Embedding font (word)

[theducks]

Originally Posted by rubeus

Doesnt importing something in Calibre makes it more worse than before? SCNR

You know about rumor...

in 15 Minutes you could download Calibre, import a TEST BOOK.
Then look at it and see exactly what changes are (not) being made, then decide if they are unacceptable.

Now if you consider yourself a total control freak, you should probably skip computers all together. Something might get changed like a timestamp, permission...

IMHO the real question: "Does the book still render the same?" So what if a program puts in a fingerprint (Metadata entry...)

[DLSieving]

I meant that I imported the DOCX into Calibre, then went through the input and output options and exported to EPUB. I then opened the EPUB in the Amazon Kindle Previewer, which automatically converts it to MOBI. I then opened the MOBI in the Kindle App and in Apple iBooks. After several iterations I saw everything I wanted to see, including my embedded fonts. I had to remove section and chapter numbers from the DOCX to get *whatever* to not try to infer structure and indent them; it could have been Calibre but was more likely the Kindle Previewer conversion.

I have not yet looked at the EPUB directly because I'm focused for the moment on KDP. EPUB is just an intermediary output for me right now, though I realize the EPUB will be the branch point for other platforms when I get to that point. I'm focused now on proofing and font licensing.

I have not yet started the Kindle upload process, though I have been through it in the past. I'm hoping my fonts won't be stripped or anything else dumbed down. If they are then I'll be back to the forums and if nothing changes at KDP then I'll be off to another platform starting from the EPUB output.

[Hitch]

Originally Posted by DLSieving

I meant that I imported the DOCX into Calibre, then went through the input and output options and exported to EPUB. I then opened the EPUB in the Amazon Kindle Previewer, which automatically converts it to MOBI. I then opened the MOBI in the Kindle App and in Apple iBooks. After several iterations I saw everything I wanted to see, including my embedded fonts. I had to remove section and chapter numbers from the DOCX to get *whatever* to not try to infer structure and indent them; it could have been Calibre but was more likely the Kindle Previewer conversion.

Yes--I admit I don't understand what you mean. The Section and Chapter numbers--these are headings and subheadings, is that right?--should render exactly how you styled them, I'd assume. For you and for anyone else that comes along: if you didn't style them as headings, but instead as PARAGRAPHS, and used styling to make them look like headings, then...yes, the KP or KG conversion would assume that you wanted the typical cascading style. That paragraph style has a base style that has a first-line indent. It's not hard to get rid of, but you'd need to either a) style it in Word as an HTML heading or b) you'd have to regex them in HTML, in Sigil or the Calibre editor.

You didn't answer my earlier question, about whether you used only the Stylesheet, or if you used his templates As you used something that you got from Kawasaki. I may know the source of the issue, now that I cogitate on it a bit. Did you use his APE template? And you used his styles? Like the ChapterNumber style? And "Chapter Title?" This APE template is one of my grievances with that book and his templates--they used paragraph classes (styles) for what should be HEADINGS. And thus...your file is inheriting the default, first-line indent for paragraph styles. This problem in rendering can be fixed in 5 minutes or so. You can skip the next paragraph if you don't want to fix them now, which I can certainly understand, but for anyone who comes along later:

Regex the paragraph class and replace it with an H class, e.g., (p class = "section" to h2 class="section") give the H class a suitable name (same name, for that matter, as shown above) and then make sure you use the correct class in the CSS. The styling may even already be available, if it outputs as .section. You may only have to add the heading class to the CSS. Bada-Bing, it ought to be fixed. I'm guessing, of course, because I don't have your book in front of me, but...I've seen this before. With folks that use named PARAGRAPH classes for what should be structural heading classes. They confuse what something looks like--a chapter head--with what it is--a chapter head.

I have not yet looked at the EPUB directly because I'm focused for the moment on KDP. EPUB is just an intermediary output for me right now, though I realize the EPUB will be the branch point for other platforms when I get to that point. I'm focused now on proofing and font licensing.

Did you use a foundry font? I'm merely curious; I'm not the font police. Although, of course, I do indeed like to see IP licensed properly. Did you subset your font?

I have not yet started the Kindle upload process, though I have been through it in the past. I'm hoping my fonts won't be stripped or anything else dumbed down. If they are then I'll be back to the forums and if nothing changes at KDP then I'll be off to another platform starting from the EPUB output.

