I’ve just bought a loupe so will read the sheets through that once it comes.
not sure I follow?
-
I want to thank all the members that have upgraded your accounts. I truly appreciate your support of the site monetarily. Supporting the site keeps this site up and running as a lot of work daily goes on behind the scenes. Click to Support Signs101 ...
Discussion ICC profile making
- Thread starter Mata
- Start date
in building the profile, you printbthe first set of patterns, set the ink limits (throttle back the ink LIMITING EACH COLOR) to the practical amount the substrate can handle.
then you print the test patterns, scsn them with the xrite i1, and the software makes small corrections for each color (comparing what you printed to what it is supposed to be. with the first set of changes, you start zeroing in, moving closer to the correct colors
then you print a more elaborate set of patterns and the software makes final changes. (not a perfect description, trying to simplify a complex topic)
at this poinpt your colors SHOULD BE printing out properly ASSUMING there are no flaws in the printer
then you print the test patterns, scsn them with the xrite i1, and the software makes small corrections for each color (comparing what you printed to what it is supposed to be. with the first set of changes, you start zeroing in, moving closer to the correct colors
then you print a more elaborate set of patterns and the software makes final changes. (not a perfect description, trying to simplify a complex topic)
at this poinpt your colors SHOULD BE printing out properly ASSUMING there are no flaws in the printer
ColorCrest
All around shop helper.
As a comparison using Onyx, this first step is regarded as Ink Restrictions. A media model is chosen and the ink restriction file is printed and read via a spectrophotometer. The ink restrictions are then automatically optimized to accommodate the media.
The original media model could be a material such as a typical SAV which can support only a certain amount of ink load. Thus, the software had a clue and a range of where the ink restrictions should be set. Onyx handles this first step competently. The software further offers a visual graph of where the individual color channels landed so far as their color gamut. One may see if over or under inking has occurred. Again, the software usually performs perfectly well.
The notion of ink restriction is to restrict the ink to a setting of ~100% level for the media tolerance because almost always the printer can deliver much more ink than the media will practically allow. So, with the setting now made, a graph will display properly along with headroom which may accommodate future adjustments for re-calibration, etc.
What does not necessarily occur automatically is gray balance which is another topic.
ColorCrest
All around shop helper.
What has happened with the original Onyx setup of this printer, exactly, and why have you not presented photos of control images so far?
Onyx was fine and prints were good using my custom profile but my design to print process was poor.
If I was printing wraps, posters etc, onyx was fine, when I was printing cut-contour graphics or just standard cad color it was slow so I’ve now changed to Flexi 22.
My graphtec FC9000 works better with Flexi meaning I don’t have to have cutting master installed as well and now I have flexi designer which will help setting my cut-contour graphics especially on detailed graphics.
Just to clarify, I’m now trying to setup a color profile in Flexi 22.
natallica
Graphics Guru
There is a technique I learned at an SGIA Color Management class that has helped me out immensely.
I wanted to take the 'visual' ink limit check out of my process, and actually SEE via readings how to get my ink limits perfect and here's what I was instructed to do.
• Be sure 'NO COLOR CORRECTION' is selected when sending charts etc. (via Rip & Print dialog) it will screw up the readings if it is enabled on tests.
• When doing the initial 'INK LIMIT TEST' Print the chart with all channels at 100%
It entails using your spectro and suitable software that will allow you to get LCH readings. We have an I1Pro2 but I've been using a ColorMuse (about $60 on Amazon) to grab readings. It comes with an app that allows you to remotely read a color and compares close matches to a variety of products (Avery, Sherwin Williams, Baer tec.) even tells you DeltaE differences from the swatches you scan. But at the bottom of the screen there is a 'compare' function that allows you to take a reading, and slide the box above through RGB, Lab, CMYK and most importantly for these tests...LCH.
After I've printed an Ink Limit test (using Flexi22) I scanned the two cyan values at 95% and 100%, 95% yielded a chroma value of 56.59 and 100% has a value of 57.39, so I'm good to leave it at 100% because there was not drop in chroma. If there were a drop in that number I would choose 95%.
I've even done this on Multi-Ink Limits. The 'properties' of the ink are what's to be evaluated, not the actual color. So my theory is this should work, no matter what. If Chroma value drops, you're just wasting ink, which is why the canned profiles from Epson, HP, Mimaki, Mutoh etc. are a waste IMO. They sell more when you pile it onto your vinyl.
