Digital BW, The Print

Hello,

First off, I wanted to thank Paul Roark for all the work he's done with getting this cheap printer working with the carbon inks!

Anyways, I just set up my Epson Workforce 1100 and EB4 inks but I've run into some problems.

I've tested some of the profiles Paul made with my cheap epson presentation matte paper to see if I can get something close, but I'm getting this weird problem when printing the 21 step wedge. Somewhere between step 7 and 8, the shade dips, i.e. darker, darker, lighter, darker, darker. 

I thought maybe it was just caused by paper so I decided to create my own profile using the aurora curve that Paul mentions as working with most papers. 

While following the instructions, I began getting more and more confused. I applied the curve to the PS raw file. Then I also applied the same curve to the step wedge. How does an RGB curve work with a grayscale file? How does it know which ink color to use? Also, this made the step wedge white from steps 1 through 10. Is this right? 

I wasn't sure what profile I was supposed to print with, but I chose Gray Gamma 2.2 (same as my working space). Is this right?

I used my spyder3print sr to take the linearization measurements. First I used steps 1-21, even though 1-10 were white.

I dropped the raw file and the exported txt file onto the qtr-icc-rgb file and it created the icc. It also gave me a warning in the text file that a soft proof could not be created because the rgb curves (raw file I'm assuming) werent in ascending order. The resulting profile gave me a printed step wedge of middle gray to black but spread out from 5% to 100% then white on 0%.

I also tried taking measurements of just steps 10 through 21 where step 10 was the last white step before it started turning gray. It resulted with a printed step wedge of something similar to Paul's aurora profile I just printed with a little dip in step 7. It also still doesn't look very linear. 

The last thing I tried was printing the step wedge with no curve applied using the gray gamma 2.2 as the printer profile. The printed wedge though not linear, did not have any dip in the gradation. I then took the measurements with the spyder. I dropped just the text file on the qtr-icc-rgb creator (no raw file used). The resulting profile gave me a nice linear step-wedge. The problem is that the middle grays look more coarse like it's using more of the K inks.

I thought maybe I received the EZ inks by mistake, but from looking at the coating ink leaves in each bottle, the shades look different.

Thank you in advance!!

"mcrib13" <edmondkim13@...> wrote:

>... I just set up my Epson Workforce 1100 and EB4 inks 
> but I've run into some problems.
> 
> I've tested some of the profiles Paul made with my cheap epson presentation matte paper ...


I've never tested or profiled that paper, but I've added a curve and ICC for Ultra Premium Presentation Matte ("UPPM") paper that might be close.

Go to http://www.paulroark.com/BW-Info/MIS-WF1100-Profiles.zip to download the new curve and profile.  

UPPM took a lower ink limit than the other papers.  I'm guessing the UPPM profile &/or curve will work better.

 
> ... I applied the curve to the PS raw file. Then I also applied the same curve to the step wedge.

To make your own ICC profiles, you will need to print the appropriate curve on both a 21-step test file (in Adobe RGB (1998)) as well as the QTR test file, saved as a raw file.

> How does an RGB curve work with a grayscale file?

Badly. That is the main problem you're having.

You need to convert the 21-step to an Adobe RGB file, then apply the curve.  Then when you print the 21-step, be sure the profile in the print preview is also Adobe RGB.  In older systems you can also use "no color management" in the preview.  CS5 will not allow that, however.


> I wasn't sure what profile ... same as my working space ...

The concept is correct.  But, use Adobe RGB (1998) for the 21-step when you make a profile.  

(For printing through a final ICC, you'll leave the files grayscale, and the ICC will do the converting to Adobe RGB.)


> I used my spyder3print sr to take the linearization measurements.

That is the unit I use and prefer.

 
> I dropped the raw file and the exported txt file onto the qtr-icc-rgb file and it created the icc.


