How good is the Z3100 spectro?

[jhein]

Recently I got access to the Z3100 Software Developers Kit.  I am using it to develop some interesting apps that hopefully will be of interest to all Z2100/3100/6100 users.

While developing my apps, I discovered what I thought was a bug but turned out to be a "feature" in the Embedded I1 spectro aka ESP.

Here is what I sent to the HP developer support:

Basically the embedded spectro in the Z3100 is advertised as having a range of 400nm to 700nm in 20nm increments for a total of 16 measurements.  It is also advertised as using a UV cut filter.  What I am finding out is that what I really get are readings from 420nm to 700nm.  400nm readings are always 0.  This is either a bug or a feature depending on your point of view.  Ultraviolet light is generally considered light with a wavelength less than 400nm.  So depending on the strength of the UV filter in the spectro, not having a reading at 400nm could be normal.  But if it is, then all of HP's documentation is wrong!  The specs should say 420nm to 700nm in 20nm increments for a total of 15 measurements.   The reason all of this is important is because to generate accurate color profiles, the profiling package needs to know if the paper being profiled is using Optical Brightening Agents (OBA's).  If OBA's are present then the profiling software may need the information at 400nm to properly compensate for them.

Here is there response:

The integrated UV cut filter (or actually, bandwidth limited illumination)
eliminates all signal up to and including at 400 nm, which is why that value
is always set to zero. The reason for this is that 400-700nm is a standard
spectral reporting range, whereas 420-700 would not be. We agree that this
can be confusing, but once you know what to expect it shouldn’t be a
problem. Regarding OBAs, there is no way to detect them with ESP, exactly
because of the integrated filter (or illumination).

They also sent me a wonderful PDF  that explains in great detail how the ESP compares with other spectros.  It also goes into detail some of the design limitations that had to be overcome.

Bottom line is the spectro in the Z series is very good indeed.

Enjoy
Jim

[attachment=6026:attachment]

[Colorwave]

Thanks for forwarding this, Jim.  
It doesn't really seem to be directed at color scientists or developers, so I wonder why one needs to ask about this subject to have HP provide this document?  One place (hey, similar to the Z3100 wiki!) for all of HP's technical documents related to the printer would be helpful.
-Ron H.

[Ernst Dinkla]

The integrated UV cut filter (or actually, bandwidth limited illumination)
eliminates all signal up to and including at 400 nm, which is why that value
is always set to zero. The reason for this is that 400-700nm is a standard
spectral reporting range, whereas 420-700 would not be. We agree that this
can be confusing, but once you know what to expect it shouldn’t be a
problem. Regarding OBAs, there is no way to detect them with ESP, exactly
because of the integrated filter (or illumination).


[attachment=6026:attachment]
[{POST_SNAPBACK}][/a]

There's an essential difference between filtering the light source on UV illumination versus blocking the sensor for UV reflected light when OBA and non-OBA papers are measured-profiled. The bump of more whiteness still happens with the last concept as the UV is already translated to longer spectral waves that should reach the sensor despite the UV block at 400 Nm.

Whether any UV filtering is the better concept is still in debate. For art prints the consensus is no OBAs, no UV light when displayed, framed behind glass (normally cutting at 350 Nm) so the Z3100 meter suits that approach.

[a href=\"http://www.pusztaiphoto.com/articles/printing/spectrums/webchart.shtm]http://www.pusztaiphoto.com/articles/print...s/webchart.shtm[/url]

the Moab Entrada Bright White paper may not suit that approach.


Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/

[neil snape]

It's good that they sent you this (these PDFs). You might want to try to find the white papers too.
Not only do OBAs influence the readings regardless of the intended filtration, but also that of super reflectance if you like. I can't remember the exact terms but between the substrate and certain inks themselves can reflect more light than is emitted.

The main reason why HP decided to do it the way they did was to comply with ISO 13655:1996 standards that to date no one can comply with> eliminating UV while maintaining correlation to D50 spectral power distribution.
There is an alternative: to use a second array of UV emitting LED and different band width LEDs to have with and without UV readings but that would go so far beyond the needs of the average user.

What makes this interesting is HP's research into this and making the Z so easy to use, goes beyond belief! All I can say is Bravo x 1000 for HP engineers.
Even if there are some concerns with image quality on certain substrates no one can fault the excellence of the ESP, and the application of it in use which is one of the most forward thinking innovations in LFP devices in recent history.

And keep at the project Jim , as marrying Argyll with the Z will make without a doubt the best profiling package in a printer bar none.

[Ernst Dinkla]

And keep at the project Jim , as marrying Argyll with the Z will make without a doubt the best profiling package in a printer bar none.

[{POST_SNAPBACK}][/a]

That's good news. Neil, I should have known that a week ago :-) Anyway it probably will take some time.

I have asked Graeme Gill to add the SpectroCam to the supported spectrometers a long time ago but as far as I know it never happened.

Is there a to do list where QTR profile creation + targets could be included in Argyll's support of the Z3100 ?


Ernst Dinkla

Try: [a href=\"http://groups.yahoo.com/group/Wide_Inkjet_Printers/]http://groups.yahoo.com/group/Wide_Inkjet_Printers/[/url]

[neil snape]

That's good news. Neil, I should have known that a week ago :-) Anyway it probably will take some time.

I have asked Graeme Gill to add the SpectroCam to the supported spectrometers a long time ago but as far as I know it never happened.

Is there a to do list where QTR profile creation + targets could be included in Argyll's support of the Z3100 ?
Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/
[a href=\"index.php?act=findpost&pid=188388\"][{POST_SNAPBACK}][/a]

I have no clue as to how much time it would take to write the code to use the ESP and an external application. So far , EFI did and GMG too.
The integration with the Z and APS is outstanding. It's just that the profile calculation using Logo PM 6 library isn't as smooth as Argyll nor Monaco. There are a lot of us using PRofile maker 5 still so it's not that bad, just the Z would benefit from smoother profiles.