The Future of CCD Sensors

[Gel]

I do like the look of the A7S, I was most pleased when they said it was 12mp.

[hjulenissen]

I find it quite interesting that many both say that they don't need many megapixels and that they don't want any AA filter.

Perhaps (some of) those people are judging their images unsharpened at 1:1 on a computer screen, and finding that the larger sensel, non AA filtered images look better in that condition than cameras of smaller sensels featuring AA filters?

It would be a strange (myopic?) setting to base ones purchase of expensive image-producing equipment on, but I have heard of stranger things.

-h

[Doug Peterson]

FWIW I tested a Phase P25 back against a 5D3.

Same relative lens setup (primes) and found that whilst the colours were exactly the same the phase was slightly sharper. Probably because there was no AA filter to screw things up.

[...]

If someone came out with a CMOS version of the P25 (same 9 micron pixel count) I'd be all over that in a heartbeat and I really think the sensor manufacturing is focused on the wrong thing.
I don't need 50mp for my line of work. Even 22mp is overkill.

Did you open the P25+ in Capture One or in ACR/LR? The overwhelming majority of my customer interactions have leaned towards Capture One giving color more desirable to the customer for P1 backs, especially the earlier backs.

Also, when you hear about color magic in medium format it's more often associated with the Dalsa based backs. Whether that's generational (more recent backs are more likely to have been Dalsa) or causal I couldn't say.

Anyway, the back you are asking about seems to already have been here for a while: a P65+ or IQ160 in sensor plus mode uses 10.4 micron pixels, produces a very very sharp 15mp file (assuming of course sharp lens and good technique), shoots quickly without buffer, has a native ISO of 200 and is very nice even at ISO1600, has a great reputation for color, and is easier on lenses than a very high res back.

It also has a nice bonus feature you can turn on where it shoots 60mp raw files in case you find yourself in a situation where you do need more resolution.

I'm not kidding; I've shot weddings with the 65+ where I used sensor+ at least 80% of the time.

[Gel]

I opened the Phase one in C1 Doug. Lightroom for the 5D3.

[eronald]

Perhaps (some of) those people are judging their images unsharpened at 1:1 on a computer screen, and finding that the larger sensel, non AA filtered images look better in that condition than cameras of smaller sensels featuring AA filters?

It would be a strange (myopic?) setting to base ones purchase of expensive image-producing equipment on, but I have heard of stranger things.

-h

Actually, it's more fun to interpolate *up* from 1:1 and look what the image looks like.
Some cameras have "enlargeable" pixels, some don't. I guess we would all like to know why ....
I just did this test on a Sigma DP3 and I think something like 1.5 linear is what it can take before images fall apart.
I once did an enlargement to 44" from about 1/4 frame of my P45+. Nice file.
I don't think it's idiotic to prefer a camera like the M8 which can take a huge amount of uprezzing for some reason to one which has a similar number of sharp pixels but cannot do crop/enlarge.

Edmund

[ErikKaffehr]

Hi,

I made a test/demo of this a while ago, the image below was shot on a P45+ with a Sonnar 150/4 at 3.8 m. The P45+ has 6.8 my pixels and no AA-filter.

another shot was made with a 70-400/4-5.6 G at 150 mm on a Sony Alpha 77 also at 3.8 m, the Sony has 3.9 my pixels and may have AA-filter. So the focal length and distance is the same. The difference is essentially the size of the pixels.

The two lenses are about the same quality.

Below, an enlarged P45+ image for better viewing.


Finally, this is the 3.9 image downscaled to same size as the P45+ image:

P45+	3.9my (Sony Alpha 99)

Full article is here: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/78-aliasing-and-supersampling-why-small-pixels-are-good

So, I feel there is a merit to small pixels. On the other hand looking at individual pixels on screen is a bit different from viewing a print.

Best regards
Erik

If you want larg-ish sensor, low-ish MP count _and_ state-of-the-art sensor tech, then the Sony A7S might be something to investigate?

When I listen to music I don't strive for 44100 Hz native D/A converters using no interpolation filter. Rather, I want the best D/A there is, and this usually means a very high sample rate, combined with noise-shaping/dithering and resampling to give good performance in the audible band. Same with image sensors, I expect that we will at some stage have camera sensors that have "spatial resolution" way in excess of what we need (or what lenses can realistically resolve), or what there is normally photon counts to excite at normal noise level. The I expect this sensor to provide very good image quality at sensible image reprodution sizes. See http://ee.usc.edu/faculty_staff/faculty_directory/fossum.htm

-h

[Telecaster]

I don't think it's idiotic to prefer a camera like the M8 which can take a huge amount of uprezzing for some reason to one which has a similar number of sharp pixels but cannot do crop/enlarge.

