Why do some journals change the final software and do not provide a template for the final draft(e.g. contains the same font as in the gallery proof)?

I often hear that some of my colleagues get unacceptable proofs (often even from Elsevier journals).

I am checking an Elsevier paper (https://www.sciencedirect.com/science/article/pii/S0141029619305024) which was in the state of Galley proof (after acceptance)

  • letters in math equations look different: (converted all math equations to vectorgraphics)
    In this paragraph, b is shown in the normal font, but b_R is in a heavier font and aliased.

  • letters have inconsistent size even in the continuous text
    The letter epsilon is printed in two different sizes.

  • made a screenshot of the rendering of jpeg-graphics if the jpeg-image contained two images (a) and (b)

  • where they converted all simple *.eps-graphics (nothing complex) to *.jpeg with 110dpi (text is almost unreadable) also declare a min of 500dpi in the author guidelines
    The word "reinforcement", badly aliased.

  • changed the font from the font declared in the template("Times") to "Charis SIL", now the continuous text is in Charis SIL, but the equations, the variables in the text and the figures are in Times
    The font color is generally dark gray, but in math equations it is black
    The digit 2 in two different fonts.

  • the variables in micro-environment are not aligned with the surrounding text (sometimes smaller letters are even higher then capital letters)
    In this paragraph, math text is aligned with the top of the text rather than with the baseline.

  • several manual line-breaks are awful (some lines contain more space than letters) also a short word is following
    Justified text with too few words per line, causing ugly space stretching
    Another example of poor text justification
    Third example of poor text justification

  • some lines start with "," or "."
    start with a comma
    start with a dot

  • converted a table (conatining text and two eps-graphics) to jpeg with obvious compression artifacts The word "cases", showing jpeg compression artifacts

  • added space between "mm" and "²"
    Two areas listed as "mm ²" instead of "mm²"

  • removed spaces
    Integrals without a space before the differential form "dz"

  • some citations are outside the page border
    A list of citations runs off the page: citation 28 is cut off by the bottom of the page, citations 29 and 30 are not visible, and the next page begins with citation 31

According to the pdf metadata, it was created with Prince 10 rev 7, which only supports SVG as a vector format.

Why don't they convert all eps/pdf-files with inkscape to svg (can be done fully automatic as a batch-process e.g. pdftocairo.exe -svg "Input.pdf" "Output.svg")?

In my field, if I use a space instead of a small space, it is unacceptable or at least unprofessional. (As a reader, such mistakes are evident and disruptive.) And it is so super-important to add it as vector, but then the proofs ruin everything! Why do they have guidelines for authors if the journal introduces illegal (according to the guidelines) mistakes on their own?

Why do some journals change the final software and do not provide a template for the final draft?

  • 9
    Then there are those journals where they accept only Microsoft Word...
    – GEdgar
    Nov 20, 2019 at 13:28
  • 5
    @GEdgar MSWord would be ok for me. Yes it is annoying to transfer LaTeX to winword.exe (MSWord), just because you decided/found a better journal. Some Autors/Editors prefer Word, others LaTeX, but if I send them the paper in there template using there specified program (MSWord,dvi-latex,pdflatex,HTML-Online) then the proofs should look the same. But they use different font, smaller font-size, other program, converted eps2jpeg,... Nov 20, 2019 at 14:07
  • 47
    This seems more like a rant than a question. The lengthy examples make it a lot of work to wade through the question, and don't add anything to the question. The specific examples here probably can be interpreted in terms of the workflow of this journal, the software you use, the format you submit, and the software they use. But none of that is of any interest to the general user-base of SE. If you really want a sociological or economic analysis, then please cut all the material that's irrelevant or reads like a rant, and phrase this more like an actual question.
    – user1482
    Nov 20, 2019 at 19:43
  • 12
    They get money for each typo introduced; plus kickback from blood-pressure medication companies. Nov 20, 2019 at 21:55
  • 3
    The worst experienced I had was with IEEE Access. The person typesetting my paper had poor English and no idea how academic papers are put together. Nov 21, 2019 at 2:35

6 Answers 6


Because they are incompetent.

That's about it, really. But don't lump all publishers/journals as one - the typesetters for one journal might not be the same as that for another journal, even one published by the same publisher, and of course there are good and bad employees everywhere.