The goal here is to achieve the maximum 'Chroma' value for a given channel. Any more ink added after that maximum value is wasteful and it will not improve gamut at all. If you've ever seen a web view of a profile and saw the ends 'curve' from a given point...that is where chroma was at it's highest...if the tail curls from that point you're just wasting ink.
So the theory is, set the Ink Limit to the maximum chroma value by scanning it, not just looking at it. Visually the test will show you some modelling or pooling, but the only way to know for sure if you're wasting ink is to read the values given by the spectro.
Doing this eliminates the need to stare at it with a loupe or a magnifying glass because mathematically you'll know that any darker swatches you'll get are a waste of ink and not necessary to go any higher.
Flexi's little swatches make it a tad difficult to read because of the triangular voids, so I have to take a few readings. It's hard to get the reader right onto the solid area.
I had a trial of a program called SpotOn! that you can run tests of your profile by printing Idealliance Control Strips and uploading the results and surprisingly my profiles have been PASSED and VERIFIED so I'm assuming this works pretty well for creating a pretty 'on the nose' profile.
Some may say this is too much work, but I prefer to remove 'human error' variables when I can, and I think looking at a swatch is like throwing a dart... however I prefer hitting a bullseye on the first throw, instead of going back in an tinkering with settings, not knowing what else could change. And even it's not 'the way' real color guys do it...I'm always willing to learn but it's working for me, so far.
Hope this helps you Sir,
-N8
I wanted to take the 'visual' ink limit check out of my process, and actually SEE via readings how to get my ink limits perfect and here's what I was instructed to do.
• Be sure 'NO COLOR CORRECTION' is selected when sending charts etc. (via Rip & Print dialog) it will screw up the readings if it is enabled on tests.
• When doing the initial 'INK LIMIT TEST' Print the chart with all channels at 100%
It entails using your spectro and suitable software that will allow you to get LCH readings. We have an I1Pro2 but I've been using a ColorMuse (about $60 on Amazon) to grab readings. It comes with an app that allows you to remotely read a color and compares close matches to a variety of products (Avery, Sherwin Williams, Baer tec.) even tells you DeltaE differences from the swatches you scan. But at the bottom of the screen there is a 'compare' function that allows you to take a reading, and slide the box above through RGB, Lab, CMYK and most importantly for these tests...LCH.
After I've printed an Ink Limit test (using Flexi22) I scanned the two cyan values at 95% and 100%, 95% yielded a chroma value of 56.59 and 100% has a value of 57.39, so I'm good to leave it at 100% because there was not drop in chroma. If there were a drop in that number I would choose 95%.
I've even done this on Multi-Ink Limits. The 'properties' of the ink are what's to be evaluated, not the actual color. So my theory is this should work, no matter what. If Chroma value drops, you're just wasting ink, which is why the canned profiles from Epson, HP, Mimaki, Mutoh etc. are a waste IMO. They sell more when you pile it onto your vinyl.
The goal here is to achieve the maximum 'Chroma' value for a given channel. Any more ink added after that maximum value is wasteful and it will not improve gamut at all. If you've ever seen a web view of a profile and saw the ends 'curve' from a given point...that is where chroma was at it's highest...if the tail curls from that point you're just wasting ink.
So the theory is, set the Ink Limit to the maximum chroma value by scanning it, not just looking at it. Visually the test will show you some modelling or pooling, but the only way to know for sure if you're wasting ink is to read the values given by the spectro.
Doing this eliminates the need to stare at it with a loupe or a magnifying glass because mathematically you'll know that any darker swatches you'll get are a waste of ink and not necessary to go any higher.
Flexi's little swatches make it a tad difficult to read because of the triangular voids, so I have to take a few readings. It's hard to get the reader right onto the solid area.
I had a trial of a program called SpotOn! that you can run tests of your profile by printing Idealliance Control Strips and uploading the results and surprisingly my profiles have been PASSED and VERIFIED so I'm assuming this works pretty well for creating a pretty 'on the nose' profile.
Some may say this is too much work, but I prefer to remove 'human error' variables when I can, and I think looking at a swatch is like throwing a dart... however I prefer hitting a bullseye on the first throw, instead of going back in an tinkering with settings, not knowing what else could change. And even it's not 'the way' real color guys do it...I'm always willing to learn but it's working for me, so far.
Hope this helps you Sir,
-N8
Thank you so much. I will order that and let you know how I get on. I ended up creating a profile in i1Profile and importing it for now. I I still think I can improve it but I needed to get some jobs out the door.
