Print ouput file but not as EPS or TIFF. Is it possible?

[SylviaW]

I am a student and have been taught that for a properly done print file (acceptable for pre-press people), the pictures must be 300ppi CMYK and imported into the layout program as EPS or TIFF files.

I wonder is it possible to create proper print file acceptable for pre-press for, say, a booklet, where pictures are 300ppi CMYK but PDF or even high quality .jpg, and NOT EPS or TIFF?

If yes, how to do it in, for example, InDesign?


Thanks for any practical advise !

Sylvie

[eugenetyson]

I don't know who is telling you to use EPS files? I suggest that using Illustrator images to be placed in InDesign that you save as .ai not eps.

For raster images, yes TIFF is better, I wouldn't go near EPS from Photoshop.

It all depends on what the content of the image is. If it's just raster then TIFF or PSD would be the preferred choice. You can maintain layers and edit the image completely again when you need to. TIFFs and PSDs always output as just raster, i.e., any vectors (like text, vector masks or vector shapes) are rasterised on output.

If the raster image contains text layers, vector masks or vector shapes then you're better off saving as PDF from Photoshop, this can retain layers for editing later on in photoshop and it will output raster and vector data when mainitained in photoshop as raster and vector data.



If you're supplied a JPEG file that is 300 ppi, then it's fine to use that. There's no advantage to resaving it to TIFF, all you get is a larger file.

If you have TIFFs, PSDs, PDFs etc. , don't resave these to JPGs, jpgs are destructive file formats, once you save as a jpg you damage it, although it's barely noticeable.


And 300 ppi is a bit of a prepress myth, well in lithographic printing it is.

When you are printing using Lithographic Prinitng, they use a Linescreen called LPI. There are usually

133 lpi for newspaper (the paper is thin so basically less ink etc.)
150 lpi for normal usage, like books, brochures, business cards etc.
175 lpi for high-end printing, like artbooks and such like


Now when your image is exposed onto a Plate, it's done in screen angles, and in printing 45 degrees is the achievable angle technically, the other angles are theoretical, bare with me.

When you print a halftone cell, as you're doing lithography you deal with halftone cells. The cell is made up 16 squares x 16 squares (256 all together)

This halftone cell is 1 x 1 in size, and it's rotated at 45 degrees. At this stage you have to think about the size of the cell when it's on it's tip, so taking basic Hypothenuse theory - a^2*b^2=c^2

Or to make it simple - the square root of 2 which is 1.41 (a very long decimal) this is how big your halftone cell is when printed lithographically.

You take this figure 1.41 and multiply it by the LPI which for most printing is around 150 LPI

So you're ideal PPI for printing at 150 lpi would be 211.5 ppi

The equation is simple LPI X 1.41 = PPI for your images

So you can multiply 133 x 1.41 for newspaper PPI for images etc.


Yes 300 PPI is a figure bandied about - but it's not entirely accurate.



So to sum things up

EPS - they are fine nothing wrong with them, I avoid them when I can, because I know that the EPS format is there as a legacy format for older workflows, they really arent' good for modern workflows (Dov Issacs of Adobe said specifically that EPS format is available as a legacy format, it's not just me saying it)

EPS from Illustrator - fine to use, I use them, but I avoid them when I can and I prefer .ai files

EPS from Photoshop - really not a good idea, it supports vector and raster images on output, but Photoshop will flatten the EPS when you open it thereby rasterising the entire file. Not a great idea. If you must save as EPS from Photoshop make sure you also make a PSD file with layers, edit the PSD rather than the EPS and resave as EPS instead.

JPEG - fine to use, if you get a file sent to you like this you can use it. There is no need to resave it as EPS or TIFF or PSD or PDF. However, if you do make changes to the JPEG then you're better off resaving it as TIFF, PSD or PDF (If you have vector elements added then stick with PDF for the save format) JPEGS don't support transparency and are a FLAT file, so you lose Layers and things like that when saving as JPEG.

TIFF - thes are fine, supports layers, supports future edits, supports transparency and is losless file format. Outputs everything as raster (at the native resolution of the file)

PSD - same as TIFF really

PDF - Supports everything the TIFF and PSD do, and they output raster and vector data and it's fully editable. So that's the best one in my opinion when dealing with vector and raster data.


