Fuji Frontier generic profile

I was about to take a couple of prints into my local print shop who use a
Fuji Frontier 370 Minilab. I've been reading up on the process involved in
preparing an image, and they all seem to assume that I have a colour
profile for the exact machine/paper used. Well, I don't and I'm not sure I
can get one. I'll ask at the lab tomorrow.

In the meantime, I kind of assumed that a generic profile for the Frontier
370 would be the next best thing, but I can't find one of those either.

What are my options here?

 

The Frontier is set to expect sRGB images, because that is what 99% of
all customers will bring.

 

G'day!
Colab in the UK offer their Fuji 370 ICC profile for softproofing. You
can get a copy here:
http://www.colab.co.uk/digital/colour.html and select the link 'ICC
Profiles'. A javascript window will then open with the link contained in
a spiel about colour management. If for any reason you still can't
d/load it, then give us a shout and I'll email it.

HTH

 

The Frontier is set to expect sRGB images, because that is what 99% of

That would imply that the profile I just downloaded (thanks to those who
pointed - Google let me down!) is just a renamed version of the sRGB
profile. A little experimentation with soft proofing shows that the sRGB
and Fuji profiles are very similar in most areas except for blue. Changing
the soft proof profile changes the colour of the sky on my test image quite
notably.

 

No. it does not. It only means that the Frontier is set to get the best
results with sRGB images. So if you cannot get hold of a Frontier
profile, use sRGB.

The Frontier profile is a device specific profile however, sRGB is an
abstract color space. The chance that the Frontier is EXACTLY the same
as sRGB is vertually zero. It's always better to use a device specific
profile if you have one.

BTW, do not forget that you still do not have a profile of the Frontier
*you* use! You only have a profile of *one particular* Frontier. It may
be a better option that using sRGB, but that is not true by definition.
If your lab uses different paper and/or different chemistry in its
Frontier, the results may still be quite different.

 

That would imply that the profile I just downloaded (thanks to those who

Oh, yes. I'm with you. }

I've now found 2 profiles for the Frontier 370. Soft proofing suggests they
are slightly different.

Yes, I understand. The machine I'll be using isn't in the DryCreek database
so I have to go with generic. I'm going to get 3 sets of prints done: one
for each of the Frontier Profiles I have, and an sRGB set. I suspect this
experiment will just end up providing me with more information but it could
be interesting nevertheless.

 

http://drycreekphoto.com/Frontier/FrontierDatabase.htm

 

That's what I'd expect.

Indeed. The results *should* be very similar, but you'd be surprised
what some labs can do to screw up their settings! ;-)

 

Yeah, but they don't have one for the the machine I'm going to use (or in
fact any in Perth). If the boss of the lab is in tomorrow I'll make sure he
knows about DryCreek.

 

From: Derek Fountain

Hey Derek, I'm sure a lot of us would be interested in hearing if you see any
differences in these.

As mentioned, very few of the people who use these labs even know what an ICC
profile is and have no idea how to convert to it. Most of their cameras
default to sRGB so that's what the labs expect. Would be interesting to hear
if you see any differences between sRGB and the converted files

Bill

 

Then download the test targets from Dry Creek and have them printed up. It
is free, how hard can that be?

It won't help you tomorrow, but then neither will any generic 370 profile as
they are, in my experience, variable to say the least.

For now, just use sRGB and look at the output. For real control, print the
Dry Creek targets and have Nathan profile the printer for you.

Tom

 

Then download the test targets from Dry Creek and have them printed up.

I was under the impression they'd stopped doing that?

http://www.drycreekphoto.com/Frontier/Targets/TargetPreparation.htm

 

That would imply that the profile I just downloaded (thanks to those who

Having had chance to sleep on this, I'm still a bit puzzled. There's a
colour space in Photoshop called sRGB IEC61996-2.1 which I understand isn't
a real profile, but a "fits nothing quite perfectly" abstract colour space.

Q1) Is that colour space the same one as the images from my digital camera
(which tags images as "sRGB")? Or are there "flavours" of sRGB?

Q2) Assuming the answer to Q1 is that there is only one definition of sRGB,
why isn't the generic Frontier just a copy of sRGB? That is, why can't the
machine be calibrated so its idea of colour is identical to that of sRGB? I
understand that each machine would ideally still have a specific profile to
adjust for local conditions (temperature, etc.). I guess it's a case of
changing the universe to fit the equation, but as a generic starting point,
wouldn't having the machine really set to sRGB be best for all those people
taking sRGB images in?

