If i can get this to work, this should be an example of the results.

OK how the hell do I attach pdfs to this site? Everytime I attach the file and hit submit, it disappears in the post.

I can’t get a pdf to post. you’ll just have to try it for yourself to see for a little bit.

Wouldn’t it be a lot less trouble to just generate the PDF in Indesign and convert the text to outlines?

Yepp.

j a m e s

I agree - if one used InDesign. This process is just as easy, however, once all the initial settings are employed; and perhaps more viable since Photoshop seems to be more prevalent than InDesign. The big question, however, is why more people don’t know or do either of these things. I would for my own edification, however, like to see a pdf created in the manner you describe.

I don’t question its applicability in the least - only if the level of visual quality is equal in both scenarios, and/or if there are any backdoors, so to speak, to extract fonts from a pdf created in the manner you describe. Whatever the case, given the degree of pilfering out there, this is knowledge every typographer should have and employ when producing these samples, regardless of the method used. Apparently from what I’ve been told, there is a way for someone to obtain from the myfonts.com site the entire list of kerning pairs for the fonts sold there and if a sample pdf has an embedded font and not merely a subset, the result is far more usable than most here seem to think.

Given that just about every font ever produced can be found online somewhere with a lot less effort than it takes to dig it out of a PDF, this is all sort of a moot point. I think that anyone willing to get into extracting characters from PDF specimens and combining them with kerning tables extracted from web apps probably knows enough about computers to get a font from a torrent aggregator or usenet, which takes a lot less time and effort.

Absolutely no doubt about that last comment. Still, it does not apply to those fonts yet to be distributed or released. Including those being critiqued on this very site. I believe you’ll find that some of the greatest typefaces - which haven’t been distributed in an extractable pdf form - are nearly impossible to come by in any way other than purchase.

There has been much debate over the functionality of extracted fonts, and most see it as no big deal because of a lack of kerning, etc. I’m convinced, however, that 75 percent of font purchasers either don’t know what kerning is, don’t care, or both. It’s a subtle yet definite epidemic and costing typographers exponentially more revenue than people here seem to think. A typographer with expert knowledge on the issue of kerning will more than likely believe a font is worthless without it and doesn’t see the danger. Most buyers, however, couldn’t define what kerning is much less insist on a font having impeccable kerning - or any at all for that matter.

Not to mention if given the choice, in my opinion, between buying a beautifully kerned font for $35 or extracting an unkerned font for free, someone will be losing that much money probably 90 percent of the time. I doubt people who spend money on fonts then turn around and give it away to a bunch of people unwilling to pay for it.

For example, I believe Font Farm has beautiful typefaces, including Kofi, etc., but few if any of their fonts are being circulated by pirates. Why? Check out the website. Download the pdfs. Google the font name, designer’s name, foundry’s name, etc., and look for some. Does this mean bigger profits for Font Farm? Who knows? But I am relatively certain this kind of unavailability in the long run convinces a would-be torrent downloader or pirate to break down and buy the font if they really love it. The semantics are arguable, but the frame of mind of potential buyers with some computer knowledge can’t be. He or she invariably will look first to see if it can be obtained for free, even spending months at a time doing so, before considering a purchase if they do at all.

I’m no zealot or anything, but for Pete’s sake, when extracting can be made nearly extinct in a few basic steps like the two methods described above, it should be protocol.

If this is proved true I want to post this has a hot topic!

You could just make a JPG and throw it in a PDF document…

But really, the reason for keeping it vector is to be able to zoom into the finer points and it not be pixelated, which you lose in a raster conversion. Wether it’s outlined or an embedded font. Personally, I think outlining is a much faster and straightforward process, and if people are going to go through the pains of unlocking and extracting fonts, you bet they’d sooner turn to usenet or a torrent site to grab it.

I don’t see unreleased fonts in PDF’s being a big market for someone to waste time on. For one they’re unfinished, and why waste time on an incomplete product when you can wait to get it in a few months. Considering the limited scope in which a test PDF would even be available to, I’d trust they weren’t trying to rip you off. Aside from here, friends, and a few other type designers I don’t see a test sample document really being out there to the general public, and one would hope that in that small group of people that font pirating from a PDF would not be an issue.

> I believe you’ll find that some of the greatest typefaces - which haven’t been distributed in an extractable pdf form - are nearly impossible to come by in any way other than purchase.

