Pages: [1] 2   Go Down

Author Topic: PS curves for Epson ABW printing without the use of ICC profiles.  (Read 13196 times)

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575

Madmanchan peaked my curiosity about his idea of using curves profiles instead of ICC profiles when printing in ABW mode. He hinted of sharing his method to the world but has not to my knowledge. His website has no contact info, and he has not posted to LULA in over a year. Does anyone know how to do what he speaks of?

Madmanchan:

"A more flexible way to build black and white profiles for the ABW driver is to measure a grayscale stepwedge target and then use the measurements to build output curves in Photoshop (instead of using an ICC profile). This method also allows you to soft proof the image (including color toning). One advantage of this approach is that it's easy to make small tweaks to the curves to fine-tune the results (e.g., open up or compress shadow detail). It's like having a simplified profile editor right in Photoshop.
I've written a custom software program to generate Photoshop curves automatically from a set of LAB measurements and am considering releasing it to the public."

https://people.csail.mit.edu/ericchan/dp/Epson3800/faq.html#abw_custom_profiles
« Last Edit: July 19, 2017, 08:11:49 pm by texshooter »
Logged

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #1 on: July 19, 2017, 10:23:57 pm »

Nevermind for now. Upon further reading, I found  Madmanchan's instructions on how to build .acv gray curves from scratch. It looks complicated with a bunch of quantum physics equations, but i hope after a looksie I can learn how to import a data table into PS so to generate a curve profile.  He wrote a program to convert the .acv curve profile into an ICC profile, but it appears he's kept that one under lock and key. I like his idea of .acv gray curves because it allows one to tweek it to taste, something I doubt you can do to a standard Colormunki ICC profile without the aid of software or coding skills.

https://people.csail.mit.edu/ericchan/dp/Epson3800/bwcurves.html
Logged

richardboutwell

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 56
    • Black and White Mastery
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #2 on: July 20, 2017, 04:36:42 am »

I was working on something over the weekend for digital negative curves that is similar to what I think you're after.

A few years ago I put out a free correction curve tool for the gray_curve setting in QuadTone RIP ink descriptor files, but was never able to make .acv curves from a text file. I tried looking into it further, but it was easier to edit the quad curves directly than making Photoshop adjustment curves layers from the spreadsheet tools. Then over the weekend I remembered Photoshop Color Lookup Tables, and it was easy enough to figure out how to get my correction calculations into the .cube format.

I worked it out for digital negatives a few days ago, and just now tweaked it for straight inkjet prints. After a few quick tests printing with the P800 (using ABW) from Photoshop I can say that it works pretty dang well (if I do say so myself...)  I put up a quick webpage and download option. Now, it isn't as good as making Piezography curves from scratch, but for people with a colormunki or densitometer wanting to do a little better customizing ABW then it should be a step up from the canned media settings.

Here is a page on my site with a little more info an option to download it: https://www.bwmastery.com/lut-printer-correction-curves
Logged
Personal Site — http://www.richardboutwell.com
Black and White Aesthetics, Editing, and Printing — http://www.BWMastery.com

Eric Myrvaagnes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 22814
  • http://myrvaagnes.com
    • http://myrvaagnes.com
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #3 on: July 20, 2017, 09:07:57 am »

I was working on something over the weekend for digital negative curves that is similar to what I think you're after.

A few years ago I put out a free correction curve tool for the gray_curve setting in QuadTone RIP ink descriptor files, but was never able to make .acv curves from a text file. I tried looking into it further, but it was easier to edit the quad curves directly than making Photoshop adjustment curves layers from the spreadsheet tools. Then over the weekend I remembered Photoshop Color Lookup Tables, and it was easy enough to figure out how to get my correction calculations into the .cube format.

I worked it out for digital negatives a few days ago, and just now tweaked it for straight inkjet prints. After a few quick tests printing with the P800 (using ABW) from Photoshop I can say that it works pretty dang well (if I do say so myself...)  I put up a quick webpage and download option. Now, it isn't as good as making Piezography curves from scratch, but for people with a colormunki or densitometer wanting to do a little better customizing ABW then it should be a step up from the canned media settings.

