What to profile? [Archive]

Rooster

09-27-2010, 01:55 PM

Yes, you absolutely need a unique profile for each resolution. You can change the number of passes and head speed without effecting color. But resolution has a big impact on the way the ink goes down.

I find using the same profile for different number of passes very definitely changes the color of your prints. My ink limits vary considerably when using the same resolution settings with a different number of passes. You'll find a higher number of passes will allow you to increase your ink limits for a particular media and expand the available color gamut.

Head speed and uni/bi directional print settings seem to not have any real discernible effect by comparison. Although I haven't messed around with it enough to really verify. I just lock it into fast and bi-directional and leave it there.

kffernandez

09-27-2010, 02:02 PM

@rooster: thanks for the info! with the expansion of the color gamut at least i now have an incentive to profile at higher resolutions other than the "you just have to do it or else" reason. :)

Rooster

09-29-2010, 01:25 PM

Here's a sample of how resolution can affect color.

This is 720x720 8 pass:
http://www.signs101.com/forums/attachment.php?attachmentid=38464&d=1241384323

This is 1440 x 1440 32 pass:
http://www.signs101.com/forums/attachment.php?attachmentid=38463&d=1241384314

Reds are the worst on a inkjet, so don't expect the same boost of color across the board. Reds always show the most problems and are the most responsive to slowing the printer down and laying down more ink. The improvement in color is worth the wait on some jobs. Other projects might see no discernible improvement.

kffernandez

09-29-2010, 02:11 PM

wow. look at those differences.

but are those differences caused by improperly using the same profile across 2 different resolutions, or is it because only the higher resolution one was able to reach the particular color gamut that the art was looking for? [due to it's higher ink limits]

Rooster

09-29-2010, 05:02 PM

Different profiles for each image, same media, same inks, same printer.

What you see is just the difference between laying down less ink really fast (720-8 pass) and laying down more ink really slowly (1440-32 pass). Each profile has been set to use the maximum amount of ink that can be handled at that print speed.

eye4clr

09-29-2010, 05:17 PM

1440/32 pass....holy crap that's gotta be slow.

Rooster

09-29-2010, 06:46 PM

1440/32 pass....holy crap that's gotta be slow.

Yup, it's still the same speed as 16 pass uni though.

I only use it for fine art reproductions on canvas where color is most important. I save those jobs up and let it run overnight so it doesn't impact the daily schedule.

720 x 1080 24 pass is pretty close in color gamut and quite a bit faster. It's sure nice to be able to create custom profiles and have all these options!

kffernandez

09-29-2010, 10:28 PM

i don't know whether i'm supposed to be sad that my usual 720 x 720 would be limited to print at such mediocre quality, or that i now have the potential to print such vibrant colors after the proper profiling of my 32 pass resolutions.

:P

GAC05

09-29-2010, 10:40 PM

Rooster is that you in the photo?
nice dress......

wayne k
guam usa

eye4clr

09-30-2010, 11:25 AM

720 x 1080 24 pass is pretty close in color gamut and quite a bit faster. It's sure nice to be able to create custom profiles and have all these options!

Amen brother.

kffernandez

09-30-2010, 12:21 PM

Rooster

09-30-2010, 01:06 PM

@rooster - i just made my first 720-8 pass profile. it took me 5 whole hours to do it. but at least i think i did it right. my prints look waaaay better than the generic versaworks profiles. in fact, i think they look almost WYSIWYG after i profiled my monitor.

now, i hope it's not too much to ask, but could you please send me the original "seeing red" file that you used? i would love to see how my printer/ink/media comes up beside yours'. thanks!

on a side note, i think it would be a great topic if we could use a "good" sample print having most if not all of the CMYK color gamut... and then have everybody post their printouts at a given resolution while indicating their machine and inks. has anybody done something like that here?

Here's the file I used in the original thread:
http://www.signs101.com/forums/attachment.php?attachmentid=38462&d=1241384299

As I mentioned previously, reds show the most problems and react the most favorably to higher resolutions and ink densities. If you use another image that isn't as vibrant and doesn't contain deep saturated reds you might not even see a difference between different settings. That image was chosen specifically to highlight the problem with reds and show what our equipment is capable of compared to the average shop out there.

