mikael Risedal said:
Have we now reached the limit of pixels on a 24 x 36 mm chips
area ???.
Click to expand...
Will the real 100 please stand up?
Click to expand...
--
DaveA
 
Ross said:
Mikael Risedal said:
Already now on a 6MP Camera and F-stop 8 and above that,
causes diffraction problems.
Click to expand...
Click to expand...
Ross25784 said:
If I can join in this small discussion (so far) I am not sure
smaller pixel areas yield better results.
Click to expand...
We are all saying the same things. Also please consider that 14MP on a full frame has about the same pixelsize of the D100. I don't think is an accident.

I would keep 14MP for the next 100 years (OK, I'm exagerating), there is so much to do beside the pixel count. Sensitivity for one. The Kodak doesn't have a dramatic sensitivity. One day we will have ISO6400 and no noise at all. Triple layer that increase the quality without increasing the diffraction, the moiree and with limited low pass filter.

We can prepare our wallets and toss money every 2/3 years even if pixel count will not increase. We will have other reasons for and they are all good!

--
Regards
Gabriele Sartori
 
Also, some lenses are designed specifically for 24x36 - eg. the Nikon 200mm f/4 micro actually crops out the parts of the image circle that are unused.

I noticed this when looking down into the lens barrel.
  • Andrew
 
Why is the image diffraction limited at only F6? I understood that diffraction depended on the physical size of the diaphragm opening. A full frame SLR will use exactly the same lenses as film and that is not generally regarded as diffraction limited until about F16....
Mikael Risedal said:
Have we now reached the limit of pixels on a 24 x 36 mm chips
area ???.

A 14 mp chip have a pixel size of about 6 micron. Because of lens
difraction problems the maximal f -stop will be F6. This means
also that all f-stops above ( example F-stop 11 ) will not gain
a higher resolution or better results from the CCD or CMOS today .
Because of the smaller iris ( F-stop 8. 11. 16.) the light beam
will bend more than 6 micron and not only reach one pixel at the
time.. This phenomenon is caused by the lens difraction . F-stop
8 are similar to 8 micron difraction and so on.
Shall all new lenses have the maximum resolution and contrast
up to about F-stop 6 on a Kodak 14 mp camera body ???.
And what about the future , let us speculate in a 22mp 24 x
36 mm chip , now the pixel size is about 3.8 micron and a lens
at F-stop 4 produce a maximum result ???? What kind of DOF
does this camera have?
Difraction problem or not, does the physical law of light put an
end of more smaller pixels on a 24 x 36 mm area.??

Can some psysical lens and ccd experts help me out here
Best regards
Mikael Risedal
Photographer
Sweden
Click to expand...
 
Hi Mikael:

The good news is that your numbers are slightly off. The Kodak 14mp chip uses 8 micron pixels, not 6 micron. 6 micron pixels would result in a 24mp camera.

6 micron pixels equates to about 83 lp/mm. A diffraction-limited lens at f/8 has an MTF of more than 50% at 83lp/mm. At f/16 the MTF at 83lp/mm drops to about 15%, which is still detectable. Therefore, a 24mp chip would be completely compatible with good 35mm lenses.

If you drop down to 3.5 micron pixels, which is the size used on most consumer digicams, then you get more than 70 megapixels on a 24x36mm chip. In order to take advantage of this you need to use a diffraction-limited lens stopped down to no more than about f/8, which gives about 25% MTF at the resulting Nyquist frequency of 142lp/mm. This is possible when using the best lenses available, but flawless technique will also be required.

Given all this, I'd be very surprised to see more than 70 megapixels in a 24x36mm format, so this might be thought of as an upper limit.

Brian
Mikael Risedal said:
Have we now reached the limit of pixels on a 24 x 36 mm chips
area ???.

A 14 mp chip have a pixel size of about 6 micron. Because of lens
difraction problems the maximal f -stop will be F6. This means
also that all f-stops above ( example F-stop 11 ) will not gain
a higher resolution or better results from the CCD or CMOS today .
Because of the smaller iris ( F-stop 8. 11. 16.) the light beam
will bend more than 6 micron and not only reach one pixel at the
time.. This phenomenon is caused by the lens difraction . F-stop
8 are similar to 8 micron difraction and so on.
Shall all new lenses have the maximum resolution and contrast
up to about F-stop 6 on a Kodak 14 mp camera body ???.
And what about the future , let us speculate in a 22mp 24 x
36 mm chip , now the pixel size is about 3.8 micron and a lens
at F-stop 4 produce a maximum result ???? What kind of DOF
does this camera have?
Difraction problem or not, does the physical law of light put an
end of more smaller pixels on a 24 x 36 mm area.??

