Ink Splotting - ImagePrint - Roadmap for further testing.
Ink Splotting - ImagePrint - Roadmap for further testing.
- Subject: Ink Splotting - ImagePrint - Roadmap for further testing.
- From: Nick Wheeler <email@hidden>
- Date: Wed, 29 Jan 2003 09:49:18 -0500
>
Eric Bullock wrote:
>
>
> I had one tinceey question about the System tab in ImagePrint. If you are
>
> intending to send a RGB chart through with no color mamangement, what do you
>
> choose for "printer" in the system tab? There is no "none", and the manual
>
> simply alludes to the fact that maybe that is only for screen matching. I
>
> also don't see how you can just send the chart through without regard to the
>
> media type. Doesn't some kind of ink limiting need to occur? Sending some
>
> paper to ColorByte is starting to sound appealing after all of this
>
> discussion.
>
>
Eric:
Yes by all means get ColorByte to make you a custom profile (another teensy
problem is you have to send them your printer, ink, temperature and humidity
as well). The problem has been that at the moment this takes time which a
lot of us don't have. Searching for a viable alternative is what has us
spinning round and round.
In ImagePrint the "profile" is not really working the way we normally think
of profiles working. There have been many posts to this list on this
subject. Essentially it appears that the "profile" is a fundamental piece of
the linearization in IP. Check out my post to this list yesterday. Think of
it as maybe like a "media" setting. Just my best guess at the moment.
So the idea is find a ColorByte profile that gets you close and then do your
own profile with that ColorByte profile in place. Just like finding the best
media setting in a printer driver.
The Neil Barstow quote in my post yesterday represents the only way I know
of to do a profile now. But in that scenario you would have to profile for a
given RGB workingspace. ImagePrint documentation tells you to turn off the
workingspace which is what I have done in the past, but I now wonder if this
is really working properly.
Essentially you are profiling the IP profile, which seems bizarre to me.
Can't see how this could be then used in IP to replace their profile. How do
you profile with a profile and then toss out the old profile and substitute
the new one, that does not make any sense to me but I haven't tried it yet.
Wouldn't that be like profiling with a given media setting and then tossing
out the media setting and putting your new profile in its place? Also I'm
not seeing how the IP supplied profile could be used in view proof setup in
Pshop. Don't know the answer to any of this, I have not tested any of this
stuff.
This is how the mind of an idiot works - you ask a dumb question, come up
with the wrong answer, test and decide you were right after all. Use your
great discovery for a year and find out how really wrong you were to begin
with.
Get drunk,
Repeat.
One can get through an entire lifetime that way. Works for me.
You can select "Select Printer Profile" in that drop down dialog you
describe and print. See Andrew Rodney's post. You will quickly see what I am
talking about. Also print a tagged file and then convert that same file to
Lab and print. Observe the results carefully. Try it with real pictures and
with ramps. Let me know what happens. I'll be interested to see if your
conclusions match mine from yesterday. (Incidentally don't toss out that
select printer profile profile, it croaks the app).
>
Andrew Rodney wrote
>
>
> It has to be an untagged file AND in BitMap, you have to have RGB set to OFF.
>
> That is a similar setting to "same as souce" (but nowhere as obvious) whereby
>
> the data gets pasted through to the printer without the profiles being used.
>
> The actual setting "Select Correct Printer Profile" is itself just a profile
>
> with the internal named set as such!
>
>
Problem here is if John P was giving me the right scoop "it all goes to Lab"
(and my tests from yesterday confirm this) how is IP handling that untagged
data. According to the information that John gave me as well as my testing
from yesterday - the incoming data has to be converted to Lab somehow. Uses
the source profile as Andrew suggests (same as source)? But I have done this
with untagged tiff files - what then? Some sort of default RGB or CMYK. I
haven't got a clue. That is why I am saying you have to profile for a given
source space. With IP a source profile of some kind would seem to be
essential in either RGB or CMYK.
One test might be to convert your test chart to Lab and print it and then
profile. That is the next idea I plan on testing. I think this is why folks
are getting good results with either RGB or CMYK profiling. You can see how
either might work better depending on individual circumstances. RGB should
be better in theory, but CMYK might be better as a practical matter due to
more appropriate test charts, profiling software peculiarities etc.
If I am right about the Lab thing then it's how you best get into Lab that
is important. Ultimately the answer of which working space to use is
probably far more dependant on which space you are used to working in.
It's a conundrum. Henrik is right.
>
Henrik Holmegaard wrote:
>
>
> Eccentricity in UI design locks out the user, making it a toss how to deploy
>
> the software. It should be simple to make an ICC profile and simple to deploy
>
> it.
>
>
There have been many posts to the colorsync list now on ImagePrint and I
would imagine the subject is getting tedious for most readers and
contributors to that list. There is an ongoing discussion here:
>
on 1/29/03 12:17 AM, Thomas Fors at email@hidden wrote:
>
>
> Hi Nick,
>
>
>
> Very informative thread. I like to see people getting down to doing real
>
> testing of the product because the answers colorbyte gives me always leaves
>
> me wondering.
>
>
>
> I've started an ImagePrint yahoogroup at
>
> http://groups.yahoo.com/group/ImagePrint for the discussion of the product,
>
> and I was hoping I could perhaps persuade you to join and cross-post your
>
> test results there as well. We have about 80 members at this time.
>
>
>
> Thanks --Tom
>
>
I will be posting there as well as some other places in the future - no more
bugging the colorsynchers.
Best wishes
Nick
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.