Why are the tests of IQ250 against CCD-backs taking so long?

[Jim Kasson]

I don't know if anyone has ever designed a CFA with a spectral response identical to the response of the cones in the human eye or if there are maybe even reasons against that though.

When I was working at the IBM Almaden Research Laboratory in the early 90s as a color scientist, I consulted with Fred Mintzer and his group in Yorktown Heights who developed a scanning camera with the objective that the wavelength-by-wavelength product of the camera's RGB filters, the IR-blocking filter, and the CCD's spectral response would be close to a 3x3 matrix multiply away from human tristimulus response. The camera was used to digitize Andrew Wyeth's work, to capture artwork in the Vatican Library, and for some other projects where color fidelity was important.

Here is a a paper that has some system-level RGB responses. You'll note a high degree of spectral overlap of the red and green channels, just as there is overlap in the medium and long cone channels. You'll also note an absence of the short-wavelength bump in the red channel; this camera didn't do violet. Because the illumination was part of the camera system, the camera did not have to deal with illuminant metamerism the same way as a human.

There have been papers that indicated that more accurate color encodings can be obtained with four capture channels, which are then processed to yield something close to a tristimulus encoding. Here's a reference that discusses several color capture strategies.

Jim

[bjanes]

If the CFA would be able to exactly mimic that process in a way how it is done in the system of trichromatic vision in the human retina we'd have at least
something equivalently powerful as human trichromatic colorvision, I think.
I don't know if anyone has ever designed a CFA with a spectral response identical to the response of the cones in the human eye or if there are maybe even reasons against that though.

Having the sensor response identical to that of human vision would not be desirable, since the L and M cones (long and medium wavelength, red and green respectively) are too close together and this would lead to coordinate transformations with large coefficients, exacerbating noise. One can move the red and green responses further apart without sacrificing color accuracy as long as the responses of the sensor are linear combinations of the human response curves (Luther-Ives conditions) are met.  However, there are no sensors meeting this condition, so compromises are necessary. See Doug Kerr's excellent article for more details.

Bill

[bcooter]

except for tiresome mantra about "CMOS colors" with the word CMOS in it... it is the time to really drop it... it is a blood libel really in camera's world  

I don't know, or care about the science, but do care about the look and usability.

I think these samples look over sharp, like trying to match the sharpness/depth look of the ccd cameras.

Really It's hard to tell unless you test it yourself, but most of the article seems to me to be a push to say, "yea I know we used to say ccd was superior, but we've done so much special work we can make the cmos images looks as good".  

Personally it's just hard to tell, though the couple on the sofa kind of freaks me out, but that's more the image than any capture device.

Once again, it would have to be something you test yourself, but I have a lot of 1000 iso and higher cameras so tacking on 25 grand to get another stop or two would take some serious soul searching.

For the first time buyer, it might make sense. but there are so many interesting cmos cameras out today, it would take some thought.

IMO

BC

[BernardLanguillier]

They are walking on a very narrow ridge here.

Once you admit that CMOS is just as good as CCD, you quickly reach a point where the only differentiation with the upcoming next gen high megapixel bodies from
Nikon (and - who knows - Canon) is resolution, but even that is unclear.

Everything is being played within a few % of absolute performance and ends being mostly in the head of the buyers and about how they perceive their end to end experience with the equipment.

So comforting buyers in the belief that their check is well worth it remains essential.

Cheers,
Bernard

[narikin]

The article is interesting reading Doug, thank you.

Sadly the one thing I wanted an answer to is not there, namely: why the heavy crop factor? I cannot be the only one for who this is a deal-breaker, as there is no way I will go back to a 1.3 crop.  Would have accepted 1.1, but no more than that. It certainly curtailed my interest - I might have bought it as a second back, for high ISO and the improved LiveView,  if it was full frame or very near.  All that interest evaporated with the heavy crop factor.

Doug, if Phase One, had such an early input in the sensors design/specification, why didn't they make this an absolute must-have requirement from Sony?

[eronald]

Once you admit, yes.
But exactly why should we admit that?
Because Sony in Japan makes CMOS sensors, or because someone can show us the images?

Edmund

They are walking on a very narrow ridge here.

