Re: Do we convert soft proof when converting to profile?
Re: Do we convert soft proof when converting to profile?
- Subject: Re: Do we convert soft proof when converting to profile?
- From: John W Lund <email@hidden>
- Date: Thu, 19 Feb 2009 18:38:00 -0800
Hi John,
Putting my responses inline below:
> Isn't the goal of convert to profile to bias the file so that the final output
> approximates what actually exists in the file? ( if the printer is +5G the
> convert removes 5G)
-- No. At least I don't conceptualize it that way. Convert to profile is an
attempt to maintain the image's color appearance *to the extent that the
destination device is capable*. If the source color cannot be produced by
the output device, it gets remapped - this is where rendering intents come
in, with their varying strategies, and where the profiling software's
algorithms ("secret sauce") determine what the converted color recipe is.
> Doesn't soft proofing with the profile show us the charateristics of the file,
> on the display only? ( Printer is +5G so we now see the files as +5G)
-- Soft proofing shows you how the file will appear if converted to the
chosen "device to simulate" profile (typically, how a print would look if
you send it *with color management enabled* to the printer represented by
the selected profile).
If the profile for that printer/paper makes the image look +5G compared to
what you see viewing the file on your monitor *without Proof Setup enabled*,
then viewing a soft proof should show that +5G; printing the file with color
management enabled (working space > output profile) will not compensate for
this - you will still see that +5G in the print. The point of soft proofing
is to alert you to potential problems before you output; it's still up to
you to make the correction before you send the file to the printer.
Typically an issue like this +5G will not happen if you have quality
profiles for your monitor & printer/paper combo, though. Current profiling
software generally does a pretty good job of maintaining neutrals/gray
balance, and "in gamut" colors from display through print. More likely
issues come about because of "out of gamut" colors - e.g., images with those
vibrant blues that monitors can display, but don't stand a chance of being
produced by any ink-on-paper reproduction.
> Softproofing with profile --> now thanks to both profiles(display and printer)
> we see what the print might look like if sent straight to the printer.
-- not to nitpick, but when you say "sent straight to the printer" - you
don't mean "without conversion to the printer profile" do you?
Because that is not what the soft proof predicts *unless "Preserve RGB
Numbers" is selected* - and that is generally not a recommended way to print
to e.g., inkjet printers. As David Tobie said, usually the file is
maintained in working space RGB, and Photoshop's color engine is used to
convert to the printer's device space profile (whether RGB or CMYK) on
output.
> Convert to profile --> biases the file to compensate for the output
> characteristics shown in the profile.
-- To convert from the working space to the output space (profile),
*remaps* many, if not all colors - depending on how many image colors in the
working space correspond exactly to same colors in the destination output
device profile + your selection of rendering intent & possibly black point
compensation. The goal is to maintain consistent appearance as much as
possible, but of course compromises are forced when image colors simply can
not be produced by the output device.
> So if the convert to profile does the biasing to get back to the values of the
> source file (what we see on a calibrated display without a softproof profile)
> why would one softproof AND convert? This seems counter intuitive. A potential
> for doubling of corrections if you will. Once by the operator ( he/she saw +5G
> and corrected to neutral) and once during the conversion.
-- No, convert to profile is not meant for purposes of color correction.
Again, assuming accurate profiles - if the softproof shows that +5G, it will
appear in the print. It's up to the operator to get the desired appearance
in the soft proof - then the print will show the same appearance.
(theoretically, of course like everything else in this world, it's not
absolutely perfect)
> If we need to softproof the convert to profile, then this would imply to my
> that the original profile is "broken" as a convert to profile should in theory
> provide all the correction we need to match the source file.
-- See above. I find it helpful to get my clients to reverse their way of
thinking from:
"I want the print to match my monitor"
to
"Soft proofing will force my monitor to match the print"
Generating & using accurate profiles will help users to learn to trust the
soft proof (I'm sure you know that I am leaving out a lot of other vital
steps like viewing conditions, print viewing booths, etc. - be that as it
may).
I'm going to send this back to the list as well. Others may care to add or
clarify.
HTH,
John
John Lund
JWL Images
Emeryville, CA
On 2/19/09 3:07 PM, "Digital Division - Reed Photo-Imaging"
<email@hidden> wrote:
> John W Lund wrote:
>>
>> Hi John,
>>
>> The Proof Setup function is intended to show you how the file will appear
>> when converted to the device profile ("Device to Simulate") - although it's
>> not limited to device profiles, you can use working space profiles, for
>> example.
> Just to clairify up front, I am not trying to argue my point, just trying to
> get the right answers to correct any error thinking I may have.
>
> Isn't the goal of convert to profile to bias the file so that the final output
> approximates what actually exists in the file? ( if the printer is +5G the
> convert removes 5G)
> Doesn't soft proofing with the profile show us the charateristics of the file,
> on the display only? ( Printer is +5G so we now see the files as +5G)
>
> If this is the case, please consider this:
>
> Calibrated display --> shows us what is actually in our file (within the
> limitations of the display of course)
> Softproofing with profile --> now thanks to both profiles(display and printer)
> we see what the print might look like if sent straight to the printer.
> Convert to profile --> biases the file to compensate for the output
> characteristics shown in the profile.
>
> So if the convert to profile does the biasing to get back to the values of the
> source file (what we see on a calibrated display without a softproof profile)
> why would one softproof AND convert? This seems counter intuitive. A potential
> for doubling of corrections if you will. Once by the operator ( he/she saw +5G
> and corrected to neutral) and once during the conversion.
>
>
> If we need to softproof the convert to profile, then this would imply to my
> that the original profile is "broken" as a convert to profile should in theory
> provide all the correction we need to match the source file.
>>
>> Anyway, think of it as an on-the-fly profile conversion from the
>> image's color space to that "...Simulate" profile, which is then displayed
>> via another conversion to your monitor profile. So yes, soft-proofing can
>> help in deciding about applying corrections to a file before you actually
>> output it.
>>
>> To display how the file would appear if sent "as is" (no conversion to the
>> "...Simulate" profile, enable the "Preserve RGB Numbers" (or CMYK, if that
>> color space) checkbox.
>>
>>
> This is beginning to make sense in regards to the Preserve RGB. But I am
> still lost as to why we would need the softproof for any other reason than to
> identify gamut issues. The convert to profile should be handling all of our
> color and density for us.
>
> Thoughts?
>>
>> ...<snip>...
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden