Pages: [1]   Go Down

Author Topic: Determining optimal ink load  (Read 4250 times)

Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4237
    • waynefox.com
Determining optimal ink load
« on: February 23, 2012, 09:24:56 pm »

Mark mentioned in his recent article about increasing the ink load (other articles and discussions have mentioned this as well), then creating profiles to maximize the printers output.  I've done a few things in the past, but I don't recall anyone actually discussing exactly how they determine which is optimal, and how far is too far.  I'd be curious as to any advice or links to ideas on the best way to go about this.
Logged

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4344
    • A Goldhammer Photography
Re: Determining optimal ink load
« Reply #1 on: February 24, 2012, 08:05:23 am »

+1 to this suggestion.  I think I even mentioned this when the Camera to Print and Screen tutorial was in the planning stages (but it was not covered).  About the only thing that I've done in this respect to to experiment with the various Epson paper settings in the driver (3880) using Scott Martin's suggested approach.  With the fine art matte papers that I use (Canson Rag Photographique and Hahnemuhle Photorag Ultrasmooth) I cannot see any difference in the selections and use the Velvet Fine Art setting for these two papers.  It would be nice if one of the printing experts would write an article for LuLa on their approach to this as it would be helpful to the rest of us.

Alan
Logged

Scott Martin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1315
    • Onsight
Re: Determining optimal ink load
« Reply #2 on: February 24, 2012, 09:58:59 am »

Good question Wayne! One of the secrets to optimal RIP calibration is to determine the optimal ink limits by looking for the chroma (saturation) peak, not maximum density, as many assume. As we increase inkload on pigment printers, density and chroma climb together. But at one point the chroma peaks and then starts to decline as density continues to climb. For pigment printers, we get the best results if we set the ink limiting right at the beginning of this peak. Then perform per channel linearization, then combined linearization, then determining the combined total ink limit and, finally, profile. High quality papers will easily hold this ink and allow for a very high total ink limit. Of course lesser quality budget papers will have a problem with this much ink and you'll be forced to back off from these peaks by an overall percentage (often just ~5%) and that's life with cheapo papers. Solvent printers will have problems with this much ink so I back off by 10% as a rule of thumb for the sake of keeping the TIL high. For photographic imagery we must prioritize keeping the TIL high even if it means backing off on the per channel limiting a little. Anyway...

The challenge here is using software that allows us to visually graph out these things as we play with ink load. For years I've encouraged various developers to come up with a simple application that would do this for us. I know I'd pay for that but I'm sure I'm a small minority. :-] I wish Curves2 would do this. If one is really familiar with the Mac version of ColorBurst RIP it can be used for this purpose. I use CB to analyze these things when I'm calibrating any other RIP. I don't want to take the time to write an entire article on this but you get the point, and something tells me Wayne, that you already know how to do this.
Logged
Scott Martin
www.on-sight.com

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4344
    • A Goldhammer Photography
Re: Determining optimal ink load
« Reply #3 on: February 24, 2012, 11:42:50 am »

Scott,

Thanks for the explanation, that helps.  For my 3880 I can adjust the overall ink limit in the 'Paper Configurations Option' and if I'm reading you right then take a look at adjusting the three color channels in the 'Color Control' panel.  Is this correct or are you really talking more about a RIP beyond the Epson driver?

Alan
Logged

Scott Martin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1315
    • Onsight
Re: Determining optimal ink load
« Reply #4 on: February 24, 2012, 12:00:23 pm »

Thanks for the explanation, that helps.  For my 3880 I can adjust the overall ink limit in the 'Paper Configurations Option' and if I'm reading you right then take a look at adjusting the three color channels in the 'Color Control' panel.  Is this correct or are you really talking more about a RIP beyond the Epson driver?
Above I was referring to calculating the per channel ink load in a RIP.

If you're using the driver, todays drivers tend to have incredible built in calibration settings for matte, RC and Fiber base papers. There's not a whole lot you can do to improve upon what's built in. I would never adjust the color control but, by all means, play with the Color Density slider in the Epson driver and the Special 1-10 media settings in the Canon driver. Feel free to use my media evaluation image for this http://www.on-sight.com/2008/04/04/how-to-determine-the-optimal-media-selection-for-any-paper/ or simply make profiles with various settings and compare the results.

Again there's not a whole lot one can do with a driver and increasing density beyond a certain point will lesson saturation.
Logged
Scott Martin
www.on-sight.com

Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4237
    • waynefox.com
Re: Determining optimal ink load
« Reply #5 on: February 24, 2012, 01:39:05 pm »

Thanks Scott.  I also just use the Epson drivers.  My concern is the manufacturers of non-Epson papers really don't test this at all, they just pick a media setting they think should work, and crank out a profile. 

The test file looks very helpful .. just what I was looking for.

wayne
Logged

TylerB

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 446
    • my photography
Re: Determining optimal ink load
« Reply #6 on: February 24, 2012, 04:12:08 pm »

Good question Wayne! One of the secrets to optimal RIP calibration is to determine the optimal ink limits by looking for the chroma (saturation) peak, not maximum density, as many assume. As we increase inkload on pigment printers, density and chroma climb together. But at one point the chroma peaks and then starts to decline as density continues to climb. For pigment printers, we get the best results if we set the ink limiting right at the beginning of this peak...

