Pages: 1 [2]   Go Down

Author Topic: ColorTRUE and CamRanger  (Read 6679 times)

keith_cooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 473
    • Northlight Images
Re: ColorTRUE and CamRanger
« Reply #20 on: June 16, 2014, 05:22:52 am »

So did you get a match between iPad/iPhone and a MBP @D65? I wasn't that lucky.

It depends on what you want a match for... (there is no yes/no pass/fail answer to this IMHO - others disagree I know, so YMMV ;-)

Given that the iPad is a glossy screen and my MBP has a matte one (I didn't say it was new ;-) and they have different primaries, there are going to be differences in any direct side by side comparison.

Looking at some of the old Kodak sRGB test images (lady, tailor, yarn etc) they are a reasonable match...
...but, and it's a big but - this depends greatly on what is reasonable for the devices and what you use them for. I'd not consider doing image editing work on my MBP screen for example (image too bright? - move your head up/down). Nor do I have a serious use for my iPad, other than as an extension of my camera interface (let's not even go close to the colour accuracy of screens on the back of cameras).

I can tune my  car engine as much as I like, but it will still be a 2003 Ford Focus afterwards.

ColorTRUE makes a positive difference to some aspects of my use of the iPad. If you were expecting it to turn the iPad into some mini reference monitor, then you will indeed find it worth every penny you paid... ;-)


« Last Edit: June 16, 2014, 05:57:22 am by keith_cooper »
Logged
bye for now -- Keith
[url=http://www.nor

Czornyj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1948
    • zarzadzaniebarwa.pl
Re: ColorTRUE and CamRanger
« Reply #21 on: June 16, 2014, 06:31:06 am »

I'm only talking about the colour of white, it doesn't match the wtpt of my glossy rMBP, not to mention my WG CCFL and GBr LED NEC PA-series displays. It's not a problem to add custom x,y target, or a slider for manual fine tuning. It's also not that difficult to make ambient light compensation that actually works. Why there's no option to use my own ICC profile to soft proof rather than the selection of a few obsolete CMYK profiles? Why the profile of the mobile device is NOT an ICC profile, and why it's hidden in a goddamned "X-Rite Cloud", so I can't make any use of it?

I'm only expecting it to work as advertised - is it too much?
« Last Edit: June 16, 2014, 06:37:59 am by Czornyj »
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

keith_cooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 473
    • Northlight Images
Re: ColorTRUE and CamRanger
« Reply #22 on: June 16, 2014, 07:31:35 am »

I'm only talking about the colour of white, it doesn't match the wtpt of my glossy rMBP, not to mention my WG CCFL and GBr LED NEC PA-series displays. It's not a problem to add custom x,y target, or a slider for manual fine tuning. It's also not that difficult to make ambient light compensation that actually works. Why there's no option to use my own ICC profile to soft proof rather than the selection of a few obsolete CMYK profiles? Why the profile of the mobile device is NOT an ICC profile, and why it's hidden in a goddamned "X-Rite Cloud", so I can't make any use of it?

I'm only expecting it to work as advertised - is it too much?

No  - I'm sure that many people must have drawn broadly similar conclusions (and expressed them) when first shown ColorTRUE, I know I did.
That's partly why it fell into the 'nifty idea' [remotely accessing my i1 Pro 2] but 'so what' category.

The reason that I delayed writing up any info was that I just couldn't see a real use for it on my iPad. It was actually the combination with the CamRanger that made a difference to me using it (and to be honest, the iPad too, since phones/tablets just don't feature in my day to day activity very much at all).

The criticisms are warranted, but I still see it as a worthwhile first step.



Logged
bye for now -- Keith
[url=http://www.nor

Czornyj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1948
    • zarzadzaniebarwa.pl
Re: ColorTRUE and CamRanger
« Reply #23 on: June 16, 2014, 08:18:11 am »

The criticisms are warranted, but I still see it as a worthwhile first step.

If further steps will be made in the same pace as in case of i1Profiler, I won't be holding my breath.

From my perspective mobile devices (tablets, smartphones) had always been a pure waste of quite decent displays, that could be useful in photo editing and exhibiting. ColorTRUE is another missed opportunity to make them more usable in this area, which is disappointing. I would't whine if it was "ColorMunkiTRUE, just another feel-good gadget for amateur photographers... But this product seems to be intended for i1Display Pro, and i1Pro 2 users, and there's even such a welcome message:
"So, you're a color perfectionist. We get it. You want your images to be as colour accurate on your mobile devices as they are on your desktop or laptop screens. ColorTRUE is the perfect solution for you".

