Welcome to OnyxTalk - Global Support Community, please Login or Create an account to get full access to the forums.
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5

Backlit Dye Sublimation Profiling

Offline MarkPrints Posted 05-03-2015 - 01:46 AM
Post: #1
Member
9 Posts
Member

Onyx Version:
Thrive 12
Printer(s):
Arizona250GT,JV5320DS,EpsonFS7100/F7200,SP540V

Having a bit of trouble profiling a backlit material printed on Epson F7100 to transfer paper and pressing to woven material.

i am printing 720x1440MD and using extended gamut preset for channel ink limiting. this reduces channels to about 35-40%, I have altered the black up to 70% to get some extra density.

Lin strip prints fine and looks nice over lightbox. there is a nice black at 100% black swatch

using the advanced ink limit print I limit to average of 230. Visually over light I have nice rich red green and blue at 200. Again the black patch looks nice even at 100K only. I notice a desaturation after about 250 which I attribute to overinking.

D and E sliders adjust automatically to 100% and 25% respectively

from here I print the standard ink patch set for profiling and again they look bright and punchy over light.

read patches with Barbieri LFP in Tranmissive mode and build profile with the no light inks preset.

the resulting profiles are effectively accurate but it appears the gamut is compressed. the bright punchy colours from the patches is gone and the black point is not as rich as the 100K only patches prior to producing ICC.

Any suggestion would be appreciated
Back to top Find Quote
Offline Correct Color Posted 05-04-2015 - 11:47 AM
Post: #2
Frequent Flyer
128 Posts
Frequent Flyer

Onyx Version:
10
Printer(s):
Epson

Mark,

Well, it's impossible to say from this distance, but if I was to hazard a guess...

Profiling is first and foremost about building machine states. And one of the critical aspects of building machine states is the correlation between single- and multi-channel ink limits. If they're not right, you can get into issues similar to the one you're describing.

And my guess would be that the problem is somewhere around your using so much black compared to your actual chroma primaries, and such a low multi-channel limit.

Black gives you no gamut at all, and doesn't really add nearly as much density as many think it does. In fact, I'd say that in probably 90% + of the profiles I write, I set the black as the lowest single channel restriction.

So while it may or may not be the issue, if it was me, I wouldn't create a condition such as the one you describe. I'd bring the black down, your chroma primaries up -- if you can -- and shoot for a multi-channel density of over 300.

Of course, I'd also do all this myself, and not rely on any presets. But that's another story.



Mike Adams
Correct Color
(This post was last modified: 05-04-2015 11:49 AM by Correct Color.)
Back to top WWW Find Quote
Offline Mikulas Posted 05-04-2015 - 05:52 PM
Post: #3
Banned
55 Posts
Banned

Onyx Version:
10.2
Printer(s):
L26500, W8400+Onyx / Mimaki JV4, JV5 +TexPrint

Hi Mark,
just a small remark - color separations in dye subli are quite different from what is common in ink-jet !!!
It is not uncommon to have the profile´s total ink limit at 380, C/M/Y at 95-100% and more limited black.
Onyx advanced ink limiting has a bit different approach but the strategy is the same - lot of CMY, and a bit less K.
And don´t forget - when generating ICC with Onyx, leave the total ink limit at 400% !!! Wink

Btw.: I would be very interested how are you doing the linearisation and profiling with Barbieri !
We do a lot of backlite on textile, but we only use our i1 Pro2 (reflective) and our eyes.
Our very common technique is using usual reflective profiles (ink limits are already maxed-out,
so no additinal saturation gain possible), and pushing the midpoint of the curves 10-30% higher.
How are you linearising in transmissive mode with woven fabric?
Is there any option to set-up some frontlight-to-backlight ratio?
From my experience this is one crazy area where one cannot get perfect results....
(This post was last modified: 05-04-2015 06:03 PM by Mikulas.)
Back to top Find Quote
Offline MarkPrints Posted 10-22-2015 - 04:20 PM
Post: #4
Member
9 Posts
Member

Onyx Version:
Thrive 12
Printer(s):
Arizona250GT,JV5320DS,EpsonFS7100/F7200,SP540V