Really? Then why can H&FJ and Enschede’s fonts be found on warez sites. AFAIK, these foundries don’t provide PDFs on their sites. (Enschede actually has sample sheets in PDF format, but these are made of 600dpi bitmaps)

> For example, I believe Font Farm has beautiful typefaces, including Kofi, etc., but few if any of their fonts are being circulated by pirates. Why?

Maybe because Font Farm is not as know as other foundries...

> But I am relatively certain this kind of unavailability in the long run convinces a would-be torrent downloader or pirate to break down and buy the font if they really love it. The semantics are arguable, but the frame of mind of potential buyers with some computer knowledge can’t be. He or she invariably will look first to see if it can be obtained for free, even spending months at a time doing so, before considering a purchase if they do at all.

Generalization with too many assumptions.

> I’m no zealot or anything, but for Pete’s sake, when extracting can be made nearly extinct in a few basic steps like the two methods described above, it should be protocol.

If your goal was to protect the outlines, I’m sorry to say you’ve failed miserably. Just try opening your PDFs with Illustrator.

“Then why can H&FJ and Enschede’s fonts be found on warez sites. AFAIK, these foundries don’t provide PDFs on their sites.”

No, their fonts just appear in extractable PDFs from Newspapers and Magazines from India to Indiana. Additionally, these two foundries have much longer histories and far more experience ... and thus have had many typefaces out there in some fashion for a lot longer than Font Farm. I think Textaxis - which indisputably produces incredible work - also can be put in this category.

“Maybe because Font Farm is not as know as other foundries...”

Have you ever once taken the time to visit the font forums on any of these “warez” sites? Have you ever looked through some of the requests posted on usenet? More people are clamoring for fonts from Textaxis and Font Farm than any others out there. In fact, these requests have decreased over time because people are aware of the difficulty in ever obtaining fonts from these foundries. Unfortunately, some of the work on the Textaxis site was stripped from the flash pages there, and for a brief period, a few Font Farm typefaces were “posted.” However, the reproduction of all these fonts was so poor, they hardly resembled the original font.

“Generalization with too many assumptions.”

Well, Miguel, I will go ahead and tell you this, then: if you find no value in using alternate methods when producing font samples/books in pdf format, by all means continue along your present course. And that goes for pretty much everyone. I’m certainly not trying to introduce a new bill to Congress here, I’m simply describing what I’ve encountered in my experiences and shining a little light on the “real world” that some typographers seem to believe doesn’t exist.

Secondly, what you identify as generalization with too many assumptions is in fact a direct statement made after actually witnessing it all. So I’m going to guess, Miguel, that when it comes to just about anything you either use or consume - or even enjoy - you’d turn your nose away from what to the best of your knowledge is a free, equal-quality, identical version of an item in favor of paying in some cases exorbidant prices for the “real thing.” To many, these extracts are just that - exactly the same. So what if they don’t understand kerning ... no one’s obliged to appreciate its benefits.

“Just try opening your PDFs with Illustrator.”

And then do what? Assuming you produced a pdf correctly using either of the two above methods and you open it in Illustrator, how would you then go about extracting the fonts? What I’ve described here is one potential method, and I believe a second also was provided.

In case you have any further doubts of the seriousness of this, let me say here that in the future you are alll but assured of running into more and more PDF samples that look beautiful but under document properties in Acrobat have no fonts listed.

This certainly wreaks of deja-vu, as I seem to remember several other occasions in which I’ve been called to the carpet by Miguel. To which I can only say, Miguel my original intent was to be of some kind of assistance to other typographers and relate to them what I’ve run into when going through the warez sites and usenet.

For you to disparage my comments as assumption, from where I sit, would first require you to a.) have spent the kind of time I have researching these things and b.) KNOW in a factual sense I was “assuming.” I sincerely doubt that, so rather than accuse someone of making assumptions, perhaps you could refrain from making accusations that in fact are only YOUR assumptions. Forgive me, but it makes me dizzy to hear one essentially say “I’m assuming your making assumptions.”

Like I said, however, by all means - go right ahead and continue embedding your fonts in those pdf samples if you elect to create them. Spread them around the Internet. I’m sure you do some beautiful work. While doing so, however, consider how you’d answer this question: How much more time and work would it be to protect fonts in pdfs, and if the answer is little to none, then tell me which pdf creation has the potential of resulting in more harm to you? Once this method of protection is routine knowledge and unassailable, what sensible person would even consider creating an extractable pdf? It’s blue pill-red pill stuff at its most pure and only someone ignorant would base the choice of what they take on the color of the pill rather than its content and potential benefit.

I apologize for being long-winded here, but what was the epitome of a take-it-or-leave-it post and observation - especially one presented for the purpose of assistance - certainly doesn’t warrant such disparaging, cynical response.

Look Christopher, I appreciate you coming here and share a procedure that you make sound as if is a no-brainer that will solve all our problems, when in fact it isn’t, IMO. The reason I’m dismissive is because I’ve seen one too many threads like yours, started by people that all of a sudden find out that it’s possible to extract (crippled) fonts from PDFs, SWFs, etc., and make alarm-sounding posts as if they were on a quest to save the World.

If the procedure you described is the one that works best for you, great! But I think that, if you’re worried about people making crippled versions of your fonts by extracting data from PDFs, I’d say you’re better off reducing the number of characters used when composing the typeface sample, and then generate the PDF using subsetting embedding. This is what many foundries do.

“But I think that, if you’re worried about people making crippled versions of your fonts by extracting data from PDFs, I’d say you’re better off reducing the number of characters used when composing the typeface sample, and then generate the PDF using subsetting embedding”

If it first required me to carry a Volkswagen Beetle up a hill before I had the ability to produce unextractable pdfs, I wouldn’t be dumb enough to start scouting hills. However, when it requires pushing a few buttons almost the same number of times to create one or the other, I not only can produce a secure .pdf, there would be no need to reduce anything I wanted to present, including the number of characters. That’s my point, Miguel.

Let’s for argument’s sake say neither of the above methods is foolproof. I’d still swear that either of them makes extracting a font 10 times harder. And when you consider the disparity in visual quality is microscopic, you could create a pdf the same way you’ve always done it and instead of hitting Button X six times to produce a pdf with embedded fonts, you could hit button Y six times and produce a pdf just a nice looking but 10 times more difficult from which to strip fonts.

“if you’re worried about people making crippled versions of your fonts”

This is the least of my concerns. My only worry is about the people who don’t buy them. I could feel like these people are getting crippled, worthless versions of my font, but what I think of it doesn’t mean crap when those people couldn’t even point out what makes it crippled in the first place. At that moment it goes from being what you’ve identified as crying wolf straight to being a loss of $35 for the good guys, if that’s what my font cost.

By the way, I’ve known for years that “crippled” fonts can be extracted from PDFs. However, I also know that a.) you still haven’t described how to extract a font once you’ve opened “my” .pdf in Illustrator and b.) these “crippled” can definitely be made to walk again by the automatic metrics and kerning functions in Font Lab in minutes ... even if it is with a small limp.

I must say I am confused by your approach and reasoning. To me this is like hearing a bad-sounding noise in your engine. Yes, you could keep driving, save time, get there faster, and eventually grow to block out the noise. Thing is, that noise is not going to correct itself ... it will continue to get worse.

As an aside, discovering this method was far from a “no-brainer.” Just ask my recycle bin and the literally hundreds of pdfs it was fed for a couple of days. Even then I stumbled on the method purely by luck and still have to read through and perform the steps myself when testing it further. As for solving everyone’s problems, recent research has shown me a few people here know at least one way to make extraction nearly if not completely impossible. Font Farm has BEEN doing it. I believe even FontShop is beginning to produce sample pdfs in this fashion. What I know for certain, however, is none of those people were disclosing their secrets to me, this site or any other site that I’ve come across and/or been able to comprehend. I can’t say I’ve looked under rocks on each of the four corners of the Earth, either.

Curves (vector) in PDFs are extractable. You can safe a PDF and edit properties, so that the document can only be modified by providing a password... No matter what you do, there is always the someone knowledgeable who can open, extract...

A good way could be just low res grey scale tifs (200 dpi or so with compression). Problem is, a pdf file with a bunch of images is big file...

Regards

-Andre

> This is the least of my concerns. My only worry is about the people who don’t buy them.

OK, so generating the PDFs like you suggest is the key to make a given person buy the fonts instead of extracting them from PDFs, assuming that he/she knows how to do it? Please, give me a break...

Christopher, this discussion is going nowhere. You have your convictions, and I have mine. Many people that run foundries read Typophile. Many of these foundries put out PDFs to show their typefaces. If what you say is true, we shall see a shift in the way they generate their PDFs, and consequently a reduction on the frequency that their fonts show up on sharing sites, particularly the new releases. So, let’s wait and see. With that said, I rest my case.

Fair enough. One other quick observation.

