Gigapixel Camera

[MrMelon]

Check out some of the photos from this camera, the detail is awesome!
http://www.gigapxl.org/gallery.htm

[White trash]

I'm gobsmaked. Farkin'ell!

Aarons gonna need one of them to see which way I went next month.....

[riffer]

Its a film camera buddy.

You still can't get that sort of performance with a conventional style CCD array or even with CMOS arrays.

They can't overcome physics just yet.

Think of a CCD or CMOS as a large bucket which can be filled up with light.

Then think of the light going straight through the centre of the lens. If its path is straight it will easily fill the bucket. Conversely if the path of light is angular then it causes a problem. It takes more light to fill the bucket.

This then gives you a major problem. How you counter the halation effect of the brighter light in the centre of the array causing overfilling of the buckets, leading it to bleed electrons, causing an effect called CCD bloom, which is the digital equivalent of Chromatic Abberation. You need an incredibly complication conversion algorithm to allow for the varying light levels hitting the array. With a gigapixel array you have so much calculation to do you can not at present perform the calculations fast enough to complete all of them in the time necessary to take a photograph that does not contain motion blur. Not to mention that although an array is a (relatively) precisely sized piece of machinery it still can't be made precise enough to counter the circles of confusion that arise from even an optically as perfect as possible lens. So therefore the larger the array the more likely to be aberration introduced on the edges.

Nice scans though from the film.

[Sniper]

Bloody hell, very nice!!

[ching_ching]

Its a film camera buddy.

You still can't get that sort of performance with a conventional style CCD array or even with CMOS arrays.

They can't overcome physics just yet.

Think of a CCD or CMOS as a large bucket which can be filled up with light.

Then think of the light going straight through the centre of the lens. If its path is straight it will easily fill the bucket. Conversely if the path of light is angular then it causes a problem. It takes more light to fill the bucket.

This then gives you a major problem. How you counter the halation effect of the brighter light in the centre of the array causing overfilling of the buckets, leading it to bleed electrons, causing an effect called CCD bloom, which is the digital equivalent of Chromatic Abberation. You need an incredibly complication conversion algorithm to allow for the varying light levels hitting the array. With a gigapixel array you have so much calculation to do you can not at present perform the calculations fast enough to complete all of them in the time necessary to take a photograph that does not contain motion blur. Not to mention that although an array is a (relatively) precisely sized piece of machinery it still can't be made precise enough to counter the circles of confusion that arise from even an optically as perfect as possible lens. So therefore the larger the array the more likely to be aberration introduced on the edges.

Nice scans though from the film.

Ahhhh... NOWWWW I understand!!! :spudwhat:

[riffer]

Ahhhh... NOWWWW I understand!!! :spudwhat:

Shit bro, you're the programmer in waiting.

I'm just a lowly amateur photography buff who used to run a digitisation facility (150 megapixel camera).

[Skunk]

Its a film camera buddy.

You still can't get that sort of performance with a conventional style CCD array or even with CMOS arrays.




Nice scans though from the film.

Well put. I don't see much sharpening there. But then I didn't read the whole site to see exactly what they were doing...

[riffer]

Well put. I don't see much sharpening there. But then I didn't read the whole site to see exactly what they were doing...

Unsharp masking is the last preserve of the incompetent scanner operator.

[Skunk]

Unsharp masking is the last preserve of the incompetent scanner operator.

I disagree as an outright statement, but view it as helping with smaller images. We should discuss this over alcohol...

[riffer]

I disagree as an outright statement, but view it as helping with smaller images. We should discuss this over alcohol...

Agreed. Actually, I should pop down and see your work facilities one time. I haven't been in there for yonks.

Actually I think the last time I was in there you were still outputting film from the back of the Dainippon.

I'm a print buyer/print broker for Learning Media.

Simon

[Skunk]

CtP now. Just about the lot. Give the stuff to BFG or Service. Keep me employed!

[jimbo600]

Its a film camera buddy.