Drop the 21-step Lab file and the qtr-icc-rgb image file (with *.acv curve on it and saved as a *.raw file) into QTR's Create ICC-RGB.  By the way, I usually re-name the qtr-icc-rgb file to the same as the text file -- aside from extension -- so that I don't mix up these files or save over the original qtr file.


> It also gave me a warning in the text file that a soft proof could not be created because the rgb curves (raw file I'm assuming) werent in ascending order.

To be compatible with soft proofing, one of the R, G, B curves must go from 255 to 0 with at least some minimal positive slope throughout the range.  This usually is not the case with partitioned RGB curves. 

> The last thing I tried was printing the step wedge with no curve applied using the gray gamma 2.2 as the printer profile. The printed wedge though not linear, did not have any dip in the gradation. I then took the measurements with the spyder. ...(no raw file used). The resulting profile gave me a nice linear step-wedge. The problem is that the middle grays look more coarse like it's using more of the K inks.

What you did there is make a non-partitioned profile.  Your main problem, above, was not converting the 21-step file to RGB before putting the curve on it.

There is likely another problem that you may not have noticed.  The dmax is probably reached too early -- before 100% input.  UPPM appears to hit the dmax at 91%.  The new partitioning curve in the Zip file, above, corrects this.

I hope this helps.

By the way, I'll be making more profiles for the 1100 Eboni-4 system.  I now have it installed in my 1100 and think it's an important platform to support.  Being able to the most lightfast B&W prints for the least amount of money seems like a significant niche for the B&W market -- and it's what I like.

Paul
www.PaulRoark.com

> "mcrib13" <edmondkim13@> wrote:
> 
> >... I just set up my Epson Workforce 1100 and EB4 inks ...



> 
> > It also gave me a warning in the text file that a soft proof could not be created because the rgb curves (raw file I'm assuming) werent in ascending order.


 <roark.paul@...> wrote:
> To be compatible with soft proofing, one of the R, G, B curves must go from 255 to 0 with at least some minimal positive slope throughout the range.  This usually is not the case with partitioned RGB curves. 

The Zip file now has both a "partitioned" ICC and curve for UPPM and also what I call a "sequential" ICC and curve.  What I do with these latter types of curves is start the inks sequentially, so that they are smoother than a straight grayscale, all ink firing equally, profile, but the curves have no negative slopes.  As such, the soft proofing works and they are also very tolerant of ink and paper batch differences.  The more fully partitioned curves and profiles are a bit smoother, but most will not see the difference in actual photos.  With many papers, including Ultra Premium Presentation Matte paper (UPPM), you can use a sequential curve for soft proofing even while you use more fully partitioned curve/icc for printing.  

Paul
www.PaulRoark.com

Thank you Paul for all your help.

> 
> I've never tested or profiled that paper, but I've added a curve and ICC for Ultra Premium Presentation Matte ("UPPM") paper that might be close.
> 
> Go to http://www.paulroark.com/BW-Info/MIS-WF1100-Profiles.zip to download the new curve and profile.  
> 
> UPPM took a lower ink limit than the other papers.  I'm guessing the UPPM profile &/or curve will work better.
> 
> 

I just tested the UPPM partition profile but I'm still getting that weird dip. The blacks stopped blocking up though like you said.

I converted the step wedge to adobe rgb 1998 and applied the uppm partition curve. That definitely helped alleviate my steps 1 through through 10 being white problem. Unfortunately, this printed curve also showed a dip. Undaunted, I took the measurements and tried creating a profile. The profile created a posterized step wedge. 

Oddly, the 1100-Eb4-PaSmBW-1-rgb profile gave me the most accurate prints, though the blacks blocked up more quickly. I think it's because the curve was the flattest?

I have some RR Aurora natural and RR premium matte double sided ordered, so I won't have to worry about this epson presentation matte anymore. The epson paper was much thinner than I expected, but the images were nice nonetheless. I keep worrying that I put the wrong dilution in the each ink tank.

