Phase One P65+ raw issue

[undavide]

Hello,
being this my first post I'd like to briefly introduce myself: my name is Davide Barranca, I'm from Italy and I mainly do post-production and digital printing for photographers involved in the contemporary art business (a strange business indeed )
I've recently run into a peculiar problem dealing with P65+ files, more specifically developing them - I hope this one is the correct forum section: while the issue I'm reporting is about raw processing (and a couple of raw processing softwares), I suspect it may affect P65+ digital backs tout-court.

The problem is about a highly geometric non-uniformity of color rendition (especially in the green/magenta axis) - something that's generally very subdued, but that can explode when you deal with grayish images that you may want to push towards a more colorful interpretation.

There are several ways to make this deliberately happens: one particularly rude way is as follows:

1) Bring .IIQ (PhaseOne raw files) directly into Adobe Camera Raw (ACR), no matter whether CS4 or CS5 version, zeroing all the sliders, and with a decent (not necessarily perfect) white balance.
2) take them in Photoshop, it doesn't matter whether in 16bit or 8bit, sRGB or Adobe1998. Convert to Lab color mode.
3) select the "a" channel (the one that is about green/magenta opponents) and do a Image-Adjustments-Equalize command.

I've tested two different P65+ digital backs with two different Phase-One cameras and the results are always these: the colors (obviously) become a mess of green and magenta noise when I look at the composite image, BUT, if I inspect the "a" channel, it's always divided into 4 quadrants (like cartesian axis) with different weight: say, the top left is lighter, the top right more dark, the bottom left the darker, the bottom right the lightest. The effect's magnitude depends on the image (photographs of colorful object show less, let's call this "quadrant posterization", while grayish images like rocks, are clearly more offended.)

I understand that no one is using the equalize command on a Lab channel on purpose, but I assure you that if you do a couple of standard tricks in Lab to boost the colors, it's easy to end up with a picture that shows a more magentaish bottom left side and a greener bottom right one (just to let alone the top).

I've run into this issue almost always with ACR, far less with Capture One - let's say more properly that the effect is less obvious with C1 - but I did not extensive tests because I prefer to use ACR on productions for several practical reasons.
By the way, the Adobe answer about this one is that P65+ files, while ACR "sees" and opens them, are not officially supported (P45+ is, P65+ is not). I'm not in touch with PhaseOne support, but I'd like to have more data on this issue.
I'll be glad if some P65+ owners would gently do a quick test with their files - I'm not sure whether is an ACR issue or a digital back... feature 
The steps could be the same: with a picture of a fairly neutral subject, ACR zeroing everything, convert to Lab, equalize the "a" channel (the "b" is, strangely, unaffected). Believe me, if the "a" channel shows the problem, you will recognize it.
Thanks in advance for your help,

Davide

[Steve Hendrix]

Hello,
being this my first post I'd like to briefly introduce myself: my name is Davide Barranca, I'm from Italy and I mainly do post-production and digital printing for photographers involved in the contemporary art business (a strange business indeed )
I've recently run into a peculiar problem dealing with P65+ files, more specifically developing them - I hope this one is the correct forum section: while the issue I'm reporting is about raw processing (and a couple of raw processing softwares), I suspect it may affect P65+ digital backs tout-court.

The problem is about a highly geometric non-uniformity of color rendition (especially in the green/magenta axis) - something that's generally very subdued, but that can explode when you deal with grayish images that you may want to push towards a more colorful interpretation.

There are several ways to make this deliberately happens: one particularly rude way is as follows:

1) Bring .IIQ (PhaseOne raw files) directly into Adobe Camera Raw (ACR), no matter whether CS4 or CS5 version, zeroing all the sliders, and with a decent (not necessarily perfect) white balance.
2) take them in Photoshop, it doesn't matter whether in 16bit or 8bit, sRGB or Adobe1998. Convert to Lab color mode.
3) select the "a" channel (the one that is about green/magenta opponents) and do a Image-Adjustments-Equalize command.

I've tested two different P65+ digital backs with two different Phase-One cameras and the results are always these: the colors (obviously) become a mess of green and magenta noise when I look at the composite image, BUT, if I inspect the "a" channel, it's always divided into 4 quadrants (like cartesian axis) with different weight: say, the top left is lighter, the top right more dark, the bottom left the darker, the bottom right the lightest. The effect's magnitude depends on the image (photographs of colorful object show less, let's call this "quadrant posterization", while grayish images like rocks, are clearly more offended.)

I understand that no one is using the equalize command on a Lab channel on purpose, but I assure you that if you do a couple of standard tricks in Lab to boost the colors, it's easy to end up with a picture that shows a more magentaish bottom left side and a greener bottom right one (just to let alone the top).

