Yahoo Groups archive

Digital BW, The Print

Index last updated: 2026-04-28 22:56 UTC

Thread

Re: [Digital BW] Re: My workaround for the PS6 -> PS7 posterization problem...

Re: [Digital BW] Re: My workaround for the PS6 -> PS7 posterization problem...

2002-07-13 by Martin Wesley

----- Original Message -----
Show quoted textHide quoted text
From: "jimhayes361" <jimhayes@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Saturday, July 13, 2002 8:12 AM
Subject: [Digital BW] Re: My workaround for the PS6 -> PS7 posterization
problem...


> --- In DigitalBlackandWhiteThePrint@y..., "coveylaw" <brad@c...>
> wrote:
>
> > Sounds great, but i don't have PS6-just PS7.
> > Any thoughts.  I am very frustrated with the posterization and it
> > makes my prints unacceptable.
> >
> > btw has anyone tried nik color efex! to convert to b&w?
>
> So someone else IS getting posterization.
>
> The real answer is to simply have the curves redone by Paul or whoever
> is handy with a spectrophotometer. This is a lot of work for all the
> pritner/OS combos though. The curves should be redone in Adobe98
> rather than sRGB. I'd respect Paul he he refuses to go through it all
> again.

Jim,

Then what would happen to all the people who do not have this problem? This
is something that is showing up in 1 or 2 systems so far, redoing the curves
would only be justified if the problem were universal.

>
> He might be able to do it with a transfer curve, I don't know. He
> hasn't responded to any posts on or off list from me on this topic,
> unless I missed it.
>
> Also, there is still the unresolved issue of whether I get
> microbanding even using Adobe98 with PS7. I can't test it right now
> any further, as I'm short on ink to squander. My test print also still
> shows just a tiny bit more posterization in PS7 vs PS6 when using
> Adobe98. And of course the tones are different using this space. It's
> really crying out for a curve tweak.

I am printing just fine from PS7 on my 1280. No microbanding and the quality
is unchanged from PS6. Any tweak would then give me and everyone else
trouble. Why not tweak the curves yourself for your own use?
>
> Then the other solution is to just use PS6 until the 2200 comes out
> and go for it. That is a big unknown as well.

BTW, I have seen some B&W 7600 output which uses the same inks as the 2200.
The results are similar to the 2000P. Odd tonal splits and a print that is
fairly neutral in tungsten lighting is yellow/green in daylight. I don't
know if these were good or poor examples of what the machine can do. I want
to see some more but if what I saw is "good" then I doubt you would be happy
with the hue or metamerism.
>
> You could buy an older computer and network it up on the OS of your
> choice with PS6 and dedicate the printer to it. A bit of bother there.

Since I can't duplicate your problem using very similar equipment (I think
the only difference is that you are running Win98SE and I am on Win 2000), I
would urge you to continue to look for something in your set-up that is not
quite right.

Martin

[Digital BW] Re: My workaround for the PS6 -> PS7 posterization problem...

2002-07-13 by jimhayes361

--- In DigitalBlackandWhiteThePrint@y..., "Martin Wesley" 
<mwesley250@e...> wrote:
> ----- Original Message -----
> From: "jimhayes361" <jimhayes@j...>
> To: <DigitalBlackandWhiteThePrint@y...>
> Sent: Saturday, July 13, 2002 8:12 AM
> Subject: [Digital BW] Re: My workaround for the PS6 -> PS7 
posterization
> problem...
> 
> 
> > --- In DigitalBlackandWhiteThePrint@y..., "coveylaw" <brad@c...>
> > wrote:
> >
> > > Sounds great, but i don't have PS6-just PS7.
> > > Any thoughts.  I am very frustrated with the posterization and 
it
> > > makes my prints unacceptable.
> > >
> > > btw has anyone tried nik color efex! to convert to b&w?
> >
> > So someone else IS getting posterization.
> >
> > The real answer is to simply have the curves redone by Paul or 
whoever
> > is handy with a spectrophotometer. This is a lot of work for all 
the
> > pritner/OS combos though. The curves should be redone in Adobe98
> > rather than sRGB. I'd respect Paul he he refuses to go through it 
all
> > again.
> 
> Jim,
> 
> Then what would happen to all the people who do not have this 
problem? This
> is something that is showing up in 1 or 2 systems so far, redoing 
the curves
> would only be justified if the problem were universal.