I gotta say Scott, just for argument's sake, this has not proven the case for me at all. I've played with this a lot for years. I know some recommend this and it no doubt results in well behaived conversions. But I find this approach to over limit, always at the expense of some gamut. Even though Chroma may fall off again for that ink, there are some dark hues you are not going to reach with an ink limited at the chroma peaks and then other color and K being utilized to reach some of the denser hues that ink may have supplied unleashed a bit more. Also, my Epsons so far, the yellow chroma never shoulders over and yet yellow is the worst contributor to bleed in high percentages with other inks, so requires a different approach.
I wish there was a rule of thumb to hit the sweet spot but I have found none. Generally chroma limited profiles sacrifice gamut, in my experience. ALso, the hooking is always dealt with nicely these days with good profile conversions.
My 2 cents, and of little use to most here.. probably totally irrelevant to those working in the RGB drivers.
TYler
Logged

Scott Martin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1315
    • Onsight
Re: Determining optimal ink load
« Reply #7 on: February 24, 2012, 05:12:03 pm »

But I find this approach to over limit, always at the expense of some gamut. Even though Chroma may fall off again for that ink, there are some dark hues you are not going to reach with an ink limited at the chroma peaks and then other color and K being utilized to reach some of the denser hues that ink may have supplied unleashed a bit more.

I'm not sure I understand what you're saying here. If each channel is limited to it's maximum saturation (which is close to it's max density BTW) then it can be *mixed* with the other inks to get even darker hues.

Also, my Epsons so far, the yellow chroma never shoulders over and yet yellow is the worst contributor to bleed in high percentages with other inks, so requires a different approach.

I'm not seeing bleeding in the yellow channel. Is it properly controlled with separation parameters? So what do you do?  

Generally chroma limited profiles sacrifice gamut, in my experience.

What??? Chroma peaked per channel ink limiting is part of the secret to gamut maximization! Where are you coming from? Give us something constructive here. Instead of just saying 'nah that doesn't work' tell us what does. How are you monitoring chroma? I'd love to have some intelligent dialog here.

And you're right, Wayne is working with an RGB driver as it turns out, so this RIP talk is off topic.
« Last Edit: February 24, 2012, 05:20:26 pm by Onsight »
Logged
Scott Martin
www.on-sight.com

TylerB

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 446
    • my photography
Re: Determining optimal ink load
« Reply #8 on: February 24, 2012, 06:20:02 pm »

...
And you're right, Wayne is working with an RGB driver as it turns out, so this RIP talk is off topic.

my apologies, I did not start that discussion about chroma limiting. No doubt best to leave it there.
Tyler
Logged

Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4237
    • waynefox.com
Re: Determining optimal ink load
« Reply #9 on: February 24, 2012, 09:50:21 pm »

And you're right, Wayne is working with an RGB driver as it turns out, so this RIP talk is off topic.
Not at all.  Many that use RIPs are focusing on workflow, and often just use profiles provided by the RIP maker. 

I am all for keeping posts on topic, something which rarely happens on the internet and thus waste a great deal of human time (wonder what value that might add up to ... and here I go getting off topic :) ) So while I am asking about a way to make sure I can get the most out of the standard driver, especially when using non-epson papers, the information is quite informative, and certainly applies.  I've used RIP's in the past, but my current workflow is a few prints a week, single prints on fine art paper, so the RIP didn't seem worth the effort. Maybe I'll take a look at this aspect.  I know Mark mentioned this about RIPS in his article a while back.

So this brings up another question.  A big part of the output quality has to do with the dither/screen of the driver/printer.  I assume a RIP uses it's own algorithms, and Epson for one has touted the substantial increase in quality of this part of their driver.  Any thoughts on this... might you lose a little of this using a RIP instead of the manufacturers dither/screen?
Logged

Scott Martin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1315
    • Onsight
Re: Determining optimal ink load
« Reply #10 on: February 24, 2012, 11:38:30 pm »

Sorry if I took everyone down the RIP ink limiting rabbithole...

So this brings up another question.  A big part of the output quality has to do with the dither/screen of the driver/printer.  I assume a RIP uses it's own algorithms, and Epson for one has touted the substantial increase in quality of this part of their driver.  Any thoughts on this... might you lose a little of this using a RIP instead of the manufacturers dither/screen?

Yes, screening can have a really significant impact. 12 years ago some RIPs actually had better screening than printer drivers but the tables quickly turned. Some RIP's have an OEM agreement and utilize a printer's screening while other RIPs use their own. I don't want to name names but I prefer to stay away from some RIPs because their screening isn't particularly impressive by todays standards...

The bottom line is that inkjet drivers offer really incredible quality these days and you have to really know what you're doing to get a RIP to look as good. For workflows like yours Wayne I think it's smart to keep it simple. Making sure you've got top notch profiles is often a better place to put one's effort.

Of course some printers don't even have driver and must use a RIP. And some workflows need RIP-only features like nesting, large length support, tiling, etc. ImageNest is a cool, hybrid RIP in that it uses a printer's driver to send images to the printer.
Logged
Scott Martin
www.on-sight.com
Pages: [1]   Go Up