 

> 
> By the way, I'll be making more profiles for the 1100 Eboni-4 system.  I now have it installed in my 1100 and think it's an important platform to support.  Being able to the most lightfast B&W prints for the least amount of money seems like a significant niche for the B&W market -- and it's what I like.
> 
> Paul
> www.PaulRoark.com
>

I love this printer so far. My first "good" epson printer was the epson 2200, and it was $700 I think. I thought the build quality of the 1100 would be worse, but it wasn't! It's amazing that it is so inexpensive. And let's not talk about my $35 CISS. This creates an amazing system. 

Thank you Paul for leading the way!

"mcrib13" <edmondkim13@...> wrote:
>
> 
> ... 
> 
> Oddly, the 1100-Eb4-PaSmBW-1-rgb profile gave me the most accurate prints, though the blacks blocked up more quickly. I think it's because the curve was the flattest?
> 

Try one of the  "sequential" curves sets.  It sounds like the curve needs to stop at 91 to hit the dmax and not overload the paper.

Profiles are paper specific.  So,  I try to, ultimately, set up profiles for  the most popular papers.  Then hopefully the underlying curves will be with linearization distance of being OK.



> I have some RR Aurora natural and RR premium matte double sided ordered, ...

Those will print about as warm as the Epson paper.  If you want more neutral, be sure to try the Premier Art Smooth BW and the Epson Hot Press  papers.

...
> 
> I love this printer so far. My first "good" epson printer was the epson 2200, and it was $700 I think. I thought the build quality of the 1100 would be worse, but it wasn't! It's amazing that it is so inexpensive.


I agree.  I really appreciate the high speed text printing.  I do  a fair amount of that, so the 1100 is my default office printer.  It's rather amazing that a cheap printer ($130 when on sale) can be both the fastest text printer of all the inkjets I've had, but also capable of outstanding 100% Eboni carbon pigment fine art prints.  

> And let's not talk about my $35 CISS. This creates an amazing system. 

With your CIS, you'll want to agitate the inks about once a week.  The ink in the carts gets agitated continuously.  The CIS tanks, however, do need weekly agitation to have the most even output.

Be sure you're getting a perfect nozzle check. 

Paul
www.PaulRoark.com

Thanks Paul!

The sequential curve worked beautifully with the epson presentation matte. No dip in the outputted curve, and final profile printed a nice step wedge with no blocked blacks.


> Try one of the  "sequential" curves sets.  It sounds like the curve needs to stop at 91 to hit the dmax and not overload the paper.
> 
> Profiles are paper specific.  So,  I try to, ultimately, set up profiles for  the most popular papers.  Then hopefully the underlying curves will be with linearization distance of being OK.
> 
> 
> 
> > I have some RR Aurora natural and RR premium matte double sided ordered, ...
> 
> Those will print about as warm as the Epson paper.  If you want more neutral, be sure to try the Premier Art Smooth BW and the Epson Hot Press  papers.
> 
> ...
> > 

You sure are right about the warmth of the epson presentation matte. Feels as warm the epson K and LK only prints from my 7600. It's amazing that paper can affect the neutrality of the inks.


> 
> I agree.  I really appreciate the high speed text printing.  I do  a fair amount of that, so the 1100 is my default office printer.  It's rather amazing that a cheap printer ($130 when on sale) can be both the fastest text printer of all the inkjets I've had, but also capable of outstanding 100% Eboni carbon pigment fine art prints.  
> 


It never occurred to me use this printer to do text printing!! I have an aging laser printer I can finally retire. More room saved!


> With your CIS, you'll want to agitate the inks about once a week.  The ink in the carts gets agitated continuously.  The CIS tanks, however, do need weekly agitation to have the most even output.
> 
>

I've been somewhat obsessive about the agitation of the cis tanks. I try to give it a gentle shake whenever I walk by the printer, so about 4 times a day. hah. Could there be a downside to doing that? I hope not.

It is g ood to hear that the "seqential" curve is flexible enough to handle the paper.  

With respect to agitation, I would avoid actions that might cause foam.