Here is a page on my site with a little more info an option to download it: https://www.bwmastery.com/lut-printer-correction-curves
Thank you, Richard! That looks quite promising. I have just downloaded it to give it a try, since most of my printing is ABW black and white. With many years of darkroom printing behind me, I have tended to push the sliders in LightRoom for each image until it looks the way I would expect a darkroom print to look. Your system should make that much more reliable. I often do one quick jump from LR into PS, so once my curves are set up this won't really add any time to the process.

Eric
Logged
-Eric Myrvaagnes (visit my website: http://myrvaagnes.com)

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4344
    • A Goldhammer Photography
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #4 on: July 20, 2017, 10:22:26 am »

If you are running Windows OS you can still use icc profiles to soft proof and print from.  QTR allows you to make these profiles easily enough and then use a normal workflow through light room.

Alan
Logged

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #5 on: July 20, 2017, 06:26:22 pm »

If you are running Windows OS you can still use icc profiles to soft proof and print from.  QTR allows you to make these profiles easily enough and then use a normal workflow through light room.

Alan

I had tried Quadtone RIP several years ago thinking that it would improve dark shadow separation of my ABW prints.   The result of my experiment was not what I had expected. The before and after prints of a 21-step wedge showed no change. Every patch on the step wedge showed differentiation except the three darkest patches, with or without the Quadtone ICC profile. Identical patches mean loss of shadow detail.  My conclusions were that

1) The Epson's ABW driver is already so well linearized that it can't be improved upon. I wasn't aware that Quadtone RIP was designed for either custom inks or more primitive printers than the Stylus Pro 3800.

2) I gave up hope that any Epson  printer/paper combination could be capable of delineating the last three patches of a 21-step gray wedge and so moved on with my life. If a glossy RC paper can't resolve that shadow detail then nothing can, short of piezography or perhaps Epson's new Ultrachrome Pro inkset.

Several years later I read about customizing Photoshop curves adjustments to do what ICC profiles can't do and that is to force some light into those blocked up shadows by distorting the linearized curve to suit the eye rather than a logarithmic equation. I'm skeptical it can work because I expect to end up with all gray shadows. But I guess muddy gray shadows are better than pure black shadows because then I could at least dodge and burn with some artistry a little bit of edge detail back into those deepest of blocked shadows.  Although the information is there in the digital negative, it's getting it to show up on the printed page is the hard part. If I were printing to screen there would be no problem. And I'm not ready to make the piezography leap yet.   One last college try.

« Last Edit: July 20, 2017, 06:30:00 pm by texshooter »
Logged

Ferp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #6 on: July 20, 2017, 08:46:38 pm »

If you are running Windows OS you can still use icc profiles to soft proof and print from.  QTR allows you to make these profiles easily enough and then use a normal workflow through light room.

I've seen this statement on Lula often enough that it's time that I point out that it can be done on OS X as well. If you print using Roy Harrington's Print Tool software then you can print to ABW using an ICC created via QTR.  This is different to what used to be possible back in the day, when Eric Chan's ICCs could be used directly in ABW, but the principle is the same. 

Of course if you had the equipment to make such an ICC you'd be more likely to print via QTR rather than ABW, or at least be well advised to, but for those OS X users who want to print with ABW using an ICC, yes you can.
Logged

Ferp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #7 on: July 20, 2017, 09:12:55 pm »

I had tried Quadtone RIP several years ago thinking that it would improve dark shadow separation of my ABW prints.   The result of my experiment was not what I had expected. The before and after prints of a 21-step wedge showed no change. Every patch on the step wedge showed differentiation except the three darkest patches, with or without the Quadtone ICC profile. Identical patches mean loss of shadow detail.  My conclusions were that

1) The Epson's ABW driver is already so well linearized that it can't be improved upon. I wasn't aware that Quadtone RIP was designed for either custom inks or more primitive printers than the Stylus Pro 3800.

2) I gave up hope that any Epson printer/paper combination could be capable of delineating the last three patches of a 21-step gray wedge and so moved on with my life. If a glossy RC paper can't resolve that shadow detail then nothing can, short of piezography or perhaps Epson's new Ultrachrome Pro inkset.

Something doesn't sound right to me here.  I've never found ABW to be anything like remotely linear, although I haven't bothered to try every paper on every printer using every paper type in the ABW settings. 