So, as a colour perfectionist I dare to say they apparently don't get it. The images on my mobile devices were not even close to what I see on my desktop or laptop screens. ColorTRUE was a perfect waste of time for me.
« Last Edit: June 16, 2014, 08:20:07 am by Czornyj »
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

keith_cooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 473
    • Northlight Images
Re: ColorTRUE and CamRanger
« Reply #24 on: June 16, 2014, 10:26:19 am »

Does anyone actually know to what extent Android and iOS actually support meaningful colour management, in a way that you'd expect as a modern win/mac developer?

What could be done?

...and from my own POV, what else might I find a use for a tablet for, that really needs good colour management? (genuine question - I just don't find them any significant use)
Logged
bye for now -- Keith
[url=http://www.nor

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: ColorTRUE and CamRanger
« Reply #25 on: June 16, 2014, 01:10:06 pm »

Quote
Quote from: Tim Lookingbill on June 15, 2014, 04:09:44 PM
Can someone here show us what "a huge mismatch" looks like by posting a photo of this mismatch so we know just how useless this product is?

Quote from: Czornyj
Quote
You'd need sort of multispectral camera to make such a photo.

Yeah, I have one those cameras. It's called eyeballs which are inextricably connected to the whole process of color matching that allow the user possessing such powerful multispectral perception to edit an image to get a match when the subject being matched to is right in front of their eyes next to there calibrated reference display.

But I guess no one here wants to go the extra mile to do some selective editing to demonstrate the white balance differences and possible matrice driven color errors between a calibrated iPad and their primary editing workstation.

How big are the differences/mismatches? Does cadmium yellow turn lemon yellow? Oranges/Reds look a bit magenta? Skin tones look reddish/greenish?

Could this be used as Keith indicated for use just as a reference display like say for online retail businesses whose sales are primarily driven by customers using mobile devices and maybe would like a reasonable match to the actual product viewed on such devices? Who cares about color perfection by those attempting to sell their fine art prints hanging on walls that barely garner a 3 second glance by patrons who most likely don't even devote that amount of time to even consider the creative efforts of the artist's intended color match.

It appears the lengths gone to splitting hairs on this topic always stop at providing any evidence. At least Keith provides photos to give some idea.

Keith, were you able to analyze the profile created for the iPad to get an idea of its color gamut? Are you able to generate a 3D color map in any of the provided software?
Logged

Czornyj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1948
    • zarzadzaniebarwa.pl
Re: ColorTRUE and CamRanger
« Reply #26 on: June 16, 2014, 02:36:19 pm »

Differences of white balance are quite small for standard D65/D50 targets, but noticeable. It's getting much worse when we calibrate the desktop display to a custom wtpt to get a match with specific paper in specific condition - why not make custom white point choices with fine tuning tint sliders?

It's even worse when we want to use Print Emulation feature - for example I never use Fogra39 profile created with ProfileMaker, I prefer ECI ISO Coated v2 which is most popular in EU, and of course my own profiles created with i1Profiler, which have slightly different perceptual rendering. And if I would like to simulate the print from my Canon iPF8300, I'd be completely in a forrest!

If you're calibrating a desktop display, you can arrange the surround to reduce the influence of viewing conditions, in case of mobile  devices it may be impossible, so the Ambient Compensation feature could be very useful in such cases. ColorTRUE is calibrating the mobile devices to a TRC gamma 2.2 - at least I suppose so, as the created profiles are hidden in a cloud and not ICC compliant, so you can't really be sure, nor make 3D gamut charts or use mobile device characterisation for soft proofing on desktop displays (which could be useful for mobile device application developers/designers). What you actually perceive on a mobile device in a darkness is visually more like - let's say - gamma 1.6-2.0, while in bright, outdoor condition it's more like gamma 2.6-3.0 calibrated display... But the Ambient Compensation function doesn't really work, it only changes the chrominance slightly and doesn't change the TRC at all.