What, in short, is the difference between having a Frontier "set to get the
best results with sRGB images", as you say, and actually setting the
machine so its profile actually *is* sRGB?

 

You cannot "calibrate a machine to sRGB". What you can do is set the
machine to match sRGB as closely as it can, but there will always be
differences between the abstract sRGB color space and the actual color
space that a Frontier can handle. The Frontier profile describes that
actual color space. That's why the Frontier profile can be *very close*
to sRGB, but the chance that it is an *exact copy* is very small.

 

From: Derek Fountain

If you want to invest in a book I'd suggest "Real World Color Management" by
Fraser, Murphy, Bunting or "Real World Photoshop" by Fraser and Blatner.
Couple of good chapters on CM in the second book, and the first one covers
pretty much the entire field well.

There are actually several kinds of profiles. Ones for a specific monitor or
scanner of printer/paper/ink combo are called "device specific". Then there
are abstract ones like sRGB, AdobeRGB and Ektaspace that don't match to any
specific device but instead define a gamut range that encompasses various
device types or targets. Roughly, sRGB was meant to match the average
uncalibrated monitor, AdobeRGB was I think originally matched to HDTV
(according to "Photoshop Artistry" author Barry Haynes), Ektaspace is wider and
tries to match the gamut of Ektrachrome color slide film. You can make your
own custom working space very easily in Photoshop if you want something else,
it's simple to do, and you can name it 'Derek-Space' .

The advantage of using an abstract space as your working space is that it's
grey-balanced (device specific profiles MIGHT be but rarely are) and it's
perceptually linear. This helps in editing. Then when you're ready to output
to a particular device you can convert to that profile to get the best match
between the two dissimilar gamuts. Usually you pick a working space to
encompass the gamuts of the devices you plan on outputting to.

Each device has its own gamut or range of colors it can reproduce. For sure
the Frontier can reproduce many more shades of colors (has a wider gamut) than
sRGB. As a gray-balanced example, in sRGB if R=G=B then you have a shade of
gray by definition ... but if you get a proof color reading from say
127/127/127 from a typical inkjet profile it will be something wildly
different.

They just assume the input files are sRGB and then convert them when they
print. In theory they could convert from any profile (AdobeRGB or a wide gamut
one like Ektaspace) but they are trying to automate everything to the lowest
common denominator like the one-hour C-41 photo print places they've largely
replaced, and sRGB is the lowest common denominator.

It's not possible to set the machine so it "is" sRGB since the printer has its
own native gamut or range of colors. Probably there are a couple of shades in
sRGB the printer can't match, and there are thousands of shades the printer can
print that are out of gamut for sRGB. The best you can do is convert to the
actual profile that matches the printer.

"Real World Color Management" is highly recommended.

Bill

 

In the meantime, I kind of assumed that a generic profile for the Frontier

I suspect a lot of people are fed up of my honking about colour management
these days, but I thought I'd finish the story, both for the benefit of
people who are still interested and for those who might someday find this
thread via google or similar and are considering treading the same path.

I downloaded a generic Fuji Frontier profile from here:

http://www.popphoto.com/article.asp?section_id=4&article_id=480

then another one from here:

http://www.colab.co.uk/digital/colour_management/icc.html

I didn't know the difference at the time, but it appears they are for
different paper types. The first is for Fuji Crystal Archive paper, the
second for what appears to be a paper named something like "Lustre".

I got 3 photos, all in sRGB colour space from my camera. One is a shot of a
walking trail here in Western Australia and consists mostly of reddish
coloured soil and rock. The second is a shot taken from a boat at sea and
is mostly light blue water. The third is a park scene in England and is
mostly green trees and grass.

I converted each photo 3 times, according to this page:

http://www.drycreekphoto.com/Frontier/using_printer_profiles.htm

Each photo was converted to sRGB (which it already was, but I did the
conversion anyway), then each of the 2 Fuji profiles. I had all 9 photos
printed at a lab on a Fuji Frontier 370 with no colour corrections. The lab
used Crystal Archive paper, so all other things being equal, that profile
should have been the best. (Of course, all others things are not equal, as
has been pointed out in this thread.) I then printed each photo on my
Epson895, allowing Photoshop to do the colour management and having the
printer driver set to no colour adjustments.

