I tried a Custom Profile

[bwana]

I have produced my own custom profiles for close to 15-years -- long before paper manufacturers started supplying them. (I have regularly been a beta paper tester and have to produce my own profiles since none usually exist.) Even though the canned profiles are sometimes quite good, almost without exception I find my own are better simply because it takes into account my printer, my inks, my papers.

You may find your next step is not to pay someone else, but to produce your own so you don't have to wait for someone else to produce it. The freedom it gives you will be equal to what you experienced going from canned to custom.

How can a custom profile made by one person be any 'better' than a custom profile made by anyone else? After all, isnt the process of making a profile supposed to be objective and designed for accurate reproduction of colors? You scan a bunch of color patches and the software makes a table of numbers. Why does the ability of making your own profiles yourself give you more 'freedom'?

Could you share your process about making custom profiles? What tech do you use? How can you 'quantify' how much 'better' the new profile is? After a few hours of looking at a screen, my critical viewing  (which relies on my brain) is fatigued and almost useless.

But then I guess quantifying anything that is aimed at triggering our emotional state is near impossible. How do you quantify good taste? or a good recipe? or good clothing?In each of these cases the example is somewhat abstracted from the sensory input. Our tools of mass production are geared towards measurements of physical properties and optimizing these measurements for some property or other (linearity, tonal distribution,etc). Quantitative measure has been excluded from our subjective experience (thankfully) so we can still say that 'humans can still do somethings machines cannot'. Of course with the advent of AI (where we can train machines to generate 'good' output) as well as the pervasive nature of the net with its ability to create a unique information stream for each person ('individual filter bubble'), we may soon lose our souls to the machines.

The corollary to the above digression is - will an AI generated ICC profile (optimized for the most 'attractive'  change of an image printed on a particular paper) be 'better' than an ICC profile designed to reproduce colors faithfully. Is that what you mean when you say that making your own profile gives you more 'freedom'?

[Doug Gray]

Advantages of DIY with I1Profiler.

1. You can create a myriad of different, perceptual intent, profiles from the same target measurements. This lets you select profiles that print most to your liking with different tone cures and color mapping.

2. You can easily adjust to your printer/paper with fewer patches by running an optimizer pass. Because this is printer specific, you get better results than a fixed set of patches from a purchased profile.

3. You can use the same spectrophotometer to cross check your printer workflow. This simplifies things like watching for changes as ink and printer age.

[Rhossydd]

How can a custom profile made by one person be any 'better' than a custom profile made by anyone else?

Replace the word 'person' and replace it by 'system' as a starter.

Andrew earlier on outlined the variables here.

the color engines used by various products, let alone how they produce a perceptual mapping (no rules), number of patches (and where in color space), the Spectrophotometer used, the options set in the software for the creation of the profiles. post optimization IF possible, all play a role.

All correct.

Where it gets a little trickier is that many of these variables don't make dramatic differences once you're on the plateau of high quality kit.
Whilst Andrew will never commit himself to any sort of helpful generalisation, it actually takes some ham-fistedness to deliver significantly bad results once you're on that level. Hence why I think that you won't see a massive step above a remote service to a DIY profile.
The vast majority of remote services supplying our market (RGB inkjets) use similar kit, almost all from X-Rite so quality of results is pretty consistent from supplier to supplier.

[digitaldog]

Whilst Andrew will never commit himself to any sort of helpful generalisation, it actually takes some ham-fistedness to deliver significantly bad results once you're on that level. Hence why I think that you won't see a massive step

You don't read or comprehend text outside your bias too well:
All generalizations are false, including this one.-Mark Twain

The vast majority of remote services supplying our market (RGB inkjets) use similar kit, almost all from X-Rite so quality of results is pretty consistent from supplier to supplier.

As Doug now, and I have attempted to explain to most everyone here, that's rubbish and yet another silly generalization for others reading here to hopefully ignore. 

[rabanito]

Advantages of DIY with I1Profiler.

2. You can easily adjust to your printer/paper with fewer patches by running an optimizer pass. Because this is printer specific, you get better results than a fixed set of patches from a purchased profile.

Sorry I didn't understand that.
If I print the patches with MY paper and MY printer before sending them to the "professional profile maker" then I regard this as "printer specific".
Isn't it?
Thanks

[nemophoto]

Thank you Rodney. That is precisely the point I was making -- the difference between canned and custom. And custom can mean doing it yourself of sending out. My preference has always been to do things myself, otherwise why do I go to the trouble of even printing my photos myself. My desire to produce my own quality prints goes back into the 90's but really was fueled in 2000 by the desire to find a way to produce better B&W. Up to that point, any time you printed B&W you had a magenta cast. I ended up going with Cone's Piezography B&W system for many years.