Well, true. We haven't asked everyone to do the test where you do the 
gamma 2.2 to sRGB conversion in each version and write down the 
hitogram mean, SD and median and see if they differ a little. It did 
for all of us involved in the discussion so something is probably 
happening-

OTOH, you would then be right to question whether the amount of 
difference affects the actual print. So far only I and apparently 
coveylaw see a real difference on the print.

But let's say that a good samaritan has lots of time and patience and 
prints out a step wedge in  PS7 AND, and this is important, uses the 
space Adobe98 this time instead of sRGB for his (her) curve 
generation. Now the corrected curves will give the same values as the 
old sRGB curves did in EITHER PS6 or PS7. Therefore it is a useful 
thing to do since 1) the better color space is now used instead, 2) 
the tones are corrected for using the new color space so it matches 
tones when the old sRGB conversion was used, 3) it is indeed more 
universal...everyone can use it, whether in PS6 or PS7 with less error 
in conversion. Whether they had a problem or not originally, it would 
work for everybody, since it now uses Adobe98 space- which gives 
almost identical results in PS6 or PS7 (barring that strange 
microbanding I had and a bit more posterization which so far I haven't 
gone further to investigate).


Should Paul be asked to do the pushups? Only if he wants to. He's done 
alot already for us. And it does depend on how many others are 
affected by the discrepancy how important it is to switch to Adobe98 
officially for all the curves.

 
> 
> >
> > He might be able to do it with a transfer curve, I don't know. He
> > hasn't responded to any posts on or off list from me on this 
topic,
> > unless I missed it.
> >
> > Also, there is still the unresolved issue of whether I get
> > microbanding even using Adobe98 with PS7. I can't test it right 
now
> > any further, as I'm short on ink to squander. My test print also 
still
> > shows just a tiny bit more posterization in PS7 vs PS6 when using
> > Adobe98. And of course the tones are different using this space. 
It's
> > really crying out for a curve tweak.
> 
> I am printing just fine from PS7 on my 1280. No microbanding and the 
quality
> is unchanged from PS6. Any tweak would then give me and everyone 
else
> trouble.

Not if the tweak gave you identical tonal results- just using Adobe98 
instead of sRGB. All I'm saying is keep the output the same, just 
strengthen the reliability of the output by changing the workflow such 
that the conversion is to Adobe98 instead of sRGB. Paul shoots for the 
same step wedge- just with the Adobe98 conversion. In fact, on the MIS 
site he has already done this for the 1270 and a couple of others- and 
the 1280 if used with a MAc, if memory serves.

At least one other person is having trouble, that's why I responded to 
this post.  If you are using sRGB as your conversion space and you try 
a relatively dark print and you have a PC- well then I guess you DON'T 
have a problem, what can I say. 

If then you are happy with things as they are, then sure, keep on 
trukin'. Having mutiple curves, PS versions, and spaces for Paul's 
workflow might get a bit confusing for me though, ie, "try this in PS7 
and if it posterizes try PS6, or, maybe just convert to Adobe98 
instead of sRGB, unless you get banding..."  I'm just after 
standardizing the curves as much as possible, probably to Adobe98, but 
maybe this just isn't possible.


 Why not tweak the curves yourself for your own use?

This is a really good idea, and probably the best advice to give 
anyone using MIS VM with all the variables of paper. I think we can 
fairly say that now just two people are starting to see a new 
variable...using PS6 vs PS7.

But using one's own tweak, the correct values can be arrived at. The 
only problem is it may take quite a bit of trial and error to get the 
tonal changes even. Right now it's easier to just go back and use PS6.

> >
> > Then the other solution is to just use PS6 until the 2200 comes 
out
> > and go for it. That is a big unknown as well.

> 
> BTW, I have seen some B&W 7600 output which uses the same inks as 
the 2200.
> The results are similar to the 2000P. Odd tonal splits and a print 
that is
> fairly neutral in tungsten lighting is yellow/green in daylight. I 
don't
> know if these were good or poor examples of what the machine can do. 
I want
> to see some more but if what I saw is "good" then I doubt you would 
be happy
> with the hue or metamerism.


