JPEG vs EPS in a workflow

Question

The work flow is this:

JPEG orignal-> Edited in Photoshop-> EPS-> Imported into a InDesign document-> Distilled to PDF-> Sent to printer

My question is it necessary to save it as the EPS in the step between Photoshop and InDesign? It would seem to me that keeping it as a JPEG at this step would be fine (and much smaller), and the conversion to EPS doesn't really gain anything.

Am I missing something other than lossy/non-lossy about an EPS that makes it better for this?

EDIT: Lots of good information in the answers.

My reason for asking is that I'm concerned that a 4mb original expanded to a 40mb+ eps times many files equals a lot of additional storage for marginal additional benefit.

To look at this another way, if the original JPEG is otherwise already perfect is it necessary to convert to a non-lossy format first?

The RAW vs JPEG debate among photographers is ongoing, but when you start talking about press ready material, it seems to be "JPEG lossy therefore JPEG evil!" Beyond this there often doesn't seem to be much thought about what lossy means in practice. From what I've seen, a single edit to a high resolution JPEG isn't going to make any practical difference.

Answer 1

EPS is not a lossy format. But the format you use for your raster portions can be (the JPGs you are using are always lossy, and you lose something with each successive save).

In either case, you should not be using JPGs for print production. Stick with TIF for raster to ensure you don't introduce unwanted compression artifacts. TIF does not have to be lossy to save on space, you just won't get the same space savings as you would with JPG. I would argue that those benefits are negligible at most. Just go with TIF for print.

Unless you are applying text to an image in Photoshop, exporting to EPS is completely unnecessary, and I would recommend going with TIF. EPS adds a wrapper of code that isn't required here.

If you are adding text to the image, I would suggest placing the image into an Illustrator document, adding the text there, and then saving that as an EPS. There are clear benefits to doing this: you get to keep editing control over the text and you get all the benefits of vector (extra sharp text for one).

UPDATE based on OP's edit:

A JPG is not "already perfect". JPG is a lossy format that the image needs to be converted out of, even the lowest compression setting is still compressing and leaving an artifact somewhere. But EPS isn't the correct one even you are adding text to the image later. For examples of how the compression works and just how bad it can get please see the following:

• www.cs.sfu.ca: JPG Compression Example
• Wikipedia: JPG, Example Photographs
• photo.net: Jpeg compression

You are right that a single edit to a high resolution file may not make a practical difference. But what happens if the application you are working on saves the JPG in a different, higher compression than the original? Those artifacts will compound over time.

You need to convert your art out of JPG and into something more appropriate format like TIF. It's not hard, and setting up a batch conversion in Photoshop is easy.

Just because the stock photo companies store their content in JPG does not mean you can actually use it in that format.

Update to answer the questions directly, per the OP's comment upon acceptance:

...is it necessary to save it as the EPS in the step between Photoshop and InDesign?

If you are not adding text to the image when editing it in Photoshop, then no, it is not necessary to save the image as an EPS. If you are adding text, then that work should be done in Illustrator and saved out as an EPS.

Am I missing something other than lossy/non-lossy about an EPS that makes it better for this?

Yes. I think you are misunderstanding the roles the various file formats play in print production.

JPG was created in 1986 to offer a means to store a lot of image data in a smaller space; JPG was created in a time when disk space was small and very, very expensive per KB (note not MB because that was a very rare metric in the late '80s). JPG was adopted by the web because it offered distribution of reasonable quality art and photos with minimal data volume, which is just as important today as it was in the 80s and 90s, and is part of the reason why the stock agencies still use it for distribution over the web. JPG was never intended, however, to be used in high quality print workflows and for multiple iterations of edits on the same file specifically because of the lossy compression that comes with the JPG standard.

EPS is a subset of the full Postscript language, and is primarily geared towards vector art. Raster art (i.e., photos) is non-native to EPS, and subject to all sorts of glue code to include it. Since Postscript is a "fat" language, inclusion of art can be done in several different ways. Since Postscript is holding the art in its original form, it is not guaranteed to be non-lossy on lossy formats since the content can be, and usually is, refactored at one point or another during a print workflow. It does make sense to import raster art into a an EPS when there is text or important line art is to be added (i.e., labels and leader lines).

TIFF was created for variety of reasons, but since it is non-lossy by default—a TIF can still utilize JPG compression but that needs to be a deliberate (and foolish) action on the part of someone—it became the de facto format for graphic designers since multiple edits didn't result is an actual loss of integrity of the core image data.

...if the original JPEG is otherwise already perfect is it necessary to convert to a non-lossy format first?

