The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Capture One or LR6?

M

mjr

Guest
I admit to finding this thread way beyond me, I don't understand it!

My idea of a RAW converter is pretty simple, it has to take the image I captured and allow me unhindered ability to create the final image, nothing more, nothing less. What I want that image to be depends on a myriad of factors, personal, business, creative, the mood I'm in, the list is pretty long. That being said, how the image looks when I open it is far less important to me than how the image looks when I am finished and ready for print or delivery to the client. As the way an image looks on opening is infinitely adjustable, what comes standard from C1 means nothing to me personally, it's just a starting point. For some images it could of course be perfect, for others, miles off, the point for me is that when I want to make minor or major alterations, the software should allow me to do that.

For me, C1 has far more options for alterations, colour editor, RGB curves, layers, masks from colour selections etc. etc. happen to work well for me, that doesn't make it better than LR, it just makes it better for me. If opening an image in LR you get exactly what you want then that's right for you. In C1 if the image is too contrasty I can do global, local and individual colour changes, same for saturation and all other parameters, I personally find the sharpening excellent and as I am experimenting more with grain on images, the C1 grain options are miles ahead of LR which are all reasons why C1 works for me.

As with all software, it takes getting used to, but I don't think it's possible to judge it's performance if you don't know how to use it properly, also even if you judge it as being poor for you, doesn't mean it is for anyone else. After reading Anders comment "As I've said before it's more difficult to get highly saturated colors right from a technical aspect, but still purple should not become blue, that must be some odd design choice by Phase One in this case. I really wonder what they have been thinking. In many genres you rarely come across colors as saturated as this though, so I guess that's why Phase One has gotten away with it." I shot a purple flower outside and opened it in C1 and LR and the flower is purple, or at least as purple as the original is, but more important than that, if I want it to be bluer or redder, then that is my choice in the colour editor, I can leave it as shot or use my own creativity to make it what I want, there isn't that level of control in LR.

Anyway, these are personal opinions, we are able to chose everything from subject to lens to camera to software to print or web, as long as those things work for the individual then that's all that matters isn't it?

Mat
 

Jack

Sr. Administrator
Staff member
Erik,

If you have time, could you dropbox me the original file with the blue/purple flowers? I'd like to take a closer look at it.

And just curious where and how you chose the WB for each of the processed versions, as they all seem pretty significantly different looking at the pot, wall and floor surrounding the flower area?

Thanks!

(jack-at-getdpi-dot-com)
 

Pradeep

Member
Anyway, these are personal opinions, we are able to chose everything from subject to lens to camera to software to print or web, as long as those things work for the individual then that's all that matters isn't it?

Mat

And that sums it up very nicely.
 

torger

Active member
I guess they've reduced those hue oddities for their later backs. After-all "common knowledge" has been that "Kodak" had less accurate color than "Dalsa", eg P45+ has been considered to have some issues, while P65+ and later was better, and from a hardware point that is actually true, it is easier to make a profile that produces a realistic look with a Dalsa than the Kodaks, but the profile has incredible power to mangle the colors and clearly purple flowers don't have to be blue even with the P45+. That is the hardware is not necessarily the reason.

In the film days the film itself had to make the "post-processing", so we got films like Fuji Velvia with a distinct saturated contrasty look popular for landscapes. When cameras became digital post-processing can be made in raw converters or by look profiles, but it only half-heartedly turned out that way -- still much of the film thinking has been kept with more or less strong subjective looks in the default camera profiles. This makes it a bit of a mess to change camera brand or raw converter.

Personally I like the default starting point to be neutral and realistic, and then I add my personal look on top. It's a logical and practical way to work, and regardless of camera used I can use the same methods and replicate the same looks. A common argument is that if you change the colors to your liking anyway, then it really doesn't matter which colors you start with. Maybe if you're a color and post-processing wizard with a super-memory, but I find it hard to start with any starting point and end up with the same end result, thus I think the starting point matter.

