How to test if a print was made at 16 bit?

[faberryman]

Do you understand what a deltaE difference is before I explain further?

Thesse were the quotes I was inquiring about:

This indicates to me, the 16-bit check box is a marketing created option in the driver. Of course if anyone finds a difference in their testing, I'd love to know. I didn't try 8-bit per color with the check box on, seems pointless. And it makes sense that the driver wouldn’t need the check box as unseen on the P800. Not super happy Epson has some drivers with and some without the option or that the option even exists based on what I see. Which again is why I think the check box is almost certainly a marketing generated 'feature' and nothing more. 

It didn't do a damn thing on this end. ON, or OFF, with 16-bit data, I get the SAME results. I also get the same results in the P800 which has NO such options.
You think that if you take 16-bit data and turn the check box off, something converts it to 8-bits per color? That's absolutely not what I see testing with a 3880 on this end. And why assume it does? Convert the data to 8 bits per color THEN make a print. There's zero question it's 8 bits per color now. I see the difference too, on screen which I reported for those who read what I reported.

If you're asking me, I turned it off WITH 16-bit data to see if it did anything with actual 16-bit data. It did absolutely nothing. It's not an option on my other printer and appears to be unnecessary; 16-bit data printed as expected with no such option.

[digitaldog]

Thesse were the quotes I was inquiring about:

Do you understand how Doug's target was created and what it doesn't show? Hence the question you didn't answer about deltaE.

So, in honor of Bill Atkinson, I have created an image I call Bills Black Ball. The reason that Bill's set of color gradient balls were created was because Bill felt that banding was easier to see in circles rather than linear gradients.
The attached tiff image, in ProPhoto RGB, has RGB values ranging from RGB(20,20,20) to RGB(35,35,35) at the outer edges. It is quite suitable for demonstrating 8 bit or higher paths for both monitors and printers.

It was a great target to examine banding, it's not at all good, nor designed to show visually or numerically anything about RGB values outside of 20/20/20 to 35/35/35/. The test I provided yesterday used colors all over color space. And thanks to 3-4 people here (Doug, Mark, Elliot), after viewing Doug's work, a bug in the driver for what appears to be only Mojave, released September 24, 2018 (not that I upgraded on that day, I did not) was found.
Nothing therefore tested solely with Doug's target over the last day on my end, on Mojave can now be accepted although in terms of banding, I still see no differences ON DOUGs Target in terms of banding!
RGB 23/23/23 THEN run through my output profile and measured show a dE 2000 of 2. Do you or do you not understand what that metric indicates in terms of the difference between the 16-bit check box being on or off for that one target? IF SO, you'll see that prior to the  bug found, my statement that I see no difference, prior to 24 hour dry down was that there was NO visual difference with either on or off.
You asked repeatedly about should the check box be on or off. IF THER IS NO BUG, it would and should make no difference with 16-bit data hence. It's ALWAYS on for me and probably why, I didn't see any issues with the output from Mojave in the few weeks it's been installed on this Mac. Until this set of posts which if anything, shows that there's a nastily bug under Mojave for the 3880 yet NOT for my other printer that gets more use; my P800. No check box, no bug.

In terms of the results of my testing, WITH A DRIVER bug, I stand by everything I wrote. But now some of us know, the analysis is based on a buggy driver.
Now, once I finally get a driver that supports the check box toggling without a bug, I'll revisit Doug's target again.

[Doug Gray]

For anyone wishing to use the image for other purposes, it is in Adobe RGB (1998). The CC's LAB values, were from measurements on a 3 y/o ColorChecker with an I1 Pro2 so there will be small differences from the CC images at places like BabelColor.

Also, there is a useful neutral gradient I attached to the side. It runs from white to black then back again. The steps are highly precise with the L* stepping from 1 to 99 then back to 1 in increments of 1. It's a good check on the ability of a printer to print good neutrals. For those with an iSis, if you save a measurement file in CGATs, the Lab readings of the last 198 entries (out of 957 total) will show the measured LAB values in sequence. Color vision is quite sensitive to small chroma and hue shifts on the neutrals so I often use this part of the image to check a printer profile.

[faberryman]

Understandable.
 Do you understand what a deltaE difference is before I explain further?

Yes, and that the deltaE differences you are seeing are clearly visible.

[digitaldog]

Yes, and that the deltaE differences you are seeing are large.

No, not with Doug's target.

[digitaldog]

The bug is invoked with the check box ON not off. Just printed my Printer Test File both ways. With it ON, it almost looks like the output profile wasn't used (dark, very ugly).
I converted this test file to 8-bit per color, but had the check box ON (which seems pointless and counter intuitive). Wanted to see if the bit depth was an issue or just the check box. Output was wrong. So it doesn't matter the document bit depth, it's the check box under Mojave that's the bug.

