Profile diff kind of machines to same colors

[raychoi]

Hi all,

I am going to profile a few diff model machine which all have diff print patterns and etc, the purpose here is to have these machines profile to the same color, which I think it may involve similar procedure as profiling a proof machine against the final output machine. could anyone advise how or where I could find out more information about this?

Printers we are using in the workshops are Mutoh R-900(dye), Valuejet(dye), Hp5500(dye), & 9000s (eco-sol).
Rip: Onyx
Profile equipment: DTP70, eye1.

Many many thanks for any info or leads.

Cheers.

[eye4clr]

The short list is...

- Profile everything as individual printers
- compare gamuts of each to find the weakest gamut
- setup all other systems to "proof" to that weakest gamut via Relative Colorimetric rendering intent

What happens is all your systems get pulled down to the lowest common denominator. No more big color from the 5500/dye.

in all my time consulting and training others to setup system, this is a very common request. It is dropped 99% after you see the results and how painful it is to maintain.

Ask yourself, does everything really need to match? Can you change your goal from tight match across all systems to sell-able color on each? If so, just profile each and they'll be close enough for the huge majority of your needs, assuming the profiling process is done well.

[raychoi]

Hi eye4clr,

thanks for the reply, could you explain more detail about this procedure? Just not quite sure what you mean. many thanks

"- setup all other systems to "proof" to that weakest gamut via Relative Colorimetric rendering intent"

You reckon the 5500 produce the best largest color gamut as well, do you? =) yes, pretty much we will lose some good color by doing so, this is just for some client's request and for project base only.

Cheers

[eye4clr]

Well, it's a bit complicated if you're not already familiar with the fundamentals of a color managed system and how to do the profiling. This is definitely not a 101 level goal.

First, learn to make profiles for each machine and each material that needs to be involved.
Second, learn how to setup, manage, and tweak a color managed workflow for every day jobs.
Third, create an additional workflow that takes advantage of the parts and pieces you've created to get each to look alike. This is the 'proof' part and is where things get more complicated and challenging.

Again, i'll wager lunch that you'll completely abandon this as a goal once you start to really understand what's involved to make it work. It has been my experience that simply making custom profiles for everything brings each printer close enough for most situations. Think of it this way, you can't get a print system that has the weakest color to look like the one that has the best color. You can only bring the good one down to the weak one.

To fully explain this process in detail would take an hour or more to write and would likely not be understood through an email. To be trained to set this up would take 2-3 days and to have it done as a service would take 1-2 days. It is not a simple process, nor will it stay in place once it is done. It would have to maintenanced.

Sometimes reality is harsh.

[raychoi]

Hi eye4clr,

thanks for the response, yes I fully understand what you mean, in fact I have very proper color profiles being done already for all the machines, and yes they do all look close enough for sale ability and no one ever complaint about them.

What however I am looking is really more advanced experience from others (such as you) on this forum about the idea and concept (not necessary the detail procedures) to profile one machine against another machine's output. For example a small proof machine against huge scitex machines, in that case definitely will bring the good one down to the weak one.

I understand it is difficult, that is why I try to find out more from here =) Many thanks for you and/ whoever able to provide me some sort of directions.

Cheers.

[M.A.C.]

Why dont you begin by describing your needs and your current workflow?

[eye4clr]

Ah, I wrongly assumed you were not currently making your own profiles.

Ok, here you go.

First I'll make some more assumptions...
- that you have the ability to setup input profiles and simulation (sometimes called proofing) profiles within your RIP and are not having to do this in your desktop applications.
- you understand the pros and cons of selecting rendering intents and how icc conversions work in a superficial sense.

The diagram of the workflow on the proofer goes ( "->" means a conversion is happening)...

File's working space -> profile for what you want to simulation -> proofer's profile

Most folks would setup the rendering intents as
File -> (perceptual for rasters and relative colorimetric for vectors) -> what you want to simulate

then

what you want to simulate -> (absolute colorimetric) -> proofer's profile

Sometimes you're better off to set the proofing step to relative colorimetric to avoid the color of the simulation profile's white being printed. It's a matter of what you're trying to achieve and how fussy you are about the overall match.

I hope I didn't assume too much now. Really you're just adding an additional conversion on to the end of an established workflow that converts from what you're trying to simulate to the icc for the output your trying to imitate it on.