A neutral and realistic starting point suits me (and I do leave some images at that, depending on subject and context), but I perfectly understand that one might prefer the experienced folks at Phase One choose a starting point for us. I find it odd though that for cameras at this level often actually lack the choice of a realistic look, there's only subjective ones to pick. The only way to achieve it is by making a custom profile using some obscure open source software ;)

AFAIK DCamProf is the only software available for users that can make general-purpose profiles (that is embed a contrast curve and actually take that into account, not just slap it on top of a reproduction profile), and it's certainly not for all in its current state. So most users have to live with what comes bundled, it's not too bad though as the profiles aren't that crazy.

Profile making is no money-making business so we haven't seen much third-party products as said. The myth that camera color is mostly about hardware (rather than mostly about profile) is still strong and photographers in general seems pleased with the situation that cameras produces rather different looks for no reason, probably because most still believe that it's hardware-related rather than profile-related. I'd love to make a change, to give power of color to the photographers. The next step would be to make DCamProf a user-friendly product and make people aware that there is a different approach to color. It's unfortunately a huge project though.
 

ErikKaffehr

Well-known member
Hi Jack,

I would suggest that you use this image instead: https://www.dropbox.com/s/a9l5e4f99on072v/20160707-CF047205.eip?dl=0

The main reason is that image contains a grey card, it is exactly the same flowers, but another day.

The image used in the original posting is here: http://echophoto.dnsalias.net/ekr/Articles/Shoots/DivSamples/RAWS/20160722-CF047246.iiq, but that image was not intended for colour comparisons so it does not contain a grey card.

The grey card is the small one on the left of the fake ColorChecker. It is the WB side of my ColorChecker Passport.

Both images look very bright in Capture One at default settings, but they are correctly exposed to the right as checked with RawDigger (please note that I use log values on Y-axis)

Part of the reason the images look quite different is that the different profiles have different tone curves. My ambition was not to make the images similar but show the differences with minimal manipulation. DCamProf profiles have a different look from C1 profiles.

Here is an earlier image that I took in connection with a discussion with Tim Parkin of OnLandscape com that shows a similar issue: https://www.dropbox.com/s/zyzz57livf1i4ne/20150107-CF046070.iiq?dl=0, that one was shoot with studio flash.

Measured colour samples on that shot:


Screen Shot 2016-07-26 at 17.51.33.jpg
Screen Shot 2016-07-26 at 17.53.12.jpg

Best regards
Erik



Erik,

If you have time, could you dropbox me the original file with the blue/purple flowers? I'd like to take a closer look at it.

And just curious where and how you chose the WB for each of the processed versions, as they all seem pretty significantly different looking at the pot, wall and floor surrounding the flower area?

Thanks!

(jack-at-getdpi-dot-com)
 
M

mjr

Guest
We all work in very different ways Anders, I guess that as I spend at the very least, 2hrs a day, every single day on C1, often much longer than that, and have done for the last 18 months, I understand exactly what I'm getting when I import an image. I have preset styles I have created depending on the work I'm doing, portrait is obviously very different to landscape etc, I even have styles that I can use depending on which lens I am using, some I just prefer to tone down or increase contrast/saturation etc. I guess these are just mini profiles based on the tools within C1. This all means that with Phase files, I can never get close to a pleasing look with LR but that is as much down to familiarity as anything else. I am not necessarily always looking for the same output, in fact personal work I am always trying to find something new that appeals to me so I make use of lots of functions. I very rarely have to alter colours, almost never in fact but I know that should I feel something is over saturated or would be more visually pleasing at a slightly different hue then it is easy to do that.

Bottom line for me is that I am far happier with an image I process in C1 over the same image in LR for whatever reason and have never had an issue with the default opening image in C1 or the way colours are handled, I have far more problems with the image content or lack of!

Mat
 

Jack

Sr. Administrator
Staff member
Here is my 2 cent look and observations, FWIW.

