Digital BW, The Print

please forgive a very basic question. i suspect it's been answered in previous postings, but 
a quick search didn't reveal any obvious ones.

i'm trying -- like most of you, presumably -- to get a nice B&W print using an inkjet 
printer. my problem is that the prints just look flat: they don't have the bold look of my 
old wetroom prints. the dark tones don't seem dark enough, and the light tones don't 
seem light enough. in short, instead of black and white, i'm getting a print whose overall 
appearance is rather drab and grey.

fyi, here's my setup:
-- images taken on canon 10d, carefully exposed so that entire tonal range is within 
displayed histogram, pushed up to the right as much as poss (that is, without clipping 
highlights)
-- my standard workflow in photoshop CS: adjust levels to ensure that full dynamic range 
is being used; convert to B&W using Miranda's BW Workflow Pro; readjust levels if nec (ie, 
gap at top or bottom where histogram at zero); applying Roark curves appropriate to 
paper;
-- printing on EEM, and also Hahnemuhle Photo Rag, using MIS UT7 inks on Epson 2200.

here's what is _not_ wrong:

-- the printer setup seems fine. i don't think i have any problems with inks. the prints 
certainly look good; they just don't look great.
-- i am getting a full dynamic range. if i print out a step wedge, i do get very good black 
at 100%  and completely clean white at 0%.

my suspicion is that what's wrong is that too much of the print is in the midrange. if i 
tweak the image with curves, and use an S-curve to increase the gradient in the midrange, 
there is a notable improvement in the overall appearance of the print (at least to my taste), 
but of course a loss of detail at the extremes.

it doesn't seem right to be doing this adjustment, because presumably Paul Roark could 
have incorporated it into his curves. and the images look far more vibrant on my monitor 
than they do on paper. maybe i'm just expecting too much from an inkjet print? or am i 
missing something really basic? for example, i've been making the assumption that 
preserving the entire dynamic range is a good thing, but perhaps i should be more willing 
to 'squeeze the levels' and lose detail at both ends for better midrange contrast? also, i'm 
shooting in jpeg; maybe switching to raw would give better results.

any comments and pointers would be very welcome.

/daniel

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "daniel" 
<dnj@m...> wrote:
...snip...
>my problem is that the prints just look flat: they don't have the 
bold look of my 
> old wetroom prints. 
...snip...
> if i 
> tweak the image with curves, and use an S-curve to increase the 
gradient in the midrange, 
> there is a notable improvement in the overall appearance of the 
print (at least to my taste), 
...snip...

I apply an 's-curve' to probably 50% of my images, so it doesn't 
sound too strange to me that someone else would do this. Most modern 
B&W films do this, I believe. Don't fight it.

...Ben

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "daniel"
<dnj@m...> wrote:
> 
> shooting in jpeg; maybe switching to raw would give better results.
 
Before you change anything else in your workflow, try shooting raw and
converting your images to 16 bits. When you apply your image
adjustments later on, (levels, curves, etc) image data may not
deteriorate as fast.

Andre

First, think of making a levels adjustment as deciding paper exposure to 
get good blacks and whites in the darkroom. Then, think of curves 
adjustments as deciding on paper grade. Creating a toe and shoulder 
aren't bad things - your darkroom prints certainly have toes and shoulders.

If you want to maximize your results, you'll probably have to improve 
your capture by using raw mode and keeping your images in 16-bit as long 
as possible.

Other than that, it takes a lot of practice. Not a lot of your darkroom 
skills translate directly into lightroom skills. You have to learn it 
all over again. Patience, and lot's of captures/prints, and lots of 
reasearch are what lead to improvement, IMHO.