The font retention/rendering is one of my curiosities, indeed--if your upload retains the fonts. There have been murmurings that Amazon fixed the issue, but we're not taking any chances at my shop, we're still using our secret sauce workaround.

I look forward to hearing back on your upload results. Was I clear enough about the indentation thing?

Hitch

[DLSieving]

Hitch: I keep getting the error "The message you have entered is too short" when I try to reply to your post. I have no idea why this is happening. I've saved off my full reply and will post it on another day when the site is working better.

[Hitch]

Originally Posted by DLSieving

Hitch: I keep getting the error "The message you have entered is too short" when I try to reply to your post. I have no idea why this is happening. I've saved off my full reply and will post it on another day when the site is working better.

DL:

It means that you've accidentally typed all of your replies, INSIDE the Quoted material. When you do that, the text you enter isn't counted. The site is working fine.

If you're trying to INTERSPERSE your replies, type open-bracket /Quote close-bracket to close the section you're quoting. Otherwise, the forum software thinks that all you are doing is putting the previous message (mine, in this case), in a quote.

So, put open-bracket QUOTE close-bracket in front of any section you want to quote, and open-bracket /QUOTE close-bracket behind it. I tried to do it here in the QUOTE tags, but it rendered the BBC (old BulletinBoard Code, IIRC) anyway. If you pretend that my () are [ and ], it would look like this:

(Quote)This is the quoted material.(/Quote)

Right? The square brackets, at or near just above your "enter" key.

Hope that helps.

Hitch

[PeterT]

(Time for Hitch to learn about the [NOPARSE]...[/NOPARSE] BBcode I think

So, put [QUOTE] in front of any section you want to quote, and [/QUOTE] behind it. I tried to do it here in the QUOTE tags, but it rendered the BBC (old BulletinBoard Code, IIRC) anyway. If you pretend that my () are [ and ], it would look like this:


[Quote]This is the quoted material.[/Quote]

[Hitch]

Originally Posted by PeterT

(Time for Hitch to learn about the [NOPARSE]...[/NOPARSE] BBcode I think

Thanks, Peter. I freely admit, I don't think I've noticed those before. Although hopefully, DLSieving can read my intent. But, hey, just for fun:

@DLSieving:

If you've tried to intersperse, you need to distinguish between the quoted material, and your response, like so:

[quote]This would be the originally-quoted text, to which you are responding. Let's say that you want to reply to this sentence, you'd type:[/quote] That would put those two sentences inside the quote. You'd type your response, and then, for the chunk that follows, that was part of my original post, you'd type [quote] and [/quote] to distinguish the parts I'd written. OR, you can use the QUOTE icon, above--the one that looks like tiny text in the cartoon dialogue icon.

What I typed, above, inside the noparse tags would look like this:

This would be the originally-quoted text, to which you are responding. Let's say that you want to reply to this sentence, you'd type:

That would put those two sentences inside the quote. You'd type your response, and then, for the chunk that follows, that was part of my original post, you'd type

and

to distinguish the parts I'd written. OR, you can use the QUOTE icon, above--the one that looks like tiny text in the cartoon dialogue icon.

Does that help you? The parts of my original post to which you are responding need to be wrapped in quote tags; your response is not. Lather, rinse, repeat.

Hitch

[DLSieving]

Originally Posted by Hitch

Yes--I admit I don't understand what you mean. The Section and Chapter numbers--these are headings and subheadings, is that right?--should render exactly how you styled them, I'd assume. For you and for anyone else that comes along: if you didn't style them as headings, but instead as PARAGRAPHS, and used styling to make them look like headings, then...yes, the KP or KG conversion would assume that you wanted the typical cascading style. That paragraph style has a base style that has a first-line indent. It's not hard to get rid of, but you'd need to either a) style it in Word as an HTML heading or b) you'd have to regex them in HTML, in Sigil or the Calibre editor.

I did use paragraph styles instead of headings. Knowing how intrusive some conversions tend to be, I was trying to hide my numbered headings so that they would be left alone. When I saw that they weren't, I just removed the numbers. I like the result for a few reasons.