What I did to create these samples was to take the original file and convert it to sRGB which is what most browsers expect to find for an input profile. Then I did a "convert to profile" in photoshop to the different custom icc profiles. After converting them to the smaller CMYK gamut I did another convert to profile back to sRGB so the browsers would know what to expect again. There's no need to print anything with this method and trying to compare scans of prints posted online is a lesson in futility. All you need for a comparison is photoshop, the profiles and a reference image.

You'll need to put a copy of your custom profiles into the right spot where photoshop can access them. Then you can pre-flight your images to see onscreen what the color will be like at different printer settings. It's unnecessary for the most part but when you run into trouble it can be a lifesaver.

kffernandez

09-30-2010, 02:09 PM

You'll need to put a copy of your custom profiles into the right spot where photoshop can access them. Then you can pre-flight your images to see onscreen what the color will be like at different printer settings. It's unnecessary for the most part but when you run into trouble it can be a lifesaver.

gee. now that i have a decent profile.. i don't need to actually print a file to see how it's going to look on vinyl?!? wow. haha. this just keeps geting better and better!

@rooster/eye4clr - one thing i forgot to mention tho... i think i skipped a very important part of the profile process. i wasn't able to get a proper limit for my Yellow [i left it at 100 for my reference data printing]. i've seen the both of you mention printing out the CMY ramp to get the proper warm gray balance. but i've searched all my apps, and i can't find a "CMY ramp" anywhere. :'( i'm using profile maker 5.0.5 and Production Manager 8.1v1. could you please tell me how i could do that? btw, this means i need to redo my profile, right? :P

Rooster

09-30-2010, 03:08 PM

Yup, change the ink limits and you change the profile.

All I do is print an CMY gray ramp without a profile while I adjust the yellow ink limit. I've found doing it after the linearization is best, but it requires re-linearizing afterward since you've changed the ink limit for the yellow channel Just create a series of patches in a CMYK file in Illustrator or photoshop with equal parts of CMY (10%, 15%, 20% etc.).

I've found when I maximize the ink channel limits, the linearization process will increase the mid-tone density. So neutralizing the gray ramp prior to linearizing will likely cause you to pull too much yellow from the ink limits. The first few times I tried it without linearizing first caused the neutrality to shift into the cooler blues after linearization. With the linear curves applied to the CMK channels while I adjusted yellow ink limit yielded a quicker and more accurate result with fewer test prints.

I'm not convinced neutralizing the inks prior to profiling is necessary. I have many profiles where I didn't do it and haven't had any issues. The profile will neutralize the result anyhow. It just takes some of the load off of the profiling software and prevents the need to re-linearize as often as your baseline settings are neutral. If your machine is in a controlled environment it's even less of an issue. If there's a better reason to do I haven't heard it yet. If your having trouble with neutrals shifting it's probably metameric in nature anyhow, so just use more GCR and use it further down the ramp.

eye4clr

09-30-2010, 04:49 PM

I'm not convinced neutralizing the inks prior to profiling is necessary.

100% agree.

But here are numerous upsides to achieving the warm grey balance, especially after linearization (before icc). You can output CMYK files without the ICC in play and get REALLY poppy color. Granted it will not be technically accurate. But for somethings, poppy is more important than accurate.

kffernandez

10-01-2010, 04:25 AM

@rooster
All I do is print an CMY gray ramp without a profile while I adjust the yellow ink limit. I've found doing it after the linearization is best, but it requires re-linearizing afterward since you've changed the ink limit for the yellow channel

this might sound like a really dumb question, but i'm a bit confused about the "after linearization" part. in profile maker, from my limited understanding, i assume that the linearization would happen during the construction of my profile. now, how do i get the max Yellow limit "after linearization" if i don't use any profile? ok, under flexi production manager, i could print the CMY ramp right after linearization, but under profile maker, there isn't any option to print a ramp within the appilcation. isnt profile manager supposed to be better? :(

with regards to the actual limiting of my Yellow... since i've made my CMY ramp already. all i need to do is to print different sets using yellow as the variable, and holding my predetermined CM limits constant, right? and the CMY ramp set that gives me the "warm gray" shade would give me my Yellow max ink limit?