That is discouraging, you would think that Epson would know they have 
to give everyone else a run for the money...all those $$$ carts should 
print out something worth printing, oh well...thanks for the advance 
opinion Martin. I ordered up a sample print from Blair (tssphoto) and 
I'll have a look for myself too.

> >
> > You could buy an older computer and network it up on the OS of 
your
> > choice with PS6 and dedicate the printer to it. A bit of bother 
there.
> 
> Since I can't duplicate your problem using very similar equipment (I 
think
> the only difference is that you are running Win98SE and I am on Win 
2000), I
> would urge you to continue to look for something in your set-up that 
is not
> quite right.

Yes, I will do that after two weeks or so. Thank you for all your 
generous help in trying to isolate the problem last week or so. I can 
use PS6 for now on my main machine. I don't need to add another 
computer myself. Again, you see, I was only responding to "Coveylaw" 
with possible options- he has MAC OSX I think.
Jim H.
 
> 
> Martin

Re: [Digital BW] Re: My workaround for the PS6 -> PS7 posterization problem...

2002-07-13 by Ricardo Lagos

I am a bit confused by the discussion .. so let me state what my setup and
workflow looks like -- it doesn't seem to be the same as the one eveyrone
else is talking about

OS: MacOS X

Scans are done to Grayscale (2.2 gamma) tif files .. (using vuescan) from a
Canon fs4000 (35mm) (16 bit scans)

All manipulation is done in the GS 2.2 gamma space. using PS7

----------------------

Using PS6 i convert to RGB (adobeRGB -- per the new PaulRoark workflow) (for
a 1270 printer)

----------------------

Back to PS7 - apply curves (i like the nw curve) ...

print on 1270 using MIS VM-----------

------------------------

this solved the posterization i was getting -- which i belive was caused by
the translation of GrayScale to adobe RGB by PS7..  (the thread i think was
talking about translating Greyscale to sRGB -- which i was not doing)

------------------------

-- ricardo

[Digital BW] Re: My workaround for the PS6 -> PS7 posterization problem...

2002-07-14 by jimhayes361

I'm sorry- I didn't see that you were using an 1270- which uses 
Adobe98 already. So that renders my other post about sRGB not 
applicable.

I am troubled though that you can detect posterization with a 
conversion to Adobe98, as it is much less of a discrepancy than sRGB 
creates from PS6 to PS7. It means that even a small change in the file 
conversion engine shows up visibly for you. I hope I'm not correct in 
that assumption. I had a tiny amount appear myself when I used Adobe98 
on my Win98SE/1280, enough to make me prefer PS6 over 7 for printing, 
but barely. So this makes two of us, but it is apparently much worse 
for you. I can live with mine, although I seem to have an odd 
microbanding fluke to investigate.

In your file where it shows up, if you follow the workflow through in 
both PS6 and PS7 what numbers do you get for the histograms in each 
for mean, SD, median? It would be interesting to see how big a change 
happens. Martin, Carolyn, etc and I were getting no or at most very 
tiny discrepancies with Adobe98, a little more with sRGB. Also folks 
suggested very good things to check for in the previous thread on this 
topic- such things like using "print with preview" in PS7 to make sure 
destination space is Adobe98...
Jim H.




--- In DigitalBlackandWhiteThePrint@y..., "Ricardo Lagos" 
<ricardo@s...> wrote:
> I am a bit confused by the discussion .. so let me state what my 
setup and
> workflow looks like -- it doesn't seem to be the same as the one 
eveyrone
> else is talking about
> 
> OS: MacOS X
> 
> Scans are done to Grayscale (2.2 gamma) tif files .. (using vuescan) 
from a
> Canon fs4000 (35mm) (16 bit scans)
> 
> All manipulation is done in the GS 2.2 gamma space. using PS7
> 
> ----------------------
> 
> Using PS6 i convert to RGB (adobeRGB -- per the new PaulRoark 
workflow) (for
> a 1270 printer)
> 
> ----------------------
> 
> Back to PS7 - apply curves (i like the nw curve) ...
> 
> print on 1270 using MIS VM-----------
> 
> ------------------------
> 
> this solved the posterization i was getting -- which i belive was 
caused by
> the translation of GrayScale to adobe RGB by PS7..  (the thread i 
think was
> talking about translating Greyscale to sRGB -- which i was not 
doing)
Show quoted textHide quoted text
> 
> ------------------------
> 
> -- ricardo

Re: [Digital BW] Re: My workaround for the PS6 -> PS7 posterization problem...

