Digital BW, The Print

So I'm new to this but am now at least knee deep in seeing what is 
possible w/ modern inkjets for b&w.  At the moment I have an Epson R280 
loaded with the 6 cartridge Eboni inkset from MIS.

I shoot two films, Ilford HP5 and Delta 3200 in both 35mm and 120mm. 
I'm quite happy w/ my film and processing methods.  Scans look great. 
Traditional wet prints look great.  Some of my new inkjet prints look 
great, others look very bad.

As a generalization it seems like in some circumstances and with some 
photos the printer is able to convincingly render the film grain and 
it's gradations smoothly and in other cases it can not.

Where I notice this most is usually with people's hair or beards.  They 
often appear very splotchy as if the printer is printing straight black 
or white and doesn't understand that there should be a gray transition 
in between.  These are mostly natural light candids on 35mm where grain 
is fairly visible.  I like grain.

What confuses me somewhat is that if I take the identical file and print 
it out with QTR in Black Only mode, QTR seems to handle these 
transitions in tone much more smoothly and hair is rendered much more 
convincingly as opposed to splotchy. I would just print with QTR but I 
do not like the brown look and I greatly prefer the overall tonality of 
the 6 cartridge setup.

So, if the above makes any sense, am I running into a limit of the 
technology or am I doing something wrong?  Why is QTR rendering hair 
much more convincingly using only one cartridge?  What should I be 
experimenting with?

David Murphy wrote:
> So I'm new to this but am now at least knee deep in seeing what is 
> possible w/ modern inkjets for b&w.  At the moment I have an Epson R280 
> loaded with the 6 cartridge Eboni inkset from MIS.
> 
> I shoot two films, Ilford HP5 and Delta 3200 in both 35mm and 120mm. 
> I'm quite happy w/ my film and processing methods.  Scans look great. 
> Traditional wet prints look great.  Some of my new inkjet prints look 
> great, others look very bad.
> 
> As a generalization it seems like in some circumstances and with some 
> photos the printer is able to convincingly render the film grain and 
> it's gradations smoothly and in other cases it can not.
> 
> Where I notice this most is usually with people's hair or beards.  They 
> often appear very splotchy as if the printer is printing straight black 
> or white and doesn't understand that there should be a gray transition 
> in between.  These are mostly natural light candids on 35mm where grain 
> is fairly visible.  I like grain.
> 
> What confuses me somewhat is that if I take the identical file and print 
> it out with QTR in Black Only mode, QTR seems to handle these 
> transitions in tone much more smoothly and hair is rendered much more 
> convincingly as opposed to splotchy. I would just print with QTR but I 
> do not like the brown look and I greatly prefer the overall tonality of 
> the 6 cartridge setup.
> 
> So, if the above makes any sense, am I running into a limit of the 
> technology or am I doing something wrong?  Why is QTR rendering hair 
> much more convincingly using only one cartridge?  What should I be 
> experimenting with?


In your case it is more  likely the better tonal 
distribution through QTR than what is possible (or harder to 
achieve) with the normal driver. But it could as well be one 
of the issues below.

At some stages in the process aliasing can occur. At the 
scanning phase the actual grain can more or less accumulate 
to larger digital grain. Grain effect (and noise in digital 
images) can also increase when downsampling is done without 
proper anti-aliasing filters. If the software you print from 
leaves the downsampling to the printer driver and the last 
has inferior downsampling routines then aliasing can happen 
there too. There are ways to avoid that. Lanczos 
downsampling shows the effect less for example. Using a high 
quality printer setting requires a higher native resolution 
and the chance your data is downsampled is less likely.

There's another issue with grain and pixels: to the eye both 
grain and pixels have the tonal information and they 
interfere with one another. For example sharpening will 
increase general contrast much more in a digital image with 
grain than one without grain, the other way around a general 
contrast change will increase grain. If the software applies 
a smart print sharpening on the fly when printing, grain 
will become more visible and contrast increases.


-- 
Met vriendelijke groeten, Ernst


|  Dinkla Grafische Techniek  |
|     www.pigment-print.com    |
|             ( unvollendet )            |

David Murphy <dm1000@...> wrote:
>
> ...I have an Epson R280 
> loaded with the 6 cartridge Eboni inkset from MIS.
> 
> ... it seems like in some circumstances and with some 
> photos the printer is able to convincingly render the 
> film grain and 
> it's gradations smoothly and in other cases it can not.
> 
> Where I notice this most is usually with people's hair 
> or beards.  They often appear very splotchy as if the 
> printer is printing straight black 
> or white and doesn't understand that there should be 
> a gray transition in between. ... 
> 
> ...QTR in Black Only mode ... seems to handle these 
> transitions in tone much more smoothly and hair is 
> rendered much more 
> convincingly as opposed to splotchy. 
> I would just print with QTR but I 
> do not like the brown look and I greatly prefer the 
> overall tonality of the 6 cartridge setup.
> 

I can't tell from this just what workflow you're using.  

In addition to what Ernst noted, one difference between the QTR black 
only and the full inkset is the amount of ink (water) that is being 
put on the paper.  Blotchiness can be caused by overloading the paper 
with too much light ink.

