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!

IQ180 - Remote Shutter Release not working anymore

anwarp

Well-known member
A few weeks ago, I was trying to profile my printer through C1.

I loaded a greyscale tiff with known grey values. The RGB values shown by the dropper in C1 were off in Luma. E.g. A patch that should be 20, 20, 20 for RGB values showed about 14, 14, 14 in C1, but the correct values in affinity photo.

I believe the problem is that the linear curve that C1 uses has a slight curve to it, changing the luminance, especially at the far ends. Linear scientific is only available with some cameras.

The LED lights might add other issues due to the colour spikes?
Can you retry the same experiment using daylight instead? To eliminate the light source from the equation.


Anwar
 

Ray Harrison

Well-known member
Just a quick note before starting my day.

I've taken an initial look at the two raw files you provided and am still in the middle of things. To answer the one question of what it looks like doing just white balance: "bloody awful" - super saturated oranges, reds, browns, etc. The challenge with this particular generation of CCD backs (I used to own the Credo 80) is that they don't have a ton of tolerance for long exposure and less so when they are underexposed. These are both under by about 3/4 of a stop, maybe a bit more. You're still within the technical range of the back, time-wise, at ~30 seconds for each image, but something to consider, especially underexposed. The base characteristics ICC profile and base curve for these backs also tend to heavily saturate reds even on the best of days so at the very least, use one of the daylight profiles and away from the "Flash" ones.

There's another thread here from a person back in 2011 that talks about this in particular, and there's some interesting information there: https://www.getdpi.com/forum/index.php?threads/iq180-with-phase-one-df-camera-colour-issue.28156/

I haven't been successful yet at creating a useful profile. I've got a couple of more ideas I'll try. In short, though, this generation of 80MP CCD backs are stellar when everything is perfect but even then require attention. They can be more challenging when conditions aren't perfect and can run off the road and into a ditch pretty quickly.

More to come...
 

KEVINS

Member
A few weeks ago, I was trying to profile my printer through C1.

I loaded a greyscale tiff with known grey values. The RGB values shown by the dropper in C1 were off in Luma. E.g. A patch that should be 20, 20, 20 for RGB values showed about 14, 14, 14 in C1, but the correct values in affinity photo.

I believe the problem is that the linear curve that C1 uses has a slight curve to it, changing the luminance, especially at the far ends. Linear scientific is only available with some cameras.

The LED lights might add other issues due to the colour spikes?
Can you retry the same experiment using daylight instead? To eliminate the light source from the equation.


Anwar
I attempted to take new pics using natural sunlight and the results weren't much better. I tried taking them outside but the painting reflects everything too much so I brought it inside where I was able to control the reflections better.

I saw Rays posting above about the long exposure so I changed the camera settings from F8 and 20-30 second exposure times to F4 and 2.5 second exposures.

WB Correction still produced a orange hue to the complete painting.
ICC profile still dulled all the colors.
The LAB colors get all stupid where some are supposed to be 60-ish but after correction end up at 80-ish.

Ray, I sent you this new indoor file with a 2.5 second exposure.

ks
 

Ray Harrison

Well-known member
I attempted to take new pics using natural sunlight and the results weren't much better. I tried taking them outside but the painting reflects everything too much so I brought it inside where I was able to control the reflections better.

I saw Rays posting above about the long exposure so I changed the camera settings from F8 and 20-30 second exposure times to F4 and 2.5 second exposures.

WB Correction still produced a orange hue to the complete painting.
ICC profile still dulled all the colors.
The LAB colors get all stupid where some are supposed to be 60-ish but after correction end up at 80-ish.

Ray, I sent you this new indoor file with a 2.5 second exposure.

ks
Thanks, I'll take a look!
 

KEVINS

Member
geezz... I have been trying all kinds of different things. Based on the thread above that mentions a possible issue how the reds are treated I decided to try a different painting that is mostly grey/green/blue.

The ICC profile that was generated through the ColorChecker software and applied in C1 and again it just dulled the colors.
Doing a WB correction also screwed up the image.