Once you admit that CMOS is just as good as CCD, you quickly reach a point where the only differentiation with the upcoming next gen high megapixel bodies from
Nikon (and - who knows - Canon) is resolution, but even that is unclear.

Everything is being played within a few % of absolute performance and ends being mostly in the head of the buyers and about how they perceive their end to end experience with the equipment.

So comforting buyers in the belief that their check is well worth it remains essential.

Cheers,
Bernard

[henrikfoto]

They are walking on a very narrow ridge here.

Once you admit that CMOS is just as good as CCD, you quickly reach a point where the only differentiation with the upcoming next gen high megapixel bodies from
Nikon (and - who knows - Canon) is resolution, but even that is unclear.

Everything is being played within a few % of absolute performance and ends being mostly in the head of the buyers and about how they perceive their end to end experience with the equipment.

So comforting buyers in the belief that their check is well worth it remains essential.

Cheers,
Bernard

I absolutly agree on this.
My first thought when I heard about the CMOS-back was that this will be the beginning of the end for
the the expensive backs. If the only difference between the MF backs and a Nikon/Canon/Sony is a bigger
sensor I think it will be even harder to justify the huge price-difference. Just my thoughts..

[eronald]

Phase seem to disagree, with their back at 2x the price of the Pentax

Edmund

I absolutly agree on this.
My first thought when I heard about the CMOS-back was that this will be the beginning of the end for
the the expensive backs. If the only difference between the MF backs and a Nikon/Canon/Sony is a bigger
sensor I think it will be even harder to justify the huge price-difference. Just my thoughts..

[synn]

Looked at the images today. Still not seeing the "CCD look". If I can, I will try to do a side by side test.
Until then though, all I am seeing is D800++.

[Doug Peterson]

I absolutly agree on this.
My first thought when I heard about the CMOS-back was that this will be the beginning of the end for
the the expensive backs. If the only difference between the MF backs and a Nikon/Canon/Sony is a bigger
sensor I think it will be even harder to justify the huge price-difference. Just my thoughts..

Normally photographers do their work by taking a naked sensor and hold it in the air towards the thing they want to photograph.

So it's natural to assume that nothing but the sensor matters – lens technical quality, lens look, feature set, tethering speed/stability, flash sync speed, features, review speed/ease/power/accuracy, tactility of the capture process, mechanical precision, ergonomics of the body, brightness of the viewfinder, color accuracy, color pleasantness, etc.

Then again, the sensor is 1.7x larger and has

[eronald]

Normally photographers do their work by taking a naked sensor and hold it in the air towards the thing they want to photograph.

So it's natural to assume that nothing but the sensor matters – lens technical quality, lens look, feature set, tethering speed/stability, flash sync speed, features, review speed/ease/power/accuracy, tactility of the capture process, mechanical precision, ergonomics of the body, brightness of the viewfinder, color accuracy, color pleasantness, etc.

Then again, the sensor is 1.7x larger and has

Doug,

 I have to agree with you

 This is why now that everyone has the same sensor I'm so looking forward to a your publishing comparison of the Hassy and Leica and Pentax and Phase bodies and optics. I am sure the Phase solution will prove to have peerless files and color, C1 will beat Lightroom for crispness, the back will tether wonderfully, and the Phase body will focus better than the Hassy, be more ergonomic  and have lenses that are sharper wide open than the Leica; unfortunately it looks unlikely that Phase will beat the Pentax on price, but at least one pig has been known to fly

 Ok, I'll stop making fun, everybody here WANTS IMAGES! Lots and lots of images made by the new hot Phase IQ 250.

 Until these images come out, any rep who passes by this forum will be treated with all the attention and courtesy which alligators reserve for fly fishermen

 I can't resist a quote from my link:

“We can confirm that the pig traveled, and we can confirm that it will never happen again,” US Airways spokesman David Castelveter said. “Let me stress that. It will never happen again.”

Sources familiar with the incident told the Philadelphia Daily News for Friday’s editions that the pig’s owners convinced the airline that the animal was a “therapeutic companion pet,” like a guide dog for the blind.

Owners Had Doctor’s Note

The pig was traveling with two unidentified women, one in her 30s, the other a senior citizen. An internal US Airways incident report said the owners claimed they had a doctor’s note that allowed them to fly with the animal.