So whether one has a service do the profiles or DIY, the results are usually better than canned.

[nemophoto]

Sorry I didn't understand that.
If I print the patches with MY paper and MY printer before sending them to the "professional profile maker" then I regard this as "printer specific".
Isn't it?
Thanks

Yes, your result is "printer specific".

[digitaldog]

Sorry I didn't understand that.
If I print the patches with MY paper and MY printer before sending them to the "professional profile maker" then I regard this as "printer specific".
Isn't it?
Thanks

Some, absolutely not all products (hence the reason generalizations are worthless) provide pre and post optimization.
In a prefect world, we'd actually produce nearly all visible colors as patches to measure. We can't produce, measure or build profiles with millions of patches. Even if we had super fast Spectrophotometers and people were willing to make that many prints, the resulting profile would have a document size that was likely larger than many images we wish to print. So all profile making software has to extrapolate based on the number of patches it is feed. More isn't necessarily better either! Those products that allow the creation of custom targets, pre and post optimization targets and who's users are good at producing such targets allow for (in theory and often in action), superior quality profiles. I routinely make optimization targets with a custom patch set for all the profiles I create. Sometimes the added work shows no difference, sometimes it does! And it's quite easy to compare the original profile and the optimized profile visually with good test images and colorimetrically by running tests with the correct software products. Because I never know if the optimization will show no or slight improvements, I just make the optimization steps as a basis for all profile creation. It only takes 3 8x11 sheets of additional paper and with an auto Spectrophotometer, about 10 minutes of work.
So here is yet ANOTHER area where DYI or custom profile services WILL differ. Do they provide the ability to optimize the original profile with additional measurements? Will they provide colorimetric reporting IF the customer desires to see additional non ambiguously data about the profiles? Will they build profiles with differing white point assumptions (instead of the usual default of D50)? Do they offer OBA compensation? The answer is yes and no. So much for silly generalizations about all DYI users and all outside profile making services. THEY DIFFER.
In addition to all the above, there are tools like ColorAnt that allow us to 'massage' the measured data. Again, in some cases, this can produce a superior profile.

[Slobodan Blagojevic]

... generalizations are worthless...

THAT one surely is.

[digitaldog]

THAT one surely is.

You're finally getting it bud. I'll state this for the third time. For someone who's posts here illustrate he isn't interested or too knowledgeable about the subject of creating profiles:
All generalizations are false, including this one.-Mark Twain
 
I think maybe you're understanding the point: ... generalizations are worthless...

[rabanito]

Some, absolutely not all products (hence the reason generalizations are worthless) provide pre and post optimization.
...
...

I think I get the idea . Thanks

[Slobodan Blagojevic]

... All generalizations are false, including this one.-Mark Twain
 
I think maybe you're understanding the point: ... generalizations are worthless...

It is, as usual, a matter of degree, bud. But subtlety has never been your strong suit. In spite of being an expert in millions of colors, your logic seems limited to just two: black and white*.

* Spare me the debate whether black and/or white are actually colors... it is irrelevant for the turn of phrase above

[digitaldog]

It is, as usual, a matter of degree, bud. But subtlety has never been your strong suit. In spite of being an expert in millions of colors, your logic seems limited to just two: black and white*.

* Spare me the debate whether black and/or white are actually colors... it is irrelevant for the turn of phrase above

No generalizations, a fact of your posting history below that you have no idea what this topic is about nor have experience or anything to add here other than a posting agenda to go off topic and be a cesspool of disingenuous nonsense. The paper trail below is clear to me and I'd suspect everyone reading these two pages.
Are you so lonely and devoid of forum attention you have to go to such lengths? Seems you do:

Well, share it then.

I was hoping to see before/after comparison, alas...

It is much simpler. Print before/after and take a snap with an iPhone. Easy.

The OP reported enormous difference. It would surely be visible in an iPhone snap.


In related story, I got today a new shipment of wine. I am simply reporting how happy I am with it. Maybe I should post it in a separate thread. Too bad we do not have here a Wine Corner, just The Coffee Corner.


I am also preparing chili for lunch. Again, simply reporting how happy I am with it.


I'll let you know about the dessert. In a separate thread.

As opposed to clogging the forum with posts related to photography, but actually meaningless? As in "I got a new lens/camera/printer and I am happy with it"?