• The headings look cleaner without numbers.
• The numbers can be seen just by clicking the heading, which takes the user back to the TOC.
• No conversion process can possibly figure out that they're headings and so will have no excuse to make assumptions and change them.

You didn't answer my earlier question, about whether you used only the Stylesheet, or if you used his templates As you used something that you got from Kawasaki. I may know the source of the issue, now that I cogitate on it a bit. Did you use his APE template? And you used his styles? Like the ChapterNumber style? And "Chapter Title?" This APE template is one of my grievances with that book and his templates--they used paragraph classes (styles) for what should be HEADINGS. And thus...your file is inheriting the default, first-line indent for paragraph styles. This problem in rendering can be fixed in 5 minutes or so. You can skip the next paragraph if you don't want to fix them now, which I can certainly understand, but for anyone who comes along later:

I'm importing directly from DOCX into Calibre without any intermediate steps. Whatever stylesheets Calibre gets were created by Word without my direct participation, except by defining and using styles as much as possible to reduce the complexity of the document.

Regex the paragraph class and replace it with an H class, e.g., (p class = "section" to h2 class="section") give the H class a suitable name (same name, for that matter, as shown above) and then make sure you use the correct class in the CSS. The styling may even already be available, if it outputs as .section. You may only have to add the heading class to the CSS. Bada-Bing, it ought to be fixed. I'm guessing, of course, because I don't have your book in front of me, but...I've seen this before. With folks that use named PARAGRAPH classes for what should be structural heading classes. They confuse what something looks like--a chapter head--with what it is--a chapter head.

Given that there could potentially be a large number of cycles in the polishing phase, I prefer to assert my wishes in the Word source document. The only manual remedial steps I'm willing to consider at this point are ones that I can't achieve any other way.

• One of these exceptions is that I post-process the ePub using BBEdit to lock in my font-family definitions using the !important qualifier.
• Based on something you said about guarding against font stripping, I'm anticipating perhaps one more such exception but don't yet know what that might be.

Did you use a foundry font? I'm merely curious; I'm not the font police. Although, of course, I do indeed like to see IP licensed properly. Did you subset your font?

Right now I'm using the fonts included in Word.

• Do I need to license those separately if I'm going to use them in an eBook? Monotype Corsiva, for example.

I will use foundry fonts where needed by law. Yes, I subset my embedded fonts using Calibre.

The font retention/rendering is one of my curiosities, indeed--if your upload retains the fonts. There have been murmurings that Amazon fixed the issue, but we're not taking any chances at my shop, we're still using our secret sauce workaround.

Due in part to tax season, I’ve had to push this project down a few frames on my stack lately but will be happy to let you know the upload results when I have them.

I look forward to hearing back on your upload results. Was I clear enough about the indentation thing?

Yes, I understand now about the indents that I never asked for. Besides taking the numbers out altogether as I did in one of my 2 books, another way to outmaneuver automated structuring is to use words in place of numbers - One, Two, Three, etc. - which I was already doing in the other of my 2 books, again because it looks better in that book. I know it sounds extreme but as I said, I do like the result and there's no way the automated format-bots in the conversion code can figure out what I'm doing and change it.

Hitch

[DiapDealer]

Are we still Sigiling at all here? Things wander all the time, so it's no big crisis either way, but if there's nothing Sigil-specific really happening with this discussion anymore, I'd prefer to move it to the Workshop or some other apppropriate venue. It seems to me that things are being handled exclusively (RE @DLSieving's project) with Word/calibre/BBEdit now, no?

[Hitch]

Originally Posted by DLSieving

I did use paragraph styles instead of headings. Knowing how intrusive some conversions tend to be, I was trying to hide my numbered headings so that they would be left alone. When I saw that they weren't, I just removed the numbers. I like the result for a few reasons.

• The headings look cleaner without numbers.
• The numbers can be seen just by clicking the heading, which takes the user back to the TOC.
• No conversion process can possibly figure out that they're headings and so will have no excuse to make assumptions and change them.

Sorry--I don't think I understand why "numbered headings" are involved here. Do you mean, when you say that, "Chapter ONE," for example, or is your document using Outline numbering, perchance?

Also, what do you mean when you say that you know how "intrusive" some conversions can be? What are you talking about? Which conversions, and the like? I find that generally, even the worst automatic converters out there recognize headings set in a source document, even if they don't recognize anything else. As you said, initially, that you were learning ePUB, what other methods have you used, to which you are referring?