  • 3
    I was about to say: because they add no value other then "cite points" to be added to CVs.
    – Ian
    Nov 21, 2019 at 9:33
  • 3
    @Ian I do not understand what you mean.
    – Allure
    Nov 21, 2019 at 9:41
  • 1
    @Allure I think he means, that you are going to publish your work in that journal even if they have crappy typesetting, so why should they care about not being crappy?
    – user253751
    Nov 21, 2019 at 12:59
  • I wish I can upvote your answer for every single sentence.
    – Crowley
    Nov 21, 2019 at 14:00
  • 4
    @WolfgangBangerth: One of my papers was published in a leading journal in mathematics, and there too the typesetting was thoroughly incompetent. I had to explain, among many other things, that the meaning of a fraction changes when you move symbols from the numerator to the denominator. Sadly, this phenomenon is not limited to less reputable journals.
    – academic
    Nov 21, 2019 at 14:17

Publishers will adhere to their own style sheet. Just because you supply a photo-ready version of your document, there should be no expectation that it will be published as is. There should be absolutely no expectation that your font choice is followed, as fonts are a matter of journal style. Further, the journal fonts might not even be open source or free, and journals would not be entitled to distribute them. Same with math layout. Same with choice of emphasis. So, a typeset error might not be an error at all -- it might be a copy editor's correction to make the manuscript match the journal's style sheet.

Real errors do happen, though. Anybody can make a mistake, or have a bad day (there are EASILY a half dozen errors in the question here!). Automatic conversion algorithms can mess up. A typesetter might not be working in their first language.

Publishers are famous for errors in tables, as tables often involve hand entry.

As for the things mysteriously captured as images, I can't imagine why. The instructions for elsarticle say

Keep it simple. (Advanced constructions with for example TikZ or pstricks will be rendered as images.

Since you specify in comment that your Latex table had eps images in it, it makes some sense that your table was rendered as an image. While the layout staff probably should have done better on this, and given you a high-resolution rendering, it really is out of their normal workflow, and violates the "keep it simple" directions in the instructions to authors. I suggest that you could have submitted your own rendered artwork for the table, following guidelines for figure submissions (which is likely how you will resolve the issue with the typesetters), but also point out that your resulting published table style might not perfectly match the journal's style sheet.

This is why you get galleys prior to publication. Take the opportunity to correct the galleys very seriously, and certainly address all author queries.

  • 4
    Yes true, totally agree, but why do they do they change the font from there template to the proof? They should name there font "CharisSIL", then all figures and the JPEG-screenshot of the LaTeXtable would have their specified font. The thing is that the proof introduces mistakes that do not agree with their author-guidelines (much to low resolution, vector should be preffered,..) . Errors in the question: I suffer from dyslexia, I know it is not an excuse, but mistakes in the question here are not that important to me as in the article, so I did not double-check. Nov 20, 2019 at 14:37
  • 4
    They may not care. They don't want to be in the business of supporting every Latex installation of every author. They are probably lifting the text and putting it into a professional non-Latex typesetting environment. The Latex template just needs to work reliably enough to support the submission, it doesn't need to be perfect. I'll point out that copy editors and layout staff might not be perfect either. They lay out a galley that's neat enough to support correction. Nov 20, 2019 at 14:44
  • 2
    @ScottSeidman: Do you think your answer is the most likely explanation for any any of the typesetting mistakes in the question? To which one do you think? I think incompetence and carelessness is the most likely explanation for each. However, your answer can be a likely explanation for some other (hypothetical) smaller mistakes.
    – pts
    Nov 20, 2019 at 20:53
  • 4
    > So, a typeset error might not be an error at all -- it might be a copy editor's correction to make the manuscript match the journal's style sheet. This can be true for some other small mistakes, but I think it isn't true for any of the serious mistakes mentioned in the question.
    – pts
    Nov 20, 2019 at 21:39
  • 6
    Yes and the importance of the galleys is why they usually give you something between 24 and 48 hours to check them. Nov 20, 2019 at 21:58

One issue that I have run into is the outsourced typesetting staff ignoring the provided images and using Acrobat to clip the image out of the author PDF, which rasterizes it at what seems to be the screen resolution.


Here is another complaint. The typesetters are simply not doing their job.

This was the email that the editor of the journal, where our paper was accepted, sent to the chief of the Elsevier production team (and put us in cc) as a result of our complaint regarding the mess that the typesetters induced on our manuscript.