First, I noted you captured the image with your back WB set to 3000 Color Temp. In itself, this is trivial as we work with raw files, however it will affect histo readout and may cause an exposure quirk if you rely on the histo -- so my reco is to set your back to flash (or daylight, but not auto) as it generally renders more accurate histos.

Next, re the file,

1) I undid your reduced exposure adjustment and returned it to zero, just to have all adjustments zeroed;
2) I set cam profile to Flash Easy Gray;
2a) I used the film std curve;
3) I set WB to Flash.

That's all I did for this result. I do not know if this rendering is a more "accurate" rendition of the petal actual color, but it appears to my eyes it is? Also note the WB readouts on your gray-card with these settings (Edit: I just noted the top flag on the checker -- I obviously slipped while clicking and grabbed a 4x4 pixel split between the top gray patch and dark gray border for the averaged 195 readout), and then I included flags for the flowers and wall just for reference and your comparisons. Here's a screenshot crop of my C1 and your image, but note that I flagged it while in working RGB but converted it to an sRGB jpeg for posting:



For whatever reason, the P45+ back color responded much better to the flash WB and profile, why I do not know. (This was not the case with the later Phase backs I've owned.) Anyway, it's the reason for my earlier reco you set the cam profile to Flash Easy Gray and set capture WB to Flash for that back.

Cheers,
 

ErikKaffehr

Well-known member
Hi Jack,

Thanks for checking out the file and for your 2 cent look.

Yes, I have noticed the WB-setting on the back and I think fixed it before the next shot. Must have pressed the wrong button. But normally use white balance card based WB. So WB setting on the back doesn't matter at all. In this case the exposure was pretty optimally ETTR as I could see in the histograms from RawDigger.

Choosing Phase One P45+ Flash - Easy Gray in combination with WB set to Flash seems to solve the problem. I used the WB card for white balance, and doing shifts colours to the blue.

Using FLASH for WB makes the violet petals OK but turns the Pelargonium leaves into yellowish green. So we are trading one problem for another one.

I would also say the -0.3 EV adjustment is needed, the image is far to bright. The buckets holding the flowers are painted black. Also if you check my "fake" ColorChecker the 4-th grey field should be around Lab 52 but your processing gives 79, so it is far to bright.

I don't really understand how you got 240,240,240 on the WB card as I got different numbers. If I use FLASH WB, I get 243, 240, 237. If I click on the WB card I get 241, 241, 240 and blue flowers

I enclose a screen dump. Settings on the left are yours (I hope) and on the right is a clone of your settings, but setting WB on the WB card. So you see the flowers turn blue. I guess it is good that your newer backs don't show this behaviour.

Incidentally, I had some lengthy discussion with Tim Parkin on the issue and he essentially said he "wouldn't touch the P45+ with a barge pole" and he was equally negative on the Hasselblad H3D39 that uses the same sensor chip.

On the other hand Anders Torger really likes the Kodak sensors. He says it's tricky to get good colour out of it, but can be done.

Screen Shot 2016-07-26 at 23.53.45.jpg

Best regards
Erik






Here is my 2 cent look and observations, FWIW.

First, I noted you captured the image with your back WB set to 3000 Color Temp. In itself, this is trivial as we work with raw files, however it will affect histo readout and may cause an exposure quirk if you rely on the histo -- so my reco is to set your back to flash (or daylight, but not auto) as it generally renders more accurate histos.

Next, re the file,

1) I was able to undo your reduced exposure adjustment and return it to zero as you can see;
2) I set cam profile to Flash Easy Gray;
2a) I used the film std curve;
3) I set WB to Flash.

That's all I did for this result. I do not know if this rendering is a more "accurate" rendition of the petal actual color, but it appears to my eyes it is? Also note the WB readouts on your gray-card with these settings, and then I included flags for the flowers and wall just for reference and your comparisons. Here's a screenshot crop of my C1 and your image, but note that I flagged it while in working RGB but converted it to an sRGB jpeg for posting:

For whatever reason, the P45+ back color responded much better to the flash WB and profile, why I do not know. (This was not the case with the later Phase backs I've owned.) Anyway, it's the reason for my earlier reco you set the cam profile to Flash Easy Gray and set capture WB to Flash for that back.