I then took a photo using my 5DSr and generated an ICC and it looks pretty decent although the LAB colors are still off a bit.
5dsr.JPG

I then opened up the 5DSr file in C1 and created an ICC profile and it wasn't awfully bad but there is now a red tint on the frame but it doesn't look washed out:
5DSr opened in C1.JPG


No idea what I concluded other than Photoshop gives me far better results than C1 even tho the LAB colors are still pretty far off.
The below is a screenshot with the same PhaseOne IIQ file corrected in Photoshop on the left and C1 on the right.
IIQ file Color Corrected in C1 and Photoshop.JPG

*
*
I haven't compared my 5DSr against the IQ180 and color corrected in PS yet...

ks
 

Ray Harrison

Well-known member
geezz... I have been trying all kinds of different things. Based on the thread above that mentions a possible issue how the reds are treated I decided to try a different painting that is mostly grey/green/blue.

The ICC profile that was generated through the ColorChecker software and applied in C1 and again it just dulled the colors.
Doing a WB correction also screwed up the image.

I then took a photo using my 5DSr and generated an ICC and it looks pretty decent although the LAB colors are still off a bit.
View attachment 201020

I then opened up the 5DSr file in C1 and created an ICC profile and it wasn't awfully bad but there is now a red tint on the frame but it doesn't look washed out:
View attachment 201021


No idea what I concluded other than Photoshop gives me far better results than C1 even tho the LAB colors are still pretty far off.
The below is a screenshot with the same PhaseOne IIQ file corrected in Photoshop on the left and C1 on the right.
View attachment 201019

*
*
I haven't compared my 5DSr against the IQ180 and color corrected in PS yet...

ks
I have to say thank you for sending the raw files over, it's been fun to work on this! In the past, I've been pretty successful generating profiles for use in C1 when I've needed to with my Phase backs for product shoots. I haven't really been successful here :) . I can sort of get a sense of what the painting should look like from the above. I find the last image you sent is still underexposed by maybe 2/3 stop. The illumination consistency on the color checker card is a little off too, as far as uniformity is concerned. Interestingly, I also find that choosing the Phase IQ4150->Neutral profile and popping the exposure 1/2 a stop or so is more pleasing to my eye and not an unreasonable place to start. That's entirely subjective of course. I'm sure you've done similar experiments with other profiles. I don't know if this is something you're already doing, but I find AWB to be hit and miss and it's better to choose something like daylight (or whatever - something "known") to start from. Not that it specifically matters in raw files, but just starting 5k (or your choice) lets one (or at least me) have "more knowns than unknowns". Also, what camera body are you using? The XF or something like the DF+? How are you doing the metering? Do you have the option of controlled lighting like quality strobes, maybe off at an angle that minimizes reflections but providing a uniform lighting on the picture? That gives you controlled exposure in the parameters of the back and the ultimate quality.

The challenge with creating profiles is that you're not profiling the camera, you're profiling the camera, the lens used (at least to some degree) and the lighting in use at the time. These work OK in controlled studio environments but even then you may have multiple profiles covering lenses and lighting scenarios.

CF007745 2.jpg
 
Last edited:

KEVINS

Member
Thanks Ray for wanting to help/play with the images. It's a bit reassuring that it's not completely me that is having issues, tho I was REALLY hoping that you'd have it figure out with your software/process in 10 minutes. I had had my CC ready to purchase software if you were successful. Your rendering looks pretty close so you probably have a decent idea of what the painting looks like.

Lighting is from two large LED lights one on each side of the painting at a specific distance and angle to eliminate any glare off the canvas weave. I've tried settings from 5600K to 5200K and power from 70% to 100% with no difference. The lights also have linear polarizing filters on them along with the camera lens.

Darkened room (no lights on) with 10ft high, grey colored walls - the orange wall in the photo is the only wall that is not grey.
If you have Instagram I made a quick little video on photo-day and you can see how things are setup and probably get a decent idea of what the actual painting looks like:

http://instagr.am/p/Cok9IV9PhZn/
I do have strobes I can setup but I don't think it will solve the issue at hand. If the color was close and I wanted to make Giclee's from the images then I would try other lighting but since I just need images that I feel represent the painting, the images I have created with my 5DSr/Photoshop are close. But I was hoping to snap the colors in better with a better ICC profile especially now that I am getting international recognition.

