Indesign Color Management
ScottyV
that the CMS was turned on? We have just done our first job in the
program(I am a Prepress manager up here in New Hampshire). At first i
really like some of the features, lie trapping and image editing, but
when we output our film, I realized there was a big problem with the
colors. once I turned of the CMS, everything seemed to get better.
David Kilpatrick
Strange indeed. I run a 100 per cent RGB workflow only interrupted when
people who don't really understand photography send CMYK files (having
effectively discarded much of the colour data by doing so, leaving me
with a printer-and-ink specific file). We've run RGB for over a decade,
and still do with InDesign, for all photo images. InDesign does the CMYK
conversion only at the moment of output (proof printing, pdf writing, or
sending to the imagesetter).
What this means for us is that we can have good screen pdfs, decent
Canon CLC working proofs, and good output no matter who prints - and
small file sizes with very accurate colour, and the potential for
adjustment. In fact we long ago abandoned any kind of proofing except to
identify the pages to the printers. What I see on the screen, and what
was in the original image, is what we get on the page - give or take the
natural shifts a small sheetfed short run mag suffers.
You do have to get the CMS workflow correct, and matched, in every
application. One reason we stipulate RGB for source files here - Adobe
RGB or sRGB, but the former preferred - is that European presses do not
use SWOP inks (the cyan ink is quite different) but Photoshop and just
about every system CMS installation will default to SWOP colours for
CMYK. We actually need Euroscale inks instead. XPress offers to 'manage
CMYK sources to output' which in theory will concert properly
identified, tagged CMYK files in the wrong ink space (SWOP) to the
correct one we want (Canon 700/800 for proofing, or Euroscale v2 Coated
for the printers). But the prefs are stored in the document, which means
docs were always coming in with colour management turned off, and
incorrect CMYK would get sent to the press - incorrect from the source.
InDesign colour manages by default, both CMYK and RGB objects, through
to the final destination CMYK. If you set the View preferences to HIGH
QUALITY you will see accurate CMYK colours on screen, if you use TYPICAL
DISPLAY you will see good RGB, but CMYK PDFs and images placed on pages
will look desatured and incorrect.
Another benefit is that many CMYK files are not properly ICC identified
- they have no profile. Using InDesign, you can select the image, and
APPLY a profile to it, and in HIGH QUALITY view you can see a simulation
of the effect. Most often, it's just necessary to identify the image
as Photoshop 4 or 5 Default CMYK. But with so many images coming to us
from Japan (camera industry) it's surprising how many are set up for Dai
Nippon CMYK inks, which are different yet again, or the American SWOP inks.
While the differences might appear subtle they are big enough to give
British, American and Japanese printing different colour gamuts and
typical skin, sky, grass, etc tones. SWOP files sent to a European
sheetfed coldset printer without CMS or conversion print rather garishly.
I would never turn CMS off today. Since no computer is capable of
viewing a CMYK file except by simulating the result, and no RGB file can
be printed in CMYK without CMS at some stage (even old progs like
PrePrint were a form of table-based CMS) - best to use CMS rather than
attribute anomalies in printing to 'problems' with CMS. Wrong set-up
maybe, or a misunderstanding of how to use the CMS workflow - but
working entirely without it in a program like InDesign is throwing away
a huge chunk of the program's power.
David
MSD
icon...@btconnect.com wrote on 5/8/03 3:36 PM:
>
>
> ScottyV wrote:
>> How many people here have had poor output from Idesign only to find
>> that the CMS was turned on? We have just done our first job in the
>> program(I am a Prepress manager up here in New Hampshire). At first i
>> really like some of the features, lie trapping and image editing, but
>> when we output our film, I realized there was a big problem with the
>> colors. once I turned of the CMS, everything seemed to get better.
>
> Strange indeed. I run a 100 per cent RGB workflow only interrupted when
> people who don't really understand photography send CMYK files (having
> effectively discarded much of the colour data by doing so, leaving me
> with a printer-and-ink specific file). We've run RGB for over a decade,
> and still do with InDesign, for all photo images. InDesign does the CMYK
> conversion only at the moment of output (proof printing, pdf writing, or
> sending to the imagesetter).
---
What do you do about your Black RGB type
going to a CMYK build instead of 100 K ?
MSD
David Kilpatrick
What? Black is specified as a CMYK colour by default; take a look at the
InDesign swatches menu, and the icon next to it is the CMYK one, and you
can't replace the colour - same for Registration. Colours for mechanical
work do not require colour management except for screen display purposes
- in fact, colour managing any Pantone stuff etc is destructive.
The above paragraph says very clearly 'for all photo images' which is
what I'm talking about. The earliest and crudest page design packages
could do perfect CMYK native type and tint work without needing even a
colour display screen (I don't know how we did it, but the first DTP I
did was all in mono on Mac 128s and XLs - the Lisa successor - and
somehow it was much quicker and more efficient than we are now...)
I don't know if the terminology is the same everywhere, but when I
learned this stuff in newspapers all type, rules, tints, borders, stock
blocks etc were called 'mechanical' and only work which had gone through
a process camera or a clichograph was a 'halftone'. I guess I should
have qualified more clearly and said that our *halftone* workflow is
entirely RGB up to the point of writing pdfs etc. Of course, we don't
specify spot colours, process tints, or anything like that using RGB.
There is one exception - we do use InDesign's eye-dropper sample
function to grab key colours from photos, and apply these to headline
type and tints. This lets us integrate the colour theme of an article
really precisely, echoing something like a clothing colour or a
background in the colour of type of tint work. So far the colour
management has been accurate enough using this. It is less so when
sampling from a CMYK source, because it is possible for a CMYK picture
to be assigned a specific profile, while the rest of the doc is
Euroscale Coated v2. By checking carefully, we've been able to do things
like sampling a colour from an ad placed on a page, and making a patch
to amend a coupon or something - and as long as the CMYK of the original
is in our normal output CMYK, it's seamless. We convert all RGB to Adobe
RGB before placing on the page; otherwise the same caveat would apply,
that the placed image being sampled must be in the same colour space as
the main default for the document.
David
Lee Blevins
> What? Black is specified as a CMYK colour by default; take a look at the
How do you make a silhouette with a black only drop shadow in rgb?
And if you do the image in cmyk and color manage it how do you keep the
black only drop drop shadow from turning into cmyk?
David Kilpatrick
Er, in InDesign you just activate the clipping path of the RGB JPEG or
TIFF (no need to prat around with redundant-data EPS files just to get
workable clipping paths, you can have 'em in any flavour efficient file
format - or even use InDesign's slightly crude 'find edges'
self-generated cut-out function).
You apply Drop Shadow from InDesign's menu.
The shadow is applied as a pure black ink gradation with transparency,
none of that ShadowCaster TIFF generation stuff. Doesn't matter whether
the source object is RGB, CMYK, type, graphic, rules or whatever.
I did a nice ad for someone last week and produced a table with drop
shadows accidentally left on for the rules. It look incredible so we
left it!
I take it you didn't really mean silhouette, as a silhouette would be a
plain black solid shape.
InDesign does not turn cast default Black shadows into CMYK though you
can of course go to the swatches list in the Drop Shadow dialog, and
pick any other colour (or create your own). I have never tried casting
green shadows though. Might be a fun thing to try - cast a coloured
shadow with no direction, and loads of spread, and you would get a
coloured glow round the object.
David
tsucker
data? Yuk. Scanned images from RGB most of the time needs to be
tweaked to get the best cymk conversion. Of course you have to know
the final destination. Usually if you have a highend flatbed or drum
the program will get the best cymk conversion for you. I use C-Scan.
A step above Photoshop. But InDesign, you make me laugh. You wouldn't
get any of my work.
Tsucker
MSD
icon...@btconnect.com wrote on 5/9/03 1:59 AM:
---
I missed the little bit about "photos only".
Your workflow is probably the best idea for workflows
including photos.
Now InDy needs to manage grayscale.
MSD
David Kilpatrick
tsucker wrote:
> You have to be kidding right? You let InDesign convert all your RGB
> data? Yuk. Scanned images from RGB most of the time needs to be
> tweaked to get the best cymk conversion. Of course you have to know
> the final destination. Usually if you have a highend flatbed or drum
> the program will get the best cymk conversion for you. I use C-Scan.
> A step above Photoshop. But InDesign, you make me laugh. You wouldn't
> get any of my work.