It is meaningless without a visual comparison. And I am here reading it as I am interested in the visible difference. After all, this is a photography site, not a poetry site, where words would suffice.

But not before I post a photography-related dessert:

A combination I used to print hundreds of images. Never had a problem with print not matching the screen.


Of course, when one uses a different profile, there must be some change visible somewhere. The question is, does it matter? If "very slight," probably not, but if "enormous"... well, that one I would like to see.


It is equally possible that you and I were lucky because we used a manufacturer's paper with their own profile. The OP was using a mismatch, so a custom profile might have indeed resulted in a significant improvement.

THAT one surely is.

At least you now understand (?) the vast differences between desert and dessert.

[Rhossydd]

The answer is yes and no. So much for silly generalizations about all DYI users and all outside profile making services. THEY DIFFER.

I'm not suggesting there aren't differences. The issue is how significant or dramatic they are.

I can't be bothered to back reference the many times you've commented on the sliders that make hardly any difference in i1Profiler, or discussion back when i1Profiler first replaced PMP5 which was more evolution rather than revolution in the results it gave. An awful lot of these 'differences' are pretty subtle and even experienced photographers can be pressed to see them.

Sure the people that want ultimate control and are happy and well enough resourced to pay for it or whose interest extends to measuring the effect of drive wheels on printing paper will always buy the kit and DIY.
For photographers just wanting to make good prints who don't have unlimited funds, DIY profiling may not be a good use of their time and money.

[digitaldog]

* Spare me the debate whether black and/or white are actually colors... it is irrelevant for the turn of phrase above

There is NO debate. Since you appear to be utterly uninformed about how profiles are created, or what colors are, I'll simply provide you more facts to ignore. As you can see from an actual profile target, indeed black and white, as defined numerically below with the two sample points ARE used to build a profile.
These facts are your just desserts. If if you desire, your just deserts:

[digitaldog]

I'm not suggesting there aren't differences. The issue is how significant or dramatic they are.

Once you actually run colorimetric tests, instead of assuming, you'll know. By a metric known as deltaE.
The differences are visible JUST between making a profile with fixed attributes in i1P versus making one using a carefully constructed optimization target. So yeah, well over a dE of 1.
It's why some of us actually use the tools that were actually developed by color scientists and color engineers. They can work.

I can't be bothered to back reference the many times you've commented on the sliders that make hardly any difference in i1Profiler,

You can't because you misunderstood, again, the comments. They can make a difference; depends on which sliders and what kind of profile is being built. You of course misunderstood the specific comments you can't be bothered to find and assume that means ALL the sliders and controls. That's absolutely not true.

For photographers just wanting to make good prints who don't have unlimited funds, DIY profiling may not be a good use of their time and money.

Yet ANOTHER false generalization where you speak for all photographers. Again, I'm sorry the DIY possibilities for some people upset your business model. 
Are you even a photographer? Or ever paid to make a photograph?

[Rhossydd]

Once you actually run colorimetric tests, instead of assuming, you'll know. By a metric known as deltaE.
The differences are visible JUST between making a profile with fixed attributes in i1P versus making one using a carefully constructed optimization target. So yeah, well over a dE of 1.

When you start to be patronising you loose a lot of credibility.

We've crossed paths on measurements before.

Are you even a photographer? Or ever paid to make a photograph?

Wow, I've had this id for 13 years and another before that, made over 500 posts in this sub-forum alone, but you can't bothered to click on my profile.

I can remember when you had a lot of credibility here, made sensible, helpful and constructive posts. Now you just stomp around arguing with everyone. Disappointing.

[digitaldog]

When you start to be patronising you loose a lot of credibility.

Speak only for yourself.
When you are patronizing and speaking in generalizations to support your business model, you might lose creditably to your readers. 

We've crossed paths on measurements before.

Terrific.

Wow, I've had this id for 13 years and another before that, made over 500 posts in this sub-forum alone, but you can't bothered to click on my profile.

Doesn't answer the question.

I can remember when you had a lot of credibility here, made sensible, helpful and constructive posts

San's generalizations. 

Now you just stomp around arguing with everyone. Disappointing.

Use the ignore button/option. That's absolutely fine with me.

[Slobodan Blagojevic]

Andrew, a friendly advice: stick to color theory. Your excursions into logic are typically disastrous. Not to mention verbose.

[digitaldog]

Andrew, a friendly advice: stick to color theory. Your excursions into logic are typically disastrous. Not to mention verbose.

Slobodan, a friendly (piece of) advice: stick to topics you understand and hijacking posts in "The coffee corner" where anything goes.