Putting that aside, how were you using QTR to try to improve the shadow separation of your ABW prints?  Were you printing via QTR or were you just converting to a QTR-generated ICC and then printing via ABW?  The ICCs that the QTR utilities generate are based on a perceptual rendering intent, and much as happens when printing color with this setting, this will tend to depress shadow details.  Converting to a QTR-generated ICC is not what you'd want to do if you want to retain and enhance shadow detail, although the effect of this is less pronounced on PK papers than MK papers.

I'd suggest printing your images in either gray gamma 2.2 or AdobeRGB via QTR (from Print Tool if on OS X, to avoid Apple and Adobe silent color profile conversions).  I'd also suggest that matte papers will be better at preserving shadow detail than papers that use PK.  Those two in combination will give you quite a different look and you may initially find your shadows too weak, so you'll need to adapt, but it's an easier starting point that trying to recover shadows.
« Last Edit: July 20, 2017, 09:16:21 pm by Ferp »
Logged

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #8 on: July 21, 2017, 12:41:33 am »

Here is a page on my site with a little more info an option to download it: https://www.bwmastery.com/lut-printer-correction-curves

Thank you for the link to your curve tool.   Correct me if I'm wrong.  The curve generated by your tool should be applied to an adjustment layer AFTER I've finished my editing and not before?   Does your tool convert the gray curve into an ICC profile For soft proofing?
Logged

Ferp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #9 on: July 22, 2017, 07:26:01 pm »

Suggestions to change your workflow and give QTR another try don't seem to have been well received.  How about using those QTR-generated ICCs in another way?  You can use them in Photoshop to soft-proof using the preserve numbers setting in the soft-proofing dialog box, which will allow you to soft-proof how the ABW print will look when printed without any color management by Photoshop or the ABW driver, and will enable you to edit the image to preserve and enhance shadow details.
Logged

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #10 on: July 22, 2017, 08:32:55 pm »

Suggestions to change your workflow and give QTR another try don't seem to have been well received.  How about using those QTR-generated ICCs in another way?  You can use them in Photoshop to soft-proof using the preserve numbers setting in the soft-proofing dialog box, which will allow you to soft-proof how the ABW print will look when printed without any color management by Photoshop or the ABW driver, and will enable you to edit the image to preserve and enhance shadow details.

Interesting. Let me read up on the preserve numbers setting option (I'm not familiar).   My gut tells me that optimal results would be achieved by letting the  ABW  driver do the managing. (As a side note, I would never print B&W on matte paper--Defeats the purpose I think, so using matte paper to open my shadows is a bit desperate in my case.)

  I was also brainstorming the idea of making a contrast curve in PS which will show how my image ought to look when printed. I would stick that curve on an adjustment layer before doing any edits.  Then when I was happy with how my edits  looked on screen, I would delete that curve layer. Then the last step would be to add a different curve layer, namely the one created using Eric Chan's method or Richard Boutwell's tool.  In short, the first curve is to soft proof with and the second curve is to print with. 
« Last Edit: July 23, 2017, 09:20:49 am by texshooter »
Logged

Ferp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #11 on: July 23, 2017, 07:02:48 pm »

Interesting. Let me read up on the preserve numbers setting option (I'm not familiar).   My gut tells me that optimal results would be achieved by letting the  ABW  driver do the managing.

I don't agree about "optimal results", but putting that aside, that's exactly what is happening if you select that option - ABW is doing the managing.  It is a bit confusing the first time you notice it, because it's not something you would use for printing color, as it gives odd results. 

Under normal (color) soft-proofing, you are seeing a simulation of what the print should look like if you print as normal using the relevant ICC.  Preserve numbers does the opposite - it shows what the print should look like if you print without any color management.  Which is what you do when you print using ABW and the printer manages colors setting in the print dialog. 

Of course this assumes that your monitor is calibrated and profiled.

My experience is that preserve numbers is simpler and safer (i.e. less risk of an error) than mucking around with curves in the manner you describe.  YMMV.


(As a side note, I would never print B&W on matte paper--Defeats the purpose I think, so using matte paper to open my shadows is a bit desperate in my case.)

