Transparency and rasterised text

Just realised I don't know the answer to a basic problem.



Having inadvertently put text below a transparent object and suffered the emboldening caused by the rasterisation of the

text along with the transparency I was explaining to someone that the problem comes when the transparency is flattened

to make a PDF to 1.3 and below, but doesn't happen if the transparency is retained in the PDF to 1.4 and above.



But then I thought, is this so? Flattening of the transparency is pushed back to become the responsibility of the RIP.

But when the file is RIPped, does the problem of rasterising the text below transparency still occur?



k
Transparency and rasterised text
My GUESS is probably, yes, but keep in mind that everything is rasterized in the RIP (R is for Rasterized) in order to make those little dots with the laser. The difference is probably in HOW and what resolution is being used.



And are you sure it was emboldened (lovely term, by the way) through rasterization and not because it was converted to outlines? Did you check the flattener preview to see?



Peter
Transparency and rasterised text
There seem to be 3 distinct possibilities:



1. Rasterization is a cop-out for fast display on slow devices. Since RIPs can take the time, all transparent objects are mathematically clipped against eachother and rendered in their respective correct colors -- much like the 'expand objects' option in Illy. The results are mathematically flawless.



2. Rasterization is something to be done at device resolution, so the machine happily creates bitmaps -- with the same (native) resolution as the rest of the vectors are rasterized. The results are technically flawless.



3. Rasterization is a poor-man's solution to a real world problem without a solution: whatever the software specifies, paper can never be inked with the requested transparency. Whatever the software does, it's wrong, innit?



Are there RIP programmers amongst us?

This is the much-discussed ''text darkens near a cut out picture'' syndrome. Flattening a transparency on output

rasterises the text below and can make it look darker - hence emboldened - on the PDF, particularly at low

magnification. Yes, you can see it on the flattener preview as raster-fill text and strokes. And this preview actually

demonstrates quite neatly the standard work around - to lift the text to a layer above, or to simply raise it to the top

of the stack.



In print there is really no such thing as true vector - any vector image is ultimately converted to raster to enable the

plate to transfer ink to paper. I guess the same applies on a screen in that the vector image is a series of pixels, the

difference being that pixels are contiguous and dots on a printing screen aren't. But I was wondering if delaying the

flattening of an image which causes this phenomenon to RIP stage actually got rid of the problem such that any

rasterisation of the type becomes no more than the rasterisation which is going to be applied by the RIP anyway.



The difference I guess is that even with the flattener set at the standard for high resolution - 1200 ppi for text -

that is around half the resolution that one would expect the RIP to flatten to: 2,540 springs to mind.



k

%26gt;even with the flattener set at the standard for high resolution - 1200 ppi for text - that is around half the resolution that one would expect the RIP to flatten to: 2,540 springs to mind.



Perhaps 1200 dpi is enough for sharp text. I can hardly see the difference between offset and laserprinted material at (I think) %26gt; 800 dpi. When asked, I usually cheat and check if the book contain grayscale images -- now

i these

require the highest possible output resolution of a RIP.

This is why the Adobe PDF Print Engine exists. No postscript, no flattening.



Bob

Bob is just kidding. He knows as well as I do that PDF is a subset of PostScript Languagelevel 2. He also realizes that transparency has to be flattened

i somewhere

in the process, before it is turned into ink (or toner) onto paper.



PS. Not a very flattening remark, Bob.

So what is this

i flattening

? Do we know?



Al

It's an archaic method of dealing with transparency that dates back to %26lt;br /%26gt;the time of the dinosaurs. %26lt;g%26gt;%26lt;br /%26gt;%26lt;br /%26gt;Bob
Bob, handling documents with transparency to your printer does not imply they can print it with transparency. I repeat: your transparency has to be flattened

i somewhere.



Flattening: replacing transparent objects with non-transparent ones, simulating the effects of the transparency by calculating their effectual color and applying this to non-transparent objects. I i

think

InDesign can handle vector-over-vector transparency (with aforementioned clipping), but as soon as any bitmap effect is used -- glow, shadow -- it reverts to converting everything to bitmaps.

http://blogs.adobe.com/indesignchannel/2008/04/pdf_print_engine_james_wamser.htm l



Bob

PDF hasn't been a
subset of PostScript language level 2 for nearly 12 years now. With PDF 1.3 (Acrobat 4), PDF became a
superset of PostScript language level 3. And since then, with the addition of transparency and layers, PDF was well exceeded the PostScript language level 3 imaging model.



There is also a very big difference between flattening non-opaque PDF into opaque PostScript objects and reconciling the transparency directly at RIP time.



- Dov

%26gt;[...] especially true if you're using any transparency effects like a blend mode, drop shadow, gradient feather, etc. If you use PDFX-4, which is based on the 1.4 PDF spec, not only do you retain live, unflattened transparency...not only do you preserve live transparency when you submit the PDF, it's also still there after we RIP it [...]



Cool. Never too old ta learn. I was

i way

out of line accusing poor Bob of bein' old fashioned an all.



So ... raw PostScript is on its way out. Time to unlearn stuff again (*sigh*).