This appearance can also be caused by a profile that is not 
appropriate for the paper-ink combination.

If the profiles are made right and the printer is working correctly 
(perfect nozzle check, for example), the full inkset should render 
the gradations as well as the black only.  In fact, many have found 
that the graininess of the black only approach combines with the film 
grain to make an image rougher than if it were a darkroom print.

Testing the system with 21-step test prints makes it easier to spot 
were the problems are.  One such test strip is posted here:
http://www.paulroark.com/BW-Info/21-Step_0-100_L16_16-bit-top.zip

Paul
www.PaulRoark.com

Ernst, 

I was thinking in a similar line.  If the films scans were ever converted to Jpeg files, many of these subtle tones could be lost in the compression, then decompression of the file.  Also, he didn't mention files size.  Enlarging a small file could also make details blotchy.  David didn't mention file type and size.

Ben




      

[Non-text portions of this message have been removed]

pr_roark wrote:

> I can't tell from this just what workflow you're using.

Not sure that I am following one of your pdf workflows precisely but the 
process goes like this:

1. Scan film negatives w/ Silverfast SE at 3200 dpi(usually 3200 for 
35mm and 1600 for 120mm), using Silverfast's internal 16bit>8bit 
conversion, saving as uncompressed TIFF.

2. Import into Lightroom 2.2, crop, spot, sharpen, minor tonal adjustments.

3. For QTR printing I export to TIFF w/ Lightroom.  For Epson driver 
printing I am using Lightroom's internal print module set to 360 ppi,
usually standard print sharpening, and have it set to have the profile 
managed by the printer.

4.In the Epson Driver I am using the settings that are outlined here:
http://www.paulroark.com/BW-Info/R260-BW-GS-Eps.pdf

I'm proofing using Epson Ultra Premium Presentation Paper Matte.

I think the nozzles and such are likely ok as I can take a non difficult 
image and get a great print and then immediately go back to a grainier 
image and have it look bad.  I made 4x4 and 8x8 prints of a well 
exposed, natural light head and shoulders portrait I shot on HP5 120mm 
on both the Epson paper and PhotoRag and it turned out fantastic. 
Intricate details of hair were rendered well but there was fairly little 
visible grain in the negative.  My success with this image gave me the 
impression that I couldn't be too far off with workflow/settings.

35mm negatives of HP5 that have a fair amount of grain seem really 
problematic.  For example, a man's 5'oclock shadow ends up looking more 
like he has polka dots or chicken pox than a beard when printed with the 
Epson driver but not in the scan, not in a wet print and not when 
printed with QTR BO.  Really well exposed 35mm with less apparent grain 
seems less problematic.  I've printed some of the problematic images 
from 4x6 up to 8x10 and it behaves the same way.  Very distracting.

I don't know if it worthwhile but is there a way to try printing with 
all 6 cartridges under QTR rather than just BO?

David Murphy wrote:
> 
> 
>  I made 4x4 and 8x8 prints of a well
> exposed, natural light head and shoulders portrait I shot on HP5 120mm
                                                                ^^^^^^^^^

That's an unusual film format.


;-)

Dana H. Myers wrote:

>  > exposed, natural light head and shoulders portrait I shot on HP5 120mm
> ^^^^^^^^^
> 
> That's an unusual film format.

Indeed.  Not sure what I was thinking.  Maybe I had it confused with the 
120mm artillery I have in the back yard.

In any case, is it possible to use QTR to print with all 6 cartridges? 
I searched the list archives and found a similar query regarding the 
Epson 1400 here:
http://tech.groups.yahoo.com/group/DigitalBlackandWhiteThePrint/message/93208

to which Mr. Roark supplied a file but both the original poster and 
myself received an error message when attempting to import this into 
QTR.  The thread seemed to die after that.

Anybody else try using all 6 cartridges w/ QTR?  Any hints on how to 
make it work?

David Murphy <dm1000@...> wrote:

> ... is it possible to use QTR to print with all 6 cartridges? 

I missed your 12/27 post.

You're apparently using the UT-RC (260) inkset that I wrote up at 
http://www.paulroark.com/BW-Info/R260-BW-GS-Eps.pdf 

That inkset was made to be a very simple one with 2 levels of midtone 
density and 3 hues each.  I did not make QTR profiles for it, but it 
can be printed with QTR.

The problem you'll have with QTR if you use the standard profiling 
procedure is that you don't have 6 different density levels that QTR 
can sequentially partition.  There are several ways around this.

Probably the easiest way is to make a profile that uses only one 
light gray, one dark gray, and the K position inks.  Once that 
profile is done but before it's linearized, copy the light gray ink 
to the 2 other light gray positions and cut the ink limits to 1/3 
what they were for those inks. Copy the dark gray curve to the other 
dark gray ink and cut those 2 ink limits to half what the one was.  
This will use all the inks, but the cross-overs will be at the same 
places.  

If you see these cross-overs there are other refinements that can 
spread them.  But each refinement gets more complex.  So, try this 
approach first.

Hope this helps.

Paul
www.PaulRoark.com