ICC Comparison: IQ180 vs. P65+

[alan_w_george]

Interesting, I'm not sure what you can conclude from this, but it seems the P65+ has a wider (in most areas) response gamut than the IQ180. (Daylight profiles compared).

 

[Garcia]

I think this issue is related to the profile creation parameters and how a profile renders color, and it doesn't have to do with CCD capabilities. Aptus 12 profiles are way bigger than IQ180 too, in particular, those labeled as HS, (high saturation), too much color for my taste coming from a P30+, BTW, but easy to fix in C1

 

[alan_w_george]

I think this issue is related to the profile creation parameters and how a profile renders color, and it doesn't have to do with CCD capabilities. Aptus 12 profiles are way bigger than IQ180 too, in particular, those labeled as HS, (high saturation), too much color for my taste coming from a P30+, BTW, but easy to fix in C1

I am no expert, but then it would seem to me that using this profile would reduce (rather drastically it would seem) the gamut available from the back assuming this profile is not, as you are saying, a true representation of what the back is capable.

Maybe Phase is trying to remap the available gamut from the back into a smaller but high density gamut for outdoor daylight purposes?

 

[Garcia]

A tight profile is not necessary a bad thing. Colors in real world are far less saturated than one could expect, and a small gamut implies more numbers available to draw smooth tonal transitions. Color spaces like ProPhoto waste a lot of numbers describing colors that you won't ever see.

As a matter of fact, custom ICC profiles I created for my P30+ and for my Aptus 12 are way smaller than the ones embedded in CaptureOne but they work as a charm.

I don't think Phase has deliberately reduced the size of IQ180 profiles for any reason. Maybe they used different tools ( software and test charts ) to generate them.

 

[wentbackward]

Going on the assumption that a lot of experience must have surely been gained between these profiles being produced, this is actually quite interesting from the perspective of what the profile designer was intending to do! A smaller gamut will definitely avoid quantization errors, as Garcia points out, I think though, there's some other reason as even on my meagre 33mp, 16bit back, I can't imagine quantization errors are a major concern. I think it may be related to what is more manageable from an output perspective. With the data spread across a wider gamut, you end up with more work to do in production, and/or more chance of shifting into 'illegal' colours which cannot occur naturally.

So my guess is that it's a bit like like cheese spread. If you take one cheesy triangle (raw data) and spread it across a large piece of toast until you're happy (in capture one), you might want it a bit thicker in places and thinner in others, perhaps stealing from another cheesy triangle (contrast/saturation). you then come to printing and you need scoop it off and make it fit on a smaller, different shaped piece of toast (say a printer/paper profile) you'll find that all the spread overhanging the edges, will either bunch up as you squish it on (this is what rendering intent deals with) or you'll just have to lob it off and do away with it (clipping) and where the new piece of toast is sticking outside the old one, you'll have to spread it a bit thinner (loss of saturation).

For most people (without a huge production house behind them) it is probably better to leverage the profile designers experience than rely on algorithms that perform rendering intent. This is bit like film preference in the old days, it's much easier to match films and papers than to understand chemistry. I actually think this is one of the benefits of a Leaf back over a POne back is the profiles can mostly get you in the ballpark, whereas the brief play I had with an IQ was a lot more work to get the right result.

Of course I could be completely wrong, but printing "orange" from my leaf files has left me exasperated. It must be the hardest colour to deal with.

 

[GrahamWelland]

Of course I could be completely wrong, but printing "orange" from my leaf files has left me exasperated. It must be the hardest colour to deal with.

Try the bluebell ... my personal nemesis due to the difference between human optical perception of the colour vs absolute captured colour. Tough rendering blue vs purple vs pink bells in my experience. I haven't tried it with a Phase One camera yet but both Kodak DCS and Nikon DSLR's had a tough time with it.

Sometimes a subjective rendering of colours based on the profile designer's experience is better than an objective mapping.

 

[goesbang]

Of course I could be completely wrong, but printing "orange" from my leaf files has left me exasperated. It must be the hardest colour to deal with.[/QUOTE]

Printing Orange is primarily a product of limitations in the CMYK process. It is also a major problem for inkjet printers, including the high end 8 colour printers from Canon, HP and Epson. Interestingly, HP claim to have "solved" the problem by adding an orange ink into their new inkset. These printers have more than 8 colours (10 or 12).

 

[Garcia]

Try the bluebell ... my personal nemesis due to the difference between human optical perception of the colour vs absolute captured colour. Tough rendering blue vs purple vs pink bells in my experience. I haven't tried it with a Phase One camera yet but both Kodak DCS and Nikon DSLR's had a tough time with it.

Sometimes a subjective rendering of colours based on the profile designer's experience is better than an objective mapping.

There is a very informative writing at Betterlight.com covering a similar issue with cobalt blue and how to deal with, which is related to certain spectral response characteristics of some pigments.

http://betterlight.com/downloads/whitePaper/wp_color_accurate_photo.pdf

As Siebel points out, capture and printing cause different problems. My ' nemesis ' as you call it, is printing dark saturated colors like deep reds, in 12 color Canon iPF printers.