D30 Replacement

[David SL]

David:
What you are essentially arguing for is a true medium format
digital camera. This will work better than a 35mm digital camera
just as a medium format film camera works better than a 35mm film
camera. Your idea for modifying normal 35mm lenses to cover a
larger format is similar to the old Speed Magny made by Nikon back
in the 1960's to convert F-mount lenses for use on 2.25x3.25 or 4x5
film. The Speed Magny used a relay lens to magnify the 24x36mm
image up to the desired size.

A couple of points:
1) In your proposed system the internal lens would need to have
negative power, just like a teleconverter. You show a positive
lens.

You are correct the lens should be concave(divergent), I put the shape in my graphic without really thinking about it.

2) A single element teleconverter will produce an absolutely
horrendous image, no matter how well you grind and polish the
surfaces. There is a very good reason that high performance lenses
have multiple elements.

I researched it further, you are absolutely correct. Multiple elements of both convex and concave glass are needed to produce a regular(distortion free) image, thanks for bringing it up.

3) Teleconverters disperse light, they don't concentrate it. Thats
why they cause a loss of system speed: a small image gets dimmer
when you magnify it. Front converters are afocal, so they don't
change the system speed.

Right again, I researched it and my intuition on the process was incorrect.

4) A better solution would be to develop a large sensor for
existing medium format cameras. Kodak has already done this with
their ProBack, which has an area of 35x35mm and has 16megapixel
resolution. Existing medium format lenses will work better than
modified 35mm lenses.

I agree with you here, enhancing existing medium format technology might allow a convenient middle ground to be reached as sensors move up to film size.

Thanks for your corrections.

Regards,

Dsl

--DSL

 

[brian]

Hi Andrew:

I think it should be possible to make a sensor with 5 micron pixles that produces clean images. After all, 5 micron pixels have more than double the area of the 3.45micron pixels used in most consumer cameras nowadays. Also, when Nikon sliced the D1 pixels in half to produce the D1x (loosely speaking here!) they managed to create a camera that produces less noisy images than the original. I don't think that we have seen the bottom of this particular technological barrel.

If clean 30-35mp single-shot cameras ever do appear on the market they will challenge large format as well as medium format film. Some of the 25-50mp mosaics I've put together in the last few months compare amazingly well to 8x10!

Brian

Hi Andrew:
I use a D1x, and have tested it extensively with a wide range of
lenses at all apertures. The horizontal and vertical resolution of
this camera is 84cycles/mm and 42cycles/mm, respectively. I've
photographed test charts that clearly show contrast reversal at
precisely those frequencies in both the center and all four corners
of the image field. This means that the ultimate limits of the
lens was not yet reached, even at 84 cycles/mm. Admittedly, some
lenses are better than others, but the conclusion is inescapable:
5 to 6 micron pixels CAN be realistically utilized by 35mm SLR
lenses. That means that a 30-35mp 24x36mm sensor is a reasonable
goal.

Well thats good new Brian, if they can sort the noise issue out,
pixels that size will completely kill film, bothe medium format and
35mm.

 

[JCDoss]

Hi Andrew:
I think it should be possible to make a sensor with 5 micron pixles
that produces clean images.

Actually, after musing about this some more in my own mind (which is something I do too often without getting paid for it!), the 10x10um pixels that the D30 has are actually composed of photosensitive and non-photosensitive sites. The photosensitive site is the pixel, and the non-photosensitive site is the processing circuitry that runs each pixel (well, you know what I mean).

Assuming that the actual pixels consume an area no greater than 70% of that area (ie, a 70% "fill factor" which is common for full frame CCDs), then the actual photosensitive area becomes 7x7um. If there were a way to minimize the overflow drains (ie, the non-photosensitive 30% of the pixel) to a negligible percentage of the pixel, then we could say that the whole 7x7um is photosensitive. If this is true, then the size of the pixel shrinks, and the total number of MP on a 24x36mm sensor would leap from 8.6MP (based on 10x10um pixels) to 17.1MP (based on 7x7um pixels) with no change in sensitivity, signal to noise ratio, or image quality.

In reality, the fill factor of the D30's CMOS is probably LESS than 70%. The figures I am using are based on Phil's CCD tutorial (link below), and there is no mention of CMOS. If this is true, then expect the "megapixelage" of a full frame sensor with a maximized fill factor to be even greater with no change in S/N, sensitivity, or image quality.

Maybe this is the way to go.
JCDoss

http://www.dpreview.com/learn/Glossary/Camera_System/Sensor_01.htm

 

[JCDoss]

Hi Andrew:
I think it should be possible to make a sensor with 5 micron pixles
that produces clean images.

Actually, after musing about this some more in my own mind (which
is something I do too often without getting paid for it!), the
10x10um pixels that the D30 has are actually composed of
photosensitive and non-photosensitive sites. The photosensitive
site is the pixel, and the non-photosensitive site is the
processing circuitry that runs each pixel (well, you know what I
mean).

Assuming that the actual pixels consume an area no greater than
70% of that area (ie, a 70% "fill factor" which is common for full
frame CCDs), then the actual photosensitive area becomes 7x7um. If
there were a way to minimize the overflow drains (ie, the
non-photosensitive 30% of the pixel) to a negligible percentage of
the pixel, then we could say that the whole 7x7um is
photosensitive. If this is true, then the size of the pixel
shrinks, and the total number of MP on a 24x36mm sensor would leap
from 8.6MP (based on 10x10um pixels) to 17.1MP (based on 7x7um
pixels) with no change in sensitivity, signal to noise ratio, or
image quality.

