Digital BW, The Print

The 2400 profiles are on the loose! 
http://forums.dpreview.com/forums/read.asp?forum=1003&message=13547405

Don't get too excited, they have been edited to give very large gamut 
in a profile viewer. Looks like you will need custom profiles right 
out of the gate again!

And B&W Mode is Same as Source anyway.

I have been corresponding with Bruce Fraser and he and Jeff are looking at
doing softproof profiles when they get a printer.  They will likely do for
each paper one profile for each of the main settings (neutral, cool, warm
and sepia for each of darkest, darker, dark, normal, light) so 20 profiles
per paper.  One can already use EyeOne Match to do your own profiles
following Carl Schofield's workflow (if you had a printer!).  Bruce will be
using a lot higher patch count than 21 or 51.

I still think the more interesting and relevant issue is profiling the
luminance axis.  This k3 situation is just like QTR a year or so ago (and
where black-only is today).  We found a way to soft proof but still did not
have a way of managing the difference between the workspace luminance
profile to the printer reflectance profile. The old 'why do my QTR prints
print "light" and "lack contrast" issue?'.

Epson has loaded into the driver five transforms - darkest, darker, dark,
normal, and light - but these must only be simple transforms keyed off the
paper selection.  Thus they take no account of changes in workspace nor
individual printer differences in linearity.  They certainly won't
accommodate a change to the inks.  Think of them as Black-only work flow
with five preset curves to choose from.  The smarter money will figure out
if there is a way to manage the luminance/reflectance compression with a
(greyscale) ICC output profile a la Roy's generics for QTR.  The only issue
I can see is in how Epson applies the transforms. Specifically, if they are
simply always applied to incoming file data then data already adjusted by a
ICC profile conversion that was intended to alter the file data to reflect
the print space would then still be transformed again. Then you're stuck
with what Epson has hard coded and it will become progressively out of date.

Steve

> From: Ernst Dinkla <E.Dinkla@...>
> Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Date: Thu, 19 May 2005 11:41:48 +0200
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Subject: Re: [Digital BW] 2400 profiles are on the loose!
> 
> dfaprinting wrote:
> 
>> The 2400 profiles are on the loose!
>> http://forums.dpreview.com/forums/read.asp?forum=1003&message=13547405
>> 
>> Don't get too excited, they have been edited to give very large gamut
>> in a profile viewer. Looks like you will need custom profiles right
>> out of the gate again!
>> 
>>  
>> 
> It is far too early to judge the printers by their profiles. The drivers
> show changes that may well explain why Epson decided to extend the soft
> proof gamut. There's an AdobeRGB mode in the 2400 and 4800 drivers
> (R1800 had that as a first) next to the sRGB mode that was the default
> of the older drivers. This influences the gamut possible. The gamut of
> the new Epson will differ from the R1800 as the LM and LC inks deliver
> more gamut in the light color mixes (better subtractive mixing, less
> paper white) while the R1800 has its gamut gain at the blue and red
> regions of the gamut. The CM inks of the R1800 were more diluted
> compared to the CM inks of the CcMmYKk Ultrachrome printers according to
> Robert Krawitz. A compromise between not using LC and LM and what the
> finer droplets could produce in color mixing at the highlights.
> Depending on the inklimit possible in the R1800 that could mean there's
> some gamut loss at the full C-M  hues. There's also a hue shift in the
> new 4800 CM Ultrachrome inks compared to the old C-M UC inks according
> to Joseph Holmes.
> 
> I think you need a very good 3D profile viewer to see where the shapes
> of the three printer generations differ. Not to speak of the differences
> between the soft proof part of the profiles and the actual printing
> gamuts and what is shown in a 3D viewer.
> 
> I doubt Epson is returning to inflated profiles again after their
> excellent Atkinson etc profiles for the UC printers.
> 
> How to get the new drivers is explained in the first message of that
> link above.
> 
> Ernst
>

dfaprinting wrote:

>The 2400 profiles are on the loose! 
>http://forums.dpreview.com/forums/read.asp?forum=1003&message=13547405
>
>Don't get too excited, they have been edited to give very large gamut 
>in a profile viewer. Looks like you will need custom profiles right 
>out of the gate again!
>
>  
>
It is far too early to judge the printers by their profiles. The drivers 
show changes that may well explain why Epson decided to extend the soft 
proof gamut. There's an AdobeRGB mode in the 2400 and 4800 drivers 
(R1800 had that as a first) next to the sRGB mode that was the default 
of the older drivers. This influences the gamut possible. The gamut of 
the new Epson will differ from the R1800 as the LM and LC inks deliver 
more gamut in the light color mixes (better subtractive mixing, less 
paper white) while the R1800 has its gamut gain at the blue and red 
regions of the gamut. The CM inks of the R1800 were more diluted 
compared to the CM inks of the CcMmYKk Ultrachrome printers according to 
Robert Krawitz. A compromise between not using LC and LM and what the 
finer droplets could produce in color mixing at the highlights. 
Depending on the inklimit possible in the R1800 that could mean there's 
some gamut loss at the full C-M  hues. There's also a hue shift in the 
new 4800 CM Ultrachrome inks compared to the old C-M UC inks according 
to Joseph Holmes.