Here is the story. After receiving the first proof, we rejected the correction of the manuscript because there were many systematic errors, e.g. the bold symbols were not bold, some of the mathematical symbols were wrong,.... After sending it back to the production team and putting the editor of the journal in cc, we received the above email by the editor of the journal. After spending another 2 months, we finally got another manuscript but just insignificantly improved. So we quit and spent many hours to correct all the mistakes and inconsistencies.

You can follow the same procedure but don't expect too much.

  • What above email?
    – user253751
    Nov 21, 2019 at 12:59
  • Read the beginning of my response. "Here is another complaint. The typesetters are simply not doing their job. This was the email that the editor of the journal..."
    – KratosMath
    Nov 21, 2019 at 13:36
  • 1
    I think it would be better if you formatted your first line as a quotation (with > ) rather than as bold (with **). It currently reads like the first line is a title. Nov 21, 2019 at 15:07
  • 1
    @MartinBonnersupportsMonica I hope it is now clear! Sorry for the confusion.
    – KratosMath
    Nov 21, 2019 at 15:29

It's all about how the education system treats this topic. Most people are happy with a fast-written Word document as long as what they want to discuss in the paper is there. They neglect the fact that proper presentation is essential.

Sadly proper typesetting isn't something that is actively thought at universities, which is a rather interesting situation given that universities in general aim to promote science and everything that comes with it including paper publication. So if you don't get into it on your own the chances are you will end up with a Masters or PhD degree without ever having done a proper paper in terms of typesetting.

I remember at my university that there was a workshop for typesetting using LaTeX specifically for writing proper publications, seminar papers etc. It became mandatory years later (kudos to the teachers for noticing the poor level of typesetting in assignments the students were turning in) but before that most people didn't even know it existed hence a gap in typesetting skills was present

I know plenty of people, who think decent typesetting is rocket-science. It's not. But the fact is that this air of mystery that surrounds it makes many people quiver. Quite often this results in using Word or similar word processing tools, which are easier to use for simple things but fall flat once you start getting into proper typesetting. Most teachers at the university will not have a problem with you using Word and some even require it (mind-boggling, right?).

I started learning about typesetting, LaTeX etc. quite early on during my studies even though no one required it from me. I just had it with papers with poor typesetting. Sometimes it's so bad that your whole concentration goes into analyzing the typesetting issues instead of the actual topic in the paper, which is sad since the author(s) has(have) put work into doing the research.

In addition we have to consider outsourcing. Due to lack of experience, time etc. many people outsource the writing (in terms of typesetting) to others, who may or may not have the expertise to do it.

Finally proof-reading is something that is often neglected. People concentrate on what's discussed in the paper but forget that the presentation is also very important. Just like when buying a car - will you go for the slick looking one or will you pick one, that looks as if it just got dragged out of a scrapyard because it was too much of a junk for it. If my professor saw the examples you have given, he would have told me to buzz off and redo it all.

  • Thanks for sharing your experience. But I noticed even within my field different supervisors have different opinion on importance of typesetting. But all of them really know the importance. I know supervisors that will do ~90% typesetting-errors in average corrections and other supervisors they will report almost none typesetting-errors in average corrections. Typesetting styles is something you learn from corrections, and different supervisors handle that differently. Nov 21, 2019 at 9:58
  • 1
    My professor was teaching Usability and User Experience so it would have been weird for him not to be into the typesetting side of the thesis. :D Nov 21, 2019 at 21:22

I have read your question several times to understand what you are complaining about. Many answers are discussing poor quality of published material.

I have an Elsevier paper (will add link here when published) for proofreading

This note may shine a slightly different light on the issue. Maybe the scenario was:

  • Editor has recieved .docx manuscript with MSWord's equations used and graphics embedded.
  • Editor turned the manuscript to the most-common file format or his most-loved file format using the fastest (and simplest) tool they have.
  • You, as a proofreader or reviewer, are to assess the article quality - the quality of motivation, methods, arguments and conclusions.
  • When editor decides the paper is ready for publishing, the typesetter shall get the materials for typesetting and actual publishing.

I think you are at the third bullet point now and it would be a waste of effort to typeset the manuscript that is about to be changed several times before actual publishing.

  • 2
    Editor recieved the LaTeX file and all eps-graphicsfiles and now I, as one of the co-authors, have to check if the gallery proof is still as intended in the original LaTeX-file. (Maybe I did not understand you correctly?) Nov 21, 2019 at 15:58

Not the answer you're looking for? Browse other questions tagged .