For your PPI - you can use the formula 1.41x LPI = PPI for your images.


But if you can stick with 300 ppi images then that would be best probably, simply because Prepress operators have preflights set to catch images below 300ppi and it raises flags and holds up your job, despite you doing the maths and getting the PPI exactly right for your images, most prepress people won't accept images below 300 ppi, I don't know if it's a lack of knowledge on their part or not (could be?)


Anything else?

[eugenetyson]

What do you want to know about doing it in InDesign?

[Broacher]

Thanks Eugene. All that you've said? Been saying that for years, but debunking graphic production myths, especially technical prepress ones is about as popular as challenging brand preferences in hardware purchases. People (usually) don't want to hear it! ("That's not what the profs told us! And that info was very expensive to obtain!")

On the resolution issue, Dan Margulis' book on colour correction in Photoshop has a whole chapter dedicated to the debunking of the 300 ppi commandment -with great printed supporting evidence. Another issue is photo subject. Certain details and textures just work better at lower resolutions than other ones.

Here's a Q for you: in a layout to PDF workflow, how do you optimize your photo sharpening? That is, presuming that there are (relatively speaking) OVER-rezzed images placed on the layout page?

[eugenetyson]

Interesting, I've never read any of Dan's books, I should really give those colour correction books a read and fill in the missing gaps of self-taught colour correction.

Normally I don't bother sharpening images, I plonk in 72 ppi images and scale them to 24 %, they're rgb and everything.

If I was doing High-End printing, like art prints, estate agents, anything that needed a high end finish I'd be more detailed. I'd scale the image in photoshop and I'd sharpen it using the dupe layer, high-pass filter, overlay method, so there is no destruction to the image as such and I can tweak the sharpness if needed.

I'd also do colour corrections and things like that, and make it the exact size I need at the exact ppi, and because I'm working with high end finishes here I'd be aiming for the 175 lpi which would give optimum res of 246.5 ppi for my images.

If I go over this number I could potentially lose part thereof the resolution. At least if I work with the 246.5 ppi I know that nothing will be dropped out by the RIP when plating. What I see on screen and hi-res proofs is practically what I'd get in print. And I can control it a lot better this way.



But it depends on what I'm working on and how much of a budget the client has. Hardly any point in colour correcting shots of burgers for the local take-away and sharpening them, but I'd take that time if doing something for a gallery or a museum or something.

[Broacher]

Hmm. Interesting. I guess I was trying to get to the problem that seems to have slipped through a lot of the workflow since we've adopted layout direct to PDF, where a higher rezzed image is (by default, anyhow) downsampled to (again, another default) to 300 ppi for the PDF.

The problem with this auto-downsampling route imposed on by the PDF export is that it doesn't really optimize the sharpness, or allow a way to intercept the dowmsampled image through a sharpening round.

As to when and how to do output sharpening, that's a whole other story. Or stories.

But getting the final images out of a PDF workflow can be tricky. Sometimes, for a critical shot/output situation (like you, these kind of jobs are about as common for me as client approval on the first proof) -- I've opened the PDF in Acrobat and used 'Edit Image' to open the downsampled final image into Pshop for output sharpening. BUT, the pothole is if there are any flattened transparencies or effects woven into the image -- or more likely, image strips. It then becomes next to impossible.

Another route is to downsample duplicated and relinked images automatically prior to PDF output. I've found a few scripts on the Adobe site for ID that works well for this (they combine sidetrips for each image into Pshop). You end up with a single directory of optimally resampled images and from there you can do an auto-batch action from Pshop.

The other often-forgotten (since the introduction of CMS tech) is the effect of sharpening on dot gain tweaks. I used to do dot-gain tweaks all the time but a few years back I discovered that too many 'auto-prepress' workflows (especially in newsprint, where it tends to make the most difference) read these files (even when they shouldn't) as somehow 'off' and the auto-compensation makes it even worse than if you don't bother to compensate. I rarely worry about that anymore.

But back to the image rez issue. I wouldn't get too 'hung-up' on that formula. As Dan M. shows, the optimum resolution is often more empirical than mathematical. I'd really recommend his book on colour correction in Pshop. He emphasizes the psychology and dynamics of human perception and what an important role they play as well -- even though, he is a 'master' of the numbers method himself.