I'm still not happy with the ICC profiles that the Xrite software produces. When comparing the LAB values of my 5DSr in PS after color correcting the values are still off. Some values are supposed to be 40 but end up 60+, etc. If you are willing to try one more image I'd like to send you a file from my 5DSr to see how that works in your software and see if the LAB colors are more accurate? If not that's OK too you have gone far above and beyond what I thought would be required so I greatly appreciate the time you put on this!
****

As far as exposure goes I am fighting a color rendering issue in the brighter colored areas, specifically the forehead. The image below is a side-by-side of the 5DSr image (left) and the IQ180 (right) with the closest camera settings I can use to achieve a similarly exposed image. See how the forehead looks a lot brighter on the IQ180 and is missing transistion colors but the 5DSr picks them all up? I have been trying all types of camera settings/camera positions/light settings/camera lens, etc to get the IQ180 to pick up on those colors with 0 luck. This is where I was hoping a GOOD ICC profile would fix this.

sidebyside.JPG


I did come close and have written the entire "formula" down but the gold standard is still the 5Dsr image so the IQ180 image won't be submitted to anything important.

Kevin
 

Ray Harrison

Well-known member
Thanks Ray for wanting to help/play with the images. It's a bit reassuring that it's not completely me that is having issues, tho I was REALLY hoping that you'd have it figure out with your software/process in 10 minutes. I had had my CC ready to purchase software if you were successful. Your rendering looks pretty close so you probably have a decent idea of what the painting looks like.

Lighting is from two large LED lights one on each side of the painting at a specific distance and angle to eliminate any glare off the canvas weave. I've tried settings from 5600K to 5200K and power from 70% to 100% with no difference. The lights also have linear polarizing filters on them along with the camera lens.

Darkened room (no lights on) with 10ft high, grey colored walls - the orange wall in the photo is the only wall that is not grey.
If you have Instagram I made a quick little video on photo-day and you can see how things are setup and probably get a decent idea of what the actual painting looks like:

http://instagr.am/p/Cok9IV9PhZn/
I do have strobes I can setup but I don't think it will solve the issue at hand. If the color was close and I wanted to make Giclee's from the images then I would try other lighting but since I just need images that I feel represent the painting, the images I have created with my 5DSr/Photoshop are close. But I was hoping to snap the colors in better with a better ICC profile especially now that I am getting international recognition.

I'm still not happy with the ICC profiles that the Xrite software produces. When comparing the LAB values of my 5DSr in PS after color correcting the values are still off. Some values are supposed to be 40 but end up 60+, etc. If you are willing to try one more image I'd like to send you a file from my 5DSr to see how that works in your software and see if the LAB colors are more accurate? If not that's OK too you have gone far above and beyond what I thought would be required so I greatly appreciate the time you put on this!
****

As far as exposure goes I am fighting a color rendering issue in the brighter colored areas, specifically the forehead. The image below is a side-by-side of the 5DSr image (left) and the IQ180 (right) with the closest camera settings I can use to achieve a similarly exposed image. See how the forehead looks a lot brighter on the IQ180 and is missing transistion colors but the 5DSr picks them all up? I have been trying all types of camera settings/camera positions/light settings/camera lens, etc to get the IQ180 to pick up on those colors with 0 luck. This is where I was hoping a GOOD ICC profile would fix this.

View attachment 201036


I did come close and have written the entire "formula" down but the gold standard is still the 5Dsr image so the IQ180 image won't be submitted to anything important.

Kevin
Thanks for the additional information Kevin! And absolutely, feel free to fire over a 5DSr image. I'd love to have a go.
 

KEVINS

Member
Thanks for the additional information Kevin! And absolutely, feel free to fire over a 5DSr image. I'd love to have a go.
WOW! THANK YOU SO MUCH!!!
I'm curious if your ICC profile can create accurate LAB colors with this image and my 5DSr. Using my Xrite software still doesn't produce accurate LAB colors so I'm curious how your software performs.. As you can probably guess the skin tone row #8 is real important for this piece..=)

For those curious, here's a screen shot of the painting which I'm in the middle of double-framing today.