BTW, you have calibrated your monitor, yes?
--
Hogarth Hughes

>
> my suspicion is that what's wrong is that too much of the print is in 
> the midrange. if i
> tweak the image with curves, and use an S-curve to increase the 
> gradient in the midrange,
> there is a notable improvement in the overall appearance of the print 
> (at least to my taste),
> but of course a loss of detail at the extremes.
>
> it doesn't seem right to be doing this adjustment, because presumably 
> Paul Roark could
> have incorporated it into his curves. and the images look far more 
> vibrant on my monitor
> than they do on paper. maybe i'm just expecting too much from an 
> inkjet print? or am i
> missing something really basic? for example, i've been making the 
> assumption that
> preserving the entire dynamic range is a good thing, but perhaps i 
> should be more willing
> to 'squeeze the levels' and lose detail at both ends for better 
> midrange contrast? also, i'm
> shooting in jpeg; maybe switching to raw would give better results.
>
> any comments and pointers would be very welcome.
>
> /daniel
>
>

Daniel,
I add midtone contrast by selecting the midtones (select/color 
range/midtones) then feather the selection 2 pixels. The contrast is added with an unsharp 
mask setting:
Amount 20
Radius 60
Threshold 0
I print on a 2200 with Ut7 inks.
Steve


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

> BTW, you have calibrated your monitor, yes?

... and not just the monitor!  I recently borrowed an EyeOne printer
profiler from a friend and (following the Soft Proofing instructions in the
QTR package, but I believe the process should be completely independent of
the driver) made myself some soft proof profiles for the paper/ink
combinations I use most.  The difference this makes is amazing -- what you
see is not *exactly* what you get because a monitor is not a piece of paper,
but the difference is regular enough that you learn to understand it pretty
fast; in particular, you can really see on the screen where the printed
image is going to be flat or posterized.

Regards,

    Benjamin
    http://www.cis.upenn.edu/~bcpierce/photos

For starters try using the TLR B&W Conversion action set available at 
thelightsright.com. There is an accompanying online "learning 
Gallery" walkthrough and a PDF tutorial download that will get you 
started.

I've also used Dr. Russel Browns B&W conversion technique but even 
though it is very similar, had better luck with the TLR action set.

I have also found that QTR gives me the best results when printing.

Tim

i thought it might be useful to summarize my experiences following my posting and the 
helpful responses i got. in short, my question was: how do i get more vibrant B&W prints 
using greyscale inks?

i learnt the following things that might help other newbies:

1. yes, some quite extreme S-curves are often necessary. as hogarth pointed out, the 
curves adjustment is like selecting a paper grade.

2. getting the monitor to match the print is a huge help. i've not yet managed to borrow 
an eyeone, but paul roark suggested i try his instructions for generating a softproof profile 
by eye (apparently a photoshop hack tyler boley discovered). it worked wonderfully well. 
see: http://home1.gte.net/res09aij/Monitor-Profiling.htm. i found that starting with a 
step wedge and then refining the curve as i found prints with slight anomalies helped. in 
particular, it seems to be easier to set the curve when you have a troublesome print. in my 
case, for example, i was judging tone in a girl's hair, which is much easier than estimating 
the luminance level of big blocks of grey.

3. using this technique, you don't really need to calibrate your monitor, because it's an 
'end to end' adjustment. paul says that the same softproofing profile should do for all his 
curves. as far as i can tell, the only advantages of doing a more serious callibration job (of 
the monitor and also of the printing) would be that (a) if you switched monitors but not 
printers, you could use the old softproofs, and (b) you'd get to see the tone on the screen. 
there's also a minor irritation in the boley-roark scheme for me, which is that i need to 
switch to greyscale to do the curves adjustment, since the softproofs don't work in RGB. 
before, i was staying in RGB -- using channel mixer (in the form of fred miranda's B&W 
Pro) to convert to monochrome, and then applied the roark curves directly. now i have to 
switch to greyscale and back again. i hope there's no information loss there.

kudos to paul roark and MIS. the UT7 inks are just fantastic. i'm using 
EEM and PhotoRag, and the results are far better than i ever managed to get with epson 
inks.

many thanks also to all those who responded: ben, andre, hogarth, steve, benjamin and 
tim. 

/daniel

daniel- excuse me for not reading the replies b4 replying myself, 
but I think you problem is that you're looking for a traditional B&W 
look from a digital capture. Unlike film, digital capture has a 
completely linear exposure curve; films typically have an S curve. 
So, is applying an S curve with Curves you are, in effect, 
replicating the look of film and thats where your solution lies.