US Airways and Federal Aviation Administration rules allow passengers to fly with service animals.

“According to [the] Philadelphia agent who talked to passenger over phone … passenger described pig as being 13 pounds, so based on this info, authorization was given,” the report stated. Passengers on the flight told the Daily News the pig actually weighed several hundred pounds.

Pig Goes Wild

The pig, which spent the flight in the first row of first class, went ape when the aircraft taxied into Seattle, according to the report.

It reportedly ran loose through the aircraft, squealing loudly, and even tried to enter the cockpit.

“Many people on board the aircraft were quite upset that there was a large uncontrollable pig on board, especially those in the first-class cabin,” the incident report stated.

Edmund

[Fine_Art]

So the owner won the pigs can fly bet.

It's a fairly standard strategy in luxury goods to create a mythology around the product. Customers are then not focused on the value of the item, they are looking at the value of the prestige. They buy into an association with the perceived quality, thinking it reflects on them. Maybe that works in the camera world, who knows.

CMOS has growing benefits so for medium format to pretend the current 135 sensors do not solve several problems for the pro would be "in denial". The DR of larger pixels on a bigger area sensor is the beef of the matter. You either need that extra light in one shot or you go 135.

[torger]

Backs are expensive because 1000 customers is going to pay for development costs rather than 50000. Even if components were free they would probably cost more than $20k. You are paying for Phase One expertise. Nothing wrong with that. What is wrong is pumping up the impression that you mainly pay for expensive components. Sure sensors of this size is not cheap but it's a few k tops (my guess is that the sony sensor is about 2k, it would be interesting to know exact price), not 15k.

The only way to make backs "cheap" is to make a product that can interest more customers, and have a sales model that can handle the required amount of customers.

I dont think the CMOS sensor is cheaper than the CCD. It should be the opposite if you look at the complexity, but of course the smaller formats has already financed most development cost of this sensor.

[henrikfoto]

Normally photographers do their work by taking a naked sensor and hold it in the air towards the thing they want to photograph.

So it's natural to assume that nothing but the sensor matters – lens technical quality, lens look, feature set, tethering speed/stability, flash sync speed, features, review speed/ease/power/accuracy, tactility of the capture process, mechanical precision, ergonomics of the body, brightness of the viewfinder, color accuracy, color pleasantness, etc.

Then again, the sensor is 1.7x larger and has

haha. The problem is that the lenses are not any better (except wide angle LF) and the cameras for Medium format are not even close to the best canon/nikon etc.

Don't misunderrstand. I use many different LF and Medium format systems and love the results.
BUT the difference to canon/nikon etc are getting smaller (I hate to admit this).

[Theodoros]

Normally photographers do their work by taking a naked sensor and hold it in the air towards the thing they want to photograph.

So it's natural to assume that nothing but the sensor matters – lens technical quality, lens look, feature set, tethering speed/stability, flash sync speed, features, review speed/ease/power/accuracy, tactility of the capture process, mechanical precision, ergonomics of the body, brightness of the viewfinder, color accuracy, color pleasantness, etc.

Then again, the sensor is 1.7x larger and has

My wonder is if P1 is considering to use a cropped size of this sensor down to 135 FF size and corporate with a large Dslr maker to bring it in production based on a current FF body…. Would (the same) people prefer the back, or the smaller (and much cheaper) body in such a case? I say this to back up my opinion that MFDBs should apply to different applications than DSLRs, instead of trying to only expand the image area and increase resolution both of which aren't key factors to develop one's photography further.

[torger]

As far as I know the Sony Alpha 99 was/is a DSLR where CFA is more focused on color than on high ISO performance, and should have better color than the more recent A7r. I'm not sure, but I think the A99 is pretty unique in this aspect.

I sure think that there is a market for making DSLRs with sensors designed for color, I think there are many out there that would sacriface a stop in ISO performance in exchange for better tonality at base ISO. One problem however is that the manufacturer's own raw converters are not that popular, if you have a Nikon/Canon/Sony camera you'd probably use Lightroom or Capture One rather than the manufacturer's own converter although they may have better color.

