RE: experiments with printer drivers (long)

[Ian Young <iay@3L.co.uk>]

> From: Ian Lyons [mailto:ilyons@email.msn.com]
> Sent: 31 December 1999 21:48
> To: epson-inkjet@leben.com
> Subject: Re: experiments with printer drivers (long)
> 
> A really neat piece of sleuth work Ian <ggg>.

Thanks.  Now, is <ggg> a grin or a giggle? :-)

> However, I think that if
> you followed the Colorsync list you may well have saved yourself the
> bother, and few guys on this list know this fine well (or 
> they have very
> short memories <GGGGG>). 2nd September 1999 was the day that Windows
> users pray for and most of you missed it. What happened ??, simple the
> question you asked, got no answer and then decided to experiemnt with
> was answered in PUBLIC by none other than the ICM2 Quality Assurance
> manager for the W2K project.

I hadn't expected a Colorsync list (which I equate with "Mac", of course)
would have much of general interest in it.  It turns out that I need to
subscribe to a few more lists :-(

I have read that thread now, and having read it, I would still have
performed the experiment I did.  Why?  Well, the first thing is that having
read it, I'm not sure that the guy in question uses some of the words the
same way as I do. :-(

Secondly, the question asked was whether "Windows assumes sRGB on untagged
files" (and, at least to me, that's a malformed question), and that wasn't
the question under discussion here.  The _answer_ given was far from
precise, also:

>>In ICM2 printing, using a print driver where a printer
>>profile is associated and used, this [sRGB assumed for source]
>>is true for untagged images.

It's not at all clear whether the situation we're talking about (printing to
a printer using "colour adjustment / no adjustment" rather than "ICM" mode)
qualifies as "ICM2 printing".  I don't think the above really adds anything
to our knowledge, except for another strike against Microsoft's score on
properly explaining what they are doing...

Of course, he then went into the standard Microsoft spiel about how
everything is sRGB anyway, as far as Auntie Doris was concerned, thus making
everyone on that list (except Auntie Doris) cross with him...

> But first to some of your points:
> 
> >Again, a clarification: if the Epson printer driver sees
> >some untagged RGB data coming at it, not necessarily from
> >Photoshop, and I have selected a color management mode in
> >the driver of "Color Adjustment" and a color adjustment mode
> >of "No Adjustment" all in an attempt to make the printer
> >driver a stable target for a profiling program, you seem to
> >be saying that rather than assuming that RGB data is in, say,
> >some fixed space it is in fact looking at the default profile
> >for the primary monitor instead (using Win98SE terminology).
> 
> Not quite so, it depends on the program from which the print driver is
> called. If called from within Photoshop then driver just 
> takes the data and applies its own internal processing (not
> ICM engine related) to the incoming RGB data. Depending upon
> the printer mode this can be from virtual data destruction
> (PhotoEnhance with sharpen or PhotoUnrealistic)
> to nothing, Aka No Color Adjustment mode.

I started my Epsoneering life with some of the other driver modes; I think
the original place I saw a recommendation to use No Color Adjustment mode
was in one of your articles; I'm indebted to you for that as it certainly
seems to tell the driver to treat every image the same if nothing else :-)

However, if what is actually happening is that it is telling the driver to
treat everything as sRGB then I'm a very unhappy bunny.  I want my cyans,
dammit.  I know they are in there somewhere...

> Outside of Photoshop and a very few other programs the print 
> driver MUST
> assume some "Source" Colour Space, and that it is where the real
> confusion seems to lie, at least for some.

Well, yes, although I'm not sure why you would distinguish Photoshop with
"printer color management" unchecked any other application, from what you've
said elsewhere surely they are the same from the driver's point of view?

> Seriously, the drivers work different ways because NOT everything in
> PhotoShop PCM engine worked the way Epson were told!!!. So with the
> 1200/750 they cut their losses and TOTALLY ignored the feature, with
> these printers ON/OFF it doesn't matter a jot, likewise the 
> ICM button, and I ain't getting into that debate again!!!

I haven't seen that debate, but if you are saying that the "photo" printer
drivers ignore image colour space information I'd be a little surprised if
that were the case.  If you're not prepared to go through this again
on-line, can you send me one of your old messages off-line or at least a
pointer to where the discussion happened?