Capture.JPG

Kevin
 

KEVINS

Member
Small update for those that want something to read:

I have been in contact with Ashley at Xrite and I gave her my image. She produced an ICC and gave it to me to try. It's far better than mine but the LAB colors are still off. The image looks/prints too light still but better than my ICC. All my C1 settings are at default but maybe hers are modified. Hopefully she has thoughts that could explain it.

My ICC on Left - Her ICC on the right:
compare.JPG



Between Ray's and my results I have been concerned that maybe this IQ180 is not performing correctly so I am also in contact with Steve (again) at Capture Integration and tho he won't be in the office till sometime next week he said he will try some things out on his end with an IQ180 when he gets back.

Given that Ashleys ICC seems to work better than mine for some reason maybe it's still a software/profile issue, not the IQ180.

KS
 

Ray Harrison

Well-known member
Small update for those that want something to read:

I have been in contact with Ashley at Xrite and I gave her my image. She produced an ICC and gave it to me to try. It's far better than mine but the LAB colors are still off. The image looks/prints too light still but better than my ICC. All my C1 settings are at default but maybe hers are modified. Hopefully she has thoughts that could explain it.

My ICC on Left - Her ICC on the right:
View attachment 201171



Between Ray's and my results I have been concerned that maybe this IQ180 is not performing correctly so I am also in contact with Steve (again) at Capture Integration and tho he won't be in the office till sometime next week he said he will try some things out on his end with an IQ180 when he gets back.

Given that Ashleys ICC seems to work better than mine for some reason maybe it's still a software/profile issue, not the IQ180.

KS
Great info! I spent a little more time with this today. I got a bit of a better one with some tweaking on the IQ180 which I’ll try and shoot over to you tomorrow. It did pretty well on row 8, and elsewhere. The challenge with an in-situ profile, if you are after correct color, is it needs to have careful exposure and you have to really work to get the illumination uniform. The images have been somewhat underexposed, and while it’s possible to get away with it, there can be challenges, especially with cameras that aren’t great performers with under exposure (like the IQ180). Too, these sorts of profiles aren’t going to be as versatile as the profiles available in Capture One in many situations. They’re built with more flexibility in mind. Hand built profiles really only reflect the camera, the lighting and the lens at the time. That’s why it’s a great technique for cultural heritage/museums. Change any of those things and there’s a strong possibility you need a new profile. Interestingly, I had pretty decent results with either an IQ3 or IQ4 “pro standard” profile. The hue shifts in shadows and darker areas were much more uniform, so you may think about looking at those.
 

KEVINS

Member
Thanks again Ray!
I do have images that are are centered-metered or very close if you want to play with a few of those?

I also have images where I metered the overall image, took the picture, then walked up to the card and took another pic with the exact same camera setting. I thought this would work perfectly having a larger card image to create the profiles from but the image of the card was waaaaay over exposed when up that close so the ICC was really bad.
While I was performing the above workflow I took photos at several different exposures but never thought to try and match exposures and ignore camera settings when creating the profiles. I may try this when I get home tonight and if you want I can send you a few of these images.

Yea, so far the closest corrected image is produced with the default ICC for the IQ180 but the forehead on the figure ends up being too bright and loses the color transition so this is what I was hoping a custom ICC would repair (Row 8 on the SG card).
Thinking that the painting was too brightly lit, thus causing the forehead to be over exposed, I started moving lights farther back and moving the camera back. Moving lights didn't help just started inducing glare. Moving the camera back helped a very tiny but not a lot. If I was getting decent ICC profiles similar to what Ashley at Xrite produced then I could fine tune my lighting and camera settings but I was never able to produce anything even close no matter what I did.

Thoughts on Color Cards?:
It seems the ColorChecker SG is extremely popular and most/all software recognize it. However, the one I like, and can't find a way to purchase, is the CF card from Coloraid.

I like all the minor transitions of the lighter colors and would like to try this card along with basiccolor. Any idea how to purchase this card? I have sent two emails from two different email accounts and have not had a reply.