Yeah, I'm amazed at just how well M8 files uprez & print at sizes you'd think would be beyond the safe zone. Except for the occasional bit of moiré it's been all upside with this camera. Wish I'd discovered this a year ago...coulda saved myself a lot of time, effort & $$.   

-Dave-

[eronald]

Yeah, I'm amazed at just how well M8 files uprez & print at sizes you'd think would be beyond the safe zone. Except for the occasional bit of moiré it's been all upside with this camera. Wish I'd discovered this a year ago...coulda saved myself a lot of time, effort & $$.   

-Dave-

How is the M9?

Edmund

[Telecaster]

How is the M9?

It's basically an M8 with wider coverage for any given lens...and more processing, in-camera or post, required to deal with corner/edge color casts & light falloff. I've never owned an M9 but I've used my friend Bruce's nearly as much as he has.     I think the M8 is crisper, if just, at the pixel level. And no jiggery-pokery correction needed with my lenses.

-Dave-

[Chris Livsey]

How is the M9?

Edmund

Different 

http://pskiss.com/shop/cross-camera-color-profiles/

[eronald]

Different 

http://pskiss.com/shop/cross-camera-color-profiles/

Hehe, I see yet another guy has discovered the look profile technique

Edmund

[Fine_Art]

Well,

I don't disagree with you, but one of my observations were that those images could be made pretty close, if work flow is calibrated and matched.

My guess is that the profiles provided with C1 for Phase One backs may be better than those provided for say Nikon cameras. Some posters here even found that profiles for the IQ-250 worked better with Nikon D800 than the D800 profile provided by C1.

That said, I am an engineer with photography as a pastime and quite interested in the engineering part of it…

Best regards
Erik

What kind, since you bring it up? Electrical, Mechanical, Software, Optical? The term engineer is used for so many things...

Edit - I would guess software since you seem quite skilled with it.

[ErikKaffehr]

Hi,

I have a degree in mechanical engineering, but my main area is reactor physics. Anyway, my professional education is not related to photography, and I definitively don't see myself as an artist.

Best regards
Erik

What kind, since you bring it up? Electrical, Mechanical, Software, Optical? The term engineer is used for so many things...

Edit - I would guess software since you seem quite skilled with it.

[JoeKitchen]

OH MY GOD! 

We finally brooke out the feathers again; watch out. 

[araucaria]

I tried using the phase one iq250 adobe profile with the d800 in photoshop and the color hues are all wrong, my guess the same thing will happen in C1.

[henrikfoto]

Hi,

I made a test/demo of this a while ago, the image below was shot on a P45+ with a Sonnar 150/4 at 3.8 m. The P45+ has 6.8 my pixels and no AA-filter.

another shot was made with a 70-400/4-5.6 G at 150 mm on a Sony Alpha 77 also at 3.8 m, the Sony has 3.9 my pixels and may have AA-filter. So the focal length and distance is the same. The difference is essentially the size of the pixels.

The two lenses are about the same quality.

Below, an enlarged P45+ image for better viewing.


Finally, this is the 3.9 image downscaled to same size as the P45+ image:

P45+	3.9my (Sony Alpha 99)

Full article is here: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/78-aliasing-and-supersampling-why-small-pixels-are-good

So, I feel there is a merit to small pixels. On the other hand looking at individual pixels on screen is a bit different from viewing a print.

Best regards
Erik

Hi Erik!

I am sure you will see much better results from the P45 with the best lenses like the 60cfi, 100cfi and 180cfi.
The 150mm is a portrait lens and is far from the sharpest lens. Even the 120 macro is not very good with a high res digital back.

Love reading your tests!

Henrik

[ErikKaffehr]

Hi Henrik,

Thanks for feedback!

Some of those lenses are on my shopping list. I actually replaced my Sonnar 150/4 with the 180/4 CFi and I am very happy with it.

The 120/4 is now also a CFi, and I will keep it as I want to have a macro lens. The 100/3.5 is on the long term list, but I feel that I have to many lenses in that focal range, 80 and 120. It may come that I replace 50 -> 60 and 80->100, replacing the 150/4 with 180/4 was the first step in that process.

I bought my lenses after checking the MTF curves from Zeiss, and the Sonnar 150/4 is one of the better ones, but the 180/4 is a bit better.  