[faberryman]

No, not with Doug's target.

I was referring to the differences you report in post 206.

[digitaldog]

I was referring to the differences you report in post 205.

And I was referring to your questions in post #224:

Thesse were the quotes I was inquiring about:

[Mark D Segal]

The bug is invoked with the check box ON not off. Just printed my Printer Test File both ways. With it ON, it almost looks like the output profile wasn't used (dark, very ugly).
I converted this test file to 8-bit per color, but had the check box ON (which seems pointless and counter intuitive). Wanted to see if the bit depth was an issue or just the check box. Output was wrong. So it doesn't matter the document bit depth, it's the check box under Mojave that's the bug.

I suppose this applies to the 3880 (by now a deprecated model) and not the P800, because you were saying said option doesn't exist in the P800 driver. Nonetheless I'm sure many people are still happily using their 3880s, so it should be fixed. Now who owns this problem: Adobe, Apple or Epson?

[digitaldog]

I suppose this applies to the 3880 (by now a deprecated model) and not the P800, because you were saying said option doesn't exist in the P800 driver. Nonetheless I'm sure many people are still happily using their 3880s, so it should be fixed. Now who owns this problem: Adobe, Apple or Epson?

P800 is fine. And there was a new driver for it that I installed after Mojave. Epson owns the problem IMHO thus far. I'm still waiting to hear back if there's a newer driver but their web site still shows this:

[Alan Goldhammer]

I suppose this applies to the 3880 (by now a deprecated model) and not the P800, because you were saying said option doesn't exist in the P800 driver. Nonetheless I'm sure many people are still happily using their 3880s, so it should be fixed. Now who owns this problem: Adobe, Apple or Epson?

I wonder whether Apple give hardware manufacturers enough lead time to test everything.  Of course the 3880 is an old model but it's Epson's fault for making it such a rock steady printer that hardly ever (never in my experience) clogs.  With all the installed Epson printers, it would be a tedious job to check every driver and setting.  I've not looked at the manufacturing specifications for the P800 but in the absence of the checkbox, does it print 16 bit? 

[digitaldog]

I wonder whether Apple give hardware manufacturers enough lead time to test everything.

Considering how they allow the public to test beta's of the OS, I'd have to believe the answer is they do.
Yes, it appears the P800 simply passes the bit depth in the document through the driver, based on the results I see on mine with Doug's test file. And that's HOW it should be. The check box is dumb and unnecessarily even when no bug exists. It's an extra and unnecessary item in the GUI and again, I think it's more for marketing than anything else (I'm sticking by my original opinion). You have 16-bit data, send to the printer and it should use it. Have 8-bit per color data, ditto. Have 16-bit but a HUGE file that's going to slow you down or make the driver (?) barf, convert to 8-bits per color.
What we have now is multiple Epson's with and without the check box. That's just confusing even without one driver producing a bug that affects the output.

[Mark D Segal]

I wonder whether Apple give hardware manufacturers enough lead time to test everything.  Of course the 3880 is an old model but it's Epson's fault for making it such a rock steady printer that hardly ever (never in my experience) clogs.  With all the installed Epson printers, it would be a tedious job to check every driver and setting.  I've not looked at the manufacturing specifications for the P800 but in the absence of the checkbox, does it print 16 bit?

I second Andrew's comments. My understanding from talking with various developers is that Apple gives them at least six months advance notice to come into compliance with OS upgrades before they issue them.

As for 16-bit printing in a P800 - inconceivable that it wouldn't.

As for why that 16-bit checkbox even exists for some driver versions and not others, I think one would have to ask Epson and hope they would tell. I'm not convinced it's marketing. This is too obscure an item on which to hang one's marketing hat. It's of immediate appeal to a small crew of techno-literati, but hardly to the larger constituency forming the prosumer market for these 17" printer models. They are much better off touting brilliance of colour, fine detail reproduction, media handling features and that kind of stuff. I think the explanation is more innocent - somewhere in the minds of some engineers there is a rationale that impressed management enough to include this "feature". So those minds need to be tweaked to find out what's the charm! :-) I suspect I may have an idea of this answer. They may well say that 16-bit is the default which should remain checked for all 16-bit image files; however, just in case you need to send a smaller amount of data through the print pipeline, you have the option to revert to 8-bit by turning it off, and the driver dumbs down the file - just in passing for that one print - from 16 to 8. Other than that, I think Andrew's explanation of how it should work is the most generally sensible.

[Alan Goldhammer]

I second Andrew's comments. My understanding from talking with various developers is that Apple gives them at least six months advance notice to come into compliance with OS upgrades before they issue them.