Thanks for your replies
Sorry I missed the second one and the questions posed

For our backlit work I pretty much use the barbieri in transmissive mode and follow the standard Onyx profiling route

I would only use the backlit channel ink limiting preset in onyx if I had a problem with ink bleeding but this hasn't been an issue in direct or transfer printing

this seems to arrive at accurate profiles but can't help thinking the onyx initial channel ink limiting is a bit aggressive for dye-sub inksets which don't really conform to standard CMYK values the software looks for. thus producing a compressed gamut

I am interested in your approach and also have an i1. Made excellent frontlit profiles for years with it. Do you find it a challenge to keep prints nuetral with your workflow?
Back to top Find Quote
Offline Mikulas Posted 10-23-2015 - 07:20 AM
Post: #5
Banned
55 Posts
Banned

Onyx Version:
10.2
Printer(s):
L26500, W8400+Onyx / Mimaki JV4, JV5 +TexPrint

Hi Mark,
we don´t use Onyx for textiles, we use Texprint. But I know Onyx a bit.

...but can't help thinking the onyx initial channel ink limiting is a bit aggressive for dye-sub inksets...
As far as I understood - Onyx ink-limiting presets are a general recommendation only. Nothing special but a set of hidden parameters how Onyx calculates/cuts the "reasonable" top end of CMYK channels. Instead of evaluating density (the old method) it does look at the saturation. It´s trying to cut more or less near to the saturation hook according to what media is used - better media =closer to the hook, weak media a bit earlier. With non-standard inks it´s not so clear to decide, and everyone should keep an eye on the ink-limiting graph and double-check. Wink
In dye-sub we always had a problem with strange blue-cyan. At the reasonable density limit it´s still not enough saturated - saturation hook is a quite long bow, and it is not easy to find the best limiting point.

...thus producing a compressed gamut
We push the ink limit relatively high, and we get a very dark primary cyan... Sad M and Y are significantly better but still far from normal inks - curves are very bowed. Even after linearisation the ICC gamut graph is "very round", looking like compressed (as you say). But this is not caused by low ink limits, but by the ink itself.

Do you find it a challenge to keep prints neutral with your workflow?
Neutral like neutral grey ? Hmmm ... What is a neutral grey ? LOL
I have measured most of our textiles and found out that they heavily use optical brighteners... So the white point is far from neutral (blue cast). Then I have measured 100%K and found out that it has a strong red cast. If the grey axis is not grey, but a blueish-white to redish-black... Eek! it is very problematic to get neutral grey! ConfusedCry I have tried some tricks, but none of them is ultimate. My best until now is tweaking linearisation and ICC measurements in Excell, and smoothing them out. But it cannot solve the transition between the cool white and the neutral light grey tones.



And I have some question again:

What is my current problem - I found out that our software is not able to generate an ICC with proper gamut mapping transforms. There are strange twisted irregularities in some out-of-gamut color areas. Especially deep blues which should normally be very nice are producing strange hue shifts. Sad Sad
I´m examining a way how to import my measurements into Onyx, and generate the ICC from there. Have you ever tried this ?

My second... I need to persuade my boss to upgrade from i1 Standard Profiler to i1 Publish. But I need some heavy arguments. It seems to me that the patch reading (i1 Standard) is a way more consistent than everything we have here (less errors and less erratic readings). And with it´s "double-scan" mode much quicker.
Have you ever tried the i1 Publish on textile? Reading or ICC profile generation?

And then ... after years of testing I still have a feeling that this "simple" spectrometers have some inherent problem with textile reading. As if the readings are darker and/or less saturated. And going to shadows contrary is true - it is lighter than observed by eyes.
Have you ever compared some readings between i1 and the Barbieri ? Is it giving equal Lab values in shadows/highlights/neutrals/saturated colors ?
(This post was last modified: 10-23-2015 07:55 AM by Mikulas.)
Back to top Find Quote
Offline MarkPrints Posted 10-25-2015 - 08:28 PM
Post: #6
Member
9 Posts
Member

Onyx Version:
Thrive 12
Printer(s):
Arizona250GT,JV5320DS,EpsonFS7100/F7200,SP540V

Hi Mikulas,

Could your first problem be caused by muddy 3/4 tones caused by higher channel limits? I have always pulled my channels back to as close to pure CMYK as possible as any extra Ink is not adding any extra colour (of course I am questioning this now, thus the thread)
A ways back I tried to use onyx to profile our Vutek which is not driven out of ONYX. The problem as I recall was that the patch sets are mathmatically generated out of the software. No tiffs I could find in the guts of the software. Can your ONYX actually drive your printer?