I think you need a very good 3D profile viewer to see where the shapes 
of the three printer generations differ. Not to speak of the differences 
between the soft proof part of the profiles and the actual printing 
gamuts and what is shown in a 3D viewer.

I doubt Epson is returning to inflated profiles again after their 
excellent Atkinson etc profiles for the UC printers.

How to get the new drivers is explained in the first message of that 
link above.

Ernst

BTW if anyone is interested in making a more accurate soft proof for their
B&W work I used Colorlab to make a 256 step test chart and corresponding
reference file which Eye One Match can use (Carl's method).  The EyeOne
scans the chart easily and quickly.  Let me know if this is something that
should be posted in the files section of the group.

In playing with this I noticed something which puzzled me.  When I look at
the test chart tiff in PS it is obviously the data I told it to be: equal
RGB steps (Lab a*=b*=0).  When I assign the profile generated by Eye One
Match the a* and b* of the patches alter (looking in the info palette),
reflecting the colour in the inks used to print the test chart, but for all
steps R=G=B still.  I guess it is because I assigned the profile thereby not
altering the RGB file info but "giving it meaning" with the profile.  That
meaning is reflected in  the Lab info.  That is, for example, patch RGB
80,80,80 with the profile means Lab 39,4,-3.

> From: Ernst Dinkla <E.Dinkla@...>

 
> Yes, the B&W driver may be less sophisticated than I figured out it
> could be.
> 
> BTW I'm not familiar with the R1800 driver but in the 4800 driver there
> are two B&W modes, Advanced B&W and what they simply call Black. It
> would be very inappropriate if the last isn't Black Only printing. 180
> nozzles at 3.5 picoliter and possibly 8 pass like the 10000 firmware
> upgrade had must deliver a nice BO print. Clayton are you there ?
> 
> I do not see the same option in the 2400 driver. Nor am I able to get to
> Advanced B&W in that driver but I have little time right now.
> 
> Ernst
>

Steve Kale wrote:

>And B&W Mode is Same as Source anyway.
>
>I have been corresponding with Bruce Fraser and he and Jeff are looking at
>doing softproof profiles when they get a printer.  They will likely do for
>each paper one profile for each of the main settings (neutral, cool, warm
>and sepia for each of darkest, darker, dark, normal, light) so 20 profiles
>per paper.  One can already use EyeOne Match to do your own profiles
>following Carl Schofield's workflow (if you had a printer!).  Bruce will be
>using a lot higher patch count than 21 or 51.
>
>I still think the more interesting and relevant issue is profiling the
>luminance axis.  This k3 situation is just like QTR a year or so ago (and
>where black-only is today).  We found a way to soft proof but still did not
>have a way of managing the difference between the workspace luminance
>profile to the printer reflectance profile. The old 'why do my QTR prints
>print "light" and "lack contrast" issue?'.
>
>Epson has loaded into the driver five transforms - darkest, darker, dark,
>normal, and light - but these must only be simple transforms keyed off the
>paper selection.  Thus they take no account of changes in workspace nor
>individual printer differences in linearity.  They certainly won't
>accommodate a change to the inks.  Think of them as Black-only work flow
>with five preset curves to choose from.  The smarter money will figure out
>if there is a way to manage the luminance/reflectance compression with a
>(greyscale) ICC output profile a la Roy's generics for QTR.  The only issue
>I can see is in how Epson applies the transforms. Specifically, if they are
>simply always applied to incoming file data then data already adjusted by a
>ICC profile conversion that was intended to alter the file data to reflect
>the print space would then still be transformed again. Then you're stuck
>with what Epson has hard coded and it will become progressively out of date.
>
>Steve
>  
>
Yes, the B&W driver may be less sophisticated than I figured out it 
could be.

BTW I'm not familiar with the R1800 driver but in the 4800 driver there 
are two B&W modes, Advanced B&W and what they simply call Black. It 
would be very inappropriate if the last isn't Black Only printing. 180 
nozzles at 3.5 picoliter and possibly 8 pass like the 10000 firmware 
upgrade had must deliver a nice BO print. Clayton are you there ? 