As to the optimal sharpening techniques, I learned a lot about LAB techniques in Dan's other book about working in there. The L-channel sharpening route can have some definite superiority over RGB or CMYK sharpenings-- particularly in minimizing (or in fact, eleminating) the 'white halo' effect on high contrast edges. Again, there's not always a definite 'best method' to consider-- the image analysis, as always, is just as critical.

Margulis' writing style is a little weird for many. Not dry, but a little "well, if you want to think about this, read on, otherwise just remember this..."

But it's balanced with real production evidence there, on the printed page. And he always works with REAL world images. That's not something a lot of writers in this area do. Dan routinely challenges the 'theorists' with side-by-side tune-up tasks. Can't recommend his books enough.

[SylviaW]

Thanks eugenetyson ! Your reply is really something ! So succinct and simple to understand and makes me feel much more confident and getting rid of some inhibitions like this "300dpi is a must"…

And from what I’ve learned I will prepare the following output file for prepress:

-It is a catalogue of 16 facing pages to be printed in offset 150 LPI.

-Simple b/w text, nothing fancy, in 2 columns layout.

-Has 28 raster images (photos) in CMYK, cleaned up in PShop, saved without layers (flattened) . Some of them (only four) are saved as high quality JPG -the same format as they have been given to me.
The other 24 pics were PShop files and after doing a bit retouch and Levels workout on them I saved them as TIFFs .
- All raster images have the same resolution - 231dpi (don’t know why exactly this resolution but I left it as is…).

-The completed catalogue, done in InDesign, with crop marks, colours /fonts checked, etc, I’ll save as print quality PDF and deliver to prepress.

Is this job good enough to be accepted by a prepress professional for sending to platemaking?

If not - what I need to improve?

Any advise is very welcome, please!

Big Thanks!!!
Sylv.

[eugenetyson]

Well as I said earlier if you get jpegs you can use them that's fine. If you edit them in anyway you're better off saving as TIFF, PSD or PDF.

But as long as your images are CMYK and at 211.5 ppi to 300 ppi then you'll be fine.

You say you're making them CMYK, but there are many things going on in colour management workflow. You'd be better off asking the printers to give you PDF joboptions file, so you can make the pdf to their spec.

Note that Print Quality PDFs don't include bleed or crop marks. You'll need that.

I usually use PDFx1a as the base for my PDF settings when preparing for print. But it's best to take to your print provider about this.

[eugenetyson]

Originally Posted by Broacher

Hmm. Interesting. I guess I was trying to get to the problem that seems to have slipped through a lot of the workflow since we've adopted layout direct to PDF, where a higher rezzed image is (by default, anyhow) downsampled to (again, another default) to 300 ppi for the PDF.

The problem with this auto-downsampling route imposed on by the PDF export is that it doesn't really optimize the sharpness, or allow a way to intercept the dowmsampled image through a sharpening round.

As to when and how to do output sharpening, that's a whole other story. Or stories.

This is also a problem when lo-res images get upsampled through the transparency flattener!

There's nothing wrong with letting indesign reduce the image resolution to 300 ppi on output, it's just you don't get that step to sharpen the image which is needed when reducing resolution. But 9/10 it's barely noticeable on a 133 lpi or 150 lpi, you might notice something through a loupe or in in 175 lpi prints. But generally on average images of average content of average quality the difference is negligible, in my opinion.

Originally Posted by Broacher

But getting the final images out of a PDF workflow can be tricky. Sometimes, for a critical shot/output situation (like you, these kind of jobs are about as common for me as client approval on the first proof) -- I've opened the PDF in Acrobat and used 'Edit Image' to open the downsampled final image into Pshop for output sharpening. BUT, the pothole is if there are any flattened transparencies or effects woven into the image -- or more likely, image strips. It then becomes next to impossible.

I don't really get hung up on sharpening the images, I do for high end finishes, and again I open the image in photoshop and make it the correct res at the correct size for placement and sharpen it before I put it in the document. That way I know that InDesign won't do anything with the image and I know what I've done in Photoshop is the only thing that can right or wrong with the image.