I can imagine it might be useful to edit image colours while actually watching the scene, to get the exact impression, or make colour notes. It could be even more useful in case of architecture/product/fashion outdoor photo session. It could also be used to simulate the appearance of a print. Problem is that due to the above mentioned limitations it doesn't work as it could - at least from a perspective of picky "colour perfectionist" it's advertised for.
« Last Edit: June 16, 2014, 02:39:40 pm by Czornyj »
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

keith_cooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 473
    • Northlight Images
Re: ColorTRUE and CamRanger
« Reply #27 on: June 16, 2014, 02:53:26 pm »

I'm minded to suggest that the "Color Perfectionist" referred to in the marketing, is utter tosh...  I for one have no desire to receive such an appellation - I've always gone along with the engineering adage "Perfection is the enemy of excellence" ;-) [OK wrong place to admit such heresy - I'll get my coat...]

Tim - the profiles are not accessible at all - one of my own personal gripes about the setup, not because I particularly want to make a lot of use of the info, just that I'm curious.

Logged
bye for now -- Keith
[url=http://www.nor

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: ColorTRUE and CamRanger
« Reply #28 on: June 16, 2014, 03:04:25 pm »

Tim - the profiles are not accessible at all - one of my own personal gripes about the setup, not because I particularly want to make a lot of use of the info, just that I'm curious.
I don't believe any such profile exists. Not on iOS. Doesn't matter anyway, the product doesn't produce a visual match, and that's what it's supposed to do. How far off it is, is pointless, it's obviously and visually not correct (so let's say well over dE7). Better maybe. But correct (accurate as they promise), not at all.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Some Guy

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 729
Re: ColorTRUE and CamRanger
« Reply #29 on: June 16, 2014, 10:14:43 pm »

Can you email me directly (email addr. on the Northlight site) with more precise details about what you are actually doing?

A 21 or 51 step wedge is incorporated into the new B&W test image too, and I used it during testing with ColorPort (after importing the targets). It all works fine here, so I'm keen to see what's going wrong (tested using ColorPort 2.0.5 under OSX 10.6 and 10.9  - I've never tested anything under Windows.)

The 4 xml target files for ColorPort all seem OK, but they were updated not long after the article was originally published (2012)
http://www.northlight-images.co.uk/article_pages/bw_printing/bw-print-correction.html

The profiler info is interesting too - can you also let me know what it is you are doing? I normally use just Colorport for the ramps, and have not tried with i1P for quite a while.


Keith, the i1Profiler (1.5.4) has a "Measurement Chart" reading within it under the printer.  One can set number of rows like one single row of 21 patches and do a scan there.  There is the ability to set both rows and patches (or columns) both within their range in that window.  Colorport, newer one in Windows 8, just doesn't feed into it right somehow.  That or your instructions are not clear how to get it in there to get a reading off it with Windows.  Scan of the 21 steps seems to stop at 18 for some reason.

However, my scan in newest i1Profiler differ quite a bit in the whites verses a spot scan of the whites, and by a lot too: L=8 points difference!  Their suggestion was to try a powered USB hub as the i1 PhotoPro 2 head needs more current (And why it won't work with ColorTRUE either with a tablet as will their i1 Display Pro which runs on less current off the tablet or cellphone.), but results were the same.

Also, the PSD step-wedge files in your download do not feed into QTR correctly either, or at least for the crippled (old?) Windows QTR version that seems written for Windows 98.  It only takes TIFF or JPG and not PSD which has to be converted prior to getting into QTR.  Worse is for Cone's K7 inks which likes a Gray Gamma 2.20 too and might not be caught in Photoshop on the conversion.

I've been playing around putting break-bars (white and/or black) between the 21 gray step patches and the i1Profiler got up to L=90, but still short of L=94 in spot mode.  Better than L=86 but something is really messed up in the targets somewhere - or their software isn't scanning correctly which could likely cause some color patch reading errors too in ICC profile making.

For fun, try a spot reading of a white, and then compare that to a scanned reading.

SG
Logged

keith_cooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 473
    • Northlight Images
Re: ColorTRUE and CamRanger
« Reply #30 on: June 17, 2014, 05:02:54 am »


> Keith, the i1Profiler (1.5.4) has a "Measurement Chart" reading within it under the printer.  One can set number of rows like one single row of 21 patches and do a scan there.  There is the ability to set both rows and patches (or columns) both within their range in that window.  Colorport, newer one in Windows 8, just doesn't feed into it right somehow.  That or your instructions are not clear how to get it in there to get a reading off it with Windows.  Scan of the 21 steps seems to stop at 18 for some reason.