btw, you guys lost me at the "neutralizing" part. the only "neutralize" i know is during the defining of the Black Point. and i don't think that's what you guys are talking about.

eye4clr

10-01-2010, 11:53 AM

ok, under flexi production manager, i could print the CMY ramp right after linearizationBingo.

with regards to the actual limiting of my Yellow... since i've made my CMY ramp already. all i need to do is to print different sets using yellow as the variable, and holding my predetermined CM limits constant, right? and the CMY ramp set that gives me the "warm gray" shade would give me my Yellow max ink limit?What you're striving for is to have your linearization give you a warm grey ramp from the fixed CMY ramp. IOW, don't tweak the file to get results, tweak the linearization and/or ink restriction.

btw, you guys lost me at the "neutralizing" part. the only "neutralize" i know is during the defining of the Black Point. and i don't think that's what you guys are talking about.Everything we're talking about happens PRIOR to the icc coming into play in any way. Set profilemaker aside for now and focus on getting good output established before you print and measure the icc.

As Rooster pointed out, you can still get good results without bothering to achieve the warm grey CMY output. But what you gain is more flexibility in your output options to use or not use the output icc and avoid possible problems from asking the icc to overly correct for a print process that does not have proper grey balance. From my experience, having color bands in gradients when testing your linearization results (again, prior to icc) will cause much more problems with the icc than poor grey balance.

Rooster

10-01-2010, 01:11 PM

btw, you guys lost me at the "neutralizing" part. the only "neutralize" i know is during the defining of the Black Point. and i don't think that's what you guys are talking about.

What we're referring to is bringing the CMY grayscale back into neutrality or non-casted color. In other words we're making the grays gray again and not the greeny gray you tend to get with a standard linearization. That's the reasoning for taking the yellow out as it's the ink that most affects that particular color cast.

kffernandez

10-02-2010, 01:32 AM

@eye4clr
Set profilemaker aside for now and focus on getting good output established before you print and measure the icc.

i wonder how many people actually miss this part. when i started this, i thought that the icc profile was the beginning and end of it all. you guys have been incredibly helpful in providing the much needed details. i kinda feel sorry for the other guys who started profiling with only being given the basic steps without even talking about the significance of each step. i really appreciate the help. :)

ok, i've decided that i want to have the warm gray ramp. in flexi, [after linearization] i've set up the CMY multi-ink limit at 300% and have printed out several sets of ramps using 70%-90% yellow max ink limit settings at 5% intervals. the thing is, my CMY ramp is showing up as blueish-green. from what you mentioned from your other posts, this just means too much yellow. i don't know why, but there seems to be very little difference between 70%-90% yellow. now, i'm not really comfortable with going below 70% since my Cyan and Magenta's have an 80% limit. am i doing something wrong here? or should i keep going down till i hit my warm gray? you did say from your other posts that we shouldn't sacrifice the gamut too much for the gray balance.

btw, just for kicks, i printed out my own CMY ramp using the first profile i made [100% yellow] and it had a pretty nice shade of gray. not the pantone warm gray you mentioned [yes i have swatch with me]. but a pretty decent one still. i know that this is besides the point. but how do we measure and compare gray? :) like black, it just feels too subjective. or is the warm gray balance the beginning and end of this conversation?

@rooster / eye4clr

where do you learn this stuff?!? :P after this, i might try to apply some of the things i've learned here on my dad's industrial rotagravure printing. they waste tons of money on trying to get colors right for each and every project. and i'm worried that not that many people might be able to help me out on that one.

as always, your inputs are truly appreciated.

Rooster

10-02-2010, 12:12 PM

If your gray ramp was close to neutral before you started making the adjustments then the adjustments were probably unnecessary. There's another method of setting ink limits using measured chroma values that eye4clr is familiar with. I've never tried it, but the theory is sound. Maybe he'd like to explain it to you or post a link to his previous explanation. I'm a little set in my ways so I haven't bothered to experiment with it yet.