> >The question, restated, is whether if you just chuck RGB values at a
> printer
> >driver without telling it which colour space the values are in, what
> space
> >does the printer assume?  Mr Tobie's assumption is that it is the
> monitor
> >space.  I think that this may be true for some people but 
> very much not
> true
> >for others.  If you're interested in performing an experiment to find
> out
> >how _your_ system works, and you're prepared to take a stab at some
> system
> >tweaking, read on.  Otherwise, this is all pointless ramble and you
> should
> >return to your seasonal festivities...
> 
> Not really that interested to be honest since the answer is written in
> tablets of stone (Colorsync List Archives). The source is 
> assumed to be sRGB if untagged.

I disagree with your interpretation of that discussion.  What you say may be
true, of course, but I don't see it written on those tablets you're talking
about.  I guess if you're getting results you're happy with, you would tend
to pull back from overturning every stone that comes along.  BUT, as far as
I can see, such a conclusion would imply that "printer color management"
unchecked, "no color adjustment" mode would be a guaranteed road to ruin for
profile creation.  I will, therefore, personally, not rest until I can prove
that assertion one way or another.  If it turns out to be true, I guess I
have to dump my printer and get something supported by PressReady or some
other driver-bypassing RIP rather sooner than I had hoped would be
necessary.

> Question: by Andrew Rodney, and avery long list of Mac GURU'S
> 
> "Is it true that when Windows finds an untagged file, it 
> assumes it's in
> sRGB?"
> 
> BTW: Ian,  I think thats what David Tobie meant when he wrote monitor.
> Since we PC users seem incapable of telling the difference how would a
> mere Mac user know differently <GG>.

Somehow, I rather think David Tobie knows the difference between "the
monitor space" and sRGB in particular...

> Answer:
> 
> A quote from Mr Mike Roulette ICM2 Quality Assurance /W2000 Microsoft
> Corp

[...snip a huge rant from Mike Roulette about how great sRGB is...]

Honestly, I think the only actual substantive part of Mike Roulette's answer
is the couple of lines I quoted before; the rest of it is just the Microsoft
party line and doesn't actually address the question in any useful way...

> Now in plain english what this means is - with some apps the assumed
> profile for untagged data is "sRGB" (BTW, Mike went on to 
> write that so far as Mirosoft are concernd ALL monitors are sRGB near
enough).
> 
> Back to PCM Checked or Unchecked - If Checked the RGB data  
> is Tagged so from the above discussion one would assume that the RGB print

> data from Photoshop is "Tagged" with a Profile - QUESTION, Which 
> profile?????, and it goes on and on and on and on, etc.

If the printer driver assumes untagged data is sRGB, this is indeed now the
interesting question.

> ANSWER,  it ain't what you would
> think from the spec. sheet <ggg>.  As I said at the outset Epson got
> "burned" with the 700 and EX, a screwup in Photoshop PCM meant "Double
> Profile" potential. However, with 750/1200 Epson just ignored the PCM
> tag, so unless you apply the profile in Photoshop NONE gets applied -
> END OF STORY.

Really?  Bizarro.  Of course, I don't care about those people with expensive
6-colour ink sets, so I will try to see if this is also true for my lowly
740.  With luck, I can still work round this.

> So you see Ian, someone had you chasing up ablind alley <gg>.

Yes, but he took it very graciously, didn't he?  And I have learned a lot
from it, although it may take me a couple of months to catch up on my E-mail
as a result :-)

> I hope
> you grasped all that, because its been written a million times by guys
> that have "skidded further than you or I have driven" <ggg>.

I think I understand what you're saying, if that's what you mean by
"grasped".  I still feel I need to confirm some of it, given that the some
of the implications might be quite costly by my standards...

> Appologies to all those  had to read this again!! - I promise 
> NOT to do it again, honestly.

I guess if you think people are getting bored by this, we could continue
this discussion off-list.  Unless, of course, you really believe that all of
this is so completely clear-cut that it doesn't bear discussion even there?

> (not always such an obnoxious g--)

Don't sweat it.  If that's your idea of obnoxious, you need to get out
more... :-)

	-- Ian
-
Turn off HTML mail features. Keep quoted material short. Use accurate
subject lines. http://www.leben.com/lists for list instructions.