At some point I will probably purchase basiccolor 6-Pro ($800) b/c I like the idea of creating a custom ICC profile using multiple cards, unless I misunderstood what was being said in the video. The bad thing about me buying basiccolor 6-Pro is that is will probably have sooo many settings that it's probably not plug-n-play like the Xrite. There are very few how-to vids available so the learning curve could be a frustrating one.

If there are specific camera settings/exposures/etc that anyone wants to try let me know and I can easily accommodate and get the files.

Regards
Kevin
 
Last edited:

Ray Harrison

Well-known member
Thanks again Ray!
I do have images that are are centered-metered or very close if you want to play with a few of those?

I also have images where I metered the overall image, took the picture, then walked up to the card and took another pic with the exact same camera setting. I thought this would work perfectly having a larger card image to create the profiles from but the image of the card was waaaaay over exposed when up that close so the ICC was really bad.
While I was performing the above workflow I took photos at several different exposures but never thought to try and match exposures and ignore camera settings when creating the profiles. I may try this when I get home tonight and if you want I can send you a few of these images.

Yea, so far the closest corrected image is produced with the default ICC for the IQ180 but the forehead on the figure ends up being too bright and loses the color transition so this is what I was hoping a custom ICC would repair (Row 8 on the SG card).
Thinking that the painting was too brightly lit, thus causing the forehead to be over exposed, I started moving lights farther back and moving the camera back. Moving lights didn't help just started inducing glare. Moving the camera back helped a very tiny but not a lot. If I was getting decent ICC profiles similar to what Ashley at Xrite produced then I could fine tune my lighting and camera settings but I was never able to produce anything even close no matter what I did.

Thoughts on Color Cards?:
It seems the ColorChecker SG is extremely popular and most/all software recognize it. However, the one I like, and can't find a way to purchase, is the CF card from Coloraid.

I like all the minor transitions of the lighter colors and would like to try this card along with basiccolor. Any idea how to purchase this card? I have sent two emails from two different email accounts and have not had a reply.

At some point I will probably purchase basiccolor 6-Pro ($800) b/c I like the idea of creating a custom ICC profile using multiple cards, unless I misunderstood what was being said in the video. The bad thing about me buying basiccolor 6-Pro is that is will probably have sooo many settings that it's probably not plug-n-play like the Xrite. There are very few how-to vids available so the learning curve could be a frustrating one.

If there are specific camera settings/exposures/etc that you want to try let me know and I can easily accommodate and get you the files.

Regards
Kevin
I use the ColorChecker SG and it holds its own pretty well. I have used but don't own a handful of more expensive charts, and have a few other colorchecker variations lying around. As for basiccolor, it's more straight forward and easy to use than it might first seem and super powerful. It has the occasional quirk but really, I've personally found it worth the money.
The overall process is the same to create the tiff and most of the work is still in getting a good exposure of your card. In Capture One it's Base Characteristics -> ICC Profile -> Effects -> No color correction, Curve->Linear Response (or Linear Scientific is you have it for the camera), white balance on F5 and export (embed camera profile). Drag-n-drop onto the basiccolor interface and away you go.


You may see where you get by applying one of the Phase One IQ3100 or IQ4150 ICC profiles in Base Characteristics in Capture One. If you don't see them, under ICC Profile, choose "Show All".

I haven't quite caught up with the latest files you've sent but am happy to take a look at some of the others!
 

KEVINS

Member
A question to chew on - I'm curious what the proper workflow "should" be:

Base Setup:
1. LED Lights are setup in specific areas to illuminate the painting from the left and right to eliminate glare and reflections so these should never move. Currently each light is 2ft in front of the paintng but 7ft to the left and right. This gives me great lighting with no glare on the weave of the canvas.

2. Camera Position is placed to encompass the entire piece so this won't change. Most of my paintings are 40" tall.

3. Easel is located at specific location with a black cloth behind the easel so this won't change.

Workflow ?'s:
1. With camera located in specific location as noted above: With proper exposure take picture of color card on the easel without painting, just the card.
2. Create ICC profile making sure LAB colors are as accurate as possible.

3. Place painting on easel (no color card), expose as needed then take picture. (Exposure settings may be different than settings required for color card).