2002-07-14 by Martin Wesley

----- Original Message -----
Show quoted textHide quoted text
From: "jimhayes361" <jimhayes@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Saturday, July 13, 2002 2:47 PM
Subject: [Digital BW] Re: My workaround for the PS6 -> PS7 posterization
problem...


> --- In DigitalBlackandWhiteThePrint@y..., "Martin Wesley"
> <mwesley250@e...> wrote:
> > ----- Original Message -----
> > From: "jimhayes361" <jimhayes@j...>
> > To: <DigitalBlackandWhiteThePrint@y...>
> > Sent: Saturday, July 13, 2002 8:12 AM
> > Subject: [Digital BW] Re: My workaround for the PS6 -> PS7
> posterization
> > problem...
> >
> >
(snip)

> >
> > Then what would happen to all the people who do not have this
> problem? This
> > is something that is showing up in 1 or 2 systems so far, redoing
> the curves
> > would only be justified if the problem were universal.
>
> Well, true. We haven't asked everyone to do the test where you do the
> gamma 2.2 to sRGB conversion in each version and write down the
> hitogram mean, SD and median and see if they differ a little. It did
> for all of us involved in the discussion so something is probably
> happening-
>
> OTOH, you would then be right to question whether the amount of
> difference affects the actual print. So far only I and apparently
> coveylaw see a real difference on the print.
>
> But let's say that a good samaritan has lots of time and patience and
> prints out a step wedge in  PS7 AND, and this is important, uses the
> space Adobe98 this time instead of sRGB for his (her) curve
> generation. Now the corrected curves will give the same values as the
> old sRGB curves did in EITHER PS6 or PS7. Therefore it is a useful
> thing to do since 1) the better color space is now used instead, 2)
> the tones are corrected for using the new color space so it matches
> tones when the old sRGB conversion was used, 3) it is indeed more
> universal...everyone can use it, whether in PS6 or PS7 with less error
> in conversion. Whether they had a problem or not originally, it would
> work for everybody, since it now uses Adobe98 space- which gives
> almost identical results in PS6 or PS7 (barring that strange
> microbanding I had and a bit more posterization which so far I haven't
> gone further to investigate).

Jim,

I feel a bit dim at times. The conversion between the PS6 and PS7 are easy
to compare. Just take the same 2.2 grayscale wedge file and convert it to
sRGB in each and then measure the results with the eye-dropper tool.

The sRGB conversion matches exactly in both versions of PS except at one
point, 95%. At this point the PS6 conversion gives 10,10,10 and the PS7
version gives 5,5,5. I did the same converting to Adobe RGB in PS and PS7
and they match at all points. I have posted a pdf file showing the
comparisons in folder:

Files > Message Related Files > 07_2002 uploads

I will be interested to see if you get the same values if you repeat the
exercise. In theory these should be system and OS independent.
>
> Should Paul be asked to do the pushups? Only if he wants to. He's done
> alot already for us. And it does depend on how many others are
> affected by the discrepancy how important it is to switch to Adobe98
> officially for all the curves.

I don't think that we should count on him. He has done more than his share
by far and doesn't even own a 1280. It is time for all the curve slackers
(which includes myself) to take the plunge and start doing their own RGB
separation curves and sharing them just as Paul has done. I will be happy to
host any and all curves in the Files section.
>

> > >
> > > He might be able to do it with a transfer curve, I don't know. He
> > > hasn't responded to any posts on or off list from me on this
> topic,
> > > unless I missed it.
> > >
> > > Also, there is still the unresolved issue of whether I get
> > > microbanding even using Adobe98 with PS7. I can't test it right
> now
> > > any further, as I'm short on ink to squander. My test print also
> still
> > > shows just a tiny bit more posterization in PS7 vs PS6 when using
> > > Adobe98. And of course the tones are different using this space.
> It's
> > > really crying out for a curve tweak.
> >
> > I am printing just fine from PS7 on my 1280. No microbanding and the
> quality
> > is unchanged from PS6. Any tweak would then give me and everyone
> else
> > trouble.
>
> Not if the tweak gave you identical tonal results- just using Adobe98
> instead of sRGB. All I'm saying is keep the output the same, just
> strengthen the reliability of the output by changing the workflow such
> that the conversion is to Adobe98 instead of sRGB. Paul shoots for the
> same step wedge- just with the Adobe98 conversion. In fact, on the MIS
> site he has already done this for the 1270 and a couple of others- and
> the 1280 if used with a MAc, if memory serves.