V. V.
Inkjet printing guru
isn't it enough to check the linearization graphs after preliminary Ink limit set up? it shows everything you need
natallica
Graphics Guru
It may be, but I'd assumed in my case, the linearization chart is CMYK only, and having a 9 color printer it would help during the 'total' ink limit step as well.
I've read that for these Epsons (80600's) to achieve great results you have to only build a CMYK ONLY profile and 'Precision Dot' takes care of the rest as far as adding when necessary, because I initially tried doing CMYKLcLmOR but that was way too red and no where near 'color correct' on a standard photo.
Is that truly the case, and I really only have to 'chroma check' the linearization table after initial ink limits?
Thx,
N8
natallica
Graphics Guru
It's called Color Muse, a small USB chargable bluetooth device available on Amazon (for about $60). It's about the size of a size C battery. It is used for taking color samples and matches measurements with many databases (i.e Sherwin Williams/Baer/Avery/Pantone etc.). When comparing samples the app for the device will give you RGB/LCH/Lab values of the samples and you can compare colors based on DeltaE differences in the readings. It quite a handy little item for getting a close match off of a customers car or a printed sample.
I'm sure an i1 or similar spectro would do the same, but the software for those is 'clunky' IMO when your not doing something profile specific so I just use this, seems easier and I'm not tethered to a USB cable when charged.
-N8
LiqueColor
New Member
Ok, your ink limits are way off. In Flexi, start at the 100% mark and look at the triangle part of the print. Keep going down ( To the left) until you see a difference in the triangle part. The difference being the density of the triangle. Then pick the triangle to the right of where you see a difference in the density. Your eyes need a Delta 2 or greater to detect a difference in color. On magenta’s and cyan‘s, I often go two to the right of the color change. If your blacks tend to be flooded with ink, it’s because the software is trying to do a composit black with the CMY. The GCR replacement is where you tell the software to substitute straight K for the CMY value. Often, I start at 35 and get blacks that look correct.
Don’t worrry too much on the bleeding as much as the change in the density of the color of the triangle. Bleeding can be controlled with heat and the solvent in the ink is also responsible for controlling the dot. A real good ink can control the dot to the point the ink will pool with no bleed. Coated medias assist in controlling the dot and you might look at some different medias if your not happy with the dot control.
Don’t worrry too much on the bleeding as much as the change in the density of the color of the triangle. Bleeding can be controlled with heat and the solvent in the ink is also responsible for controlling the dot. A real good ink can control the dot to the point the ink will pool with no bleed. Coated medias assist in controlling the dot and you might look at some different medias if your not happy with the dot control.
So I bought a color muse, it's quite a small funky design compared to the i1Pro2.
So I am going to attempt to create a new profile in Flexi 22.
I'll write a list of what I'll do and please correct me if I am wrong.
Since writing that, I've now printed the initial ink limits chart (not using Epson's cantone selection)
1. Click Color Profile
2. Create ICC Profile
3. Select your options
4. Device: Epson
5. Media: Adhesive Vinyl
6. Print Mode: 1440 x 1440 16 Pass
7. Resolution: 1440x1440
8. Color Mode: CMYK (2bit)
9. Dither: Enhanced Stochastic
10. Give your profile a name
11. Click Next
12. Print Ink limit Test (Colour correction will be turned off automatically)
Two options here...
13. Read the chart using Color Muse and look for the highest Chroma value and select that colour
13. Look at the chart and select your colour based on the bleed/pooling
I've got to 13. I've read the chroma values but the results change depending on your scan position, it's not the easiest to pick a spot especially as the squares are very small. I'm trying to get the sensor in the top right hand corner but I can't see. So far Cyan 55% has the highest Chroma but when I scanned it again, the value changed.
If I did this by eye, I would have selected 60% for all.
I have attached the images to get your thoughts.
natallica
Graphics Guru
Yeah, it's a little tough to get the tool directly over a good spot. I usually do a couple of readings while slightly moving around the area to hone it in as best as I can. You'l know when you're way off base because it will even show a wierd sample in the app. Just ignore that one, scoot it around and try again. That's the main problem I have with this method is getting accurate readings every time. And visually it looks as though that's probably about right by the photos. Highest chroma before dropping off is best. Any more and you're just wasting ink, and it won't improve gamut, also...it takes longer to dry, it makes some vinyls very 'gummy' even if it appears dry, if it's solvent it can (and will) affect the adhesive and make installations much tackier and harder to install.
But yeah, it would be nice to have larger swatches, or at least a few checkboxes in the software to maybe choose a different 'style' of swatch for those who choose to use a spectro.