That's appropriate but what if Apple makes a last minute change to some part of the new OS which breaks the work the developer has done.  This is not unheard of and we are seeing the same thing now with Windows 10 which MSFT is treating just like IOS updates as they do not plan to issue a 'new' Windows product.  I've avoided all that because I'm staying on Win 8.1 until support runs out or some needed app no longer runs.

As for 16-bit printing in a P800 - inconceivable that it wouldn't.

As for why that 16-bit checkbox even exists for some driver versions and not others, I think one would have to ask Epson and hope they would tell. I'm not convinced it's marketing. This is too obscure an item on which to hang one's marketing hat. It's of immediate appeal to a small crew of techno-literati, but hardly to the larger constituency forming the prosumer market for these 17" printer models. They are much better off touting brilliance of colour, fine detail reproduction, media handling features and that kind of stuff. I think the explanation is more innocent - somewhere in the minds of some engineers there is a rationale that impressed management enough to include this "feature". So those minds need to be tweaked to find out what's the charm! :-) I suspect I may have an idea of this answer. They may well say that 16-bit is the default which should remain checked for all 16-bit image files; however, just in case you need to send a smaller amount of data through the print pipeline, you have the option to revert to 8-bit by turning it off, and the driver dumbs down the file - just in passing for that one print - from 16 to 8. Other than that, I think Andrew's explanation of how it should work is the most generally sensible.

I agree with both you and Andrew.  I've not had to deal with this as my 3880 runs under WinOS and hence the only option is for 8 bit printing.  Epson engineers may have figured out how to have the driver detect 8 or 16 bit files and for the newer printers the checkbox is not needed. If you have MacOS do you still have to check the 16 bit output box in LR or is that a thing of the past (just asking for informational purposes).

[Mark D Segal]

That's appropriate but what if Apple makes a last minute change to some part of the new OS which breaks the work the developer has done.  This is not unheard of and we are seeing the same thing now with Windows 10 which MSFT is treating just like IOS updates as they do not plan to issue a 'new' Windows product.  I've avoided all that because I'm staying on Win 8.1 until support runs out or some needed app no longer runs.
I agree with both you and Andrew.  I've not had to deal with this as my 3880 runs under WinOS and hence the only option is for 8 bit printing.  Epson engineers may have figured out how to have the driver detect 8 or 16 bit files and for the newer printers the checkbox is not needed. If you have MacOS do you still have to check the 16 bit output box in LR or is that a thing of the past (just asking for informational purposes).

The P5000 is a newer, more expensive model than P800 and its driver does provide that checkbox. And yes, Lr for OSX does provide a 16-bit option checkbox in the Print module.

[Alan Goldhammer]

The P5000 is a newer, more expensive model than P800 and its driver does provide that checkbox. And yes, Lr for OSX does provide a 16-bit option checkbox in the Print module.

Thanks.  I've done some research about WinOS printing and it looks as though MSFT provided 16 bit print capability back when Vista was developed well over ten years ago now.  The specification is XPS and drivers need to be compliant with that.  It looks like only Canon has done this but the problem as I've read is that Adobe has not put the feature in LR to do this with WinOS (and I assume PS as well).  It's interesting that the ability to do 16 bit printing in WinOS has been around for so long and never implemented until recently and only by on manufacturer.  Some RIPs can do 16 bit printing in WinOS.  Curious situation.

[Mark D Segal]

Indeed, that is curious. One wonders why. Roughly half the graphic arts market these days uses Windows from what I've heard.

[Alan Goldhammer]

Indeed, that is curious. One wonders why. Roughly half the graphic arts market these days uses Windows from what I've heard.

Perhaps the driver emulation is complicated and printer manufacturers decided that 8 bits was enough for 'most' everyone.  I'm surprised that Epson have not provided 16 bit drivers for their new line of printers but of course LR would have to allow 16 bit delivery to the printer under WinOS as well.

[faberryman]

Fortunately, I am going to dodge the bullet. I am running High Sierra on my mid-2011 machine and so am not able to upgrade to Mojave, but I am able to upgrade to LR8/PS20.

[digitaldog]

Perhaps the driver emulation is complicated and printer manufacturers decided that 8 bits was enough for 'most' everyone.  I'm surprised that Epson have not provided 16 bit drivers for their new line of printers but of course LR would have to allow 16 bit delivery to the printer under WinOS as well.

LR appears to do so on the Mac, according to the output I just tried using Doug's target. Smooth as it should be. This again begs the question why a check box for 16-bit isn't necessary for the P800 but appears on others.
Like Photoshop, there's a 16-bit check box in LR. Is there one under the Windows version?