Cheers,
 

torger

Active member
So I guess the purple-became-blue flowers is a bug in one of the P45+ profiles then, or was it something more?

It would be interesting to know what Tim would think about the P45+/H3D-39 with a DCamProf profile. AFAIK the problem he had with the P45+ was green separation, or lack of green separation in landscapes? And the assumption was that it was due to the design of the color filters on the sensor, and this both P45+ and H3D-39 failed.

This is really interesting to me as I haven't really seen a camera fail to separate normal-saturation colors. I don't say it can't happen, I just haven't seen an example with screenshots etc. So it would be great to revisit if possible.

I don't particularly like the Kodak sensor's color filters as they don't make it easy for us to make neutral color, lots of non-linear corrections are then required, and as non-linear corrections can hurt gradients, we relax them and thus the colors tend to end up less on the mark than for a more modern sensor. However, I've found the problem to be small in practice, it's not like purple turns blue or anything like that. And as the Kodak was the last sensor with proper tech lens support, it's a natural choice for me. I made a tradeoff. But is it still bad at separating greens? I haven't noticed anything, but perhaps if I got to know details from Tim's observations I could find something?
 

Jack

Sr. Administrator
Staff member
Re P45+ greens...

Greens were the main reason I went to the P45+ and medium format in general -- they looked so much better than anything else I had used till then, and that was when I was using Canon.

This image was one of the first I made with the camera. Bob was with me on this trip, and if I recall correctly, it's the image that convinced him to make the MFDB move simply because of the green range:



Sometime later, I made this image on one of our workshops, also shows pretty nice green separation IMHO:



But, there is more to the story :ROTFL: It took a bit of work with WB, hue and sat sliders to get good green separation with those files, even with C1 -- in fact, I would often use some light color editor to aid green separation in my landscape images. Regardless, the fact I *could* get it was still paramount to me and I was very happy with the images I made with that back. However, as soon as the P65+ came out, I was even more impressed with native green separation. It took basically zero effort to "get it out of the file" as it was pretty much there from the start. Hence my P45+ was immediately traded in for the P65+...
 

Jack

Sr. Administrator
Staff member
I don't really understand how you got 240,240,240 on the WB card as I got different numbers. If I use FLASH WB, I get 243, 240, 237. If I click on the WB card I get 241, 241, 240 and blue flowers
What proofing profile do you have selected? See View>Proof Profile. If you have "selected recipe" which is recommended for accurate output, C1 will use the profile in your highlighted process recipe, which is not necessarily the checked recipe for actual output processing(!)

My guess is you have something limited chosen there, perhaps Adobe RGB that trims high blues and violets? Try setting it to Profoto or your back profile and you should see my same result.
 

ErikKaffehr

Well-known member
Hi Jack,

I have "selected recipe", but it doesn't really matter as I do my comparisons in Photoshop CS, with the image converted into 16 bit Prophoto RGB.

Anyway, there is something fishy. Clicking on the WB-card gives around 241,241,240 but changes flowers into blue, and using "Flash" as WB shifts them a bit off neutral.

Best regards
Erik


What proofing profile do you have selected? See View>Proof Profile. If you have "selected recipe" which is recommended for accurate output, C1 will use the profile in your highlighted process recipe, which is not necessarily the checked recipe for actual output processing(!)

My guess is you have something limited chosen there, perhaps Adobe RGB that trims high blues and violets? Try setting it to Profoto or your back profile and you should see my same result.
 

ErikKaffehr

Well-known member
Hi Anders,

I need to go back to my communications with Tim. His take on the issue was not green separation but greens having an unremovable yellow tint. Tim is good friend with British landscape photographer Joe Cornish and he also had the same issue. Of the digital cameras he tested, Tim found that the Sony Alpha 900 delivered best colour and the P45+ the worst. Tim also found a good match between DxO-marks SMI and his perception of colour rendition. Sony Alpha 900 on top and P45+ at the bottom.