>
That's why we publish some of the top photo magazines in Britain, do so
using what were once called DTP methods, and have done for longer than
any other UK publisher. But I must admit my approach is very crude. Back
in 1987 we won the inaugural Desktop Publishing Awards of the Printing
Industries Research Association outright, but it wasn't difficult, as
apart from something called 'Whalesong' no-one had ever produced
commercial colour magazines using DTP before.
When we printed the very first magazine cover ever from Kodak Photo CD,
we had not the faintest idea what we were doing. The same with the first
pictures ever reproduced, in the world, from Kodak digital cameras. And
when my wife gained one of the earliest Masters Degrees in Colour
Science for specific research into human perception and colour matching
between D2T2 and pigment-based printing systems in November 1995, she
was backed up by almost ten years of total misunderstanding of colour.
Of COURSE I let InDesign convert all my data - just as many years ago we
let EfiColor convert all our XPress data. And just as, before the ICC
initiatives if 1992-3, we used PrePrint and sundry other utilities to
create our own calibrated conversions.
I have no idea what a high-end flatbed is. In fifteen years of using
relatively modern scanners, I have marvelled at how extending the bit
depth and resolution of flatbeds makes less difference to the end result
than changing the illuminant/original/lens geometry and the RGB
filtration. While I currently scan using different cheap DTP things (we
have to test the things in order to write about them) I don't really
like many of them. I think the nicest photographic quality came from
Agfa's Focus Color Plus three-pass when used with their own profiling
Fototune. And it's still hard to beat a Leafscan 45 with a fresh set of
filters. For a couple of years we did all our print repro on a batch of
five Ł50 Agfa Snapscans. 300 dpi was plenty; we never enlarge from a
print, only reduce; all that mattered was colour quality and accuracy,
and resistance to picking up grot from photographers who sent in lustre
prints. The Snapscans did all that rather well. As each one broke, we
chucked it away - we bought five because we knew they were hopelessly
badly made scanners. But the colour was dead neutral and just right.
Drum scanners are something I absolutely detest, especially for 35mm
slides. Over the last 25 years drum scanning ruined more of my originals
than any amount of handling (only public projection over a long period
did more damage to some). Drum scanning has been the only reason we have
ever had to pay (or our repro house had to pay) sums of many hundreds
per original in compensation for the permanent, irreversible damage done
to photographs. The Crosfields etc of the mid-1980s, which were held up
as paragons of accuracy against the threat from the first DTP scanners,
were only capable of showing 64 steps per ink colour and one of the
reasons for my own success at the close of that decade was that our
'unprofessional' DTP scans blew away the repro in top magazines and
books of that era.
We ended up in Apple's international marketing brochures, I got some
nice work as a consultant to sundry outfits like Times Newspapers, and
my long-term collaborator at that period john Henshall went on to be a
keynote presenter for Seybold. John's great interpersonal skills got him
through many doors, and back out with gear or software which no-one had
yet seen, and with free hands-on control of our publications - owning
all the production system and using most of it personally - we were able
to PROVE that DTP, DT scanning and digital photography were the future.
More fools us - we conspired to wipe out half the industry which had
previously made is fairly wealthy.
Still, that's an aside. But we knew RGB was the best colour model for
photographic images then, and it still is, because RGB is basically what
vision is all about.
As for converting to CMYK - *what* CMYK? Should I seriously want to
increase the filesize of all my images, and target them not just to a
printing process, but to so properly, to a specific printing press, ink
and paper at a specific printing works? It's bad enough having to apply
USM to the actual image file - permanently damaging it, and forcing you
to resample it to a specific usage size if it's to be done well.
My perfect workflow would be one where not only did we keep all photos
in RGB, but placed them on the pages without regard to their original
file size, and without any USM applied - and where the application (say
InDesign) not only downsampled when writing the pdf or sending to the
printer, but also applied USM to the image on the fly, set globally but
with individual over-rides (as happens with colour management).
I work in a context where tweaking individual files is not desirable; we
publish very precise comparisons between, for example, the results of
applying or using various colour profiles with new digital cameras. We
can not adjust the RGB files, and we always state exactly what resizing
or USM has been applied for repro. Otherwise our readers have no idea
whether our published comparisons between pro digital cameras, different
film types, and so on, are valid. Much of the time we have to use
scanners where all automated functions can be turned off, and if any
corrections are used at all, they must be set manually.
Then again, our benchmarks are not very exact. We are in the business of
perception and not required to reproduce colours as a measurable
counterpart of some original subject. We are not doing the Land's End
catalogue with six different shades of near-dark-grey required to have
visible blue, green, mud, teal, puke or whatever matches to fabrics. I
will admit that for that job, you need CMYK files, matched to your
press, proofed wet on your press, adjusted in a way which would once
have meant dot-etching the seps selectively. Thank heavens all that is
no longer needed and three stages of wet proofing are not often called
for, nor soft-dot work on films with bleach.
These procedures were needed because photography itself is not accurate,
printing is not accurate, and human vision is both highly variable and
inaccurate. CMYK specific retouching and correction work replaces them,
because nothing has changed and digital photography (while generally
more accurate than film ever was) still isn't the same as vision.
If that is the kind of work you do, where the client is going to stick a
proof next to a fabric swatch, then CMYK without on-the-fly conversions
or CMMs is the right approach. My concern is simply to make colour
magazines full of photographs which look right to a readership of
photographers, and better than our opposition titles.
As for real accuracy? Two months ago we took a digital camera file in
two formats - TIFF and JPEG - with the JPEG converted from the TIFF,
both assigned the same Adobe RGB profile, placed as full adjacent pages
on the same running sheet, and printed using InDesign colour management.
The repro from the JPEG and TIFF is clearly different - nothing to do
with JPEG quality issues - so InDesign obviously doesn't really have it
right. But the visible differences were within acceptable limits. When
'good enough' is better than the competition, it doesn't really matter.
In your case it might well not be, and you are better taking your
position and converting photos to CMYK before placing.
David
Lee Blevins
> You can define the color of your drop shadow to be whatever you want. As
> David has explained, the CMYK colors don't "go away" if you are using RGB
> photos and CMS. So, defining a CMYK drop shadow is no problem, even if it
> overlays an RGB object.
Sorry,
After a week of drinking only water I decided to imbibe so I'm not sure
I read this right but...
Adding the drop shadow in ID is not creating a k only drop shadow in
rgb.
ID (in this case) defines colors in CMYK and the drop shadow is defined
in CMYK.
It is definately K only.
Why?
The shadow and the image are defined separately.
You cannot place that image in Quark and get the same effect.
Neil Gould
"Lee Blevins" <le...@digitalgraphics.net> wrote:
> Neil Gould <ne...@terratu.com> wrote:
>
> > You can define the color of your drop shadow to be whatever you want. As
> > David has explained, the CMYK colors don't "go away" if you are using
RGB
> > photos and CMS. So, defining a CMYK drop shadow is no problem, even if
it
> > overlays an RGB object.
>
> Sorry,
>
> After a week of drinking only water I decided to imbibe so I'm not sure
> I read this right but...
>
> Adding the drop shadow in ID is not creating a k only drop shadow in
> rgb.
>
As well it shouldn't, as there is no "K" in "RGB"! 8-)
> ID (in this case) defines colors in CMYK and the drop shadow is defined
> in CMYK.
>
Which is your choice, but is probably what you'd want to do anyway.
> It is definately K only.
>
> Why?
>
> The shadow and the image are defined separately.
>
Yes, they are separate objects, as they should be... I must not understand
what you're trying to accomplish, if not to create a K-only drop shadow for
an RGB object.
Regards,
--
Neil Gould
--------------------------------------
Terra Tu AV - www.terratu.com
Technical Graphics & Media
Lee Blevins
> Yes, they are separate objects, as they should be... I must not understand
> what you're trying to accomplish, if not to create a K-only drop shadow for
> an RGB object.
It's just my usual rant that "all the world should work in rgb" really
isn't practical.
Some things are better done in cmyk and always will be.
I use the appropriate color space for what I'm trying to do.
First let's dispell the myth that if I have my image in rgb and print it
to different printers it will come out the same. That just isn't true.
I've been working with color management as long as it's been around and
that myth never pans out in the real world.
Another myth, "I'll get exactly what I see on my monitor" should be
trashed. Even with color management and calibration and profiling your
monitor may not be able to display all the colors in your output color
space.
It's all relative (or perceptual or absolute) and there's a lot of
variables in color reproduction.
Calibrating and profiling a color space doesn't make it now possible to
have colors that were out of gamut for that space. They remain out of
gamut even after that (profiling) is done.
For example, calibrating and profiling a shitty printer still results in
a shitty printer.