I'm not sure if it is in Phase One's interest to make profiles for a Sony DSLR such that it rivals say the IQ250 in color. And Adobe doesn't seem to be have the interest in making those hand-tuned profiles that seem to be Capture One's edge.

But say if DSLR manufacturers start to think more about CFA for color in their high MP models and Adobe hire some color magician to make MF-like profiling for those key models then competition would harden further.

[Theodoros]

As far as I know the Sony Alpha 99 was/is a DSLR where CFA is more focused on color than on high ISO performance, and should have better color than the more recent A7r. I'm not sure, but I think the A99 is pretty unique in this aspect.

I sure think that there is a market for making DSLRs with sensors designed for color, I think there are many out there that would sacriface a stop in ISO performance in exchange for better tonality at base ISO. One problem however is that the manufacturer's own raw converters are not that popular, if you have a Nikon/Canon/Sony camera you'd probably use Lightroom or Capture One rather than the manufacturer's own converter although they may have better color.

I'm not sure if it is in Phase One's interest to make profiles for a Sony DSLR such that it rivals say the IQ250 in color. And Adobe doesn't seem to be have the interest in making those hand-tuned profiles that seem to be Capture One's edge.

But say if DSLR manufacturers start to think more about CFA for color in their high MP models and Adobe hire some color magician to make MF-like profiling for those key models then competition would harden further.

Why do I have that suspicion, that if one uses C1P1 with IQ250 profiles for his D800 or other Sony sensor Dslr, will benefit a lot…?   

[Paul2660]

I also have to assume the D800 as also not CFA focused on high iso as much as the smaller chip sized D4 and now D4s.  The D800's iso strengths are in the 100 to 2000 range, with 3200 being quite a push.  Since the April 2012 launch of the D800 family, Nikon has made no significant firmware updates to either camera.  This tells me that the CMOS chip/processor developed for this camera was pretty much done when first shipped, which was a bit surprising to me.  Phase One does tend to tweak their backs over time and historically have releases many imaging improvements to a family of backs.  

It's also very probable that the Sony A7r, A7 are both CFA focused for color.  Per Sony's site and info gleaned from others the 36MP chip in the A7r is not the same exact chip that's in the D800e. So Sony did continue to develop along that chip family.  No doubt larger 35mm CMOS based chips are coming in the future.

Landscape outdoor profiling for the D800 is tricky and the default Capture One profile leave much on the table.  This is even worse with the D800e.  

Nikon does have their own Capture software, and it's actually very good, but since it's a stand alone and had no plug in support I still tend to lead with LR for raw conversion on Nikon files.  

Also, I have not seen any real price quoted for the "new" 645DII, everyone seems to feel it will fall in the 10K price point, but Pentax/Ricoh may surprise a few folks here.

Paul C

[Paul2660]

Why do I have that suspicion, that if one uses C1P1 with IQ250 profiles for his D800 or other Sony sensor Dslr, will benefit a lot…?   

That is a interesting point and well worth the effort, I will try that later today as I have 7.2 loaded now. 

Paul C

[torger]

DxOmark makes some basic measurements of color response of the sensors, select one camera and look at the "color response" tab.

The A7r and the D800 does not have the same, although difference is not huge. Using the same profiles would not work well though. If you compare D800 and D800e color response they are almost exactly the same, ie you only see measurement errors and/or unit variation. Interesting to note is that the older Sony a99 with the great color reputation scores a bit higher on the sensitivity metamerism index, which may indicate that it has better color separation capability.

Anyway if to use a different camera's profiles they must have the same color response, and that won't be the case with the D800 and the IQ250. They have different CFAs. As far as I understand it's quite easy to manufacture the same sensor with different CFA tunings, so different models with the same sensor may have different CFAs. This could also be the case between IQ250, Pentax and Hasselblad's cameras all using the same Sony 44x33 sensor, but possibly with different CFAs. If that is the case it will show up in lab measurements like DxO's, but you will see it with the naked eye too of course if just doing A/B switching of files using the same color profiles.

I don't think the designers of A7r and D800 think that they have sacrificed a lot of color accuracy, but they've done some tradeoff which is different from if they would only look at base ISO performance and let high ISO become what it becomes, which I think is the design criteria for the IQ250.