Getting back to the violet flowers turning into blue the major factor seems to be white balance. Using Flash (5500K, 0) renders flowers violet and greens yellowish green.

WB on white card gets me 4719K, -2.1.

"Daylight" mode gives 4284K, 1.6, this always confused me with C1.

WB on the grey card using your DCamProf in Lightroom profile gives (5500K, 9)

The screen dump below shows the effect of WB, left is with WB set to "Flash" and right is WB on the white card. I have seen a similar problem on Sony Alpha 99 files under studio flash, so I don't think it is a P45+ issue. This page is a good demo: http://echophoto.dnsalias.net/ekr/Articles/OLS_OnColor/SimpleCase/

Screen Shot 2016-07-27 at 07.06.00.jpg

Best regards
Erik





So I guess the purple-became-blue flowers is a bug in one of the P45+ profiles then, or was it something more?

It would be interesting to know what Tim would think about the P45+/H3D-39 with a DCamProf profile. AFAIK the problem he had with the P45+ was green separation, or lack of green separation in landscapes? And the assumption was that it was due to the design of the color filters on the sensor, and this both P45+ and H3D-39 failed.

This is really interesting to me as I haven't really seen a camera fail to separate normal-saturation colors. I don't say it can't happen, I just haven't seen an example with screenshots etc. So it would be great to revisit if possible.

I don't particularly like the Kodak sensor's color filters as they don't make it easy for us to make neutral color, lots of non-linear corrections are then required, and as non-linear corrections can hurt gradients, we relax them and thus the colors tend to end up less on the mark than for a more modern sensor. However, I've found the problem to be small in practice, it's not like purple turns blue or anything like that. And as the Kodak was the last sensor with proper tech lens support, it's a natural choice for me. I made a tradeoff. But is it still bad at separating greens? I haven't noticed anything, but perhaps if I got to know details from Tim's observations I could find something?
 
Last edited:
M

mjr

Guest
I have zero issues with the 260 on colours so went back to look at some old files, I like greens so looked at some old P25+ files, obviously not a P45+ back but from the same era. These are a few years old but I still think the way greens are handled is pretty good. No idea what make of sensor was in the P25+, is it the same make as in the P45+?

This is a 50% crop from original file so quality may not look great but I think it's pretty nice, not great light either.



Here's one from home, shot in pretty crap bright light but still I think it holds up well

 

torger

Active member
Getting back to the violet flowers turning into blue the major factor seems to be white balance.
One experiment one can do is to slowly change white balance and watch how the colors gradually change. If a color suddenly changes rapidly compared to the general trend that means that there is a strong non-linear correction in the camera profile at that point. I use this sometimes to sanity check profiles. If I find that the profile has such a sudden change, I relax the LUT more.

If one prefers to have some sort of tint or otherwise "creative" white balance it's actually better to set the white balance for white first, and then use other color controls than white balance to achieve the creative tint. This is because the profile is per definition designed to make its corrections relative to white, and if there are strong non-linear corrections in the profile it can behave in unstable ways if the white balance is set differently.

What's strange here is that the white card white balance throws the profile out of balance. There's not a huge difference between the two white balance examples either, so I guess it's unstable for blue/purple there.
 

torger

Active member
I have zero issues with the 260 on colours so went back to look at some old files, I like greens so looked at some old P25+ files, obviously not a P45+ back but from the same era. These are a few years old but I still think the way greens are handled is pretty good. No idea what make of sensor was in the P25+, is it the same make as in the P45+?
The P25+ also has a Kodak sensor. The IQ260 has a Dalsa sensor. The Kodaks had quite saturated filters on the sensor, meaning that the sensor itself has quite subjective color, and as a profile designer you need to "fight" more to get the color where you want it. I don't really know but my guess is that Kodak's intention was to take some of the subjectivity from their color films into the sensor. The Dalsa has color filters similar to the modern CMOS sensors, that is lower saturation and more neutral and versatile. (Lower saturation filters require a lower noise sensor though as the profile will increase saturation and thus enlarge noise.)