I would be very interested in seeing a 6 color press using the inks
Espon uses (CMYKlclm) in a stocastic screening mode ran to the densities
that espon using in the 10K model and see how starting with rgb (wide
gamut) would turn out.
Certainly on our HP we get outstanding colors that can't be had in a
normal (SWOP) color model.
BUt...
If these inkjets keep getting faster and faster we might not need a
standard offset press in the future.
The only riddle I have left to sovle is the cutting and we can compete
with offset on short run single sided work.
Oh, and that problem with the ink being water soluable.
Neil Gould
"Lee Blevins" <le...@digitalgraphics.net> wrote:
> Neil Gould <ne...@terratu.com> wrote:
>
> > Yes, they are separate objects, as they should be... I must not
understand
> > what you're trying to accomplish, if not to create a K-only drop shadow
for
> > an RGB object.
>
> It's just my usual rant that "all the world should work in rgb" really
> isn't practical.
>
> Some things are better done in cmyk and always will be.
>
No argument, here. However, I think that David's opinion is based on his
specific application. If one has complete control over images, from original
photography to all subsequent output, it is conceivable that an RGB-based
workflow will not only be practical, but better than maintaining multiple
formats of the same images. In your business, that is simply not the case.
> I use the appropriate color space for what I'm trying to do.
>
This is not necessarily incompatible with having all of your images in a
single color space. Most images used on press will be converted from a
3-color space to CMYK at some point; it's simply a matter of when, where and
why this is done. There isn't much of a reason why this conversion can't be
handled as well at final output if the images are within the CMYK gamut to
begin with and color management settings (such as GCR and UCR) are
optimized. Again, this isn't likely to work for a print shop because of the
lack of control over the original images.
> First let's dispell the myth that if I have my image in rgb and print it
> to different printers it will come out the same. That just isn't true.
> I've been working with color management as long as it's been around and
> that myth never pans out in the real world.
>
This goes well beyond color management. One can take the same plates, put
them on the same press on two different days and get differences in the
printed results. Whether or not this is a problem depends on how picky one
is.
> Another myth, "I'll get exactly what I see on my monitor" should be
> trashed. Even with color management and calibration and profiling your
> monitor may not be able to display all the colors in your output color
> space.
>
No argument here, either. In fact, David also pointed out that he was
getting good color output results when using a monochrome monitor, so I
doubt that he's matching colors on the basis of what he sees on the monitor.
David Kilpatrick
Lee Blevins wrote:
> Another myth, "I'll get exactly what I see on my monitor" should be
> trashed. Even with color management and calibration and profiling your
> monitor may not be able to display all the colors in your output color
> space.
>
>
But if you work in CMYK, you have an absolute guarantee that what you
see on your monitor is inaccurate! Generally most monitors will exceed
the gamut of most output spaces (some inkjet printers being an exception
but slightly irrelevant) but there isn't a monitor made which works in
CMYK. From the very earliest days, colour management has always been
required to interpret a CMYK file on to a computer display, even if just
the internal look-up tables of a program.
Photographic colours, especially those from C-type prints, are
technically 'dirty'. C-types are made on the basis of using CMY inks
with added neutral density (dirt!) and use no black silver or black dye.
A true black is only produced because the yellow (especially) is very
impure, and the cyan and magenta slightly impure. Colour slide film
varies in approach - early Kodachromes enhanced what would have been a
green maximum density by leaving a controlled amount of granular silver
present, modern E6 films are supposed to remove all silver in the
interests of efficiency but still leave a trace image - the rest of the
their max density is again created by using impure dyes. However,
transmissive viewing creates a visual effect with such a large dynamic
range compared to reflective viewing, that we see slides are purer than
prints.
Colour photography, despite research which makes it highly consistent,
got there by dirty methods and a lot of empirical learning experiences.
I would say that almost any monitor can display a good representation of
a photograph on traditional colour materials.
Computer-generated artwork, digital photography, and colours defined in
a CMYK print space are a totally different matter and it takes a lot of
sophistication to get a decent visual simulation of all these -
alongside photographs - on a monitor. Especially these damn LCD/TFT
things. I would never go back to using a CRT but the inaccuracy of a TFT
system - even after the usual Pantone OptiCal tweaking - is frightening.
It probably can display everything I need to see, but getting it to do
so is a different matter.
I have to write a large review of the Pantone Spyder in its variations
from two different UK vendors, and they are fighting against a writer in
this case who's owned half a dozen different systems from Barco
Calibrator downwards, and been amazed at the huge differences which all
of them have produced when asked to calibrate to the same standards!
That alone has always made me doubt calibration as an absolute tool.
However, using one calibrator device on a room full of identical
monitors DOES result in consistent colour. That is perhaps more useful
and I guess my final article will probably be slanted in that direction.
David
Ted
>
> I have no idea what a high-end flatbed is....
Look it up, it isn't hidden.
>
> Then again, our benchmarks are not very exact. We are in the business of
> perception and not required to reproduce colours as a measurable
> counterpart of some original subject. We are not doing the Land's End
> catalogue with six different shades of near-dark-grey required to have
> visible blue, green, mud, teal, puke or whatever matches to fabrics. I
> will admit that for that job, you need CMYK files, matched to your
> press, proofed wet on your press, adjusted in a way which would once
> have meant dot-etching the seps selectively. Thank heavens all that is
> no longer needed and three stages of wet proofing are not often called
> for, nor soft-dot work on films with bleach.
>
>
> If that is the kind of work you do, where the client is going to stick a
> proof next to a fabric swatch, then CMYK without on-the-fly conversions
> or CMMs is the right approach. My concern is simply to make colour
> magazines full of photographs which look right to a readership of
> photographers, and better than our opposition titles.
I guess we are in 2 different fields. Although I would assume that your
field would want to print the best quality pics you can but I would be
wrong.
Ted
Ted
>
> I have no idea what a high-end flatbed is....
Look it up, it isn't hidden.
>
> Then again, our benchmarks are not very exact. We are in the business of
> perception and not required to reproduce colours as a measurable
> counterpart of some original subject. We are not doing the Land's End
> catalogue with six different shades of near-dark-grey required to have
> visible blue, green, mud, teal, puke or whatever matches to fabrics. I
> will admit that for that job, you need CMYK files, matched to your
> press, proofed wet on your press, adjusted in a way which would once
> have meant dot-etching the seps selectively. Thank heavens all that is
> no longer needed and three stages of wet proofing are not often called
> for, nor soft-dot work on films with bleach.
>
>
> If that is the kind of work you do, where the client is going to stick a
> proof next to a fabric swatch, then CMYK without on-the-fly conversions
> or CMMs is the right approach. My concern is simply to make colour
> magazines full of photographs which look right to a readership of
> photographers, and better than our opposition titles.
Derek Tree
<icon...@btconnect.com> wrote (snipped some):
>But if you work in CMYK, you have an absolute guarantee that what you
>see on your monitor is inaccurate!
That would depend on your, or anyone else's, definition of "inaccurate".
Speaking for our own small studio whose output is primarily geared
towards Euroscale coated 4 colour offset litho (4 colour press
advertising, brochures and magazines) we get acceptably *accurate*
screen to proof, screen to press and screen to dupont cromalin euroscale
matching working in a custom cmyk workspace.
But then we are using reasonably decent CRT's (NEC-Mitsubishi 2045u's ,
NEC-Mitsubishi 2070's and one Apple 22" ACD) and an accurate inkjet
proofer (Epson 1270 driven by PressReady).
>I have to write a large review of the Pantone Spyder in its variations
>from two different UK vendors, and they are fighting against a writer
>in this case who's owned half a dozen different systems from Barco
>Calibrator downwards, and been amazed at the huge differences which all
>of them have produced when asked to calibrate to the same standards!
>That alone has always made me doubt calibration as an absolute tool.
That I would not dispute. In many years of fiddling with various
calibration tools (mainly Optical and DTP Spectrophotometers) I have
come to the inescapable conclusion that it is perfectly possible to
calibrate a modern high-end CRT monitor (see above) to produce
*reasonably* accurate screen to press matching without anything more
sophisticated than the monitor's own controls (gamma, rgb gain controls,
contrast and brightness) and Photoshop's (6x and later) colour settings
controls. IME the differences that Optical and a spectrophotometer can
make are minimal and often not worth the effort for the generality of
work destined for 4 colour output which often ends up in the bin
anyway...