In reality, the fill factor of the D30's CMOS is probably LESS than
70%. The figures I am using are based on Phil's CCD tutorial (link
below), and there is no mention of CMOS. If this is true, then
expect the "megapixelage" of a full frame sensor with a maximized
fill factor to be even greater with no change in S/N, sensitivity,
or image quality.

Maybe this is the way to go.
JCDoss

http://www.dpreview.com/learn/Glossary/Camera_System/Sensor_01.htm

 

[Al Goldis]

The light sensitive portion of a CMOS pixel is much smaller than that of a CCD. This is why CMOS sensors are so much noisier than CCDs (without lots of additional on-chip noise reduction).

10x10um pixels that the D30 has are actually composed of
photosensitive and non-photosensitive sites. The photosensitive
site is the pixel, and the non-photosensitive site is the
processing circuitry that runs each pixel (well, you know what I
mean).

 

[vpera]

Assuming that the actual pixels consume an area no greater than
70% of that area (ie, a 70% "fill factor" which is common for full
frame CCDs), then the actual photosensitive area becomes 7x7um. If

But 7*7um is just 49%.

To have 70% of AREA (10um*10um) would mean 8.4 * 8.4 um.

Vesa

 

[Dan Wells]

Someone on this forum suggested using pixels the size used on consumer digicams to produce a camera with a 32 MP full frame sensor. The problem with that is that pixel size contributes at least as much to image quality as pixel number. I upgraded from a Pro90 IS to a D30, and there is simply no comparison in image quality. The Pro90 produces barely acceptable prints at 6x9 inches, while the D30 is still going strong (as strong as slow 35mm film) at 11x14 and probably beyond (I need to try a 12x18 or two, to get anything bigger than that, I'd have to use a bigger printer than I own.) The two cameras are within 10% in pixel count, so the count doesn't seem to be the answer. The CMOS sensor in the D30 has a very nice image quality to it-probably a combination of the sensor technology, those big pixels (less noise with larger pixels) and th e fact that it can produce a 12-bit TIFF from its raw files. I always adjust color in 16 bit mode (Photoshop can only handle 12-bit files as 16-bit), and that makes a huge difference. I've never seen a small-pixel camera that can capture enough color information to make that work!

From my results with the D30, I would guess that an 8 megapixel camera with the same sensor type and pixel size (the "D60") should be able to make a fine-art quality print of 16x24 inches or larger (with judicious and mild use of upscaling in Photoshop or better yet, Genuine Fractals)-I upscale my 11x14s so I have a 240 dpi file at print size. Does anyone ever really print bigger than 16x24 from 35mm film? Certainly not for detailed viewing-even Velvia falls apart beyond that. Most photographers who intend to produce 30x40 inch gallery prints start out with medium format or even 4x5 film- bigger camera, bigger film, almost always using a tripod. The digital equivalent to that is Kodak's Pro Back, NOT a handholdable, 35mm style camera.

-Dan

 

[JCDoss]

Assuming that the actual pixels consume an area no greater than
70% of that area (ie, a 70% "fill factor" which is common for full
frame CCDs), then the actual photosensitive area becomes 7x7um. If

But 7*7um is just 49%.
To have 70% of AREA (10um*10um) would mean 8.4 * 8.4 um.

At least for me it is! You're right, vpera!

JCDoss

 

[JCDoss]

Someone on this forum suggested using pixels the size used on
consumer digicams to produce a camera with a 32 MP full frame
sensor.

I understand your point. I wasn't suggesting that we simply "blow up" the sensors in consumer digicams to full-frame DSLR size just to get smaller pixels and higher resolution images. What I was curious about was the relationship between fill factor (ie, photon well surface area/insensitive circuitry surface area per pixel) in the D30's CMOS, and if that could be maximized.

If the D30 contains a 50% fill factor (and this is hypothetical, because I don't know), then by reducing that to near zero would reduce the physical size of each pixel to about 7x7um, thereby increasing the total resolution of the image to over 6.7MP without increasing the size of the sensor and without compromising S/N ratio or sensitivity (as the photon wells would be the same size). The net result is a doubling of resolution without detrimental effects on image quality.

From my results with the D30, I would guess that an 8 megapixel
camera with the same sensor type and pixel size (the "D60") should
be able to make a fine-art quality print of 16x24 inches or larger...

8-12MP is probably more than most of us need.

Does anyone ever really print bigger than 16x24 from 35mm film?

The higher resolutions would be useful for cropping and enlarging more than just enlarging the whole frame. Then again, with a 12MP full frame sensor, it's hard to imagine needing more for even that application.

However, those in the business suggest that pixel sizes of 5x5um are necessary to fully maximize the benefits of the best lenses. If the sensor can match those specs without compromising image quality in other ways, there's no reason to not shoot for 5x5um as an ultimate goal in downsizing.

JCDoss

 

[Steve Strawn]

Steve Bryant wrote:
.... In time, pixel

size will shrink, signal to noise ratio, speed and light
sensitivity will all improve....

First time through I read this as:

"the speed of light will improve" and I thought, Man this guy is good.