True, if the tweak is to convert everything over to AdobeRGB then I would
see this as a step in the right direction. The trend now is to apply the
curves in Adobe RGB space and to print with the Epson driver's color
management turned off. This makes the curves much less printer and OS
specific. The Dan Culbertson RGB soft proof method. A lot of work to go back
and redo all of the curves done in sRGB and with the original workflow but
probably worth it.
>
> At least one other person is having trouble, that's why I responded to
> this post.  If you are using sRGB as your conversion space and you try
> a relatively dark print and you have a PC- well then I guess you DON'T
> have a problem, what can I say.

I am not seeing it in at the print level. Not a difference in the versions
of PS or between sRGB and Adobe RGB with the same file. I do realize that if
I tried the same thing with a different print I might get a different
result.
>
> If then you are happy with things as they are, then sure, keep on
> trukin'. Having mutiple curves, PS versions, and spaces for Paul's
> workflow might get a bit confusing for me though, ie, "try this in PS7
> and if it posterizes try PS6, or, maybe just convert to Adobe98
> instead of sRGB, unless you get banding..."  I'm just after
> standardizing the curves as much as possible, probably to Adobe98, but
> maybe this just isn't possible.

I certainly agree on standardizing and having to shift back and forth
between PS6 and PS7 does not seem like a reasonable solution. The more
curves the better though. I have found that a change of curves or blending
two curves will often get me out of a posterization jam.
>
>
>  Why not tweak the curves yourself for your own use?
>
> This is a really good idea, and probably the best advice to give
> anyone using MIS VM with all the variables of paper. I think we can
> fairly say that now just two people are starting to see a new
> variable...using PS6 vs PS7.
>
> But using one's own tweak, the correct values can be arrived at. The
> only problem is it may take quite a bit of trial and error to get the
> tonal changes even. Right now it's easier to just go back and use PS6.

Along those lines when I look at how close the conversion results of Adobe
RGB and sRGB are and how close the PS6 and PS7 conversions are (at least on
my system) my instinct is that you would have a tough time tweaking in such
small differences.
>
> > >
> > > Then the other solution is to just use PS6 until the 2200 comes
> out
> > > and go for it. That is a big unknown as well.
>
> >
> > BTW, I have seen some B&W 7600 output which uses the same inks as
> the 2200.
> > The results are similar to the 2000P. Odd tonal splits and a print
> that is
> > fairly neutral in tungsten lighting is yellow/green in daylight. I
> don't
> > know if these were good or poor examples of what the machine can do.
> I want
> > to see some more but if what I saw is "good" then I doubt you would
> be happy
> > with the hue or metamerism.
>
>
> That is discouraging, you would think that Epson would know they have
> to give everyone else a run for the money...all those $$$ carts should
> print out something worth printing, oh well...thanks for the advance
> opinion Martin. I ordered up a sample print from Blair (tssphoto) and
> I'll have a look for myself too.

Well it is a single sample and if you see something different, please let me
know. B&W aside I am interested in one to replace our 1270 for color
printing.
>
> > >
> > > You could buy an older computer and network it up on the OS of
> your
> > > choice with PS6 and dedicate the printer to it. A bit of bother
> there.
> >
> > Since I can't duplicate your problem using very similar equipment (I
> think
> > the only difference is that you are running Win98SE and I am on Win
> 2000), I
> > would urge you to continue to look for something in your set-up that
> is not
> > quite right.
>
> Yes, I will do that after two weeks or so. Thank you for all your
> generous help in trying to isolate the problem last week or so. I can
> use PS6 for now on my main machine. I don't need to add another
> computer myself. Again, you see, I was only responding to "Coveylaw"
> with possible options- he has MAC OSX I think.

Got it. I seem to recall you are off on vacation and I assume with camera at
hand so good shooting.

Martin

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.