I'm anxious to see if you are happy with the results of your profile. Unless someone comes up with a more accurate way of building one, I'm sticking with this method for now.
Hope this was helpful.
-N8
Thanks, I'll read my watches with the Color Muse and take it from there and let you know how I get on.
Once you've created your profile, how do you compare it to another? I used to do this some time ago but I can't remember the software. Onyx had a pretty good compare feature which should be the profiles in numbers.
Does Flexi have a test print image? I can't find it. The I used Onyx I and the wine glass and bottle image so I can check my completed profiles my eye.
So after playing about at the weekend, I have come up with the following:
Print out a Ink Limit chart in Flexi
Leave the print to dry for a day
Scan the Cyan row using i1Profiler (Advanced options) / Measure Chart
Select 1 row and 1 column
Change selection to Spot
Scan 100% patch, then scan 95%, 90 etc etc and check the Chroma Values and look for the highest value
Do the same for all C, M and Y
Once read, change the dropdown box to CIE Lch
We’re going to be looking at the M1 results for C (Chroma)
For K, we’re going to scan the 100% patch and save the result and name it 100 black, do the same for 95, 90, etc all the way to about 30%
Now open Babelcolour Patch Tool and Open the 100% chart in M1 then click on the black colour
Now open up the other charts and hover your mouse over the black colours, we’re looking for a delta value of 1 or slightly less
My results are so..
C - 35% (Chroma Value of 66.63)
M - 35% (Chroma Value of 67.72)
Y - 60% (Chroma Value of 66.63)
K - 40% (Delta Value of 0.90 compared to 100% patch)
Open Flexi
Create a new profile
Ink Limits
change the ink limits to match the percentages you read earlier
Once done click next
Linearization
print the chart
Select your i1Pro2
Read the patches
Multi-Ink-Limit
Print the chart
But I have a problem with Multi-Ink-Limit
I don’t feel you can read the Chroma values like you would CMY, I tried this and a few came out fine but any colour with K in gave me a low percentage if I went by the chroma value. I tried looking at the Delta value but this didn’t deep to help either so…
I can check this by eye
And I would say 90% across the board
Or
I can keep the values based on the chroma and ready the charts with K in them but use the H from Lch. I then check by eye to check for pooling and it seems fine so went with that.
If I did use that way my findings would be
A: 95%
B: 85%
C: 80%
D: 90%
E: 100%
F: 75%
G: 80%
H: 70%
I: 100%
J: 100%
K: 95%
I’m now going to read a chart and see how it goes.
Print out a Ink Limit chart in Flexi
Leave the print to dry for a day
Scan the Cyan row using i1Profiler (Advanced options) / Measure Chart
Select 1 row and 1 column
Change selection to Spot
Scan 100% patch, then scan 95%, 90 etc etc and check the Chroma Values and look for the highest value
Do the same for all C, M and Y
Once read, change the dropdown box to CIE Lch
We’re going to be looking at the M1 results for C (Chroma)
For K, we’re going to scan the 100% patch and save the result and name it 100 black, do the same for 95, 90, etc all the way to about 30%
Now open Babelcolour Patch Tool and Open the 100% chart in M1 then click on the black colour
Now open up the other charts and hover your mouse over the black colours, we’re looking for a delta value of 1 or slightly less
My results are so..
C - 35% (Chroma Value of 66.63)
M - 35% (Chroma Value of 67.72)
Y - 60% (Chroma Value of 66.63)
K - 40% (Delta Value of 0.90 compared to 100% patch)
Open Flexi
Create a new profile
Ink Limits
change the ink limits to match the percentages you read earlier
Once done click next
Linearization
print the chart
Select your i1Pro2
Read the patches
Multi-Ink-Limit
Print the chart
But I have a problem with Multi-Ink-Limit
I don’t feel you can read the Chroma values like you would CMY, I tried this and a few came out fine but any colour with K in gave me a low percentage if I went by the chroma value. I tried looking at the Delta value but this didn’t deep to help either so…
I can check this by eye
And I would say 90% across the board
Or
I can keep the values based on the chroma and ready the charts with K in them but use the H from Lch. I then check by eye to check for pooling and it seems fine so went with that.
If I did use that way my findings would be
A: 95%
B: 85%
C: 80%
D: 90%
E: 100%
F: 75%
G: 80%
H: 70%
I: 100%
J: 100%
K: 95%
I’m now going to read a chart and see how it goes.