Color profiling; Things I ponder.

[Andy D]

These might be a bit out there, but I have a couple thoughts/questions regarding color profiling;

I just did a color adjustment for our high-end, professional, color paper printer (not a desktop, made for print shops) and it was so easy and fast... all you had to do was; print out a internal color chart, scan it in (using same machine) and repeat four times. It took maybe 2 minutes total, and worked perfectly.
My first question is this: Why couldn't a color profiling company create software, that ships with
hard copies that are color correct? You could adjust/dial-in whatever desktop scanner you use, using the color-correct profiling charts hard copies via their software. Next step, print out the same color profiling charts to your wide format printer and scan through desktop scanner... instead of the mind numbing and expensive way we do it now, using an eye-one.Does that make any sense? Even if it wasn't as accurate, it would be affordable and easy enough for everyone to create their own profiles.

The second thing, I have asked a couple years ago on here, but I'm reposting anyways:
Wouldn't it make more sense to import the rip software's color profiling charts into whatever design software you use, and then export them as you would do normally for a printfile? I have seen some major color shifts due to small color management changes made in the design software... by importing and then export from
the design software, it allows the rip program the opportunity to color correct them.. again, does that make sense?

 

[ColorCrest]

all you had to do was; print out a internal color chart, scan it in (using same machine) and repeat four times. It took maybe 2 minutes total, and worked perfectly

If your machine is of a type I'm familiar with, I suspect the process was that of calibrating the current state of the machine back to the baseline of when its ICC output profile was created. So long as the machine stays true to calibration, the original ICC output profile is accurate. They same goes for wide format printers.

 

[iPrintStuff]

The main issue would probably stem from any accuracy limitations. I see a lot of HP users with the built in spectro still complain about hitting neutral greys for example. Or just admit that they use the canned profiles.

(Never used a HP so could be down to poor hardware etc, just used that as an example) but even when the tools are available people either seem to not use them or still don’t seem to manage the colours they want; granted this is likely user error. So my only worry would be that if this paved the way for less accurate profiles, the colour management section on the forum would get a whole crop of newbies!

I think more people just need to be aware of colour management as a whole. We visited a sign shop close to us recently that had just started out and they ran every single media on the same colour profile. They’d bought over 150k worth of equipment and weren’t even aware that media profiling was a thing. Worrying stuff..

 

[ColorCrest]

My first question is this: Why couldn't a color profiling company create software, that ships with
hard copies that are color correct? You could adjust/dial-in whatever desktop scanner you use, using the color-correct profiling charts hard copies via their software. Next step, print out the same color profiling charts to your wide format printer and scan through desktop scanner... instead of the mind numbing and expensive way we do it now, using an eye-one.Does that make any sense? Even if it wasn't as accurate, it would be affordable and easy enough for everyone to create their own profiles

You're not far from making sense. A similar method is still in use today but it is mainly for legacy photo printers, as an example. The machine prints a certain calibration file and the operator feeds the print though an automated reader or (depending) an ordinary inexpensive desktop scanner. Done.

The "ships with hardcopy" is referred to as "Shirley." (Dying to meet her?) An actual color film negative and / or a digital file of a studio portrait is supplied for the shop to image and compare with the factory-supplied print. It's usually a slam dunk mainly because the system is using a closed-loop of proprietary color lookup tables instead of the modern method of ICC profiles. This process is performed for each "channel" of the flow combination of input-type-to-output-type of which there are very few. Think of the few color negative film types over the decades and just sRGB as the only digital color space going to only a handful of photo paper types.

20+ years ago when the method of ICC profiling came along, one would print ~768 swatches and could use a desktop scanner to produce profiles with entirely accurate and valid results. However, the norm back then was smooth, 8x10 size material. Very different media today.

Two thoughts for you...
1) If one is really making a lot of profiles, an ideal spectrophotometer is the X-Rite i1 iSis. Expensive, yes, but then the chore becomes easy.