As for learning color management, between 20+ years working in the offset business doing everything from pasting up galley type and shooting darkroom halftones to operating a drum scanner gave me a good base. The best teacher was reading the manual I got with my first profiling package (picto's color synergy) from cover to cover as many times as I needed to to understand everything in it. You won't find a more boring book to read (three times or more), but what I learned has saved me so much time and money over the years it was one of the best investments of time I've ever made.

kffernandez

10-03-2010, 05:01 AM

@rooster

i might not be able to wait for 20+ years worth of work experience.. but yes, the stuff that profile maker manual has shown me is pretty significant. it might lack the detailed explanations that i need, but at least it teaches me the basic terminologies and concepts that i need to get myself started. and yes, it really is boring. hehe. this, coming from a guy who was excited to read his Dos 3.3 and 1980's genre Hard Disk survival book. :P

hmm.. you have experience with offset? isn't that not too far off in concept from gravure printing? i might still be able to find some resources here after all! :D

eye4clr

10-04-2010, 12:40 PM

As usual, Rooster is dead on. You don't have to match the warm grey thing perfectly. Only if you were striving to make things as perfect as possible without the icc. But, you'll have the icc to get the fine tune for most jobs. Again, don't use the icc only for jobs that are not color critical and would sell better by being "poppy" color instead. In that case, if the grey balance is a tiny bit off and you're just printing GRAND OPENING in red letters, who cares!

I cut my teeth in digital printing in the early days of ColorBlind. In the days of walking up hill to school both ways in the rain and snow, Apple had a ColorSync forum that taught me a solid foundation from the best in the industry. I also had the luxury of working for an equipment dealer that allowed me time to experiment and learn while on the job. I can't see how folks really learn color in depth while having to get jobs out. It takes either time or money spent on training. Even with the training, you still need time to practice and experiment to really get a feel for it. I have also been fortunate enough to work around some very smart and articulate people that were generous enough to share their knowledge.

To do ink restrictions by chroma you'll need to use ProfileMaker's Measure Tool. Note that all of this is done PRIOR to linearization and with all restrictions set to 100. Use the spot measure function set to LCH (instead of LAB) and the delta e set to the last choice in the list (something with 2000 in the name). Create some ramps of individual C, M, Y, and K with an additional ramp of CMY combined. Start by measuring the simple ramps of each color that increment by 2-4% steps. In the Spot measure tool, measure the 100% patch of cyan as the "reference". Then measure down the ramp and watch for the delta e for Chroma to go up above 1. At that point you have the patch that is the first step that gives the Statistically Average Viewer (whoever that is) the threshold of seeing a difference in saturation (chroma). Do not get sucked into judging differences by lightness. Density does not play a role in this step, only saturation (chroma).

Now you have the ink restriction for cyan. Do the same for magenta. Skip yellow, we'll revisit this one shortly. Do black the same way but watch for delta e change in the L (lightness) channel.

To choose yellow, eyeball the CMY ramp to determine what color the cast of the ramp is. I'll guess that without restricting ink at all, it will be some degree of green/yellow. I'll also guess that your cyan will restrict more than your magenta. That will naturally help with a green cast. As a starting point, try restricting Y to be the same as C. Reprint the ramp test and eyeball the CMY ramp again for color cast. Adjust Y only and reprint in cycles until you get it as close to warm grey as practical. Voila! done.

Now move on to linearization and be prepared to tweak the linearization results for flexi. If it behaves as it did when I worked with it, the linearization results are not neutral nor linear. Having smooth linearization results is far more important than the warm grey thing.

Happy profiling!

kffernandez

10-05-2010, 12:55 PM

@eye4clr

if i got interested in large format printing 10 years ago, there is a good chance that i would have joined the same forums that you were able to join. 10-15 years ago, the only forums/lists that i joined were technical discussions on hardware, games and networking. these were forums filled with very talented, passionate, and incredibly generous people who were willing to share/defend whatever they know to whomever that needed their skills. there are still people like that nowadays, but they seem to be more of the exception, rather than the rule.

thanks for making such a detailed instruction of the steps. actually, i've seen you say the same thing, but in bits and pieces from your various posts. it makes a ton of difference when you have a single paragraph that you could read over and over again to make sure that you don't miss anything.

frankly, i think you explained it as best as anybody could - irregardless of the software they use. if you could say the same thing in a different post, i strongly feel that it should be stickied in the forum. i feel profiling is a critical element for anybody who bothers to buy such an expensive machine. now that i have a basic understanding of profiles, i'm wondering why people would even consider doing otherwise.