Paul
Www.paulroark.com 

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "mcrib13" <edmondkim13@...> wrote:

>
> Thanks Paul!
> 
> The sequential curve worked beautifully with the epson presentation matte. No dip in the outputted curve, and final profile printed a nice step wedge with no blocked blacks.
> 
> 
> > Try one of the  "sequential" curves sets.  It sounds like the curve needs to stop at 91 to hit the dmax and not overload the paper.
> > 
> > Profiles are paper specific.  So,  I try to, ultimately, set up profiles for  the most popular papers.  Then hopefully the underlying curves will be with linearization distance of being OK.
> > 
> > 
> > 
> > > I have some RR Aurora natural and RR premium matte double sided ordered, ...
> > 
> > Those will print about as warm as the Epson paper.  If you want more neutral, be sure to try the Premier Art Smooth BW and the Epson Hot Press  papers.
> > 
> > ...
> > > 
> 
> You sure are right about the warmth of the epson presentation matte. Feels as warm the epson K and LK only prints from my 7600. It's amazing that paper can affect the neutrality of the inks.
> 
> 
> > 
> > I agree.  I really appreciate the high speed text printing.  I do  a fair amount of that, so the 1100 is my default office printer.  It's rather amazing that a cheap printer ($130 when on sale) can be both the fastest text printer of all the inkjets I've had, but also capable of outstanding 100% Eboni carbon pigment fine art prints.  
> > 
> 
> 
> It never occurred to me use this printer to do text printing!! I have an aging laser printer I can finally retire. More room saved!
> 
> 
> > With your CIS, you'll want to agitate the inks about once a week.  The ink in the carts gets agitated continuously.  The CIS tanks, however, do need weekly agitation to have the most even output.
> > 
> >
> 
> I've been somewhat obsessive about the agitation of the cis tanks. I try to give it a gentle shake whenever I walk by the printer, so about 4 times a day. hah. Could there be a downside to doing that? I hope not.
>

"mcrib13" <edmondkim13@...> wrote:

> 
> ... and final profile printed a nice step wedge with no blocked blacks.


By the way, to find where the final points on the curve ought to be, use this test strip: http://www.paulroark.com/BW-Info/80-100.jpg

Print it with an RGB profile in the preview (or no color management if you have an older PS).  With a spectro of scanner you'll see where the ink limit is -- at the edge of where the density plateaus.

Paul
www.PaulRoark.com

Paul,

I just tested the Aurora Natural paper using the Aurora White profile and the printed step wedge had that strange dip again. I also tried printing the step wedge with the Aurora curve and it gave me no discernible gradation from steps 5 through 10.

I'm not sure what could be happening. 

When you print, which media type are you choosing? Ultra Premium Presentation Matte? 

>
> It is g ood to hear that the "seqential" curve is flexible enough to handle the paper.  
> 
> With respect to agitation, I would avoid actions that might cause foam.
>

"mcrib13" <edmondkim13@...> wrote:
>

> I just tested the Aurora Natural paper using the Aurora White profile and the printed step wedge had that strange dip again. ...


I just printed a 21-step test strip on Aurora-N with the Aurora-White ICC and had the same problem.  I then printed an Aurora-W test strip with the ICC and it looked fine.  It seemed weird that there would be such a difference, since the papers in the past could use the same profile.  Just to double check the results, I printed a second Aurora-Natural test strip with the Aurora White ICC and -- amazingly -- it now looks fine.

So, I'm not sure what is going on.  Does the 1100 need a warm-up print before it behaves like it usually does?  I'll see if I can isolate the variable.

Paul
www.PaulRoark.com

"mcrib13" <edmondkim13@...> wrote:

>...that strange dip again. ...


I put a new profile -- ICC as well as the curve -- in http://www.paulroark.com/BW-Info/MIS-WF1100-Profiles.zip

In examining my test strip that showed the type of error you describe, the test strip clearly used a different profile than the Aurora ICC.  So, it could have been a mistake I made, or somehow a different profile got into the workflow. 