I've run into this issue almost always with ACR, far less with Capture One - let's say more properly that the effect is less obvious with C1 - but I did not extensive tests because I prefer to use ACR on productions for several practical reasons.
By the way, the Adobe answer about this one is that P65+ files, while ACR "sees" and opens them, are not officially supported (P45+ is, P65+ is not). I'm not in touch with PhaseOne support, but I'd like to have more data on this issue.
I'll be glad if some P65+ owners would gently do a quick test with their files - I'm not sure whether is an ACR issue or a digital back... feature 
The steps could be the same: with a picture of a fairly neutral subject, ACR zeroing everything, convert to Lab, equalize the "a" channel (the "b" is, strangely, unaffected). Believe me, if the "a" channel shows the problem, you will recognize it.
Thanks in advance for your help,

Davide

Hi Davide:

Welcome to the forum.

This is a Camera Raw issue, and as such, an issue for anyone who wishes to use Camera Raw with P65+ files. As you mention, the issue is much less significant in Capture One. There are numerous issues that Capture One corrects compared to a 3rd party converter, and part of this is simply the result of Capture One's proprietary algorithm, part is that Capture One reads the embedded calibration file for that individual sensor (whereas Camera Raw does not, so many baddies that get caught/remedied in Capture One are not caught/remedied in Camera Raw).


Steve Hendrix

[BernardLanguillier]

Capture One reads the embedded calibration file for that individual sensor (whereas Camera Raw does not, so many baddies that get caught/remedied in Capture One are not caught/remedied in Camera Raw).

Hey Steve,

Is the data contained in the calibration file available for anyone to access?

If yes, do you know if Adobe plans to support them in the future?

Thanks.

Cheers,
Bernard

[Schewe]

By the way, the Adobe answer about this one is that P65+ files, while ACR "sees" and opens them, are not officially supported (P45+ is, P65+ is not).

Not sure what part of "not officially supported" you don't understand? In point of fact, a lot of work went into Camera Raw 6.2 to more fully support the P65+ but it's still not "officially" supported. And, as far as I know, ACR 6.2 DOES read and use the cal file. This is new in ACR 6.2. Adobe has worked with Phase One to bring full support to P65+ backs in the near future but 6.2 is the first version where both color and tone have been normalized...

[Doug Peterson]

Do the problems you have show up in any meaningful way when processed in Capture One?

Support in Adobe Camera Raw should only be expected to improve. However, best support will always be with the software developed (at great cost in time and money) to make those digital backs shine.

And it's not just Phase One - have you looked at a Nikon D3X file in both Capture One 5.2 and ACR lately to see what a difference a raw convertor can make in your workflow?

[Steve Hendrix]

Not sure what part of "not officially supported" you don't understand? In point of fact, a lot of work went into Camera Raw 6.2 to more fully support the P65+ but it's still not "officially" supported. And, as far as I know, ACR 6.2 DOES read and use the cal file. This is new in ACR 6.2. Adobe has worked with Phase One to bring full support to P65+ backs in the near future but 6.2 is the first version where both color and tone have been normalized...

If this is the case, then that is good news. My understanding has been that hasn't been the case in the past. A year ago, I had a client who shoots aerials show me a file of a parking lot where the specular highlights from reflective sources (car bumpers, mirrors, etc) were unacceptably large (excess blooming). By simply re-processing the files in Capture One, the blooming was dramatically reduced. Night and day. As Doug said, native software (especially in the case of Capture One) will usually provide superior results. That said, better support from Adobe is welcome news.


Steve Hendrix

[undavide]

Dear forum members,
Thanks for your answers! I'm finding my way out of the fog 

So, as far as I understand, embedded inside each raw file there's some kind of calibration info, which helps C1 to balance the sensor hue disparity, am I correct? BTW, is this "calibration" affecting something else as well or it controls the quadrants only?
And ACR partial support of P65+ files means that it reads and develops them, but currently without accessing and interpreting the calibration, right? (this is the part of the "not officially supported by ACR" that I couldn't understand)

I'm afraid I've a couple of extra questions, now: if, in C1, I export as DNG the IIQ, would this help? I mean, this way will the DNG be, somehow, normalized, will it be the result of the original raw plus the calibration applied?
So far I made some quick tests, and there's a lot of difference within ACR in the interpretation of the same IIQ and DNG (zeroing everything - the DNG is a lot darker, I can't figure out the why. I'm tempted to play with DNG profiles to correct this one - but as a retoucher I'm rarely allowed to play with backs, only with Photoshop)

Another question that may arise is: what if the P65+ "calibration" isn't accurate? I guess I've no way to test it, other than brutalize the files with all sort of manoeuvres. And I may suspect that this calibration is a factory one - so we've to rely on what we got, and there's no way to tweak it... or not?
Thanks again and best regards,

Davide

[tesfoto]

Dear Davide

You got it - and yes it is a factory calibration individual for each back to compensate for different issues - I think only C1 read this calibration.