As far as spectros If you are looking to upgrade wouldn't transmissive be desirable?
As far as side by side LAB comparisons, No I have never done them, I expect there will be differences (errors) from day to day, device to device but hopefully is averaged out by multiple readings and/or larger patch sets. And these differences should be very, very small in real terms

Our i1 only profiles monitors now. It acts as a dongle to the software and somewhere along the way the printer profiling section got greyed out!
By then we had Barbieri and X10 so didn't really worry about it. I used to use Spectral Vision Pro (comes with Vutek) for profile builds back then.
Back to top Find Quote
Offline Mikulas Posted 10-28-2015 - 05:46 PM
Post: #7
Banned
55 Posts
Banned

Onyx Version:
10.2
Printer(s):
L26500, W8400+Onyx / Mimaki JV4, JV5 +TexPrint

Could your first problem be caused by muddy 3/4 tones caused by higher channel limits? I have always pulled my channels back to as close to pure CMYK as possible as any extra Ink is not adding any extra colour (of course I am questioning this now, thus the thread)
This problem definitely lies in the profile generating engine - I have checked other profiles (ISO, Fogra, our Onyx profiles, ...) and they all give better mapping of out-of-gamut blues and violets.
And I also have set our ink limits quite "scientifically" - I have analyzed the 80% -100% measurements in Excell, and found out where to cut (saturation gain relatively compared to tone value increase). I definitely cannot imitate standard cyan with our inks. Our cyan is very "exotic", giving me max. saturation approx. at Lab 38/-1/-55. If I pull back ink limit to mimic ISO Coated 100%C (55/-37/-50) or something standard then I loose a huge part of saturation (54/-14/-40).

A ways back I tried to use onyx to profile our Vutek which is not driven out of ONYX. The problem as I recall was that the patch sets are mathmatically generated out of the software. No tiffs I could find in the guts of the software. Can your ONYX actually drive your printer?
No, for textile we use solely Texprint. I know the problem with "secret" patches very well. Also from Texprint. But after all that years I already found some tricks how to get them... Cool Before investing the effort I was curious whether it has a sense.

As far as spectros If you are looking to upgrade wouldn't transmissive be desirable?
I really don´t know. Backlight is a small part of our job. Until now we have satisfactory results with just RIP corrections and printing small trials (don´t know the right word for it). My boss will not invest 5000€+ until I´m totally sure it´s giving MUCH better reflective results on textile too.

I expect there will be differences (errors) from day to day, device to device but hopefully is averaged out by multiple readings and/or larger patch sets. And these differences should be very, very small in real terms. I´m not sure with this. My readings with i1Pro2 show clearly - on paper I get max. dE for the whole patch set not more than 3-4. On textile I sometimes get dE 30 or more. The whole reading is VERY erratic. I always need to do at least 3 readings and calculate averages. Wink
But I have read on more than one web page that people recommend more advanced spectros for textile. Confused
(This post was last modified: 10-28-2015 06:06 PM by Mikulas.)
Back to top Find Quote


Possibly Related Threads...
Thread: Author Replies: Views: Last Post
  Icc profile for dye sublimation jimmychongkc 2 1,178 10-27-2016 11:56 AM
Last Post: MaxGamut
  Backlit Light Source Question MarkPrints 5 1,271 01-22-2016 09:40 PM
Last Post: Swtor223
Question Textile Printing/Dye Sublimation how do you profile the best ???? Gunther 1 3,515 08-18-2011 08:23 AM
Last Post: robfoster
  Profiling Epson 9880 and Backlit Photoman 4 3,260 06-09-2010 08:26 AM
Last Post: Photoman


User(s) browsing this thread
1 Guest(s)
myRepono - Website & mySQL Database Backup Service
Community software by MyBB