Laying the photos out for comparison shows the following:

1) The reddish photos from the Frontier are all just about identical. I have
to study them very carefully to find any areas which look even slightly
different. The photo from my Epson is more heavily saturated in the reds,
perhaps by about 15%.

2) The bluish photos from the Frontier are all notably different. The sRGB
one has the water as really quite deep blue, while the other two are closer
to each other. The blues are still notably different though. The photo from
my Epson is again more heavily saturated than any of the Frontier ones.
It's not as overcooked as with the reds in the other photo, but it's still
maybe 8% more saturated than the Crystal Archive profiled one.

3) The green photos from the Frontier are all very similar, bordering on the
identical. They mostly differ in the yellowish areas where sunlight is
hitting the grass. Once again, the Epson photo is over saturated. The
greens are deeper by about 8%.

The real question is how do the photos look against their soft proof
versions on my supposedly (Adobe Gamma) calibrated monitor? I compared the
Fuji Crystal Archive print and the Epson print for each photo. The answers:

1) The red Fuji photo is pretty good. Subjectively about 90% accurate. The
Epson print is way off - the reds are far too saturated as usual. I gave it
70%.

2) The bluish photo from the Fuji is pretty close to spot on. I gave it 95%.
The Epson version is also pretty good - 90% accurate.

3) The green photo from the Fuji is very good. I gave that 92%. The Epson's
greens are not quite so good. Light greens are very close to the screen
image, but the dark greens are too dark on the print. Overall 85%.

My conclusion: based on how the Fuji soft proofs look, my screen is, I
think, within an acceptable area of accuracy. Better than 90% in all three
cases, and given that I don't have a proper calibrating device or a proper
profile for the exact Frontier I used, that seems reasonable to me. If I
needed professional colour matching I'm pretty sure I could get it via a
decent screen calibration system and possibly a specific Frontier profile.

I further conclude that the Epson printer profile isn't accurate enough. It
would be just about good enough for my requirements if the reds weren't
such a long way off, but as it is, it's not much use. Maybe more
modern/expensive Epsons have better profiles. I'll be sure to ask in this
newsgroup when I start considering a replacement. In the meantime I'll use
the Fuji Crystal Archive profile and the Minilab to get printed images.

 

Have you considered buying a custom profile for your printer? It
sounds like that would solve your Epson problems, or at least go a
long way in fending them off.

 

Actually, I'd conclude that you're using sRGB as your image profile.

The reason I say that is that sRGB has more out of gamut colours for
printing. Whilst the Fuji is set up to correct that, your Epson
printer likely isn't. So, when the conversion from RGB to CMYK is
done in the printer, you're not getting the same result.

I'd be interested to see what you get in the soft proof window in
Print Preview. You should be able to see the "too redness" before you
print. In which case why print it like that?

 

I further conclude that the Epson printer profile isn't accurate enough.

The thought occurred, but I think I'd rather spend the money on a more
modern photo printer with colour profiles which are reported (by folks in
this newsgroup) as accurate.

Alternatively, since I only occasionally print a photo (most of my photos
end up in web-based photo albums) common sense is telling me to shelve the
Epson and just take photos into the Minilab when I want print copies. They
did each image for one aussie dollar (cheaper than the Epson I suspect),
and I got better than 90% colour accuracy. A new photo printer might be

a Spyder for my screen seems like a much better investment.

 

Actually, I'd conclude that you're using sRGB as your image profile.

Er, in all of this I'm not sure I've come across the term "image profile".
You mean my working space? Yes, sRGB.

I check the images for out of gamut colours in the Epson-profiled Photoshop
soft proof window before sending them to the Epson. If there are out of
gamut colours (which is quite frequent, as you say) I tweak them until the
gamut warning disappears.

Which print preview? The image in Photoshop's Print Preview is far too small
to judge. I read that the Print Preview facility in the Epson driver wasn't
colour controlled in any way and therefore wasn't any use in predicting
what shades the printer will actually print. I have yet to come across any
form of soft proofing that indicates the reds will print too saturated. If
I could find such a thing I'd use that as my soft proof feature and the
problem would be solved!