Question:
Since the "exposure" of the card and the painting should be nearly equal would we expect the ICC profile to work accurately?
Because:
If the card is placed on the painting and the exposure is adjusted to get a good exposure of the painting then it's possible the card will be underexposed or overexposed depending on how bright the overall color of the painting is.. Yellow painting vs Brown painting. Correct? If this is true then I can see where the ICC won't work.

Hmmmmm.:unsure::unsure::unsure:

I'm getting quotes for basiccolor and they recommended a training course but didn't give me a price. Any thoughts on what this encompasses?

Edit: Ouch...the quote for the class is $500...

ks
 
Last edited:

Ray Harrison

Well-known member
A question to chew on - I'm curious what the proper workflow "should" be:

Base Setup:
1. LED Lights are setup in specific areas to illuminate the painting from the left and right to eliminate glare and reflections so these should never move. Currently each light is 2ft in front of the paintng but 7ft to the left and right. This gives me great lighting with no glare on the weave of the canvas.

2. Camera Position is placed to encompass the entire piece so this won't change. Most of my paintings are 40" tall.

3. Easel is located at specific location with a black cloth behind the easel so this won't change.

Workflow ?'s:
1. With camera located in specific location as noted above: With proper exposure take picture of color card on the easel without painting, just the card.
2. Create ICC profile making sure LAB colors are as accurate as possible.

3. Place painting on easel (no color card), expose as needed then take picture. (Exposure settings may be different than settings required for color card).

Question:
Since the "exposure" of the card and the painting should be nearly equal would we expect the ICC profile to work accurately?
Because:
If the card is placed on the painting and the exposure is adjusted to get a good exposure of the painting then it's possible the card will be underexposed or overexposed depending on how bright the overall color of the painting is.. Yellow painting vs Brown painting. Correct? If this is true then I can see where the ICC won't work.

Hmmmmm.:unsure::unsure::unsure:

I'm getting quotes for basiccolor and they recommended a training course but didn't give me a price. Any thoughts on what this encompasses?

ks
I should be very clear that I'm not a color or profile expert. I'm always learning and I find it thoroughly interesting. I do make profiles for certain product shots where color is important but it's only part of the equation and in general, I use Capture One profiles, certainly for my Phase gear. There are far, far more qualified people on this forum who can speak with authority on such things. I'm merely a garden variety schmuck. I can only offer my own experiences and hopefully there are nuggets worth something :D. With that...

What's your end output? I know you sent the painting image in question to a magazine for a cover recently. Is that typical? Do you send them to social platforms and such? The question is really, how accurate do you want your colors? Do they have to look good and seem right or... do they absolutely have to be spot on? The latter can be a fair amount more work, in my experience. Profiles are part of the story and help, but the Capture One color management tools are another part (or whatever your processor of choice happens to be). If the answer is the former ( look good and seem right ), I'd really recommend getting a nicely exposed shot of your painting and experimenting with non-IQ180 profiles and working with the excellent color management tools in C1 to tweak as needed. Really, if you get a nicely exposed shot, you can probably tweak the image directly from the IQ180 profiles themselves. The included ICC profiles in C1 are designed to be much more flexible than an in-situ profile. They work in a variety of lighting situations and lens choices.

Making a generalized camera profile is possible similar to what you find in C1 or other tools, but it's (way) beyond my level, for sure. There's a really in-depth article by the developer of Lumariver here which is super interesting and very technical. But basically, what we're doing with a tool like Basiccolor or the software from Xrite is we're building an in-situ profile. It works for the specific camera, the lens, the lighting, the scene and so forth. Well exposed, even illumination on the card of choice is very important, I've found. I would trade slightly underexposed for making sure it was evenly illuminated. If it's a little underexposed but still evenly lit, you can tweak the levels so that the L* value is 95-ish on, say A1 or E5 on the ColorChecker SG. Don't change the L*value with exposure as it's very non-linear. So levels work great. Basically, you want the white point of A1 or E5, say, to be in the 95 range. The steps I use (YMMV) typically are:

* Evenly illuminate the target color card with how you want to illuminate the painting and get a good exposure. Avoid excessive under exposure. Optimally, keep the image plane/sensor parallel to the card, even if the desired targeted image - say the painting - are at a different angle (on an easel).
* For the "Base Characteristics" ICC profile, I use Effects->No color correction.
* For the curve, use whatever curve you tend to apply. Capture One's default curve is "Film Standard". If you see "Auto" for curve, that's what it's using. If you use Linear Scientific or Linear Response, choose those for the curve.
* Set the white balance - I use F5 on the ColorChecker SG.
* Set the white point luminance. 95-ish for A1, E5, etc on the ColorChecker SG. Use levels or curves - not exposure.