one last question tho... i took me a while to realize that when you were talking about "poppy colors," you actually meant not using the icc profile at all! that kinda blew me away. hehe. i thought you were just skipping the gray balancing part and still completing the profile. with that in mind, i'm a bit confused as to what the icc profile does in the first place. i've always thought that it was a required step for printing, in the way that it makes sure that i am getting the right colors. reds are red, and greens are green. now, if i understand you correctly, you're saying that i could have more striking colors [poppy] with just doing ink limits, dropping icc profile altogether, although sacrificing an acceptable level of color correctness? [great for arts and signs but maybe not for photographs or logos with color specific instructions]

on a side note, why would it make the colors "poppy" in the first place? bec we are operating under the max ink limits without sacrificing any ink whatsoever to the gray balancing?

thanks again!

:)

eye4clr

10-05-2010, 02:43 PM

why use an icc at all?

Good question. Simply put, the icc gives you color accuracy. Accurate reproduction of the color in your incoming digital file based on the LAB values of that color. This assumes the file's embedded profiles are respected as the input profiles for the printer's output profile to convert from.

To not use the icc lets the printer's color gamut be fully utilized with no regard to color accuracy. An example goes a long way here. If I print 100% cyan and use the icc, i'll get a print that is true to the LAB reference for the cyan in my file. That means one thing in sRGB, another thing in US Web Coated SWOP, and another in US Sheetfed Coated. If all my output is properly color managed and has tight quality control (yeah...right), then I'd get the same output on most materials and off multiple printers. If i print without the icc, I likely will get a bit more vibrant cyan because the ink has a higher gamut than any of those source examples, it would be different on each material, and certainly different on different machines as each cyan is slightly different hue.

The trade off is accurate color for vibrant (inaccurate) color.

Other goodies of using the output icc are greater control and customization of Total Ink (black), adapting output to different lighting conditions, how K gets substituted for CMY (GCR), and when K starts getting used in the lightness scale. These are all valuable tweaks to optimize the printer's behavior for different materials and output goals.

For example, I setup my banner output on a given machine to use WAY more K and start using it sooner in the lightness scale than I do for printing Durasol tradeshow film or poster paper. This hides the black dot when I don't want it and uses it a lot to reduce overall ink use and improve dry time when print quality is not the primary goal.

I'm trying to write this with my 2 year old crawling all over me. I hope it is coherent.

kffernandez

10-05-2010, 04:18 PM

If i print without the icc, I likely will get a bit more vibrant cyan because the ink has a higher gamut than any of those source examples

if i now understand it correctly, the key concept here is gamut range. an icc profile setup has a lower gamut simply bec an icc profile will tranlate the 100% cyan into what it thinks will produce the LAB reference given my ink/machine/media/etc setup-whether it means adding/subtracting colors. while a non-icc setup simply drops the amount of ink that the CMYK color dictates without even thinking about it. my example here is when printing my CMY ramp without any icc, my machine/software just blindly drops the ink i tell it to [5%, 10%, etc]. thus, usually resulting in a "non-neutral" color. while printing the same ramp using my icc allows it to "think" first of what color i'm trying to achieve, and then drops just the amount of ink it thinks my machine needs to produce that color. [including K when needed]

if i got that one right, i think i'm really getting the hang of things. :)

----------------

i hope you don't mind. but since i already got your attention, i wanted to ask you about a post you did on another thread. you were talking about a guy's attempt at profiling at 360x720:

Start by abandoning 360x720. The printer will run too fast for decent dot control. Then if you slow it down by going to uni directional or choosing a slower head speed, you're back at 720x720 productivity or slower. Let me suggest 720x720, 8 pass, fast head speed. You'll end up with better output, more stable print production, and a larger color gamut.

now, i also feel the need to make a profile for my banners/tarps [either 450x360 or 360x720] so, are you saying that it's more practical to just do an 8 pass 720x720 as a minimum for any print?