The only area where is does benefit our workflow is in the production of
artwork for fine art prints (of which we a do a few annually). Here, the
wider gamut of the final output devices (usually high-end HP and Epson
A1 printers) does benefit from accurate calibration and profiling at our
end and we are able to achieve an extremely accurate screen to printer
match using Optical 3.7 and Pantone's latest color spyder whilst still
working in a custom cmyk colour space. But put that job on a 4 colour
Heildelberg and - well, you can guess what it looks like!
The irony for us with the majority of our press work is that as CRT's,
calibration software and 6 colour proofers have got better it is
increasingly necessary to work within washed out custom cmyk colour
spaces to match the reduced gamut of 4 colour offset litho presses!
But I doubt whether that is something the makers of these "solutions"
want to hear! I would go further - after all these years of
calibrating, profiling and tweaking I *still* go by the numbers for
colour critical work. Ironically, our Apple 22" ACD which has never been
calibrated by any external means displays as a good a screen to printer
match as our calibrated and profiled Mitsubishi CRT's. Granted the
colours are a *little* more saturated but that can be reduced merely be
using Photoshop's "de-saturate monitor colours" setting.
>However, using one calibrator device on a room full of identical
>monitors DOES result in consistent colour. That is perhaps more useful
>and I guess my final article will probably be slanted in that direction.
My only caveat with that is that I have yet to find such an animal as an
"identical" monitor. Our two NEC-Mitsubishi 2045u's are anything*but*
identical even though they were manufactured in the same plant within
weeks of each other! Consequently the settings needed to match the same
printed piece or target are very different on each monitor. So much for
manufacturing "consistency".
The bottom line, as Neil has already pointed out which the manufacturers
and purveyors of all this CM wizardry signally fail to mention, is that
press conditions *VARY*. After 25 odd years of putting ink on paper on
press I would go further and say that the possible variations (and their
complex permutations) on press vastly outweigh the subtle variations
achievable through monitor calibration and profiling software - always
assuming that the devices being profiled and calibrated are not some old
dog of a monitor picked up at a garage sale for 8/9d.
So - until someone comes up with a cmyk monitor and/or an rgb GTO there
is simply no getting away from the basic problem that mixing additive
and subtractive does not add up to anything other than a kludge.
Best wishes
--
Del Tree
Fixx
le...@digitalgraphics.net (Lee Blevins) wrote:
> If these inkjets keep getting faster and faster we might not need a
> standard offset press in the future.
jetted or pressed, it does not matter how to ink gets to paper? :-)
It is going to be a while until the price, speed and stability of inkjet
is good enough :-) -F
Lee Blevins
> > If these inkjets keep getting faster and faster we might not need a
> > standard offset press in the future.
>
> jetted or pressed, it does not matter how to ink gets to paper? :-)
> It is going to be a while until the price, speed and stability of inkjet
> is good enough :-) -F
It's already here.
On short run posters we are already higher quality and lower priced
thatn offset.
The extended gamut of the inkjet makes for far better color than
traditional CMYK printing.
Lee Blevins
> But if you work in CMYK, you have an absolute guarantee that what you
> see on your monitor is inaccurate!
That is myth #4 I think.
If you can see on a monitor an accurate CMYK preview of an RGB file then
you can see an accurate preview of a CMYK file.
It isn't the data format or color space of the file that is the
limitation here.
It might more be the limitations of your CMS.
David Kilpatrick
Derek Tree wrote:
> My only caveat with that is that I have yet to find such an animal as an
> "identical" monitor. Our two NEC-Mitsubishi 2045u's are anything*but*
> identical even though they were manufactured in the same plant within
> weeks of each other! Consequently the settings needed to match the same
> printed piece or target are very different on each monitor. So much for
> manufacturing "consistency".
>
After working for a time with different monitors including the Barco and
various large Trinitrons, and early calibrators culminating in the
LightSource ColorTron with monitor kit, we capitulated and bought three
identical Apple 17 inch Colorsync/Applevision beige case Trinitrons with
internal calibration.
I would say that we were delighted when they were first set up, and they
looked closer to each other - and brighter, and better - than any of the
previous stuff. They turned out to be the most unreliable and expensive
monitors with repairs under warranty and then out of warranty... etc.
We moved to Mitsubishi Diamontron 19 inchers, three flat screens but all
slightly different - one branded by Iyama, two by Mitsubishi, one with
extra USB ports etc. We stopped using any kind of hardware based
calibration and did it by controls, display software and 'eye'. I reckon
these were never a perfect match, but each one was set up to suit the
operator, and the profiles were carefully named so that files were never
'converted' repeatedly.
Last year I got a Formac 18 inch TFT studio screen and I guess we will
at some future stage (only two of us now working on displays which need
calibration, or colour accuracy) get a couple of much larger identical
displays, but not at present prices. I still do the Leafscan work and
some checking of material on a Diamondtron in a grey-painted special
studio room; I have learned to mistrust the TFT.
As commented, variations at the printer's end (same sort of Euroscale
small sheetfed stuff as Derek described) far exceed those at our end.
Humidity, paper stock, Friday, blanket needing replacing... you name it.
We are good customers and NEVER complain. We just understand that
printing is a less exact science than it should be, at this level!
David
David Kilpatrick
I think Photoshop's internal management is pretty good. However, I can
convert an Adobe RGB space RGB file to a CMYK file without seeing a
massive shift in appearance. A slight shift, true. In fact a lot of data
is discarded in the CMYK conversion but Photoshop's re-expansion of the
CMYK image for monitor viewing hides this.
I don't think anyone would choose to work with CMYK images if they could
really see what they look like.
I'm sure there are CMY (forget the K) models which almost match the
image we expect to see on a screen. Leaf's Catchlight digital camera
back, which used CMYT (CMY and Teal) filters was very brave and produced
wonderful colours for printing on CYMK inkjets. Leica's Digilux uses a
fairly similar approach.
But there are things present in CMYK which a monitor simply can not
show, such as the visual effect of tone breaks dependent on how UCR and
GCR are implemented, and the visual difference between a CMY grey tone
and black ink greytone, and especially the visual effect of a superblack
(say 100K + 40M + 40Y - usually the heaviest I run). Even in InDesign, a
superblack is shown as a lighter colour than pure K.
Now if you have a colour management system which will simulate on a
monitor the visual appearance of a black ink halftone, as distinct from
the same colorimetric value created by various CMY and CMYK blends, you
do have something far better than I have.
I've never seen such a system yet.
DK
David Kilpatrick
Ted wrote:
>>I have no idea what a high-end flatbed is....
>
>
> Look it up, it isn't hidden.
Well, I've used Topaz and Celsis and heavy stuff like that. Just struck
me as industrial strength scanners little better in real terms than most
inexpensive desktops. I have a friend who spend £20k on such an item and
remains pissed off for ever that it does not produce anything better
than a £150 Epson! But it does keep someone in employment since it takes
four times as long to make a scan.
>
>
> I guess we are in 2 different fields. Although I would assume that your
> field would want to print the best quality pics you can but I would be
> wrong.
>
Of course, but best quality does not always mean using the most accurate
equipment.
Sometimes it means knowing what to do with a picture to ensure it
reproduces well, even if that is not strictly true to the original.
Photographers have particular obsessions; for example, they like to see
halftones which go to a really dense black and almost be able to wipe
the ink off the page, and they don't especially mind specular highlights
which are plain paper.
If I had an objective, it would be to make monochrome pages look as
close as possible to sheetfed gravure as possible, even if that means
losing the deep shadow detail; colour repro generally must be tipped
away from neutrality towards a warm result, and again, it's better to
lose shadow detail than to have even a slightly anaemic looking image.
They don't want soot and whitewash, but they don't want literal accuracy
either. Our readers are happiest when our pages are punchy and we put
2pt black keylines round photos - or even better, stick 'em on a black
bleed background (but our printers hate that, so I use it very sparingly
and watch the rest of the running section carefully).
We tried accuracy once, trying to match proportionally on the page the
colours in their prints using measurements and press trials. It isn't
liked; it looks too flat visually. We have learned from the odd occasion
when our printers seriously over-inked an issue that the readers LIKE
that. It reminds them of a really fully developed darkroom print. Even
better if we ink heavily and then varnish seal the pages.
'Quality' in print is always a perceptual thing, not a measurement
thing. Have you seen Cartier magazine? They've got the right idea; best
of all worlds. Superb accuracy AND ink laid on with trowel on paper you
can cut yourself with.
David
Fixx
le...@digitalgraphics.net (Lee Blevins) wrote:
> On short run posters we are already higher quality and lower priced
> thatn offset.
of course, short run posters and large format jobs are right here
already. You are right.