I don't know how the color filters differ between the P25+ and P45+ (and Hasselblad's 50MP CCD backs which also use Kodak), but I think they are a bit different but still similar. The IR filter on the back can also modulate the response even if the sensor's color filters are the same.

In any case when the P65+ was released (Dalsa) many said there was big improvement in color over the P45+. This could be more about Phase One's profiles than the actual hardware, but indeed the color filters are much different between them too.
 
Last edited:

torger

Active member
I need to go back to my communications with Tim. His take on the issue was not green separation but greens having an unremovable yellow tint. Tim is good friend with British landscape photographer Joe Cornish and he also had the same issue. Of the digital cameras he tested, Tim found that the Sony Alpha 900 delivered best colour and the P45+ the worst. Tim also found a good match between DxO-marks SMI and his perception of colour rendition. Sony Alpha 900 on top and P45+ at the bottom.
The Phase One profiles have a yellow tint, but it's not only the P45+, the yellow tint / warm tone seems to be something that's incorporated to Phase One's trademark look :). Also Hasselblad adds in a bit yellow to the greens if I remember correctly, but to a smaller extent, and indeed in my own Neutral+ look I add in some yellow to the greens (in scenes with sunlit and shadow areas mixed such a change make up for a more pleasing look according to many). It should be easy to remove though, and I know Tim (and Joe) is a skilled photoshop user so I guess there must be some more to it than just the default rendering?

The SMI only says how well a camera can match X-Rite's CC24 with a linear matrix so I don't think it would mean that much. A good value does meant that less non-linear corrections is required in the normal range, but cameras can still turn out to be problematic with high saturation colors as those are not covered by the CC24, and this is indeed the case with some of the Sonys which I've had significant problems with in the saturated blue range due to very high sensitivity there, which is good for tungsten light but makes it unbalanced for daylight and dusk.

The thing is that camera profiles affect color rendition a lot so I think no color rendition analysis can skip past them. Still most do, and instead search for explanations to what they see in the hardware data, and I think that is because still most think that hardware is 90% of the color rendition and profile 10%, but I'd say it's the other way around. That said I've got the impression that my own DCamProf profiles are much more "designed" than the typical bundled profile which instead is a thinner layer on top of a linear matrix and as such let through more of the hardware's native look. It's a speculation at this point though.
 
M

mjr

Guest
Anders, what is the relationship between hardware and software when it comes to colour in your opinion? I noticed you wrote this..

"The myth that camera color is mostly about hardware (rather than mostly about profile) is still strong and photographers in general seems pleased with the situation that cameras produces rather different looks for no reason, probably because most still believe that it's hardware-related rather than profile-related. I'd love to make a change, to give power of color to the photographers."

And also in your last post this..

"The Kodaks had quite saturated filters on the sensor, meaning that the sensor itself has quite subjective color, and as a profile designer you need to "fight" more to get the color where you want it. I don't really know but my guess is that Kodak's intention was to take some of the subjectivity from their color films into the sensor. The Dalsa has color filters similar to the modern CMOS sensors, that is lower saturation and more neutral and versatile."

Is it an even split or biased towards one or the other? My personal view as a non technical understander of the design and implementation of cameras and software, is that different cameras have different looks because they are different cameras! I buy a camera because I prefer it's output over another camera but don't look much further than that. I remember opening a Leica S portrait file in C1 after I had processed it in LR as normal just to see, and immediately thinking WTF!! It was like a 10% white layer had been lifted and all this beautiful rich tone was released, did the same with Nikon files and felt I had wasted so many shots by not processing them in C1, so I understand software and profiles make a huge difference, just wondering what the relationship really is?