2) If one keeps their machine in the same state of calibration and mechanics as to when their profiles were created, there is need to re-profile. The more common i1 is a measurement device to read routine calibration strips but those strips have require few readings.

 

[Andy D]

the colour management section on the forum would get a whole crop of newbies!

Good Grief! Did I really post this in screen printing? I'm an idiot sometimes...

 

[iPrintStuff]

Good Grief! Did I really post this in screen printing? I'm an idiot sometimes...

The screen printing section would get a whole crop of newbies* lol

 

[Andy D]

The "ships with hardcopy" is referred to as "Shirley." (Dying to meet her?)

I started off color correcting 30 plus years ago in the custom photo lab gig.
I'm very familiar with Shirley, she really got around!

 

[Andy D]

All great info, what about the second question?
Why wouldn't it make more sense to run the rip software's color profiling charts
through your design software and then through your rip software.
Your rip software creates profiles on the premise that all print files are created they same,
but simple tweeks in the color management can make big changes.

 

[ColorCrest]

I'm very familiar with Shirley, she really got around!

Then realize you can create your own "ships with hardcopy / Shirley" from printing a photo file from your phone or camera to compare input vs output. Be sure to include a genuine Kodak gray scale, or the like, and the larger gray card.

 

[ColorCrest]

but simple tweeks in the color management can make big changes.

I think what you've discovered is; What you believe to be "simple" tweaks in color management, on the contrary, really can have a large impact. Know that the red color 255, 0, 0 in sRGB color space is a very different intensity compared to the same red number in other color spaces, RGB or CMYK.

Forums are not the best for discussions, however...

Your rip software creates profiles on the premise that all print files are created they same

What you're describing is a closed-loop system. The modern system accounts for print files that are not created the same; different color spaces destined to any given output space(s).

 

[Andy D]

I think what you've discovered is; What you believe to be "simple" tweaks in color management, on the contrary, really can have a large impact. Know that the red color 255, 0, 0 in sRGB color space is a very different intensity compared to the same red number in other color spaces, RGB or CMYK.

I'm not sure what your answer is... I have worked at place where the designers used different designing software and used different methods of setting up print files and it created some havoc...
By running the rip software's color profiling charts through each of the designers design software before ripping, and create an ICC for each of them, wouldn't that sync all of them up?

 

[ColorCrest]

wouldn't that sync all of them up?

The short answer for "sync" is to be sure the designers all use the same color space for their documents. Then, be sure the RIP is properly honoring said input profiles in a consistent manner and using the proper output profile. A controlled group is easy enough but PFP shops accept work from outside where the working spaces could be anything. Some of those spaces could be out of gamut for a certain printer, many are smaller gamut spaces and not taking advantage of available color, etc.

What, exactly, was an instance of "havoc?"

 

[Pauly]

All great info, what about the second question?
Why wouldn't it make more sense to run the rip software's color profiling charts
through your design software and then through your rip software.
Your rip software creates profiles on the premise that all print files are created they same,
but simple tweeks in the color management can make big changes.

It's also far more complicated than that. the "chart" is only 1 thing, but the rip also controlls ink drop sizes, ink limits and curves and more depending on your printer,

So for example, your printer is shifting colours after a month, you can re linearise and changes the ink curves back to how the output originally was and off you go again.

 

[C5 Service&Repair]

I honestly didn't read all the responses lol. But I think the answer to your question is because of embedded profiles. You really need the printer to be calibrated to what the RIP is intending to send.
Therefore the color swatches that are scanned with the i1 or whatever device you have are compared by the RIP to what the colors are intended to be. Your design software can't be involved because there really is no way to get a file to the printer without an embedded color space. The color space (SWOP or sRGB or whatever) is for the information between the design software and the RIP. Then the RIP uses the scanned swatches to convert that into what should look correct on the printer. If the printer is only printing its own swatches and rescanning them, then all it is doing is a density and linearization check on itself.