Can you mail me some details - a screen shot or two would help.
I've not used a Win PC this century, so just don't know what you're seeing - there are two articles, one using colorport and the other i1P - both have appropriate data files. I prefer the colorport one (it has spacer bars)


>However, my scan in newest i1Profiler differ quite a bit in the whites verses a spot scan of the whites, and by a lot too: L=8 points difference!  Their suggestion was to try a powered USB hub as the i1 PhotoPro 2 head needs more current (And why it won't work with ColorTRUE either with a tablet as will their i1 Display Pro which runs on less current off the tablet or cellphone.), but results were the same.

>Also, the PSD step-wedge files in your download do not feed into QTR correctly either, or at least for the crippled (old?) Windows QTR version that seems written for Windows 98.

No idea I'm afraid - I just make the measurements in colorport and drop the text file of measurements onto the QTR Profile creation script.


>  It only takes TIFF or JPG and not PSD which has to be converted prior to getting into QTR.  Worse is for Cone's K7 inks which likes a Gray Gamma 2.20 too and might not be caught in Photoshop on the conversion.

The test files are GG2.2, so should easily convert.  The files BTW were never created for use with QTR, other than the script  to make the correction profile. They were not developed at all for linearising QTR or setting ink limits.

>I've been playing around putting break-bars (white and/or black) between the 21 gray step patches and the i1Profiler got up to L=90, but still short of L=94 in spot mode.  Better than L=86 but something is really messed up in the targets somewhere - or their software isn't scanning correctly which could likely cause some color patch reading errors too in ICC profile making.

I'd suggest using Colorport and the data files in that article. The i1P article was much more a 'lets see if this works' one, although it contains targets (and data files) for both the i1Pro and i1Pro 2
Logged
bye for now -- Keith
[url=http://www.nor

Some Guy

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 729
Re: ColorTRUE and CamRanger
« Reply #31 on: June 17, 2014, 10:47:56 am »

Keith, could be you use OS X and not Windows. I see in some of the 21 and 51 step folders there are resdiual .DS_Store propraitary files used by OS and not Windows so who knows?

I'll attach a couple of screenshots off Windows 8.1.  One shows the rows and columns (patches) in the i1Profilier (newest version). It's located at i1Profiler > Printer > Measure Chart.  That is user customizable for whatever chart you feed it, although the "Scan" of whites will differ from "Spot" readings of whites for some yet undiscovered reason by me.  Been going around with x-rite in it too who thought the USB port had too little current for the i1 PhotoPro 2 head (Turned out false and not a problem, yet patches reamain different depending on scan or spot mode.).

As to Colorport 2.0.5, the screenshot shows typical results of your 21 linear step image.  If I do get it to read, it will stop at the top of the second column with a red square and only column one is read - or 18 of the 21 steps.  Don't know why Colorport will not show all 21 steps in a single row, but it show 18 and then second column of 3 in Windows.  That window is not resizable so I cannot get beyond the 18 steps.

As to the PSD files, for use with Jon Cone's K7 piezogrpahy you do not use the QTR linearization (drag and drop) or Print Tool at all.  You need just the data of L value off each of the 21 steps.  You put those values into "his linearization spreadsheet" that shows the ink placement on the straight line within Excel or Open Office spreadsheet.  QTR is merely the print driver and nothing else for piezography.  If the linearization is off, then you need for him to alter the curves of each of the 7 ink carts in varying degrees of black to fix it.  Changes with each paper too as well as the selction of B&W ink tone (Sepia, Carbon, Selenium, etc.).

One could, if your scales were TIFF and not PSD in Windows, drag them into QTR and print the step wedges direct, but they need to be converted in Photoshop to TIFF to see and work in the Windows QTR version as is.  IF you toss your PSD file into QTR in Windows, it crashes out with an Error message per the screenshot.

SG
Logged

keith_cooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 473
    • Northlight Images
Re: ColorTRUE and CamRanger
« Reply #32 on: June 17, 2014, 10:56:01 am »

Keith, could be you use OS X and not Windows. I see in some of the 21 and 51 step folders there are resdiual .DS_Store propraitary files used by OS and not Windows so who knows?

Yes, can you mail me the details rather than divert this thread?
Logged
bye for now -- Keith
[url=http://www.nor
Pages: 1 [2]   Go Up