Acrobat Output Preview

[rich apollo]

But raw values (for device colorspaces) have no meaning!

I would argue that PrePress, and especially the circumstance that we're discussing is an exception.

1) Most PDFs are 'mystery meat', in terms of color. No embedded profiles, and no output intent. Whether that is right and proper doesn't matter - these are the conditions.

2) In the circumstance we're discussing, color appearance is not what PrePress is after. If we need k-only, the output intent is not relevant.

3) Very often in PrePress it's important to know the k content of neutral tones. 3-color grays are a very common incident. 3-color gray is volatile on press - so we try to minimize the occurance. The embedded profile or output intent may be applicable to the output condition, but not optimal in black generation.

4) You are relying on the quality of the embedded profile or output intent. They may be inappropriate for some purposes.

5) Is it still possible to apply an output intent without an associated profile?

6) When viewing elements of different colorspaces in relation to the output intent, what rendering intent is utilized? For example, if I have an RGB element in a PDF/X4 file with a CMYK output intent - how are the Output Preview values calculated?

I think it'd be useful for the Output Preview to list RGB and Lab values, too. I can envision circumstances in which elements in inappropriate colorspaces could get passed along because of what the operator saw in the Output Preview window.

I understand that you could catch those with Preflight.

 

[leonardr]

If you have an object that is colored Multiple Spot Colors (NChannel) and 1 spot color "Black" with an alternate color space that is an ICC-Profile, Acrobat will render it as 4-color black on screen, whereas a RIP and Acrobat will ignore the alternate color space and produce 100% black upon output.

While it is true that we need to use the alternate colorspace for spot & DeviceN to render on screen, that's only true for the color itself. When we compute the "output plates" that are displayed in the Output Preview window, it is done using a "virtual RIP" so that the alternates aren't used in this case.

Leonard

 

[toronar]

While it is true that we need to use the alternate colorspace for spot & DeviceN to render on screen, that's only true for the color itself. When we compute the "output plates" that are displayed in the Output Preview window, it is done using a "virtual RIP" so that the alternates aren't used in this case.

Leonard

Hi Leonard,

I have to disagree: take a look at the attached PDF with output preview under Acrobat 8 or 9.
The image is NChannel with one spot color Black but an alternate colorspace that is an ICC-profile.
Output Preview will show a 4-color black unless you choose the ICC-profile of the alternate colorspace.
If you print separations or RIP the PDF (APPE 2.x or CPSI), the image will be black only (as intended).
I have no idea why the ICC-profile is "Reader 9 CMYK" or where it comes from. I extracted the profile with Heidelberg's PDF Toolbox.

Screenshots from PitStop, output preview Isocoatedv2 and Read 9 CMYK are also attached.

Kind regards,
toronar

 

[leonardr]

I have to disagree: take a look at the attached PDF with output preview under Acrobat 8 or 9.
The image is NChannel with one spot color Black but an alternate colorspace that is an ICC-profile.
Output Preview will show a 4-color black unless you choose the ICC-profile of the alternate colorspace.

That is because the NChannel colorspace in this PDF is declaring the black channel as actually being composed from all four process colourants - thus we are using them. NChannel is an interesting (and complex) colorspace, and if not used properly can lead to less than accurate results. Since there is only a single colour in this image, there is no reason to use NChannel and instead it should be a simple DeviceN which would have properly been treated as just the named Black channel.

Bottom line - Acrobat is doing the right thing according to the data in the PDF based against the PDF standard (ISO 32000-1).

Leonard

 

[rich apollo]

toronar,
How was the image in the PDF you posted created?

PitStop lists it as Nchannel (Multiple process colors) - 1 spot color. Alternate color space is ICC profile, but is there a profile embedded? What's the profile? Where's the profile?

Photoshop can't open the image when invoked with the Touch Up Object tool.

When extracted and opened in Photoshop I got an RGB image.

Is this a duotone, monotone, process image with a spot channel...?

 

[toronar]

That is because the NChannel colorspace in this PDF is declaring the black channel as actually being composed from all four process colourants - thus we are using them. NChannel is an interesting (and complex) colorspace, and if not used properly can lead to less than accurate results. Since there is only a single colour in this image, there is no reason to use NChannel and instead it should be a simple DeviceN which would have properly been treated as just the named Black channel.

Bottom line - Acrobat is doing the right thing according to the data in the PDF based against the PDF standard (ISO 32000-1).

Leonard

Interesting.
I don't want to be nitpicky, but why does Acrobat's output preview show black only if you choose the simulation profile as Reader 9 CMYK which is the alternate color space?
Shouldn't the "virtual RIP" of the output preview produce only black -- like the "regular" output (print separations, any "real" RIP) -- or always 4-color black regardless of chosen simulation profile in this case?

And if Acrobat's display is according to ISO 32000-1, does that mean that Acrobat's print output and every APPE/CPSI-RIP is in violation of ISO 32000-1?

BTW, InDesign CS2 created this PDF (with PDF-Export as PDF1.4 to keep transparency live) and chose to embed the image as Nchannel with one spot color "Black" and alternate color space CMYK.

toronar,
How was the image in the PDF you posted created?

PitStop lists it as Nchannel (Multiple process colors) - 1 spot color. Alternate color space is ICC profile, but is there a profile embedded? What's the profile? Where's the profile?

Photoshop can't open the image when invoked with the Touch Up Object tool.

When extracted and opened in Photoshop I got an RGB image.

Is this a duotone, monotone, process image with a spot channel...?

The profile is embedded in the PDF and can be extracted with the proper tools (like Coloreditor from Heidelberg's PDF Toolbox), which I did and it is attached to my original posting (Reader 9 CMYK.zip).
Photoshop can only touch up a small subset of all possible colorspaces in a PDF, and Nchannel is not one of them.

The image is a black only picture, although as Leonard mentioned it is not quite ideal to embed it this way.

A lot of this PDFs are created by our workflow, Prinect from Heidelberg.
In the original PDF this image is the same (Nchannel, 1 spot color "Black") with only one difference: alternate color space is CMYK. After the PDF has cleared Heidelberg's colormanagement and transparency flattening, the alternate color space changes to an ICC-Profile.

We flatten these PDFs because they will be printed by another printhouse and their RIP chokes on transparency but can render the flattened PDFs fine.

toronar

 

[mattbeals]

Here's what I see. Acrobat Pro 9.2 shows the box as 50% K using the ISO Coated v2 (ECI) profile, as well as GRACoL 2006 1v2 and US Web Coated SWOP v2. However PitStop Pro 9.0.1 shows it as 15%k.

www.mattbeals.com/misc/PitStop09-Acrobat9.2.jpg

 

[toronar]

Here's what I see. Acrobat Pro 9.2 shows the box as 50% K using the ISO Coated v2 (ECI) profile, as well as GRACoL 2006 1v2 and US Web Coated SWOP v2. However PitStop Pro 9.0.1 shows it as 15%k.

www.mattbeals.com/misc/PitStop09-Acrobat9.2.jpg

Thank you for checking, but there should be an image, not a gray box.
My Acrobat 9.2.0 (Mac OS X 10.5.8) shows an image and 4 color-black with output preview.
Strange, what happened to the image?

toronar

 

[mattbeals]

Whoops, sorry about that. I had display large images off.

www.mattbeals.com/misc/ISOCoatedv2.jpg
www.mattbeals.com/miscReaderCMYK.jpg