> The extended gamut of the inkjet makes for far better color than
> traditional CMYK printing.
again right, even pigment based inks give larger gamut than trad CMYK. -F
tsucker
> >
> >
> > Look it up, it isn't hidden.
>
> Well, I've used Topaz and Celsis and heavy stuff like that. Just struck
> me as industrial strength scanners little better in real terms than most
> inexpensive desktops.
Well, your dead wrong. First off the software is about a miilion times
better. If someone wanted a scan of a transparency at 600% at 3000 dpi
could that inexpensive destop do it. Of course it couldn't. So your
either drunk or just stupid.
I have a friend who spend £20k on such an item and
> remains pissed off for ever that it does not produce anything better
> than a £150 Epson! But it does keep someone in employment since it takes
> four times as long to make a scan.
Well there is something wrong with the hardware, software, or him. I
can see it now in all the trade magazines. Don't buy a high-end
scanner because my friend can't get a good scan from one. You know how
stupid that sounds?
> >
> >
>
> > I guess we are in 2 different fields. Although I would assume that your
> > field would want to print the best quality pics you can but I would be
> > wrong.
> >
> Of course, but best quality does not always mean using the most accurate
> equipment.
> Sometimes it means knowing what to do with a picture to ensure it
> reproduces well, even if that is not strictly true to the original.
I agree with that but it depends on a lot a variables.
>
> Photographers have particular obsessions; for example, they like to see
> halftones which go to a really dense black and almost be able to wipe
> the ink off the page, and they don't especially mind specular highlights
> which are plain paper.
Most prepress and printers hate photographers for obvious reasons.
>
> If I had an objective, it would be to make monochrome pages look as
> close as possible to sheetfed gravure as possible, even if that means
> losing the deep shadow detail; colour repro generally must be tipped
> away from neutrality towards a warm result, and again, it's better to
> lose shadow detail than to have even a slightly anaemic looking image.
> They don't want soot and whitewash, but they don't want literal accuracy
> either. Our readers are happiest when our pages are punchy and we put
> 2pt black keylines round photos - or even better, stick 'em on a black
> bleed background (but our printers hate that, so I use it very sparingly
> and watch the rest of the running section carefully).
You have to go with what pays the bills but I don't believe in losing
shadow details. I wouldn't want to see a black dress or a black car
printed in one of your magazines. Yuk
>
> We tried accuracy once, trying to match proportionally on the page the
> colours in their prints using measurements and press trials. It isn't
> liked; it looks too flat visually. We have learned from the odd occasion
> when our printers seriously over-inked an issue that the readers LIKE
> that. It reminds them of a really fully developed darkroom print. Even
> better if we ink heavily and then varnish seal the pages.
>
> 'Quality' in print is always a perceptual thing, not a measurement
> thing. Have you seen Cartier magazine? They've got the right idea; best
> of all worlds. Superb accuracy AND ink laid on with trowel on paper you
> can cut yourself with.
So what is your max ink density?
Tsucker
David Kilpatrick
tsucker wrote:
> Well, your dead wrong. First off the software is about a miilion times
> better. If someone wanted a scan of a transparency at 600% at 3000 dpi
> could that inexpensive destop do it. Of course it couldn't. So your
> either drunk or just stupid.
Drunk most of the time :-) If I wanted a scan of transparency at 600 per
cent at 3000 dpi, I would have to be drunk before I started. I don't
print many pages using 2000 lines per inch screening. If I want a scan
of a transparency, I don't use a flatbed at all and never would. None of
the desktop scanners I have ever tried are suitable for transparencies,
and neither are the big rip-off flatbeds.
And I didn't think that any of them scanned at an optical 18,000 dpi
either. More like a maximum of 5000 dpi true optical, and a d-Max of 3.7
if you are very lucky. Been able to get both of those since 1990 with a
Leafscan. Can now get both from a $750 Minolta 5300dpi 35mm scanner,
plus ICE3.
> Well there is something wrong with the hardware, software, or him. I
> can see it now in all the trade magazines. Don't buy a high-end
> scanner because my friend can't get a good scan from one. You know how
> stupid that sounds?
>
No, buy a high-end scanner, but don't buy large flatbeds. They are not
high-end. They involve fairly crude transparency handling and while the
software can be pretty good, if you need that sort of thing, I have
never used any software feature on a scanner apart from:
Framing/cropping the preview
Setting the final repro size and dpi output for file
Assigning a colour profile
Setting exposure
Very rarely indeed, with bad originals, correcting colour or tweaking a
gamma curve
It has to be said that most photographs we deal with are there in the
magazines because they are superb photographs - 50 per cent are
award-winning images, from exhibitions, or just from very professional
sources. We pick photographs only on grounds of quality. Consequently,
we do not have to muck around compensating for poor work.
And we don't do CMYK conversion in scanner software, nor any USM. Nor do
we crop small parts of pictures - in fact, we breach the copyright terms
of most images we use and would be heavily criticised if we cropped them
in any way. Or made them into backgrounds, or colourised them etc.
Frankly, if a scanner is decently designed and basic software is
correctly implemented, bells and whistles don't matter for this type of
work. Quality in, quality out.
> Most prepress and printers hate photographers for obvious reasons.
Well, we don't, but photographers used to hate prepress (repro) houses
in the past for ruining their originals. Desktop repro has changed that
in a wonderful way. Some of the best work we now receive is on CD-R from
the photographer - and when we ask, we find that they have an Epson 1250
Perfection or a basic Linocolor, or a Nikon Coolscan. Yet they are
selling direct scans, made on this type of gear, through top
international stock agencies.
Some very old-fashioned repro houses (and publishers) have a problem
with accepting work from photographers in this form; they would almost
rather have them write a 10 x 8 sheet film repro, and then scan it, just
to retain the business they once had.
> You have to go with what pays the bills but I don't believe in losing
> shadow details. I wouldn't want to see a black dress or a black car
> printed in one of your magazines. Yuk
>
Photographers don't like to see 95% black as the max. If a picture is
quite soft in
the original, I will normally use a keyline round it. It provides a 100
per cent black reference. In photo printing, having the deepest tone
fall anything short of the maximum D the material can produce is an
error, so they don't like seeing softish, 'open' print work with a dot
present right to the densest halftone. European printers generally
concur and work for a deep black, and so do Japanese art printers.
American printers tend to go for very flat reproduction (or did in the
past) with a biad towards plummy colours - pink skin not golden skin,
blue-green grass not gold-green grass, purple blue sky not cyan blue.
British printers (and readers) prefer a very warm sunny look, like the
world is permanently stuck on a springtime evening.
I remember the old Agfa Gamma 1.0 repro parameters - you could see every
bit of shadow detail in uncorrected scans with displays running at that
setting. Then we moved to Gamma 1.4, and scanners which had far superior
nonlinearity (curve correction); then to Gamma 1.8. I don't like Gamma
2.2 but I guess one day we'll end up using it. It actually looks
'photographic' in the way it crushes shadow detail.
All I am saying is that losing a small amount of shadow detail -
typically in the dot range 95-99 per cent - is preferable in our market
to printing a max 95 per cent and seeing that detail. A rich, true black
comes before 3/4 tone accuracy.
I've just had to reproduce a whole set of American portrait pix from a
show at Chrysler Museum, which are supposedly the best you can get. By
our tastes, they are dark, soft and pink and the people look artifical -
like waxworks with make up on. They are quite heavily retouched, by a
fine artist (not digitally) despite being current (last six months) and
very deliberately and skilfully lit. It's amazing how they look dated to
us in Britain, because we no longer use retouching; we generally don't
want clients to be groomed or made up specially for a portrait; we like
natural lighting not strobes - etc.
There are pretty strong differences in visual taste across the pond
which render some of the color lab photo work 'wrong' to us, but the
actual standard of technical know-how in the USA is generally far
superior - more consistency, more measurement.
>
> So what is your max ink density?
>
We are only running sheetfed, cold set, and we're lucky to get much
better than 1.6 on K alone.
I think we have achieved around 1.81-82 without overinking, and since
moving to pdf workflow, the pages tend to be more like that. In an ideal
world, we would have densities for a max CMYK black of 2.1 to 2.3 on the
page. It might be possible to manage that on a heatset machine varnished
job running very slowly on superb stock. But we are on the usual trade
gloss 110gsm (130gsm for some mags), most pages unsealed. That's just
economics. On a bad day our printers can drop to 1.35 or so... which I
really hate.
Just occasionally we get some money to do something special. I love
doing sealed duotones on special stock. Next month we don't print at all
on our cover, except the masthead and stuff - Kodak are putting a
tipped-on 8 x 6 inch real print, shot on the new DCS14n, and printed on
their D2T2 biggie - the pro version of the 8500 dye sub. Got to sent the
sample print to the finishers for a quote today, and send them a box of
silk gloves when they do it - don't want fingerprints on every glued-on
print!
My guess, having seen the sample dye sub, is that a good printed page in
the magazine probably looks just as good as the dye sub print will. They
can only manage about 1.7 max d. We may print the image on the cover
anyway, underneath the position for the print, so that if they tear the
print off, they don't get a blank white cover.
David
Ted
>
> > Well, your dead wrong. First off the software is about a miilion times
> > better. If someone wanted a scan of a transparency at 600% at 3000 dpi
> > could that inexpensive destop do it. Of course it couldn't. So your
> > either drunk or just stupid.
>
> Drunk most of the time :-) If I wanted a scan of transparency at 600 per
> cent at 3000 dpi, I would have to be drunk before I started. I don't
> print many pages using 2000 lines per inch screening. If I want a scan
> of a transparency, I don't use a flatbed at all and never would. None of
> the desktop scanners I have ever tried are suitable for transparencies,
> and neither are the big rip-off flatbeds.
My Fuji C-550 is just fine and does a great job on transparecies.
> And I didn't think that any of them scanned at an optical 18,000 dpi
> either. More like a maximum of 5000 dpi true optical, and a d-Max of 3.7
> if you are very lucky.
True, but I wouldn't have thought you would know that. Amazing you do know
something.
Been able to get both of those since 1990 with a
> Leafscan. Can now get both from a $750 Minolta 5300dpi 35mm scanner,
> plus ICE3.
Quality's price has definately come down but a skilled operator with a drum
still can't be beat.
>
>
> > Well there is something wrong with the hardware, software, or him. I
> > can see it now in all the trade magazines. Don't buy a high-end
> > scanner because my friend can't get a good scan from one. You know how
> > stupid that sounds?
> >
>
> No, buy a high-end scanner, but don't buy large flatbeds. They are not
> high-end. They involve fairly crude transparency handling and while the
> software can be pretty good, if you need that sort of thing, I have
> never used any software feature on a scanner apart from:
> Framing/cropping the preview
> Setting the final repro size and dpi output for file
> Assigning a colour profile
> Setting exposure
> Very rarely indeed, with bad originals, correcting colour or tweaking a
> gamma curve
Well your not even a rookie than. Why talk about something you know nothing
about?
>
> It has to be said that most photographs we deal with are there in the
> magazines because they are superb photographs - 50 per cent are
> award-winning images, from exhibitions, or just from very professional
> sources. We pick photographs only on grounds of quality. Consequently,
> we do not have to muck around compensating for poor work.
To get a good seperation you do usually have to muck around. Some
seperations may require GCR and others UCR, can you tell the difference?
>
> And we don't do CMYK conversion in scanner software, nor any USM. Nor do
> we crop small parts of pictures - in fact, we breach the copyright terms
> of most images we use and would be heavily criticised if we cropped them
> in any way. Or made them into backgrounds, or colourised them etc.
>
> Frankly, if a scanner is decently designed and basic software is
> correctly implemented, bells and whistles don't matter for this type of
> work. Quality in, quality out.
It sounds that it could be a hell of a lot better.
>
>
> > Most prepress and printers hate photographers for obvious reasons.
>
> Well, we don't, but photographers used to hate prepress (repro) houses
> in the past for ruining their originals. Desktop repro has changed that
> in a wonderful way. Some of the best work we now receive is on CD-R from
> the photographer - and when we ask, we find that they have an Epson 1250
> Perfection or a basic Linocolor, or a Nikon Coolscan. Yet they are
> selling direct scans, made on this type of gear, through top
> international stock agencies.
Still a better drum scanner with skilled operators give you a better
product.
Ted
Jono Moore
jdoh...@nowhere.null.not wrote on 8.5.03 9:14 PM:
> He said RGB for photos. Probably, his photos don't include any type
> (and why would they, after all?).
What, doesn't everybody typeset in PhotoShop?
They seem to around here...
...Jono
David Kilpatrick
Ted wrote:
> True, but I wouldn't have thought you would know that. Amazing you do know
> something.
>
When you have to test the gear for a living, you get to see what really
works, what's hype, what is marketed to suit the preconceptions of the
buyer, etc.
I also get to see trends, real trends. From the 1970s onwards, I had
several thousand transparencies in major photo libraries - Tony Stone
Associates, ACE, Images, Rex, Camera Press and (ancient history, prints
not trannies, from the 1960s when I was at school!) J Allan Cash.
Today, I don't have a single one. All the agencies I dealt with
eventually ceased to deal with miscellaneous stock photography on film,
and returned most routine material. Mine was routine - landscapes,
travel, architectural, a little bit of still life, very few people
shots. All either shut their doors, got taken over, or have a future
workflow which is digital. My trannies came back home.
None of them are of much use. Some, after 20 years of storage, show
signs that the pro labs did not process them correctly; they have colour
shifts or fading or discoloration. Others which had sold, and been used,
are hopelessly marred by unremoved scanner oil or lacquer which has been
allowed to harden (my own repro people in the 70s and 80s used an
ultrasonic bath to clean the originals AFTER scanning, and we paid extra
for this).
For all routine photography now, I shoot digital. We produce Minolta's
own magazine and have done for 23 years now in the UK, and because of
this, we use Minolta scanners; as to whether I would if we didn't...
maybe I wouldn't use the rollfilm scanner, probably would use their new
35mm. We have had some rough repro patches trying to make their early
film scanners produce good results; we always had a Leafscan (now have
three - one original, two for spares!) but sometimes we had to put at
the beginning of the magazine, 'all reproduction in this issue scanned
on the XXXXXX'. I use their Dimage 7hi camera in the studio and out, and
actually don't feel too unhappy with that. It stands up against most 6
megapixel digital SLRs for the limited range of use I require.
We have about 2,500 professional photographers who subscribe directly to
our other magazines, which are not sold by any other means. Ten years
ago we would expect 80 per cent of all commercial work to be on either 5
x 4 sheet film or on rollfilm, and we had a scanner operator full time
on the Leafscan (actually, my son, who now sells advertising instead).
We also had production contracts in from outside, with a very busy
turnover on imageset seps for two outside magazines and many catalogues.
The balance of scanning was 35mm slide (not common for pro work in
Britain) and prints (from which we always managed to get repro to match
transparencies - and there was real prejudice against prints at that time).
My use of prints goes back to 1975, when I developed a method for
cutting repro costs using large R-type assemblies of in-pro prints
mounted on Kodatrace or Melinex foils and wrapped round the scanning
drum (in fact, the very first such job I did went under a Littlejohn
process camera - that would be 1975 - by the next year, the repro house
had acquired one of the first drum scanners). In 1979-81, we introduced
in-pro flexible base repro to Eaglemoss Publications and Marshall
Cavendish, providing complete assembled spreads of 'how-to-do-it'
illustrations for their partworks. The repro managers were sceptical in
the extreme, but we were commissioned by Agfa to write the instruction
manual and do the beta test of their new C66 and colour paper. This was
one of the earliest high speed colour print processors; it gave us the
ability to make precisely colour matched, exactly scale in pro gloss
prints. The leading edge of each print is fixed to the flexible carrier,
the prints are made with a bleed allowance, and the mask is created on a
separate sheet of pin registered foil using red lith film.
I guess this all may offend traditional standards, but it was just an
extension to reflective scanning at 100 per in pro size, of a technique
already used by many UK specialist photographic houses - in pro
transparency shooting. Several of my own assistants either came from, or
went on to work for, the big mail order catalogue studios (Woburn,
Carlton etc). They were taught very economical methods. The main tool
was always the 5 x 4 camera, but film was 'work and turn' shot on half
sheets used masked carriers. A full page was scaled to fit 5 x 4. Any
smaller repros were sketched by the layout artist before they were shot,
and the photographer worked strictly in pro. Tracings scaled down were
placed on the groundglass. Each catalogue page was always assembled on a
lightbox by cutting up the 5 x 4s and making one single foil-mounted
composite, with a mask. Exposure had to be absolutely precise, and we
used the same techniques in my studio - our primary tool in the early
1980s was a flashmeter with a probe to read from the groundglass.
During the 1980s, DTP arrived, and we owned the first LaserWriter in
Britain. Even in 1984, we were trying to use it for setting, though it
wasn't possible. We used a process camera, and Arnold Cook halftone
screens, Letrape, tints, masking film, punches and pins... and we did
assemble some peculiar final film, mainly for product packaging. That
year we had a Nottingham city centre studio operation sharing space with
a graphics company who had a Diatronic with the optical effects add-on.
They made a fortune from distorted type and special fonts. Of course,
you see this happening, and look for other ways to do it.
In short, we always pushed the envelope. We had the first Nikon 3510AF,
we worked with a Hasselblad MacSIE for two or three years, we produced
books where the scans were created with a hand-held Logitek, we spent
£1800 (!) on the Microtek 300 scanner and I built a tranny adaptor which
actually worked. Digital photography started with the Mavica and Ion (on
the Ion, I removed the lo-pass filter... still hopelessly blurred) but
before that we had capture boards and adapted video cameras. The first
capture device was on a Mac 512 Plus. We even had a Truvel at one stage
for large prints and 3D objects.
In 1998 we were still in full flow with scanning, handling hundreds of
trannies every month, and producing hundreds of pages. The imagesetter
could still earn me £3k over the weekend and all I had to do with set
documents running and process the film every 10 metres. That year we
persuaded a large company (Paterson Photographic) to do their 80-page,
800 product catalogue entirely digitally. We used the very first
Fuji/Nikon DSLR (E-2? N-2? can't even remember) which took 1.3 megapixel
images! We did also have, by then, a complete Leaf Lumina equipped still
life studio with high frequency fluorescent heads on ceiling tracks, but
this was too slow. We wanted to cut their costs to £10,000 for
photography - $15 a shot. So told them the pix were only good for single
column use, sold them the idea, and went ahead. My wife did the
photography and we had zero film costs, the camera was borrowed (no-one
had ever done this before) and we completed the entire shoot in 10
working days.
I guess at this stage we knew where it was going, and in our magazines,
we said so. The rest all of you know, though you do seem still to be
living in a world of trannies and prints and scanning.
We are not. I have not seen a 5 x 4 transparency original from a
photographer for over a year now, and I think in the last year I have
scanned about 15 rollfilm originals. 90 per cent of our news, PR and
press images are delivered to us by email; all ads come in via ISDN or
email or on CD. Most photographers supply all their portfolios on CD. I
have to scan about 10 per cent of the images in any given professional
magazine, and about 40 per cent of those for 'Minolta Image' magazine -
even the amateur readers/contributors to that mainly send material on
CD. Our small ad pages are full of people trying to sell Hasselblad,
Bronica and similar gear - and darkroom systems; they can't find buyers.
They are all converting to use Nikon D1X, Canon EOS 1D, 1Ds, and in due
course they will all regret selling their rollfilm SLRS to do so,
because as the prices come down, they will be wanting to upgrade to
digital backs. The major portrait chains already have; two years ago,
the largest one bought Mamiyas with digital backs, they are swapping
this year to Contax 645s with Kodak backs.
The catalogue industry is now 100 per cent digital. The days of highly
skilled shooting in-pro matched, clip-tested E6 sheet film page
assemblies are long gone; they shoot on digital and resize as needed.
I'm contributing to photo libraries again, but now, just digital shots.
Until we get a much bigger camera I just use the Dimage 7hi. It passed
the quality criteria for Alamy.com, in comparison with drum scanned
work. Only 14.4 megabyte files, so I put them in the 'royalty free'
category (fixed fee) rather than try to sell them for usage scale rates.
Now the guys at Alamy are asking me to interpolate the files up 2X,
which in my opinion is utterly stupid - they want all their files to be
over 50 megabytes to impress the buyers, and they say it's fine to size
up digital camera TIFFs. I say they are just creating redundant data and
wasting space and time, and there's nothing wrong with selling half-page
size files; after all, libraries used to sell 35mm, and that's only
barely usable for full pages (the 5 megapixel digicam files generally
look much better than a typical scan from 35mm, at A4 size).
Our imagesetter hasn't seen use for a year; every single printer we deal
with has either converted to CTP, or installed imposition workflow to
film, and all their prices have dropped by 20-30 per cent overall in the
last three or four years. We now pay less for our entire print job than
we used to pay 15 years ago for the REPRO! What cost us £7,500 in 1983
costs us £2,950 in 2003. I used to make that from the charges for
scanning and typesetting alone on some of our contracted titles.
We used to output the film for the Adobe European software manuals which
were printed locally by McQueen (Aldus Europe) long after Aldus became
Adobe. But eventually even they got their own imagesetter. We used to do
all the film for our main contractor printer, then we sold them one of
our old imagesetters; we also sold one to the printers over the road...
and since then of course, having discovered that it was about, they have
upgraded. It has earned them NOTHING AT ALL because they have all had to
throw in the film output free, and then when faced with competition and
the free transferrability of files to different printers, to cut their
print prices. Back in the days when your type was set on a dedicated
page make-up system and ONLY the printer could edit it or use it, you
were locked in. Today, the client creates the pages, and can go ANYWHERE
for output and print. In real terms, British short run colour litho
prices have fallen by an actual 50 per cent in the last 20 years - but
in 1983, the house we owned was worth £50,000 and now the same house is
worth £350,000 (we moved long ago). Wages have increased by a factor of
four to five times. The true cost of repro and print has probably been
cut to 15 per cent of its 1983 'value'.
When we invested in scanners and setters from 1989 onwards, we were the
first in the entire country - the first company in the whole of Scotland
to offer PostScript imagesetting. We made back the entire cost of our
gear in the first year, and doubled what we made by importing complete
Mac DTP system front-ends from the USA and selling them to our new clients.
So, I'ver got no special regrets, but today I would not invest ANYTHING
in high end equipment. OK, an existing contract may justify it, but such
things are ephemeral. Spend £15,000 on a high-end flatbed repro scanner
and by next year you may find it sitting idle four days a week. It's
happened in Britain. Spend £100,000 on a large imagesetting system and
within the next year, all the printers your clients deal with could go
in-house CTP.
I am not always very savvy - I did splash out £4k on a nice used CLC800
with 5000 RIP two years ago, since there was good local business for
colour copying and output. Wrong! I did not know that colour lasers,
producing better colour than my CLC, would fall to £800 price - or that
colour inkjets would end up being used by estate agents (realtors).
Every single local client who used to buy CLC work now owns their own
colour laser or prefers to waste time with an inkjet.
So, I apologise if I offend the professional sensibilities of some by
suggesting that low end is the new high end, and that business doom
waits round the corner unless you can step out of your body and see the
perspective from other viewpoints. I found this NG by accident when
deciding we should either sell the old imagesetter or give it away,
before throwing it in a skip. I'm not wasting my time writing this,
because I write about this for a living, the feedback and response helps
me understand other people's perspective, and writing about my own
experience reminds of me of why we are now working the way we are.
This experience has helped me when advising photographers and photo
labs, who are facing just now the same sea-change that British repro and
printing appear (looking back) to have gone through around 1998-2001.
From postings on this NG, I get the impression that US repro and print
are still very traditional and profitable, with room for craftsmanship,
and there's no way that I denigrate pride in understanding precision
correction of images.
Just concerns me that what has certainly happened in Britain may creep
up on yas all unseen.
David
tsucker
> So, I apologise if I offend the professional sensibilities of some by
> suggesting that low end is the new high end, and that business doom
> waits round the corner unless you can step out of your body and see the
> perspective from other viewpoints. I found this NG by accident when
> deciding we should either sell the old imagesetter or give it away,
> before throwing it in a skip. I'm not wasting my time writing this,
> because I write about this for a living, the feedback and response helps
> me understand other people's perspective, and writing about my own
> experience reminds of me of why we are now working the way we are.
You have all this great stuff, great photographs etc and still don't
know how to make a great seperation. If you want to go all the way
please pick up Proffesional Photoshop by Dan Margulis.
Tsucker
David Kilpatrick
I think I've got it... got about 50 such books (right now I get FOUR
copies of some new ones to review - people seem to think that I'm four
people not one person editing several magazines). It was done a long
time ago.
I do know how to make good seps. When we started out, there was no ICC
colour management. I had to create all our sets of separation curves in
Aldus PrePrint, which allowed very precise control; I also wrote CMYK
conversions for Photoshop in its earliest form. Before that, when
Photoshop did not even exist, we used ColorStudio from Letraset. I still
have the original package - it was a wonderful program. I was a
specialist consultant for Letraset selling Mac systems to the publishing
industry (not for very long, things moved very fast then, but on 25 per
cent commission it was a good ride) and ColorStudio was the Letraset
program.