You still can't get that sort of performance with a conventional style CCD array or even with CMOS arrays.

They can't overcome physics just yet.

Think of a CCD or CMOS as a large bucket which can be filled up with light.

Then think of the light going straight through the centre of the lens. If its path is straight it will easily fill the bucket. Conversely if the path of light is angular then it causes a problem. It takes more light to fill the bucket.

This then gives you a major problem. How you counter the halation effect of the brighter light in the centre of the array causing overfilling of the buckets, leading it to bleed electrons, causing an effect called CCD bloom, which is the digital equivalent of Chromatic Abberation. You need an incredibly complication conversion algorithm to allow for the varying light levels hitting the array. With a gigapixel array you have so much calculation to do you can not at present perform the calculations fast enough to complete all of them in the time necessary to take a photograph that does not contain motion blur. Not to mention that although an array is a (relatively) precisely sized piece of machinery it still can't be made precise enough to counter the circles of confusion that arise from even an optically as perfect as possible lens. So therefore the larger the array the more likely to be aberration introduced on the edges.

Nice scans though from the film.

Fovian chips go some way to counter CCD/CMOS noise. Conventional film also suffered from halation to to excess light (overexposure) reflecting off the resin backing.
Gigapixel chips are possible but would be an equivilant of a 1ft sensor and trying to find a lens with that amount of coverage would be a mission.

Are you a photographer or an electronics dude celtic?

[Big Dave]

Unsharp masking is the last preserve of the incompetent scanner operator.

More like the difference between high art - and commercial reality.
bd
www.davidcohen.co.nz

[riffer]

Fovian chips go some way to counter CCD/CMOS noise. Conventional film also suffered from halation to to excess light (overexposure) reflecting off the resin backing.
Gigapixel chips are possible but would be an equivilant of a 1ft sensor and trying to find a lens with that amount of coverage would be a mission.

Are you a photographer or an electronics dude celtic?

Enthusiastic amateur photographer Jimbo. Plus a good 10 years as a Photoshop Artist and retoucher. Started with Photoshop 1.5 and was involved as a betatester for Adobe with Photoshop 3 and 4.

Currently still using film. OM-system mainly (love my black OM-1N) but also using Canon T-90 gear with some of their excellent FD lenses.

I agree with you on halation problems with conventional film methods but in my experience with ccd/cmos arrays I've found that due to the way that photosensitive electronic devices handle light on angles you will have difficulty achieving spectacular results on array edges without electronic post-processing (whether on the array itself, or in-camera processing, or in Photoshop afterwards.

The 4/3rds system along with Olympus' telecentric lenses sound like an interesting proposition, but I've not yet got my hands on an E1 to test it out yet.

Always interested in new photographic technology of course.

Are you doing a lot of digital macrophotography in your work?

[jimbo600]

Enthusiastic amateur photographer Jimbo. Plus a good 10 years as a Photoshop Artist and retoucher. Started with Photoshop 1.5 and was involved as a betatester for Adobe with Photoshop 3 and 4.

Currently still using film. OM-system mainly (love my black OM-1N) but also using Canon T-90 gear with some of their excellent FD lenses.

I agree with you on halation problems with conventional film methods but in my experience with ccd/cmos arrays I've found that due to the way that photosensitive electronic devices handle light on angles you will have difficulty achieving spectacular results on array edges without electronic post-processing (whether on the array itself, or in-camera processing, or in Photoshop afterwards.

The 4/3rds system along with Olympus' telecentric lenses sound like an interesting proposition, but I've not yet got my hands on an E1 to test it out yet.

Always interested in new photographic technology of course.

Are you doing a lot of digital macrophotography in your work?

Just remembered who you are now.

Don't do much macro on digital as CCD's hate long exposures. I still use 5x4 or 6x6 Hassleblad for that high end tech stuff as it's easier to get the scaling right and film still has higher resolution than digital. Otherwise it's CanonD1 all the way.

The T90 is an awesome camera. I still use film for my personal stuff too.