I do not see the same option in the 2400 driver. Nor am I able to get to 
Advanced B&W in that driver but I have little time right now.

Ernst

>>>>>>>>
BTW I'm not familiar with the R1800 driver but in the 4800 driver
there
are two B&W modes, Advanced B&W and what they simply call Black. It
would be very inappropriate if the last isn't Black Only printing. 180
nozzles at 3.5 picoliter and possibly 8 pass like the 10000 firmware
upgrade had must deliver a nice BO print. Clayton are you there ?
>>>>>>>>>

I'm here, waiting eagerly for what early users have to say.


Regards,
Clayton


Info on black and white digital printing at    
http://www.cjcom.net/digiprnarts.htm

They did indeed go back to the garbage profiles. I checked it in 4 
different viewers, including 2D and 3D views (where possible). Remember 
the LAB L of 3 for the gloss/luster papers that Epson reports, well the 
gamut view reports black as being ZERO. Too bad they went backwards. 
The latest profiles for the R1800 are at least believeable. And you are 
right, this is way to early to tell what is realy going on with these 
printers, I had just hoped for a more concrete performance spec from 
Epson.

On a side note, one person has tried the 4800 driver with the 4000. If 
I'm reading it correctly, the B/W driver is working, and using all 
three black inks. It might be as simple as exchanging the chips between 
the carts for the 4800 and the 4000 to use the new inks on your old 
printer. It also means that for a $500 RIP, you can do the same and 
more with the 4000. You can see this in the Epson forum at DPReview.

dfaprinting wrote:

>On a side note, one person has tried the 4800 driver with the 4000. If 
>I'm reading it correctly, the B/W driver is working, and using all 
>three black inks. It might be as simple as exchanging the chips between 
>the carts for the 4800 and the 4000 to use the new inks on your old 
>printer. It also means that for a $500 RIP, you can do the same and 
>more with the 4000. You can see this in the Epson forum at DPReview.
>
Isn't this in conflict with the information Roy gave about the 4000, 
that it is impossible to drive both black lines at the same time even 
with QTR that doesn't use the Epson driver ? 

Or could there be a new command in the printer language so far unused 
that opens that possibility ?

Ernst

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Ernst Dinkla 
<E.Dinkla@c...> wrote:
> Isn't this in conflict with the information Roy gave about the 4000, 
> that it is impossible to drive both black lines at the same time even 
> with QTR that doesn't use the Epson driver ? 
> 
> Or could there be a new command in the printer language so far unused 
> that opens that possibility ?
> 
> Ernst

I'm not really sure. I contacted the programmer for the RIP I've been 
using, and he said the same thing, it can be either matte or glossy. I 
wish I had one of these to try. I might also consider sending the 4800 
firmware to the 4000 to see what happens, but that is a bit risky, and 
could cause the printer to stop working, without a cheap fix. Other 
printers that use the same head seem to be able to print all 8 
channels, so I'm pretty sure the problem lies in the firmware. Also the 
4000 is capable of addressing all 8 heads individually, if it couldn't, 
you would not be able to run the nozzel check function. Maybe someone 
needs to contact a really good hardware programmer, and hack the 4000 
firmware. Since both the 4000 and the 4800 firmware are available, 
making the 4000 look and act like the 4800 shouldn't be too hard. At 
least then you could use the new driver, and the new inks. It might 
take something like this to force Epson into releasing an up date to do 
the same.

The thread is here for anyone that wants to keep track of this topic:
http://forums.dpreview.com/forums/read.asp?forum=1013&message=13551031

I'm wondering if the new driver is driving the black channel, and both 
inks are spraying at the same time, but not independently controllable, 
which would make a big mess.

It still looks like a solution for the 4000 is to just use better inks 
and a RIP. If time allows, I'll work on setting that up. Haven't had 
the time to try and find a discount for the RIP, but things at work are 
finally slowing down, so hopefully I'll get a chance to call. I think 
the better inks are already here, as I've had good luck with them in my 
9500, and they really shouldn't work very well with that printer. In an 
Ultrachrome printer they should work very well. A lot of that has to do 
with the dot gain, and these inks don't have much dot gain, so it's 
harder with the older printers. I did compare the new to the old last 
night, and the new ink does have a slightly "thicker" black ink as 
compared to Eboni. Dmax looked about the same visually, but I think the 
new has slightly better measured dmax. Mostly it just gives you less 
ink on the paper, so that can help when layering with colors (which are 
also much heavier pigment than other inks I've tried).