Paul
www.PaulRoark.com

Hi Paul,

I just tested the new aurora-w profile with the aurora-n. The dip is less drastic, but I'm unfortunately still seeing it.

I have a feeling my dilutions might be off a little.

Thanks again for all the help.
 


--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Paul" <roark.paul@...> wrote:

>
> "mcrib13" <edmondkim13@> wrote:
> 
> >...that strange dip again. ...
> 
> 
> I put a new profile -- ICC as well as the curve -- in http://www.paulroark.com/BW-Info/MIS-WF1100-Profiles.zip
> 
> In examining my test strip that showed the type of error you describe, the test strip clearly used a different profile than the Aurora ICC.  So, it could have been a mistake I made, or somehow a different profile got into the workflow. 
> 
> Paul
> www.PaulRoark.com
>

>
> "mcrib13" <edmondkim13@> wrote:
> 
> > 
> > ... and final profile printed a nice step wedge with no blocked blacks.
> 
> 
> By the way, to find where the final points on the curve ought to be, use this test strip: http://www.paulroark.com/BW-Info/80-100.jpg
> 
> Print it with an RGB profile in the preview (or no color management if you have an older PS).  With a spectro of scanner you'll see where the ink limit is -- at the edge of where the density plateaus.
> 

I know I said I didn't get any blocked blacks, but I guess once I printed an actual photo of something, it's a little bit of a different story.  Using my linearized profile based on the sequential curves, I'm seeing posterization in the shadows.

Once I see where the ink limit is, how do I reduce it?

Thanks again Paul!

"mcrib13" <edmondkim13@...> wrote:
>
> ...
> I have a feeling my dilutions might be off a little.

We need a way to test that.  If we use a standardized paper and separate C, M, and Y curves(straight line Red, Green, and Blue curves) with a standard profile, like Adobe RGB (1998), we ought to be able to compare.  I'll send you information off list to coordinate this.

Paul
www.PaulRoark.com

"mcrib13" <edmondkim13@...> wrote:
>
> ...
> I know I said I didn't get any blocked blacks, but I guess once I printed an actual photo of something, it's a little bit of a different story.  Using my linearized profile based on the sequential curves, I'm seeing posterization in the shadows.
> 
> Once I see where the ink limit is, how do I reduce it?

Go into the curves and change the end point of the curves.  In the latest Aurora profile I used 93 as the end point.  That appeared to be the "edge of the plateau."  In one ICC that used the curve it initially looked like it was slightly posterized, but when it dried fully it was OK.  With Gray Gamma 2.2 95% is very close to 100%.

Paul
www.PaulRoark.com

"mcrib13" <edmondkim13@...> wrote:

> ... the printed step wedge had that strange dip again. ...

We appear to have identified a pattern that may be a bug in the way the 1100 driver and Photoshop react together.  I have not seen this with other printers; input from others would be most welcomed.

The "bug" (?) occurs when I switch from another printer to the 1100.  The profile selected in CS5's print preview for the 1100 does not appear to be used with the first print.  Subsequent prints are fine. 

The following procedure appears to avoid the apparent problem:  First, as usual, I hit File > Print to get to the CS5 print preview.  Second, I switch from the Epson 1400 to the 1100 in the CS5 "Printer" box.  Third, I got to "Print Settings" and set them as needed, hit OK and return to the CS5 print preview.  Fourth, I select the "Printer Profile" in the right hand box of the CS5 preview.  ("Photoshop Manages Colors" has been selected also.)  

The next step is where the differences seem to occur.  If I push the "Print" button the print appears to have the wrong profile.  If I make a second print, it's fine -- the correct profile is used.

If, rather than hitting "Print" the first time I push the "Done" button, and then go back and hit File > Print a second time and then simply hit "Print" in the preview, the correct profile is used.

I have not seen this with any printer aside from the 1100.  I'll see if CS4 does the same thing next.  

Paul
www.PaulRoark.com