Onyx/Epson VarDot profile

[eye4clr]

pay no attention at all to the Expected/Measured thing. As long as the colors are reading in order, all is well.

Your post icc lightness evaluation needs to be as objective as possible. The icc version may be more correct. You just may like the lighter version better. This is when being familiar with your test file makes things go very quickly. Given what you've got, how does the lightness look in the rest of the test file?

As far as your icc settings, you could likely start your K earlier for canvas, especially if you've got like K inks in the epson. If you don't have light K, go for around 40, with light K you can go down to 10 or 20.

[SWCL]

Is that for the Achromatic or Chromatic black? Or should I reduce both?

The rest of the file looks ok for lightness - there is a slightly wierd pattern in th CMYK colour wheel, most evident in the Y to M section - looks like steps outwards from the white centre (will try & take a pic tomorrow to show)

K gradient is a little warm - but that ties in with your recomendation in another thread

I will do a print tommorow on our Durst Theta that we have for the photo print side of the business - give me something to compare the printed print to other than the screen.

[eye4clr]

start K earlier in the Acromatic than Chromatic. But I doubt it makes much difference.

That Y to M banding is something i've seen as a chronic problem with v7 onyx profiles. Pure evil and another reason I'm progressively more frustrated with Onyx's continuation of beta software being passed off as an industry leader.

[M.A.C.]

Originally Posted by eye4clr

use the Measure Tool/i1 to get the hue measurement in LCH if you're doubting what color the darker patches are.

think of these values on a circle like compass headings....
0/360 = red
45 = orange
90 = yellow
135 = green
180 = cyan
225 = blue
270 = violet
310 = magenta
360/0 = red

that together with the chroma value tell you what color it is and how far saturated toward that color.

Then use this little gem to adjust accordingly. The arrangement here is one of opposites.
R <-> C
G <-> M
B <-> Y

So lets say you read the H and find a value of 140. Well, that's kinda green. Since you've got CMY to work with, you're either going to add green's opposite of M or reduce C and Y, effectively having more M by proportion.

Damn, i gotta stop giving this stuff away. My wife would kill me.

So really it sounds like a struggle between keeping your greybalance very good, and limiting certain areas of your final gamut/overinking.

IE: if your CMY ramp is coming up green you are going to either have to increase your M restriction (risk overinking) or reduce Y and C channels (risk losing chroma and gamut in those areas.

[eye4clr]

You can go pretty much willy nilly with the Y since you can't really see differences very well. IOW, Y is the variable to help the grey balance along. It's rare that you really need to adjust the C or M to help grey balance very much. A little here and there may help though.

[SWCL]

Thought I'd dig up the old thread as its the same printer etc.

I'm stuck again

Trying to profile the Photo paper again (with the golden sheen)

Have done the Ink restrictions as best as I can..

Printed at 200% as 300% ran off the page, as before C1 + C2 etc. at 100% lights at zero - all transitions set to 4. Measured C & M by Chroma, K by lightness (1DE below max for each), Yellow to balance CMY grey.

Ended up with the figures

C 72 (36%)
M 96 (48%)
Y 56 (28%)
k 184 (92%)

so

Lc1 = 58, Lc2 = 64, C1 = 72, C2 = 0, C3 = 0
Lm1 = 77, Lm2 = 86, M1 = 96, C2 = 0, C3 = 0
Y1 = 56, Y2 = 0, Y3 = 0
Lk1 = 81, Lk2 = 90, K1 = 100, K2 = 84, K3 = 0

The main problem is with the black as you'll see in a minute.. may be caused by this..



Then through the next steps.. Print calibration swatch, read swatch with eye one (tried strip mode & per swatch), unless readings are out.. 'accept' readings. Target densities 'build table' (basic density curve), linearization 'build linearization'

The tweak linearization box comes up & I get this...


(all n factors at 2)

Big seagull in the black


Again, black looks way out of line, but the CMY look ok (to me)


Hit ****yze on the last shot & it comes up 'there are significant spikes in the density readings'

Don't want to hit the build button yet as I'm scared I won't be able to go back to alter something if it is wrong.

Its really getting to me that I cant sort this.. esspecially as the canvas profile came out really well & this is just photo paper!

Stress!!!

[eye4clr]

First off, you can always go back and redo the restrictions and linearization again. No matter how badly you mess it up.

Second, there may not be a problem at all. Go ahead and build like you just don't care. Then print some gradients to try to exploit any non-linear behavior from the "bad" linearization. If you get bands in the gradients, go back. If not, go forward.

BTW, that's some serious Cyan density. It looks like the grey balance may not come out so well. But I don't know a damn thing. Only your tests will tell.

[SWCL]

Thanks for the confidence boost eye4

Should I leave n factors at 2 across the board... or change them to 1.4 for C + M & 1.2 for Y + K.. as is suggested in another thread?

[eye4clr]

N factors are to be adjusted at your discretion to help balance grey and get a better lightness/contrast in the linearization before icc.

LET THE TWEAKING BEGIN!

[SWCL]

Just printed a photo quality test pre ink limit (workflow was not set up for that before)

The colours on it are great considering the ICC is not done... all the solid colours like the pantones & the CMY RGBO ramps are brilliant! nice & vibrant.

The Black.. or should I say brown until 100% in the K ramp.. is absolute crud.

The image with the flowers in the water shows this up the most.. the black (brown) background is so banded it looks like a jpg thats been saved at max compression (I.E. 1 & not 12!)

printed the test with the n factors set at 1.4 & 1.2 as above.

I have a feeling I might have to go back to the restrictions.. in the linearization swatch the blacks look warmer than they appeared in the restriction chart.

My only concern with taking out more Y.. which I asume I would have to do, is that the colour combos like R G & O wont be as vibrant.

Chris.

examples coming...





Colours of the above are as close as I can make them... without profiling the scanner first. I'm sure if reprofiling the scanner got done more often than every 5-10 years or so it would be pretty accurate :.(

Ignore the banding at the top of the second.. that is a scan line... or lines