I'm importing directly from DOCX into Calibre without any intermediate steps. Whatever stylesheets Calibre gets were created by Word without my direct participation, except by defining and using styles as much as possible to reduce the complexity of the document.

Okay...

Given that there could potentially be a large number of cycles in the polishing phase, I prefer to assert my wishes in the Word source document. The only manual remedial steps I'm willing to consider at this point are ones that I can't achieve any other way.

OK--so, really, discussing this at all on the eBook Software--Sigil==> forums is a bit of a misnomer. If you're only going to work in Word, I'd guess that this thread should move over to the Workshop, as Diap suggested. However, when it comes to embedding fonts, normally, Word can't do that for you. You cannot embed fonts into an HTML document using Word. I am inferring that Calibre, what, pulls the referenced font files into your ePUB, is that right? When you convert the Word file using Calibre, to ePUB?

With regard to the cycles and the polishing phases...didn't you say that this was a learning exercise? As opposed to a commercial book? You downloaded a txt from Gutenberg, and all that? Or have I misunderstood?

• One of these exceptions is that I post-process the ePub using BBEdit to lock in my font-family definitions using the !important qualifier.
• Based on something you said about guarding against font stripping, I'm anticipating perhaps one more such exception but don't yet know what that might be.

Well, I expect we're about to get to this.

Right now I'm using the fonts included in Word.

• Do I need to license those separately if I'm going to use them in an eBook? Monotype Corsiva, for example.

I will use foundry fonts where needed by law. Yes, I subset my embedded fonts using Calibre.

Firstly, any font that you actively put in an eBook will most likely need licensing. Most fonts are simply licensed for "desktop publishing" purposes, which is very different than "embedding" or "redistribution." Don't get confused by font licenses saying "commercial;" that is unrelated, usually, to redistribution/embedding. Be aware that licensing a font for redistribution/embedding usually increases the desktop publishing price by a factor of 10 or more. I just today looked at a font that was $79.50 for both Desktop Publishing and Webfonts (websites, essentially) licensing, and it was $795.00 for a single eBook license. That's not unusual if dealing with foundry fonts. ( As someone who messes with this commercially, I have a boatload of alerts set up for sales and deals on fonts licensed for these uses, just so I can scoop them up at a good price.)

Second--Remember that in the normal course of things, you'd have to add the font file, the entire font file, either to Calibre in the ePUB Editor or in Sigil, to make it work. Simply putting CSS into a stylesheet, calling for (e.g., "Monotype Corsiva") won't make the font display. It might display on your computer because you have Monotype Corsiva on your system. But if you send that file to someone who doesn't, they won't see Monotype Corsiva. They'll see whatever the device substitutes for it. I'm not familiar with Cailbre for conversion, which is what you must have used, but my comprehension is that it "pulls" the fonts from the system's fonts folder, that are used in the eBook file? That's the deal?

Due in part to tax season, I’ve had to push this project down a few frames on my stack lately but will be happy to let you know the upload results when I have them.

Great, thanks.

Yes, I understand now about the indents that I never asked for. Besides taking the numbers out altogether as I did in one of my 2 books, another way to outmaneuver automated structuring is to use words in place of numbers - One, Two, Three, etc. - which I was already doing in the other of my 2 books, again because it looks better in that book. I know it sounds extreme but as I said, I do like the result and there's no way the automated format-bots in the conversion code can figure out what I'm doing and change it.

Honestly, you're doing this the hard way. You seem to have started with Calibre and Sigil first, rather than cleaning and styling your Word file first. I really don't understand where you're coming from with the "outmaneuver automated structuring" part, or even why you want it. You can make ANY bit of text, any heading, any paragraph, etc., look how you want it to, in both Word and in HTML. You're making things hard on yourself for no reason.

@Diap: as the poster doesn't really want any HTML/XHTML/CSS information, but only wants to work in Word, I guess this should move to the Workshop forum. I'm sure that more folks who do the lifting in Word will be there, instead of here at the Sigil forums...that's my guess.