The white point luminance is important because you want a profile that doesn't brighten or darken your image. So getting the patches A1, E5, 10N mostly evenly illuminated and close to their L*value (95-ish for the ColoChecker SG for example) is awesome. Again, use levels or linear curve vs C1 exposure changes. Make sure your monitor is calibrated if it isn't. Angle of illumination is important, too. When you're measuring delta Es, you're basing that on an assumed angle of illumination and things like D50/D65 color temps.

As far as basiccolor training - basic BasicColor isn't too difficult when you're using known targets and following the above ideas. The Cultural Heritage folks at Digital Transitions have a number of YouTube videos that can get you going. https://www.youtube.com/@DigitalTransitions/search?query=basiccolor shows their current list.
 

KEVINS

Member
I completely understand your perspective so rest assure that I will take any nugget I can acquire.
I have no plan to use these files for reproductions. The primary goal is to create a digital image that comes "close" to what the painting looks like for publishing in a magazine and for contest entries. This painting made the March cover of The Guide Artists out of Spain:

I learned the hard way AFTER I submitted images to the ARC Salon contest that the images were not correct. Here’s a pic I submitted not knowing it was too dark b/c it looked perfect on a monitor I calibrated using the Xrite system:

bad.JPG

It's supposed to look like this:


As you can tell its WAY dark but it looked perfect on my “calibrated” monitor. I don’t want to do that again nor do I want to submit images to a magazine that aren’t close to what the painting looks like. They don’t have to be perfect files for making giclee’s.

Please don't feel that I have any sort of expectation for any help. I do have other people helping me and so far everyone has little things that are different that may help in the long run.

One gent (Eddie W.) that I chat with all the time photographs the paintings for the IBEX Collection and he's offered to take a look at my files when he has some time..

What it really comes down to for me is: Is there something wrong with this IQ180 and it won't capture the color? Or is it the profiling software I'm using? Or is it my setup, etc?

If You and Eddie and ?? cannot seem to get any of my images to produce accurate colors then it must be the camera. This is why I am happy to take different pics based on any suggestions that people have.

My 5DSr produced a damn good image the first time with no correction but doesn't have the resolution. This IQ180 has great resolution but the color is not correct..

I gotta go to a evening show so I will look over your post later..

Thanks Ray, and no worries!

Kevin
 

Ray Harrison

Well-known member
I completely understand your perspective so rest assure that I will take any nugget I can acquire.
I have no plan to use these files for reproductions. The primary goal is to create a digital image that comes "close" to what the painting looks like for publishing in a magazine and for contest entries. This painting made the March cover of The Guide Artists out of Spain:

I learned the hard way AFTER I submitted images to the ARC Salon contest that the images were not correct. Here’s a pic I submitted not knowing it was too dark b/c it looked perfect on a monitor I calibrated using the Xrite system:

View attachment 201220

It's supposed to look like this:


As you can tell its WAY dark but it looked perfect on my “calibrated” monitor. I don’t want to do that again nor do I want to submit images to a magazine that aren’t close to what the painting looks like. They don’t have to be perfect files for making giclee’s.

Please don't feel that I have any sort of expectation for any help. I do have other people helping me and so far everyone has little things that are different that may help in the long run.

One gent (Eddie W.) that I chat with all the time photographs the paintings for the IBEX Collection and he's offered to take a look at my files when he has some time..

What it really comes down to for me is: Is there something wrong with this IQ180 and it won't capture the color? Or is it the profiling software I'm using? Or is it my setup, etc?

If You and Eddie and ?? cannot seem to get any of my images to produce accurate colors then it must be the camera. This is why I am happy to take different pics based on any suggestions that people have.