These things are a matter of taste, and of course depend on the image and also the framing, but there's a heck of a lot of good B&W work on matte.
Logged

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #12 on: July 23, 2017, 08:25:22 pm »


Under normal (color) soft-proofing, you are seeing a simulation of what the print should look like if you print as normal using the relevant ICC.  Preserve numbers does the opposite - it shows what the print should look like if you print without any color management.  Which is what you do when you print using ABW and the printer manages colors setting in the print dialog. 

Sounds promising.  It's like soft proofing but in reverse.  I wonder how it measures up to Keith Cooper's workflow.

KeithCooper ABW Workflow

To summarize Keith's ABW workflow (I hope I get this right): 


Step 1)  CONVERT TO PROFILE
               (Source Space = the color working space used to do final edits to the image, e.g., aRGB, sRGB
                Destinatiin Space = The ICC profile created by QTR.)
STEP 2) ASSIGN PROFILE
               (Working space selection = same as Source Space in Step 1)
STEP 3)  Apply curves and other edits to improve upon the QTR ICC
STEP 4) Send job to ABW driver.  No need to print from QTR driver  and no need to print with ICC profiles.

I understand that Keith is working on a product review of the new Epson P5000 printer. If you're reading this Keith, I hope you go into great detail on how to "best" linearize ABW profiles. (Because ICC profiles never get the deep shadows or midtones to look exactly the way they should.)  If I were to buy this printer, it would be primarily for its new gray-expanded  inkset . So knowing the best test-proven B&W workflow would be equally as usefull as knowing the usual printer stuff. You've got the paper and Ink, let's have the scoop.
Logged

keithcooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 414
    • Northlight Images
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #13 on: July 24, 2017, 08:22:55 am »

Sounds promising.  It's like soft proofing but in reverse.  I wonder how it measures up to Keith Cooper's workflow.

KeithCooper ABW Workflow

To summarize Keith's ABW workflow (I hope I get this right): 


Step 1)  CONVERT TO PROFILE
               (Source Space = the color working space used to do final edits to the image, e.g., aRGB, sRGB
                Destinatiin Space = The ICC profile created by QTR.)
STEP 2) ASSIGN PROFILE
               (Working space selection = same as Source Space in Step 1)
STEP 3)  Apply curves and other edits to improve upon the QTR ICC
STEP 4) Send job to ABW driver.  No need to print from QTR driver  and no need to print with ICC profiles.

I understand that Keith is working on a product review of the new Epson P5000 printer. If you're reading this Keith, I hope you go into great detail on how to "best" linearize ABW profiles. (Because ICC profiles never get the deep shadows or midtones to look exactly the way they should.)  If I were to buy this printer, it would be primarily for its new gray-expanded  inkset . So knowing the best test-proven B&W workflow would be equally as usefull as knowing the usual printer stuff. You've got the paper and Ink, let's have the scoop.

The article you mention was how I printed with my old Canon iPF 8300 and the Canon BW mode (so not quite the same as ABW)

Although I had the P7000 here last year I didn't get a chance to really delve into the B&W the way I am with this P5000 (since I know I've got it here for a while)

First up, I'm using my own colour profiles and visually I'm finding the handling of shadows and mid-tones just fine.

With matte papers I'm finding that the linearity of both the ABW and ICC (relcol+BPC) based approaches (measuring the outputted prints) is somewhat off from my ideal linear starting point.

Neither are bad and could work fine for a lot of prints (heresy I know, but I don't use soft proofing much for colour yet alone B&W :-)

See the tables/curves I posted in the 'best working space' thread http://forum.luminous-landscape.com/index.php?topic=113548.0

I will admit to being slightly surprised that a simple bit of linearisation could produce such good results from 'just' profiling - perhaps those nearly 3k patch colour targets I use are useful after all... ;-)

Now to try a few more papers.

I'm visiting Epson UK HQ on Thursday and B&W printing is one of the areas I'm talking to them about. I'm also looking at a 60" P20000 printer so will be taking some very big image files to test with it ;-)
Logged

Ferp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #14 on: July 24, 2017, 09:40:39 am »

Sounds promising.  It's like soft proofing but in reverse.  I wonder how it measures up to Keith Cooper's workflow.

I see that Keith has replied, but he hasn't really addressed this question.  I think they're doing different things.