Moderator Notice
Taking the advice. Moving this section to workshop

Hitch

[DLSieving]

Originally Posted by DiapDealer

Are we still Sigiling at all here? Things wander all the time, so it's no big crisis either way, but if there's nothing Sigil-specific really happening with this discussion anymore, I'd prefer to move it to the Workshop or some other apppropriate venue. It seems to me that things are being handled exclusively (RE @DLSieving's project) with Word/calibre/BBEdit now, no?

You are correct. I'm not using Sigil, just Word, Calibre and BBEdit. I'm temporarily stalled for tax season and other priorities but unless my DTP upload reveals new problems such as font stripping, my way forward is clear.

[DLSieving]

Originally Posted by Hitch

Sorry--I don't think I understand why "numbered headings" are involved here. Do you mean, when you say that, "Chapter ONE," for example, or is your document using Outline numbering, perchance?

I'm avoiding the use of numbered sections and chapters so that no conversion algorithm for any publishing platform going forward will be able to easily figure out what I'm trying to do and impose its own structuring.

Also, what do you mean when you say that you know how "intrusive" some conversions can be? What are you talking about? Which conversions, and the like? I find that generally, even the worst automatic converters out there recognize headings set in a source document, even if they don't recognize anything else. As you said, initially, that you were learning ePUB, what other methods have you used, to which you are referring?

If my strategy seems unclear, it's because it's forward-looking. As much as possible to minimize platform-specific workload, I want to do all of my formatting in the source document and avoid the need for platform-specific work-arounds. In this case, that means removing section and chapter numbers from the body of the book for the reasons given above.

Okay...

OK--so, really, discussing this at all on the eBook Software--Sigil==> forums is a bit of a misnomer. If you're only going to work in Word, I'd guess that this thread should move over to the Workshop, as Diap suggested. However, when it comes to embedding fonts, normally, Word can't do that for you. You cannot embed fonts into an HTML document using Word. I am inferring that Calibre, what, pulls the referenced font files into your ePUB, is that right? When you convert the Word file using Calibre, to ePUB?

That is correct.

With regard to the cycles and the polishing phases...didn't you say that this was a learning exercise? As opposed to a commercial book? You downloaded a txt from Gutenberg, and all that? Or have I misunderstood?

To conserve energy, I'd like to avoid getting into project specifics that don't bear on the technical issues at hand.

Well, I expect we're about to get to this.

Firstly, any font that you actively put in an eBook will most likely need licensing. Most fonts are simply licensed for "desktop publishing" purposes, which is very different than "embedding" or "redistribution." Don't get confused by font licenses saying "commercial;" that is unrelated, usually, to redistribution/embedding. Be aware that licensing a font for redistribution/embedding usually increases the desktop publishing price by a factor of 10 or more. I just today looked at a font that was $79.50 for both Desktop Publishing and Webfonts (websites, essentially) licensing, and it was $795.00 for a single eBook license. That's not unusual if dealing with foundry fonts. ( As someone who messes with this commercially, I have a boatload of alerts set up for sales and deals on fonts licensed for these uses, just so I can scoop them up at a good price.)

• This is useful information. I gather that I'll be faced with the higher prices when licensing fonts for use in an eBook which I plan to market on the various digital publishing platforms.
• Or do those prices only apply to redistribution of the fonts?
• I plan to avoid redistribution using font obfuscation on top of the publisher's encryption

Second--Remember that in the normal course of things, you'd have to add the font file, the entire font file, either to Calibre in the ePUB Editor or in Sigil, to make it work. Simply putting CSS into a stylesheet, calling for (e.g., "Monotype Corsiva") won't make the font display. It might display on your computer because you have Monotype Corsiva on your system. But if you send that file to someone who doesn't, they won't see Monotype Corsiva. They'll see whatever the device substitutes for it. I'm not familiar with Cailbre for conversion, which is what you must have used, but my comprehension is that it "pulls" the fonts from the system's fonts folder, that are used in the eBook file? That's the deal?

• That is the deal as I understand it. I'll find out when I have time to resume the project and start uploading to the digital publishing platforms.
• If they end up stripping my fonts, I'll be interested in finding out more about your work-around for this.

Great, thanks.

Honestly, you're doing this the hard way. You seem to have started with Calibre and Sigil first, rather than cleaning and styling your Word file first.