With regard to this test I wanted to see if the relatively large pixels would cause fake patterns on a low contrast natural subject and what effect OLP filtering and small pixels had. What I found was smaller pixels give more truthful  reproduction.

A couple of observations:

- This is pixel peeping at worst, the question is if these fake patterns are visible in print? They may be, I need to print to find out.
- MTF data only goes down to 40 lp/mm, while the P45+ is around 73 lp/mm

I measured MTF for the lenses, and the Sonnar is a tiny bit better than the 70-400/4-5.6 I used at 150 mm (f/8). See enclosed figures, Sonnar on top and the 70-400/4-5.6 at bottom. The MTF measurement is a bit off-axis but still in the central area.

Best regards
Erik

Hi Erik!

I am sure you will see much better results from the P45 with the best lenses like the 60cfi, 100cfi and 180cfi.
The 150mm is a portrait lens and is far from the sharpest lens. Even the 120 macro is not very good with a high res digital back.

Love reading your tests!

Henrik

[Paul2660]

I tried using the phase one iq250 adobe profile with the d800 in photoshop and the color hues are all wrong, my guess the same thing will happen in C1.

Actually the D800e does very well with the Phase One IQ250 default outdoor daylight profile. IMO it's better than the C1 D800e profile and I tend to use the Phase profile now.

Paul

[eronald]

Erik,

(Idiotic comments on)
 Please don't take this badly, but might I suggest that if your interest is aliasing, then you would greatly enjoy reading some standard tutorial on Fourier theory as applied to signal processing?And if you are interested in self-similar scaling patterns, then Mandelbrot would be a good place to start?  Imaging a fractal pattern (feather) is an entertaining experiment, and in fact fractal targets have their proponents, but I would say that the reason they can be used as test targets is precisely the fact that images degrade quickly where the camera's limits are reached. Of course you could reply that most, in fact all, of textures in nature are of fractal nature, and I would agree, but I still think that forcing a camera to alias (Moiré) is not a good test of its look.

If I may be allowed a gedanken experiment, if you image and you upscale a blurry image zone, you get a nice pleasant big blur; if you image a fractal and upscale a Moiré, you get a large Moiré and a lot of information about your upscale algorithm

 And yes, if you believe I don't have the slightest idea what I'm talking about you're probably right, but maybe that's because we've walked off the edge of the luminous world where one looks at images and into the jungle world of finite digital numbers where digital reflections of mathematical entities are being pushed around in a brittle way, and we should just use equations to talk our way past this point. Except we mostly don't have the right maths cooked yet and we don't understand how the perceptions relate to the maths so not only do we use the wrong maths as concepts like texture are new (fractals are only 50 years old, the wavelet transform is in its infancy) but also once we're in the world of formulae we cannot find our way back - as your feather posts demonstrate.
 
(Idiotic comments off)

Edmund

Hi Henrik,

Thanks for feedback!

Some of those lenses are on my shopping list. I actually replaced my Sonnar 150/4 with the 180/4 CFi and I am very happy with it.

The 120/4 is now also a CFi, and I will keep it as I want to have a macro lens. The 100/3.5 is on the long term list, but I feel that I have to many lenses in that focal range, 80 and 120. It may come that I replace 50 -> 60 and 80->100, replacing the 150/4 with 180/4 was the first step in that process.

I bought my lenses after checking the MTF curves from Zeiss, and the Sonnar 150/4 is one of the better ones, but the 180/4 is a bit better.  

With regard to this test I wanted to see if the relatively large pixels would cause fake patterns on a low contrast natural subject and what effect OLP filtering and small pixels had. What I found was smaller pixels give more truthful  reproduction.

A couple of observations:

- This is pixel peeping at worst, the question is if these fake patterns are visible in print? They may be, I need to print to find out.
- MTF data only goes down to 40 lp/mm, while the P45+ is around 73 lp/mm

I measured MTF for the lenses, and the Sonnar is a tiny bit better than the 70-400/4-5.6 I used at 150 mm (f/8). See enclosed figures, Sonnar on top and the 70-400/4-5.6 at bottom. The MTF measurement is a bit off-axis but still in the central area.

Best regards
Erik

[jerome_m]

(Idiotic comments on)
And yes, if you believe I don't have the slightest idea what I'm talking about you're probably right, but maybe that's because we've walked off the edge of the luminous world where one looks at images and into the jungle world of finite digital numbers where digital reflections of mathematical entities are being pushed around in a brittle way, and we should just use equations to talk our way past this point.
(Idiotic comments off)

I dare you.