Can some psysical lens and ccd experts help me out here
Best regards
Mikael Risedal
Photographer
Sweden
Click to expand...
--
J. Brian Caldwell
http://www.caldwellphotographic.com
 
brian said:
Hi Mikael:
The good news is that your numbers are slightly off. The Kodak
14mp chip uses 8 micron pixels, not 6 micron. 6 micron pixels
would result in a 24mp camera.
Click to expand...
Speaking of 24mp. I have been wondering lately if a manufactor would ever consider building at 24mp bayer pattern camera that the largest output would be a 6mp full color image. I know that images that are resized 50% look fabulous. If this was done in hardware mabye the camea could still have a fast capture speed (compared to saving a 24mp image)

4000x6000 bayer pattern array 50% downsize = 2000x3000 full color image = sharp and accurate 6mp image.

--
chris robey
http://www.chris.robey.com
http://www.pbase.com/crobey/root&view=recent
 
Mikael Risedal said:
Have we now reached the limit of pixels on a 24 x 36 mm chips
area ???.
Click to expand...
I think the answer is this: we're not at the physical limit, but we're
getting into the area of diminishing returns. As we get beyond 14MP or
so, we're reducing the number of images that would be helped by
a higher pixel count. Others' calculations are probably sound theoretical
work, but every lens is an engineering compromise. Lots of lower-cost
zooms (i.e. the most popular lenses) would be hard-pressed to match
even 14MP resolution. Even the best lenses need to be focused
perfectly to meet their potential, and this doesn't always happen.

There will always be the Contax/Leica crowd who will clamor for more
resolution; some of them will actually use it :).

I suspect that we will home in on a hierarchy of cameras in the future;
perhaps 24MP for the premium crowd, 12MP for most uses, and
perhaps 6MP for lower cost. The last category may never exist,
as the price may never drop to the level of EOS Rebel and its ilk. And
there will be super-high-end alternatives like the SInar back, but with
more like > =40MP.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.
 
I suspect that that the mega pix will keep going up.

Just look at Camcorders with 300 x zooms. They're no use what so ever but it's so easy to say it's better then 200x , some how.

More pix would help with the dead pix issue and could be averaged out to give more precise colour etc

The Kodak 14n is using 8um~ photsite on a .5 fab. which is way below what possiable, I understand the D60 uses slight smaller photosite on .35 fab and Sigma may using .25 fab. So it seams likely that either next year or soon afer that 24m pix DSLR's will be out.

I full agree with those who ask the questions do we really need more pix or some of the other issues solving first.
Alex
Stephen H. Westin said:
Mikael Risedal said:
Have we now reached the limit of pixels on a 24 x 36 mm chips
area ???.
Click to expand...
I think the answer is this: we're not at the physical limit, but we're
getting into the area of diminishing returns. As we get beyond 14MP or
so, we're reducing the number of images that would be helped by
a higher pixel count. Others' calculations are probably sound
theoretical
work, but every lens is an engineering compromise. Lots of lower-cost
zooms (i.e. the most popular lenses) would be hard-pressed to match
even 14MP resolution. Even the best lenses need to be focused
perfectly to meet their potential, and this doesn't always happen.

There will always be the Contax/Leica crowd who will clamor for more
resolution; some of them will actually use it :).

I suspect that we will home in on a hierarchy of cameras in the
future;
perhaps 24MP for the premium crowd, 12MP for most uses, and
perhaps 6MP for lower cost. The last category may never exist,
as the price may never drop to the level of EOS Rebel and its ilk. And
there will be super-high-end alternatives like the SInar back, but
with
more like > =40MP.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.
Click to expand...
 
I have no idea where you got the idea the 14n had 6 micron pixels.
If it did, we would be about 24 MP and not 14. The 14n has 8 micron
pixels and will hit diffraction problems around F11-F16.

I do see this ending at around 6 micron pixels, however.

Steven
Mikael Risedal said:
Have we now reached the limit of pixels on a 24 x 36 mm chips
area ???.

A 14 mp chip have a pixel size of about 6 micron. Because of lens
difraction problems the maximal f -stop will be F6. This means
also that all f-stops above ( example F-stop 11 ) will not gain
a higher resolution or better results from the CCD or CMOS today .
Because of the smaller iris ( F-stop 8. 11. 16.) the light beam
will bend more than 6 micron and not only reach one pixel at the
time.. This phenomenon is caused by the lens difraction . F-stop
8 are similar to 8 micron difraction and so on.
Shall all new lenses have the maximum resolution and contrast
up to about F-stop 6 on a Kodak 14 mp camera body ???.
And what about the future , let us speculate in a 22mp 24 x
36 mm chip , now the pixel size is about 3.8 micron and a lens
at F-stop 4 produce a maximum result ???? What kind of DOF
does this camera have?
Difraction problem or not, does the physical law of light put an
end of more smaller pixels on a 24 x 36 mm area.??

Can some psysical lens and ccd experts help me out here
Best regards
Mikael Risedal
Photographer
Sweden
Click to expand...
--
---
My really bad Fall Adventures:
http://www.pbase.com/snoyes/fall_adventures_2002
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top