I'm afraid you've got the wrong end of the stick on this point. I learned from Guy Kawasaki (APE author) to polish my use of styles in the source document before proceeding to the publishing tools, which is what I did.

I really don't understand where you're coming from with the "outmaneuver automated structuring" part, or even why you want it. You can make ANY bit of text, any heading, any paragraph, etc., look how you want it to, in both Word and in HTML. You're making things hard on yourself for no reason.

Wrong again, I'm afraid. I'm finding the best long-term solution to each problem as I encounter it. Covering my tracks where automated conversion tools might otherwise intrude is an example of this. I like the results and it will save me a lot of work down the road.

@Diap: as the poster doesn't really want any HTML/XHTML/CSS information, but only wants to work in Word, I guess this should move to the Workshop forum. I'm sure that more folks who do the lifting in Word will be there, instead of here at the Sigil forums...that's my guess.

Taking the advice. Moving this section to workshop

Hitch

Sounds good to me! Either way, this has all been very helpful. Thanks to all for your comments and insights.

[dickloraine]

Why do you use an automated conversion and then try to trick it? How should it automate correctly, if it doesn't know, what are headings and what not? By removing semantic information you don't make your documents future proof, on the contrary, you make it less so. You can style your headings as you want and let them be headings. I'm a little bit confused, why you would do this, if you relay on an automated conversion

[Hitch]

Originally Posted by DLSieving

I'm avoiding the use of numbered sections and chapters so that no conversion algorithm for any publishing platform going forward will be able to easily figure out what I'm trying to do and impose its own structuring.

<SNIP

I'm afraid you've got the wrong end of the stick on this point. I learned from Guy Kawasaki (APE author) to polish my use of styles in the source document before proceeding to the publishing tools, which is what I did.

I've read Kawasaki, and I've seen (and used for experimentation purposes) his templates, for Word and for INDD. Now, I'm going to say this and then shut up, as you seem to think that anything I say to you that conflicts with what you've inferred, is wrong, but there's absolutely no "platform-structuring" that will be "imposed" by some retailer. NONE OF THEM do that. The only thing that I know of that does some type of "auto-structuring" is Calibre, which isn't a platform. Neither Amazon, iBooks, B&N, or any other retailer, impose structure, or do anything based on what headers you do or don't use. That's why everyone here--with probably, I dunno, 10,000 total books under their collective belts--are fuddled by your process. Because it seems to be designed to get around a system that is not only non-existent now, but has no implementation in the foreseeable future.

Wrong again, I'm afraid. I'm finding the best long-term solution to each problem as I encounter it. Covering my tracks where automated conversion tools might otherwise intrude is an example of this. I like the results and it will save me a lot of work down the road.

OK. Well, obviously, I'm not going to be useful to you. If you're going to talk about CALIBRE, as an "automated conversion tool," that's one thing. But that has NOTHING to do with the platforms. And frankly, 5 minutes of regex would resolve any Calibre-imposed structure, and lastly, even THEN, it doesn't impose it unless you don't bother to spend an hour or so learning it.

Done now. BTW, font foundries don't really make a distinction between an "ebook license" and "redistribution." You want to subset the font, fine, but you may wish to be aware that foundries don't have text in their licensing requirements exempting licensing if the font is obfuscated.

I'll leave the folks in the Workshop to it.

Hitch

[Toxaris]

You would be much better of to just use the header styles in Word. It doesn't matter how it looks after conversion to the ePUB. The looks can easily enough be changed with some styling in the stylesheet. It is all about structure. An ePUB can be split into two parts, a structured part and a part that defines the layout and looks of the structured part. In your source you want the structured part as correct as possible. The layout and looks part can then easily be adapted to the target format, whatever that may be.

You road is only giving you more work now and for sure later when corrections are needed. I know of only one 'publisher' that has a automatic conversion that is not that good, and that is Smashwords. Their process is, aptly named, Meatgrinder.

Font embedding can be difficult, but usually the hardest part is the font licensing. A license for ePUB (electronic publishing) is usually more expensive than for printing. Subsetting a font may result in lower license costs, but is no way certain. Obfuscation also will not matter, it can easily enough be circumvented. If you want a low-cost solution, try using fonts that have no license costs for electronic publishing. I believe Charis is such a font, but I am not sure at all about that.