Mat

Edit.. We were writing at the same time, you have answered by saying it's 10% camera and 90% profile.
 

torger

Active member
We were writing at the same time, you have answered by saying it's 10% camera and 90% profile.
Yes, but I can't say that I've drilled down too deeply into that. 10/90 I would guess is an exaggeration, and I also think that many profiles let the camera filters "show" more than my DCamProf profiles do. When a DCamProf profile is rendered it tries to make all cameras look the same, but as non-linear corrections are relaxed to make good gradients they will still differ a bit.

At some point I may write an article on the subject as I'm curious about this myself. My current view that it's much more about profile than camera's color filters is from stray observations here and there I've got when I've worked with profiles, but I haven't really made a focused effort to answer that specific question.

The most obvious "stray" observations:

1) two different cameras used with profiles designed for the same target renders colors highly similar.
2) the same camera with two different profiles designed differently can produce vastly different looks.

The large difference in default look between LR and C1 comes down to profile.

I have noticed hardware differences though, you can't make two different cameras look exactly the same if you want the profile to make good gradients, that is one cannot have too much non-linear corrections but instead be closer to a linear matrix which means closer to the hardware response. Anyway the broad brush strokes are decided by the profile - the overall "look" is profile, but some details in things like color separation can be hardware related.

And there's a special case about high saturation colors when you use the edge of the color filters, then cameras can behave very differently in ways the profile can't do much about.
 

ErikKaffehr

Well-known member
Hi,

This is what I got from Tim, where he referes to Joe...

"I spoke with Joe about the P45 endlessly and he had a whole workflow to try to 'fix' the colour which all went out of the window when the IQ180 came in. He now does almost nothing to the files to get what he wants. It's definitely mostly grassy stuff but quite often skies and reflected colour that are problematic. Also shadows in geology can be an issue (I presume infra red effects as well).

They're not a problem for some subjects and an absolute nightmare for others. I can still 'see' Joe's P45+ files as long as they have some greenery in them (or sometimes from the sky)."

So they found the issues went away with the IQ-180, that uses a Dalsa sensor. I don't know what conversion they used.

My main concern in that discussion was not if P45+ colour rendition was good or bad but rather if it could be fixed with proper profiles. The other question was if I can get good colour rendition out of LR by using colour profile generation tools demanding reasonable skills.

I think that Jack's suggestions illustrate one of the issues, much is counter intuitive:

  • Use an electronic flash profile for a daylight shot
  • Use WB-setting "Flash" instead of grey card included in the picture.

So, it takes some time to master a new raw converter. Also, using a set of tools for ten years introduces some bias. I was mostly happy with Lightroom/Sony rendering. So I am a bit biased to that look. But, I feel that DCamProf has great promise so I will probable shift my workflow to it.


Best regards
Erik





The Phase One profiles have a yellow tint, but it's not only the P45+, the yellow tint / warm tone seems to be something that's incorporated to Phase One's trademark look :). Also Hasselblad adds in a bit yellow to the greens if I remember correctly, but to a smaller extent, and indeed in my own Neutral+ look I add in some yellow to the greens (in scenes with sunlit and shadow areas mixed such a change make up for a more pleasing look according to many). It should be easy to remove though, and I know Tim (and Joe) is a skilled photoshop user so I guess there must be some more to it than just the default rendering?

The SMI only says how well a camera can match X-Rite's CC24 with a linear matrix so I don't think it would mean that much. A good value does meant that less non-linear corrections is required in the normal range, but cameras can still turn out to be problematic with high saturation colors as those are not covered by the CC24, and this is indeed the case with some of the Sonys which I've had significant problems with in the saturated blue range due to very high sensitivity there, which is good for tungsten light but makes it unbalanced for daylight and dusk.

The thing is that camera profiles affect color rendition a lot so I think no color rendition analysis can skip past them. Still most do, and instead search for explanations to what they see in the hardware data, and I think that is because still most think that hardware is 90% of the color rendition and profile 10%, but I'd say it's the other way around. That said I've got the impression that my own DCamProf profiles are much more "designed" than the typical bundled profile which instead is a thinner layer on top of a linear matrix and as such let through more of the hardware's native look. It's a speculation at this point though.
 
Top