But then again you don't have to let InDesign downsample your images on output, or use compression. If I had a 10,000 image catalogue I would place all the images using the correct size and resolution. Having indesign downsample all those images on export could be time consuming. Then having indesign compress images on export would be time consuming too, then having a RIP uncompress the images on RIPing would be time consuming. I'd much rather put the images at 100% @ 265 ppi and on export have no downsampling and no compressing. That's just for a largely based image catalogue with thousands of images, more or less.

But again, PDF settings are not set in stone, it's best to tailor the settings based on what's best for the output. Is this job going to hang up the RIP because so many images have to be uncompressed? Use of JPEG, LZW or ZIP? Why use JPEG for a high end job, can the RIP handle LZW or ZIP compression? Why bother with compression? You'll get a smaller file, but will a smaller file RIP faster than a file that has to be Uncompressed?

Originally Posted by Broacher

Another route is to downsample duplicated and relinked images automatically prior to PDF output. I've found a few scripts on the Adobe site for ID that works well for this (they combine sidetrips for each image into Pshop). You end up with a single directory of optimally resampled images and from there you can do an auto-batch action from Pshop.

The one I'm quite fond of is Link Optimiser, it resizes all the images in your layout to 100% at the desired PPI and gives you the option of sharpening the image in Photoshop, all the time making new images and links in InDesign. So the original image remains. It is quite nifty.

Originally Posted by Broacher

The other often-forgotten (since the introduction of CMS tech) is the effect of sharpening on dot gain tweaks. I used to do dot-gain tweaks all the time but a few years back I discovered that too many 'auto-prepress' workflows (especially in newsprint, where it tends to make the most difference) read these files (even when they shouldn't) as somehow 'off' and the auto-compensation makes it even worse than if you don't bother to compensate. I rarely worry about that anymore.

I find trying to compensate for dot gain to be a huge undertaking. The last time I did this I found out that the printers had a huge dot gain when plating to compensate for lack of dot gain on the press. In short, they didn't calibrate the press to the plate machine at all. You can't win in those situations.

[eugenetyson]

Originally Posted by Broacher

But back to the image rez issue. I wouldn't get too 'hung-up' on that formula. As Dan M. shows, the optimum resolution is often more empirical than mathematical. I'd really recommend his book on colour correction in Pshop. He emphasizes the psychology and dynamics of human perception and what an important role they play as well -- even though, he is a 'master' of the numbers method himself.

I'm not advocating that the forumla be followed step by step. But rather that the 300 ppi is number thrown about, and I found that people never questioned the number, they just repeat 300 ppi and never looked any further other than that's what they were taught. Fair enough knowing that 300 ppi is a good number, but knowing why that number is there and how you get to that number is important too.

Not that I multiply 1.41 x lpi all the time. Generally it's said that 2 x lpi is the right about (2x150 = 300). But I wouldn't be too bothered even if the ppi of my images dropped to below this, as long as they don't drop below 215 ppi, to round things off.

I'm not really going to get into psychology and dynamics of human perception, that's really something else to take into account though.

Originally Posted by Broacher

As to the optimal sharpening techniques, I learned a lot about LAB techniques in Dan's other book about working in there. The L-channel sharpening route can have some definite superiority over RGB or CMYK sharpenings-- particularly in minimizing (or in fact, eleminating) the 'white halo' effect on high contrast edges. Again, there's not always a definite 'best method' to consider-- the image analysis, as always, is just as critical.

It's true that LAB sharpening can be superior. Like if you open a copy of the image and make that LAB mode - unsharp mask the L channel this avoids noise from the colour channels, if you sharpen noise then it can be destructive to the image. Then copy the L channel back to the RGB image it can be a nice way to sharpen your image.

You can just make a simple action that prompts you at the Unsharp Mask stage to make your adjustment, the L sharpening method is just a few clicks then.

There are tonnes of ways to achieve image sharpening, and I've yet to find one that works for all situations.

Originally Posted by Broacher

Margulis' writing style is a little weird for many. Not dry, but a little "well, if you want to think about this, read on, otherwise just remember this..."

But it's balanced with real production evidence there, on the printed page. And he always works with REAL world images. That's not something a lot of writers in this area do. Dan routinely challenges the 'theorists' with side-by-side tune-up tasks. Can't recommend his books enough.

I'll look into getting the books you've recommended, certainly seems interesting.


The only books I have on the subject is from the Real World series, David Blatner and Co. who release them each year. I find them to be great technical references with of course Real World application.