If you want the best of a P65+ file you will have to use C1 - it is as simple as this. Dont try to do workarounds like DNG etc. Just get to know C1 and be happy, it is the best developer for the files anyway.

When you encounter a file that needs LCC calibration you will have to work with C1 anyway as ARC does not handle this LCC file.

It might not be your favorite workflow, but in the end you will have a much better file and less work to do in Photoshop.

Just learn C1 - LL has some great tutorial videos for download too.

Cheers

TES

[Steve Hendrix]

Dear forum members,
Thanks for your answers! I'm finding my way out of the fog 

So, as far as I understand, embedded inside each raw file there's some kind of calibration info, which helps C1 to balance the sensor hue disparity, am I correct? BTW, is this "calibration" affecting something else as well or it controls the quadrants only?
And ACR partial support of P65+ files means that it reads and develops them, but currently without accessing and interpreting the calibration, right? (this is the part of the "not officially supported by ACR" that I couldn't understand)

I'm afraid I've a couple of extra questions, now: if, in C1, I export as DNG the IIQ, would this help? I mean, this way will the DNG be, somehow, normalized, will it be the result of the original raw plus the calibration applied?
So far I made some quick tests, and there's a lot of difference within ACR in the interpretation of the same IIQ and DNG (zeroing everything - the DNG is a lot darker, I can't figure out the why. I'm tempted to play with DNG profiles to correct this one - but as a retoucher I'm rarely allowed to play with backs, only with Photoshop)

Another question that may arise is: what if the P65+ "calibration" isn't accurate? I guess I've no way to test it, other than brutalize the files with all sort of manoeuvres. And I may suspect that this calibration is a factory one - so we've to rely on what we got, and there's no way to tweak it... or not?
Thanks again and best regards,

Davide

Davide:

Think of the calibration more as an embedded setting that corrects the imperfections of the sensor itself, rather than image parameters that might be tweaked. A silicon wafer is always the sum of its parts, and rarely are its parts uniform. Not all photosites are created equal. So the calibration file is created to equalize those bad characters. It is not something you would want to tweak, even if you could. Tesfoto's advice is perfect.


Steve Hendrix

[Schewe]

I'm afraid I've a couple of extra questions, now: if, in C1, I export as DNG the IIQ, would this help? I mean, this way will the DNG be, somehow, normalized, will it be the result of the original raw plus the calibration applied?
So far I made some quick tests, and there's a lot of difference within ACR in the interpretation of the same IIQ and DNG (zeroing everything - the DNG is a lot darker, I can't figure out the why. I'm tempted to play with DNG profiles to correct this one - but as a retoucher I'm rarely allowed to play with backs, only with Photoshop)

I'm afraid that currently, there are issues with the DNG files that Capture One creates; First off, C1 incorrectly adds a space in the device model name. So, instead of the tag reading P65+ (which is the correct model name) the resulting tag reads P 65+. You might not think that would be a problem but it is...it makes DNGs coming from C1 incorrectly tagged. Phase One is aware of the problem and hopefully it'll be fixed in the near future. But if you want to convert a P65+ raw file to DNG, you should use the DNG Converter not C1 to do so. The other problems with C1's DNG conversion is that C1 doesn't create a compressed DNG and the file size goes from about 68MBs for a properly compressed DNG (note, this is lossless compression so no reason not to do so) to a 117MB DNG in size. There is zero reason for this lack of lossless compression in the DNGs coming from C1. The other issue is there are no DNG spec controls in C1. Currently, DNG Converter 6.2 allows setting the spec to the current DNG 1.3 and allows for backwards compatibility of the resulting DNG. None of that is available in the C1 DNG conversion...

So, at this stage, I would stay away from using Capture One for making DNGs from P65+ raw files. If you need a DNG, use DNG Converter 6.2.

If you have the most recent version of C1 (5.2) and are opening P65+ raw files, you would be better off normalizing the tone and color and using the correct LCC and Lens Corrections in C1 and processing out a Pro Photo RGB 16-bit tiff. Once processed out, you'll then be able to open the tiff inside of Camera Raw 6.2 (or Lightroom 3.2) using ACR's controls for fine-tuning the image including local tone and color correction C1 can't do.

Edit to correct LLC to LCC...