You want to be able to see on your monitor what you're going to get when you print using ABW under printer manages colors, right?  So that you can edit the image with confidence.  Well what I suggested above will do precisely that for you.  Keith's workflow won't.  It seems to be a way to use the inverse of the ICC to create a linear workflow.  Linearity is not the same as print to monitor matching.  Why not give his workflow a try and see if does or not?  I think you'll find that even if you adopt his workflow, you'll still need to follow my suggestion to get your monitor to accurately predict the print.

[That said, I found Keith's article confusing.  He suggests that you convert to the generated ICC and then assign back to the original ICC profile you intend to print with, gray gamma 2.2, in order to achieve linearity.  My experience is the opposite, you want to assign to the ICC and convert back to gray gamma 2.2.  His workflow would darken the deep shadows not lighten them.

This assumes that when he says "The program produces a specialised ICC profile that I can use to correct this", the ICC is coming from the QTR-Create-ICC-RGB application.  If that's not the case then I'm not sure where it's coming from.  I prefer a simple workflow, and Keith's seems unnecessarily complex.]
Logged

keithcooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 414
    • Northlight Images
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #15 on: July 24, 2017, 02:17:42 pm »

I see that Keith has replied, but he hasn't really addressed this question.  I think they're doing different things.

You want to be able to see on your monitor what you're going to get when you print using ABW under printer manages colors, right?  So that you can edit the image with confidence.  Well what I suggested above will do precisely that for you.  Keith's workflow won't.  It seems to be a way to use the inverse of the ICC to create a linear workflow.  Linearity is not the same as print to monitor matching.  Why not give his workflow a try and see if does or not?  I think you'll find that even if you adopt his workflow, you'll still need to follow my suggestion to get your monitor to accurately predict the print.

[That said, I found Keith's article confusing.  He suggests that you convert to the generated ICC and then assign back to the original ICC profile you intend to print with, gray gamma 2.2, in order to achieve linearity.  My experience is the opposite, you want to assign to the ICC and convert back to gray gamma 2.2.  His workflow would darken the deep shadows not lighten them.

This assumes that when he says "The program produces a specialised ICC profile that I can use to correct this", the ICC is coming from the QTR-Create-ICC-RGB application.  If that's not the case then I'm not sure where it's coming from.  I prefer a simple workflow, and Keith's seems unnecessarily complex.]
Thanks very much for reading this through - I just went back to the original profile and images in the article and used the same profile. You are quite right - using it the way round I said makes the darks darker - not what was actually shown in the pictures

I've realised that in the process of updating/re-writing the whole site last year, I seem to have got my assigns and converts the wrong way round - I have a vague inkling as to why this happened but need to fix it!

As to the complexity of the method - it is only so as to print with a linear output - soft-proofing has no real place in this workflow (a deliberate choice from my POV - YMMV)

I need to go back through the articles and make sure that my bout of confusion didn't spread - this is one of the ever present pitfalls of not having anyone else to proof-read stuff :-(
Logged

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #16 on: July 24, 2017, 06:47:10 pm »

So which of the two soft proofing methods discussed is better?

A) Select Preserve Numbers box, or

B) Assign to Profile then Convert to Profile

I would say (B) because it's less editing work to make your Assigned copy look like your master copy than it is to make your completely unmanaged copy (numbers preserved) look like your master copy. (The master copy being the goal or target version.) The Preserve Numbers trick gives you a farther distance to climb, while the Assign/Convert trick gives you a head start.
Logged

Ferp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #17 on: July 24, 2017, 08:14:49 pm »

First a response to Keith.  Thanks for your follow-up, as you had me questioning my sanity when I read your workflow and tried to understand and replicate it. 

A couple of other points.  I couldn't see a lot of change in your images when I moused over them, other than the check boxes and other settings, so it was hard to follow what you were trying to achieve.  This may say something about my browser.

There are a lot of profile conversions in your workflow, and several of them are unnecessary.  If you use the QTR-Create-ICC application, rather than the QTR-Create-ICC-RGB one, this will create a monochrome ICC.  You can then simply assign to the ICC, and convert back to gray gamma 2.2.  This is a lot simpler.  I'm not convinced that this is going to result in a linear workflow, but I'd need to do some measurements to be certain.  It may.

So which of the two soft proofing methods discussed is better?