It included ColorCalibrator, which was not really a 'calibrator' but
actually a very complete solution for writing your own CMYK conversions
(at that time, necessary for repro). Before that I used ImageStudio
which was a superb mono imaging program and had some quite unique
mezzotint and similar screening options which I've never seen reproduced
in any subsequent commercial prog.
ColorStudio and PrePrint both demanded that you work with the printer to
write your own conversions, which we did to a very high standard. We
went through three stages of fine tuning with wet proofs and developed
an A3 'test sheet' file which we also sold to other printers. Remember,
a lot of this was on the Linotronic 200 before the 300 became a
practical setter for Mac; 1693 dpi setting was a luxury and 1200 was normal.
I became UK agent for FontBank but didn't very much enjoy the problems
of differential pound/dollar pricing. We took on HiLine Screening, as a
user and as a vendor, and also offered a service manually calibrating
imagesetter ppds (or, again, at that time APDs - Aldus Printer
Descriptions) not using any utilities but writing them.
We had two or three years with this sort of very manual,
measurement-intensive work before the ICC introduced colour management
to Macs in 1993, and simultaneously there was a big change in
imagesetter resolutions and Adobe introduced Level 2 which altered the
whole colour dictionary thing.
I phoned Aldus in real anger when PrePrint was killed and colour
management based solutions took over. I was far from happy with
Photoshop, compared to the power of ColorStudio - which was far closer
to the high end packages on Barco systems (I had visited Barco in
Holland during the development of the first accessibly priced ones).
But... the new stuff WORKED. I got over my anger at no having a job
where I could charge large sums just for knowing which figures to alter
in a file. Just the same way I got over no longer being paid to write
bitmap fonts...
We've had some difficult run-ins with early versions of ColorSync and at
one stage, we abandoned any attempt to work that way and invested in an
entirely Agfa-based colour workflow (this also allowed you to create
your own tags and links, returning some control). And when that ceased
to be workable we tried buying an expensive large suite of Kodak
profiles (mistake - they contained so many errors it was laughable).
Ten years later I really couldn't be bothered to want to get back into
'controlling' rather than just 'using' the technology. The colour
science has been refined, the whole process is so far superior to the
way things were; even the colour spaces changed, with better gamuts.
What says it ALL to me is PIM II. I can shoot images on a digital
camera, stick the card in the appropriate Epson printer, and get prints
which so far exceed the quality of photo prints or litho repro I can't
believe it. PIM II is simply a closed CMS. I went to Milan for the
launch by Epson which first introduced this and EXIF2.
It may only apply to a few inkjets and some 'partner' digital cameras
right now, but it is a glimpse of the future. There is no reason why a
print workflow for litho should not be PIM compatible and accept a file
directly from a digital camera and go straight to a perfect sep. In
theory the camera could even analyse the distribution of fine detail,
balance of micro to macro contrast, and instruct automatic GCR and USM
levels.
In 1975-ish Krasna-Krauz interviewed me about a book proposal. My
proposal was to write a book on automatic exposure. He laughed and said
Focal Press could never publish such a thing. Automatic exposure was for
children! There was nothing to know. How could I write about it? Well I
never did write it, but of course, dozens of books have been written
about it since.
Automatic colour management is much the same. How can you rely on
something you can't control? Well I do, and like auto exposure, it seems
to work most of the time only most is 99 per cent not 60.
David
Fixx
David Kilpatrick <icon...@btconnect.com> wrote:
> For all routine photography now, I shoot digital. We produce Minolta's
> own magazine and have done for 23 years now in the UK, and because of
> this, we use Minolta scanners; as to whether I would if we didn't...
> maybe I wouldn't use the rollfilm scanner
Well if you get tired with your Minolta Multi Pro I can take it :-)
> I'm contributing to photo libraries again, but now, just digital shots.
> Until we get a much bigger camera I just use the Dimage 7hi. It passed
> the quality criteria for Alamy.com, in comparison with drum scanned
> work. Only 14.4 megabyte files, so I put them in the 'royalty free'
I have sent scans to Alamy which are 70-80 MB range, they seem to like
them...
> there's nothing wrong with selling half-page
> size files; after all, libraries used to sell 35mm, and that's only
> barely usable for full pages (the 5 megapixel digicam files generally
> look much better than a typical scan from 35mm, at A4 size).
I do get 14MB digital camera files for a magazine... I think I would
prefer 35 mm chromes as grain in 35mm scan looks better than general
softness and "digital graininess".
> So, I apologise if I offend the professional sensibilities of some by
> suggesting that low end is the new high end
All equipment prices come down and printing costs less. It is a good
thing. Progress on the other hand is slow and inhouse ctp is not for
everybody for a long time.
And you always need some good iron to get a good scan from a chrome
though and there are always jobs which must be done with premium quality.
> From postings on this NG, I get the impression that US repro and print
> are still very traditional and profitable, with room for craftsmanship,
But I admit there is work done here in Finland with amazingly low end
equipment. It shows though. -Fixx
David Kilpatrick
Fixx wrote:
> I do get 14MB digital camera files for a magazine... I think I would
> prefer 35 mm chromes as grain in 35mm scan looks better than general
> softness and "digital graininess".
>
The Minolta Dimage 7 series is a big exception to softness. I do shoot
without sharpening, but from testing the Fuji, Nikon, Casio, Canon, and
this week the Contax Digital TVS, and also Minolta's other 5 megapixel
(F300) there's no contest on the sharpness front. The 7/7i/7hi simply
has the best lens on any pro-sumer digital as well as the only one which
achieves the same wide angle as a 24mm does on a 10 x 8 print from 35mm.
They call it a '28mm' but they are calculating on the diagonal angle,
and the 5 megapixel format of these cameras (2560 x 1920, most of them
use the same final size) is a 10 x 8 ratio not a 35mm 12 x 8 ratio.
Also, the cameras are pretty much the same res as 6 megapixel SLRs -
those typically do 2000 x 3000 or a similar size, so the prosumer boxes
are only 80 pixels short of the same width.
Digital noise is undoubtedly there, but personally I find it more
acceptable than grain, especially as there is no dust and the image is
normally pixel perfect. I tend to work only at the ISO 100 filmspeed
equivalent, and the grain really is only there in certain colours - it
happens to show most in blue skies, which of course gets noticed.
Actually, it's better to have some grain. A perfect 'grad' colour is not
as easily to reproduce well on the printed page as a slightly grainy
one, and we always used to add noise to Photoshop grads to make sure
those 256 steps never produced a tone-break or banding.
Last year we ran an interesting test - Canon D60 image versus a
Fujichrome Velvia tranny scanned to the same scale on a Nikon 4000 dpi
scanner - both reproduced A3+ bleed at 200 line screen. The Canon D60
image looked SO much better than the Velvia it was hard to believe.
If you are interested I will send you a copy of Minolta Image where I
put a full page repro of a quick shot I did at a demonstration - just
set up a single light still life, to show how direct card printing
worked on an Epson 925. We had the laminated back cover of the mag going
begging, so I decided to repro the image approx 10 x 8 on that. I remain
surprised at just how good it ends up looking - probably as good as we
used to get from scanning 5 x 4 shots in the early 1980s. Send me your
mailing address to my own email and I'll shove a copy in the post; PR
for Minolta!
Incidentally - the Casio 5700 is the 'next' sharpest camera I've used in
that class. The Contax TVS Digital is so far the least sharp which
astounds me, with the huge price-tag and Zeiss lens.
David
Ted
> >
> >
> > You have all this great stuff, great photographs etc and still don't
> > know how to make a great seperation. If you want to go all the way
> > please pick up Proffesional Photoshop by Dan Margulis.
> >
>
> I think I've got it... got about 50 such books (right now I get FOUR
> copies of some new ones to review - people seem to think that I'm four
> people not one person editing several magazines). It was done a long
> time ago.
When you review it, keep it, it is the best. Can you post the review so this
NG can read it? Or at least give us a link.
>
As for the rest of the post you tend to jump to much for me. Kind of long
winded.
Ted
Ted
> >
> >
> > You have all this great stuff, great photographs etc and still don't
> > know how to make a great seperation. If you want to go all the way
> > please pick up Proffesional Photoshop by Dan Margulis.
> >
>
> I think I've got it... got about 50 such books (right now I get FOUR
> copies of some new ones to review - people seem to think that I'm four
> people not one person editing several magazines). It was done a long
> time ago.
When you review it, keep it, it is the best. Can you post the review so this