From a purely technical standpoint, you could just use the JPG if you edit once. But as you can clearly see, there are many historical reasons why I extended my answer to state that is not in any way, shape, or form, best practice.

No one has any way to accurately predict how or when a piece of art is going to be used. EPS is overkill here just by the jump in data size alone and still not a guarantee that the integrity of the image will be maintained throughout the project's lifecycle and by extension for the life of the art. Best practice clearly dictates that you convert your JPGs to TIFs before doing anything with them and use the TIFs instead.

Answer 2

If you are in an Adobe-only environment, like I usually am, I see no added value in (rather) globally accessible formats like EPS, TIF or JPG.

When I do edits in Photoshop to something that is going to be exported to InDesign I save it as Photoshop's native file format (PSD). If the edits and layers are in place, I usually save the file as flattened or all the layers merged — depending on whether I need transparency or not.

If I would be concerned in the final PDF size, usually am not, I'd separate vector and raster content from each other. And then again, if I did the vectors in Illustrator, I'd just save the file as AI.

If you need a non-Adobe file format, I'd go with @Philip Regan's answer.

Answer 3

Philips answer is great. but i would add, since you are dealing with Indesign, one of the key advantages is that you can combine different file formats from different programs and Indesign will handle them gracefully.

So choose your file format to suit your graphic infomation. And usually there is no need to save anything other than the fully editable native adobe product format. e.g:

• Vector information should be in a vector format; EPS, AI, PDF etc
• Raster information should be in a lossless raster format; TIFF, PSD etc

Answer 4

Your question is more a technical and production question, do be a bit wary with answers from graphic designers who may lack the technical expertise. The answers can be a bit like: “Eating junk food (using JPEG) will destroy your health (print job).” Like so many other things in life, the reality is, “Simply because something can go wrong, doesn’t mean it must go wrong.”

That said, JPEG can be problematic, it can also do the job. We should not forget that JPEG has been going strong since the mid-1990’s and that JPEG hit the market a good five years after TIFF and still found its way into quality print production.

The reality of the situation is pretty much any data change results in data loss, many of these transforms which take place outside of the JPEG format can be greatly more problematic then the loss JPEG introduces. I’d like to point out that many of us used JPEG to create profession work before image adjustment layers were introduced to Photoshop. This meant often multiple adjustments each being separately applied and each destroying data.

“I'm concerned that a 4mb original expanded to a 40mb+ eps times many”.

I fail to see what benefit using EPS files gain you. Can you explain?

What you should be focusing on is what does EPS gain you in your workflow (over JPEG)? When you asked: “Am I missing something other than lossy/non-lossy about an EPS that makes it better for this?” Except when using spot colors or color transformations (duotones, tritons, etc.), you gain nothing unless your alternative format or output device ignores color profiles (which neither PSD, TIFF, or PNG do (TIFF more recently supports spot color)). If you’re trying to preserve type in an image you may find your better off saving the PSD not as an EPS, but as a PDF.

Some of the answers here have unreasonable workflows. Your workflow should rarely have more than three stages for image file formats. Original files, manipulating files, and finished files (if you pass on using PSD). You should avoid manipulating original files. You should minimize all loss introduced to files you’re manipulating (use layer transforms as opposed to applying curves/levels and the like directly to the image).

My reason for asking is that I'm concerned that a 4mb original expanded to a 40mb+ eps times many files equals a lot of additional storage for marginal additional benefit.

• It generally, isn’t a marginal benefit to have access to the original image.
• It generally, is not a marginal benefit to have access to the modifications/transformations done to an image (much easier to move text on a layer then to clone it out, etc.).
• It certainly may be a marginal benefit to save the working file into a fixed format (on the other hand this fixes the image, which can prevent tampering from others (accidental and intentional). It can be more tempting for a third party to open your PSD file and change something.

I hate to poke fun at the designers, but hunt though the compression dialogs in Acrobat Pro. Do you really think that many service bureaus and printshops don’t down-sample and compress their workflow and convert all those pictures into JPEGs? Don’t quite believe me? Consider that color prepress in an Acrobat workflow uses profiles. Do you really think some poor under-paid pre-press monkey is going to customize each output? No, they have their handful of profiles which regularly include {if image data > x, then downsample and convert}.

Answer 5

There really is no reason to switch into a lossy format when it's possible to stay in lossless. Stick with the PSD. If your hardware can't handle it, then you probably just need new hardware. It's cheap these days.

But if you MUST, for some incredible and unforeseen reason, save into a web format in the middle of your print workflow (crosses self and splashes holy water) don't use jpeg. Use png-24. At the very least you will be able to retain alpha channels and the compression is a bit better.