[undavide]

Jeff,
I didn't test ACR 6.2 so I'm not sure whether it apply the calibration corrections as you told or not, but if you say so I'm confident it will - I would have used the conversion to DNG as a workaround to "stamp" the calibrations and be able to use Camera Raw avoiding C1.

Now, it may seem that I don't want to use C1 at all (which is true indeed, and for someone this may be regarded as a bizarre attitude), but at least let me explain the why.

As opposed to photographers (who are likely to play an attack game, trying to reach the maximum quality in the field) as a retoucher I'm used to "play in defense". I've to deal with images that are already shot: if they're slightly out of focus, they're out of focus; if the exposure is not perfect, it's not; if there was a cloud shading the sun, I can't take the shot again, nor send the photographer and the helicopter back to Shanghai.

If I have to process aerial images, shot with P65+ at ISO 400, f5.6 with the very good 150mm in order to get fine art prints ranging 44"x  to 60"x (short edge) that suit my client desires I have to pay attention to the workflow's strategy and be very careful - and to my experience and my taste, which may differ from yours, C1 isn't that good dealing with P65+ high iso shots. Is it the demosaicing algorithm (with this kind of images I still prefer C1 v4 than C1 v5), is it the noise reduction, I don't know and I don't particularly care, since C1 leads from my very personal standpoint to files that are likely to give a lot of troubles when they're further post-processed.
Moreover, I'm a member of the group of people who do less in the raw processor and a lot in Photoshop, so I'm not really interested in local correction features of this or that software. I'll be happy to use C1, if it would output files that fit my needs.

That said, I don't want to sound too dogmatic - again, it's my experience, my taste, and my needs, yours will surely be different and that's not a problem at all.
To date my workflow is to use ACR for noise reduction in a smart object layer applied in Luminosity mode on the top of the C1 output (and that's the why I can't open the C1 tiff in ACR: advanced noise reduction features work, AFAIK, only with raw files and not demosaicized ones - and thanks god now ACR and C1 output files that have the same pixel width and height).
All the best,

Davide

PS
Steve, "not all photosites are created equal" will be one of my favorite quotations  

[Schewe]

That said, I don't want to sound too dogmatic - again, it's my experience, my taste, and my needs, yours will surely be different and that's not a problem at all.

Hey, I'm not telling you NOT to use Camera Raw...I'm kinda a fan of Camera Raw myself (I even co-author a book on it). To their credit, Phase One has actually been quite proactive in working with Adobe to make sure P65+ output from Camera Raw and Lightroom (6.2 and 3.2) has been substantially improved...it may strike some as odd that Phase One would actually work to improve Adobe's output but if you understand the mindset of the company, unlike Hasselblad that has worked to create vendor lock in and a closed system, Phase One has worked very hard to create a really open system and that includes working with 3rd parties to get the camera back raw files to be as good as can be.

I don't have a problem using Capture One although I really only use it for tethered shooting with my P65+ more than as a import and process tool. I'm pretty good with Camera Raw and with the improved support in Camera Raw 6.2 (and hopefully "official support" soon) I'm capable of getting very, very good conversions with Camera Raw that not only optimize for global but also local corrections. And as you've indicated, the ability to use Camera Raw to create a raw Smart Object even further extends the potential of Camera Raw + Photoshop.

[tesfoto]

Davide,

What will you do when a client comes with a P65+ file and a LCC correction file due to the choice of camera being an ALPA or Cambo ?

What about stitching files with LCC corrections. I could not even imagine dealing with this in ACR.

I have no idea why you are so reluctant to C1, I dont realy get your arguments.

Hoever much I like working with ARC, I think C1 is by far the best conversion software out there for P1 back files.
You can get everything you want from C1 if you know what you are doing.

As a retousher I would assume you would use the best possible tools for your work, and not trying to fit all in an ACR workflow.

[Doug Peterson]

C1 isn't that good dealing with P65+ high iso shots. Is it the demosaicing algorithm (with this kind of images I still prefer C1 v4 than C1 v5), is it the noise reduction, I don't know and I don't particularly care, since C1 leads from my very personal standpoint to files that are likely to give a lot of troubles when they're further post-processed.

To my taste the default noise reduction settings in C1 for high ISO from the P65+ and P40+ are too high.

Go to details tab to the noise reduction tool, turn luminance noise reduction to zero and then click the arrow to the right of the word "noise reduction" and select "set as defaults".

WELL worth the few minutes it will take you to try.

[tesfoto]

To my taste the default noise reduction settings in C1 for high ISO from the P65+ and P40+ are too high.

Go to details tab to the noise reduction tool, turn luminance noise reduction to zero and then click the arrow to the right of the word "noise reduction" and select "set as defaults".

WELL worth the few minutes it will take you to try.

Absolutly - I use this setting for low ISO files too.