My 5DSr produced a damn good image the first time with no correction but doesn't have the resolution. This IQ180 has great resolution but the color is not correct..

I gotta go to a evening show so I will look over your post later..

Thanks Ray, and no worries!

Kevin
What camera body are you using the iq180 with, I forget? I’d suggest taking a well lit shot - any shot - such that when you look at image’s histogram in C1 it looks like a hump, or at the very least, it’s not bunched on the left, in particular. “Expose to the right”. That might be a good start at looking at how it performs. Those CCD backs vs the CMOS sensor in your Canon camera really truly do need a little extra attention. Maybe try and avoid AWB in camera as well, if you are using that. Use “daylight” or anything consistent.
 

KEVINS

Member
Body is a 645DF+.
I have a handful of images where the histogram was more centered and a few that were a tad on the right but the highlights REALLY blew it all up pre-ICC and after a new ICC it was real interesting. The below exposure was centered (ISO100/F8/10 seconds) and is typical after a new ICC made. In order to prevent the below from happening I have to expose to the left, under exposed, similar to the first file you played with.
boom.JPG
After seeing all this i thought maybe the lights were too close so I tried moving lights farther back another 5ft then the camera another 5ft then trying different exposures and different lighting power but nothing made any difference.

Someone else mentioned to use daylight WB but it got a little worse than AWB.

Another thing that I still don't understand is why does the image turn orange when simply doing a WB correction? This is one reason why I think the data the IQ180 creates may not be correct thus messing up all the color correction. The Before WB image appears "close", just needs tweaked with a more accurate ICC but after WB it gets destroyed:

Before WB:
b4wb.JPG
After WB:
wb.JPG

ks
 

dchew

Well-known member
A question to chew on - I'm curious what the proper workflow "should" be:
I've been quiet on all this because I don't have much help to offer. It has been several years since I did any ICC profiling. I haven't had the need since P1 came out whith their new-ish "Pro-Standard" profiles for the new backs. I do remember good exposure being important to a good profile. I don't understand this statement of yours:
"If the card is placed on the painting and the exposure is adjusted to get a good exposure of the painting then it's possible the card will be underexposed or overexposed depending on how bright the overall color of the painting is.. Yellow painting vs Brown painting. Correct? If this is true then I can see where the ICC won't work."

Assuming the card is illuminated the same as the painting, then a correct exposure on the card is the same as the correct exposure on the painting. If the painting is dark, you want it dark; if it is light, you want it light. That doesn't change the exposure settings. I'm assuming you are setting exposure manually, not using the camera's meter targeted on something with a luminance that is too light or dark.

I would take the image for the ICC with the card in front of the painting, setting exposure to get the card luminance close to what the squares should be. Bracket a bunch and choose the one that corresponds best to the white, gray and black luminance values. They all won't be right since the contrast curves are whatever they are. I used to target the patch that had a luminance around 200, and let the rest fall where they may.

I keep going back to your white balance issue, and why when the card is balanced the painting is too orange. It seems more off than what a profile would do, but maybe not?

It makes me think maybe it is the lighting. Well, I mean the way the light source interacts with the paint. Is that even a thing? I don't know. I'm off in the weeds even more so than Ray. I am way out of my league taking photos of rocks and trees. I'll retreat back into my hole.

Dave
 

Ray Harrison

Well-known member
Kevin, one thing to do to eliminate the camera itself (and I'll go out on a limb and say it's not likely the camera itself):

Take the chart outside or in some even clean light, let the camera calculate the initial exposure, metering off the card. Maybe do some brackets as Dave suggests. Feel free to shoot those over to me and I can take a look and do some test profiles. I did a similar experiment and created a profile for a linear response curve and the L*a*b* values are in the ball park, which is about all I hope.

One other observation, and this is why I think Dave may be onto something as regards the lighting and also plays into the profile itself too (Flash): if I take one of the first images you sent over as is and select a patch like F5, which should be a gray value, it's actually fairly blue and not gray. If you take a blue patch in clean light on the color checker, the white balance shifts a very interesting direction indeed -if shifts the red side of things....second image. I used a light blue patch K5.



P0001662.jpg
P0001662 1.jpg
 
Top