regarding the head speed. i've always wondered about this. my roland head/scan speed defaults were set at 1000mm/sec for my 360x720 canned profile. my roland tech said that this was very bad for my heads, and should be set at around 650-750 max. he told me this without really saying why. so, is it about the head not catching up to the electrical instructions? [some say bandings were caused by this] or is it really about the "dot control" that you mentioned? ink drying problem? how do you even know if you've reached the point where it is already "too fast for decent dot control?" what does dot control even mean? :P

actually, i just want to know how to set my head/scan speed. i don't like to just blindly following other people's word for it without really knowing the logic behind it. :(

btw, having a 2 yr old crawl around my feet would never be a distraction. well, not unless they start banging at you, try to be a stuntman, or attempts to eat the stuff around you. :P

i hope i didn't ramble too much.

thanks again!! :D

bbeens

10-05-2010, 04:52 PM

A profile is not required for printing, although in most situations this is not a good idea. The output profile is just a big lookup table of values sent to the printer and values read from the printed output - think of the profiling patches step. Along with the output profile you input profile has the same big lookup table. If you remove one of the profiles then the other is ignored and whatever values are in your file get sent to the printer. This can produce 'saturated' colors. Although the same/similar effect can be achieved by using altered input profile which adds 10%, 20% to each colorant.

It is a very tricky workflow. But for me, and a few others here, it is a very interesting topic. Luckily I get paid to think about this stuff.

Bryan

eye4clr

10-05-2010, 05:00 PM

if i got that one right, i think i'm really getting the hang of things. :)sounds like you've got it

I don't know the roland well enough to suggest specific head speeds. But the faster you go the more the dot will "splash" when it hit the media. At some point it no longer is very round on the media and becomes more tear-drop shaped. This will contribute to a rough and possibly banded appearance.

Also know that lower resolutions have fewer locations to put dots on. So it becomes hard to get decent gamut at 360 dpi.

kffernandez

10-06-2010, 03:19 AM

well, i think i now have the basic understanding i need for a decent profile. and i have you guys to thank. hehe. :) but before i stop bothering you guys could you just give me a brief and basic understanding of what the "custom color mapping" and "global color mappings" are for?

also, under flexi and versaworks, for every print project, there is a tab for color adjustment - which looks strangely like a linearization grid - with an input/output point for each channel. my grids all just have a straight line from 0 to 100. do they actually expect us to modify this by hand without any tools?

eye4clr

10-06-2010, 11:37 AM

also, under flexi and versaworks, for every print project, there is a tab for color adjustment - which looks strangely like a linearization grid - with an input/output point for each channel. my grids all just have a straight line from 0 to 100. do they actually expect us to modify this by hand without any tools?
You're spoiled already. Love it.

My guess (I don't really know) is that the mappings have to do with spot color rendering.

kffernandez

10-06-2010, 02:54 PM

heehee. i try to get away with as much as i can. thanks again! :P

kffernandez

10-07-2010, 02:43 PM

@eye4clr

i hope you don't mind. i tried to do everything all over again - the right way. but i felt that i needed to clarify some things regarding ink restrictions by chroma. [Measure tool by Profile Maker ] After using LCh color space and selecting "Delta E 2000" in the list box - you instructed:

In the Spot measure tool, measure the 100% patch of cyan as the "reference". Then measure down the ramp and watch for the delta e for Chroma to go up above 1. At that point you have the patch that is the first step that gives the Statistically Average Viewer (whoever that is) the threshold of seeing a difference in saturation (chroma).

when you say "Delta e for Chroma" do you mean "Delta C?" because there are 4 values that are provided for every spot measurement. (1) "Delta E 2000" , (2) "Delta L" , (3) "Delta C", and (4)"Delta h." when you said "Delta e for Chroma" i assume that you mean Delta C, and likewise for checking for lightness - to you Delta L. am i correct?

i got a bit frustrated earlier bec sometimes the Delta C would exceed 1, and then drop again. so, i was second guessing myself if i was reading the right measurements, or was just having a poor sample/reading.

also, regarding the CMY ramp. do i need to make the whole ramp gray? bec, the lighter ones seem to be much easier to turn into gray than the darker ones. the darker ones [ 20% and up] look more Dark blue-ish green than anything else. and i think it's near impossible for me to figure out how to make them all warm gray without drastically changing my limits.

hope you're not yet tired of clarifying things for me. :)