A) Select Preserve Numbers box, or

B) Assign to Profile then Convert to Profile

I would say (B) because it's less editing work to make your Assigned copy look like your master copy than it is to make your completely unmanaged copy (numbers preserved) look like your master copy. (The master copy being the goal or target version.) The Preserve Numbers trick gives you a farther distance to climb, while the Assign/Convert trick gives you a head start.

There are similarities and differences.  To repeat, if you are printing to ABW using printer manages colors, then a preserve numbers soft-proof will be a reliable one, as it reflects what you are actually doing.

Taking my amended version of Keith's workflow, if the image is in gray gamma 2.2 and you assign it to the (monochrome) ICC, then that is much the same thing.  It's a soft-proof of your workflow.  Photoshop is showing the image as it should appear if printed as-is without any color management.  If you look at the monitor in most cases the image will appear different (not much different for certain PK papers), but the histogram will be unchanged.  What you are doing by assigning the image to the ICC is instructing Photoshop to interpret the numbers differently. 

Open the image and duplicate it and preserve numbers soft proof one and assign the other and they should appear much the same.  The main difference to soft-proofing is that there isn't the option to simulate the black ink and paper color.  If you leave those options unchecked the two images should be identical or nearly so.  If you want those options used to proof your assigned image, you'll have to do the soft-proof as well.

So far so good.  If you continue to follow the amended Keith workflow and now convert it back to gray gamma 2.2, the appearance on the monitor won't change (you might lose a color tint), but the histogram will change, in most cases.  This is not soft-proofing any more, you've changed the image.  You can see this from the change in the histogram. The reason that the monitor doesn't change is that you have also instructed Photoshop to interpret those numbers in the file differently, so as to leave the appearance on the monitor unchanged.

This is an image editing technique, not a soft-proofing one.  Because you've changed the data in the image, it's going to print differently using your workflow, which makes no allowance for the image's ICC.   If you want to get an accurate soft-proof you're going to to have to use the preserve numbers soft proof.  You may find this image editing technique useful, because in a lot of cases it's going to lighten the shadows - as you will see from the change in the histogram  How much will depend on the image and paper.  But to simulate the impact in the print, and assess what further edits if any are required, you'll have to do the preserve numbers soft proof, because you are printing without reference to the ICC.

(The various platine papers, genuine and rebadged, are one exception where you won't see this effect very much.)

Everything I've said has been with reference to printing via ABW and printer manages colors, but the exact same logic and procedures apply to printing via QTR without using the ICC to print.   I prefer the QTR prints.  And there's more scope to tweak.
Logged

keithcooper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 414
    • Northlight Images
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #18 on: July 25, 2017, 12:48:27 pm »

First a response to Keith.  Thanks for your follow-up, as you had me questioning my sanity when I read your workflow and tried to understand and replicate it. 
A couple of other points.  I couldn't see a lot of change in your images when I moused over them, other than the check boxes and other settings, so it was hard to follow what you were trying to achieve.

Thanks again for spotting it! I've updated the text and the images, which hopefully show a bit more of what's going on.  I'm still at a loss to explain the point where I went wrong in the article - especially since the example print (on my old iPF8300) worked so well. I've checked and corrected the other few times I referenced this too.

One reason may be that I've so rarely had to use the QTR profiles with my recent printers for actual serious printing (other than tests for reviews) Canon and Epson B&W modes have improved and newer printers are just doing so much better out of the box

With the P5000 I'm finding that relatively minor corrections are sufficing - I've a huge pile of test prints and without the notes on them it's getting difficult to spot any problem areas between them.
What started as a desire to check finer details of paper/ink/printer results has led me to a major look at just how I go about B&W printing...
Logged

Ferp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
Re: PS curves for Epson ABW printing without the use of ICC profiles.
« Reply #19 on: July 25, 2017, 08:52:22 pm »

I'm still a little surprised by the difference between before and after adjustments, having looked at them now more closely in Photoshop, but there's a lot of printer and paper variation as to how this technique works.

One reason may be that I've so rarely had to use the QTR profiles with my recent printers for actual serious printing (other than tests for reviews) Canon and Epson B&W modes have improved and newer printers are just doing so much better out of the box

Perhaps that's a factor.  My printers are an older generation and perhaps they've been overtaken by recent advances in B&W modes.
Logged
Pages: [1] 2   Go Up