“so generating the PDFs like you suggest is the key to make a given person buy the fonts instead of extracting them from PDFs”

No, not the key. That given person may not buy the fonts. However, the hundreds or thousands of people to whom that person would’ve offered the free extracted version might include a few potential buyers. Foundries will produce pdf samples forever, which I think is fantastic. Manfred Klein produces pdfs and his recent fonts are all free. I just can’t comprehend the possibility of someone hitting six buttons and producing an extractable pdf when they could hit six buttons and increase the pdf’s security tenfold while retaining visual quality, that’s all.

extractable PDFs from Newspapers and Magazines
Aren’t those .pdfs encrypted? I’ve downloaded my fair share from the Newseum, but that’s as far as that went.

And what Miguel is talking about opening the file in Illustrator is that you have access to the outlines themselves, the actual postscript curves. Of course, you don’t have any of the “font” info, but if it’s the curves themselves you’re interested in...

When it comes to extracting/illegally downloading, one thing is certain: Some dork designer who actually shells out for someone’s type design WILL eventually drop it onto the Torrent networks. It’s not an IF, it’s a WHEN.

So honestly, it’s a moot point AFAIK.

“access to the outlines themselves, the actual postscript curves.”

How would that then be converted into a usable font? My only guess was that a person could save the Illustrator file as a postscript file and try extracting from there. Unfortunately I don’t believe I have Illustrator so I can’t A.) make sure the process was followed on the Photoshop end and B.) what takes place when opened in Illustrator.

“When it comes to extracting/illegally downloading, one thing is certain: Some dork designer who actually shells out for someone’s type design WILL eventually drop it onto the Torrent networks. It’s not an IF, it’s a WHEN.”

A possibility, without question. At this point, however, both the security measures and their potential insufficiencies seem to exist only in this thread and only to a degree that everyone should trust everyone else’s skill and werewithall when using various programs.

You are making a fine, valid technical point, Curioustype. It’s just that most folks agree that it’s not really an effort that produces any meaningful results in terms of the availability of fonts in the wild.

Does it hurt to do what you suggest? Certainly not. Does it help? Probably not.

There is at least one EULA which I can think of which disallows including all of their font outlines into a single PDF.

I think while this idea is interesting it is, alas, only one more solution.

The people on warez sites can be quoted as discussing re-kerning fonts and looking for a specific glyph. They do re-build fonts from outlines just as they re-build from extracting fonts from crippled (sub-set) PDFs. (They share FontLab as much as fonts.) Further if they want the font badly enough, they pool together and buy a copy and then share that among themselves. If they have a fixation nothing gets in their way.

It is a valid way to do it, but it is, IMHO, more dangerous because more people would have access to the outlines. Afterall, some design doesn’t require paragraphs, just characters. Logos for one example.

As I recall the LetTerror folks led a grass-roots effort amongst type designers to have subsetting the default in PDF-creation and Adobe eventually made this change to Acrobat. So I think it is (or at least was) possible to facilitate change in this area - you just need to go about it the right way, build support and the business case and maybe the suits will listen. The wrong approach might cause the decision makers to dig in their heels. It’s human nature.

Personally, I’m with James on this one: “Given that just about every font ever produced can be found online somewhere with a lot less effort than it takes to dig it out of a PDF, this is all sort of a moot point.”

I’ll also support pretty much everything Miguel has already said.

That aside, all the legerdemain being suggested here is not very interesting, except for limiting the charset showing and subsetting to use only the glyphs shown. Anything else will degrade the visual quality.

If the font is still a font, it is still technically possible to extract it from the PDF.

If the font is not a font but just outlines, it has suffered a significant loss of visual quality, especially (but not only) on screen. This might not be very noticeable for display fonts at really large sizes. It would be a display problem for text faces. But the outlines would *still* be theoretically extractable from the PDF by this hypothetical determined thief.

Although the PDF can be protected via encryption, if you’re talking about evildoers who are willing to ignore licenses and break the law to extract fonts, they can easily break encryption too, unless the encryption blocks actually opening the PDF at all. (Note: there are more sophisticated PDF encryption schemes from Adobe than the basic one supported by Acrobat for PDF creation, but they are not the sort of thing I can imagine most type foudries getting into.)

If the fonts have been turned into bitmaps, the loss of quality is much more severe. They may look fine if displayed at the exact zoom level the bitmaps were created at, but if somebody either zooms out until the text is small, or zooms in to see fine detail, serious visual problems will emerge.

Cheers,

T

Thomas Phinney
Product Manager
Fonts & Global Typography
Adobe Systems

I think that the big question is... do we want that?

No matter what you do, ther’s alway someone that will break it in a few time. I still rely on the peoples common sense.

-

www.nunocoelho.com