Certain letters not showing in Photoshop... unless copied (?)

I've been working with Lunatix Bold recently. It has all the accents and special characters I need.
However, when I'm writing something in Photoshop, it doesn't support certain characters, such as ¡, ¿, Ñ and the acute accented vowels (to be precise, it doesn't accept anything that is not the standard English set of characters). Then I write the text, using the same font, in Word. It's fine. I copy-paste the text into Photoshop... and it's fine!

Does anyone know why this happens? If this helps, I use FontCreator 5.6 to view the characters, and all non-standard characters are blue (the standard are green or sometimes red).

Micha Mirck's picture

Did you get the font for free somewhere or are you a version you have paid for?
The illegal versions have funny unicode values and I'm sure the legal version is correct.

Ultimate Hammer Bro's picture

It ws a free version. The weird thing is that I'd expect it not to have certain characters. It's how they show up in one program and not in other what surprieses me.

hrant's picture

On Typophile we don't condone font piracy.

hhp

Ultimate Hammer Bro's picture

I don't think the one to blame is the one who downloads it when they see it as free-to-download, but whoever uploads it as "free" even though it's not.

Micha Mirck's picture

@Hrant... absolutly right, but Google gives the illegal version as first link.... and 50% of the first 6 links are illegal versions... That doesn't make it any easier for end-users to know that it is an illegal version

@ Ult. Just buy the official version. It will work and you work in a legal way

Michel Boyer's picture

Concerning the font of the first link, I am surprised that there are applications that can actually use it at all; the "unicode codes" of its glyphs are all between 0xF020 and 0xF12A (space is 0xF020, you can guess the rest).

hrant's picture

You guys are right of course - sometimes it's pretty hard to know that a font is "hot".

I guess, since there's so much pirated stuff out there, it's best to spend a minute checking first. But when they change the name of the font too, that makes it almost impossible! :-/

hhp

Michel Boyer's picture

That font is dated 1990. Font pirates were very active years ago. That particular font was probably extracted from a Postscript and then incompetently processed with a font editor. I have the feeling that pirates have other things than fonts to play with these days. Am I wrong?

Micha Mirck's picture

I did a search with Bing an Yahoo on the font. 1ste result was Emigre, 2nd myfonts and 3rd identifont. The "funny" sites show later in the results. Is Google supporting font piracy??

@Michel My guess is that this font was converted from Mac PS to PC TT with a bad tool and/or little knowledge. That happened a lot, where a designer would give a converted font to his client...
And looking at all the sites that offer normal fonts for free, it seems to me that pirates still think fonts are interesting enough to play with...

Birdseeding's picture

@Micha: Top three for Google is Emigre, then MyFonts, then a list of images, all of which appear to be specimens. (Then two pirate sites before Fontshop and Identifont.) It's not difficult to figure out that this is a commercial font, people.

Michel Boyer's picture

My guess is that this font was converted from Mac PS to PC TT [...] That happened a lot, where a designer would give a converted font to his client...

The font has no kerning table. Moreover, Mac OS Roman contains the characters igrave, iacute, icircumflex, idieresis; those characters in the font contain only the diacritic, not the dotlessi. They are incomplete. How come if the font was obtained from a fully functional mac font?

charles ellertson's picture

I have the feeling that pirates have other things than fonts to play with these days.

Am I wrong?

Sadly, yes. Arno Pro, Caecilia, Dolly, H&FJ Didot opticals, on & on...lots of newer fonts can be found at the pirate sites...

I use to report them to the foundries, but some don't seem all that interested, and in any case they just pop back up. I'd point out that for anyone who's time has any value, it is cheaper to license them legitimately than to steal & have to create the metrical data, but I guess the amateurs don't count their time as a cost.

The situation you mention just above could have come from someone extracting a font from a PDF file. Newer PDFs have protection, but not the older ones. The font data was there in .pfa form, without metrical data.

Micha Mirck's picture

I looked into the font again. It contains all glyphs for PC, not for Mac, so it seems to be a converted version with one important setting done wrong. And old converted PC TT (with a program like Alltype) never had kerning.

@ Birdseeing
Google automatically directs me to the dutch page and there the first result is the "funny" site... It's just great that even Google is supporting certain ideas about us Dutch...

blokland's picture

Hammer: Then I write the text, using the same font, in Word. It's fine. I copy-paste the text into Photoshop... and it's fine!
  
The (Unicode entries in the) cmap subtable for Mac OS seems to be correct and (assuming you’re working under OS X) perhaps MS Word is using this? If it looks right in Photoshop after pasting, this doesn’t automatically imply that the Unicode code points are correct too then, I reckon.
   
Michel: That font is dated 1990.
   
The original copyright dates from 1990, but the ‘Name’ table shows Macromedia Fontographer 4.1 10/5/97 in the ‘Version’ string.
  
Micha: It contains all glyphs for PC, not for Mac, so it seems to be a converted version with one important setting done wrong.
  
I had a brief look at the glyph set and it looks to me that the Symbol set is missing. The original font seems to be from 1990, so it may have had Adobe Standard Encoding.
 
FEB

Michel Boyer's picture

The (Unicode entries in the) cmap subtable for Mac OS seems to be correct [blokland]

Most entries in that table after 0x82 are incorrect. Here are just a few for comparison; on the left is the Lunatix entry of the cmap for Mac OS, on the right the character expected by Mac OS Roman.

        Lunatix         Mac OS Roman

0xf9    ugrave          breve          
0xfa    uacute          dotaccent      
0xfb    ucircumflex     ring           
0xfc    udieresis       cedilla        
0xfd    yacute          hungarumlaut   
0xfe    thorn           ogonek         
0xff    ydieresis       caron          

Also the codes from 0x82 to 0x9f are control characters in unicode and correspond to glyphs in that table

        Lunatix         Mac OS Roman    Unicode

0x82    quotesinglbase  Ccedilla        uni0082
0x83    florin          Eacute          uni0083        
0x84    quotedblbase    Ntilde          uni0084        
0x85    ellipsis        Odieresis       uni0085        
0x86    dagger          Udieresis       uni0086        
0x87    daggerdbl       aacute          uni0087        

What encoding would they correspond to?

blokland's picture

Michel: Most entries in that table after 0x82 are incorrect.
 
Ah… punished for a too brief look at the font.
 
FEB

Theunis de Jong's picture

Michel wrote:

What encoding would they correspond to?

That would be Windows: Western (which is Latin-1 plus the range that Unicode forbids). Is it perhaps a duplicate of the cmap for Windows? (If it contains more than one, that is.)

Micha Mirck's picture

The font has a MS Windows Symbol encoding. I have had some problems with some TTF barcode fonts with the same encoding and only with the gylphs beyond the 'z', where gylphs would appear white. After changing the codepage everything worked fine.
My guess is that Word is copying the unicode values to PS, so the same gylphs are used. When typing in PS, PS is using the unicode values to 'find' the special characters or character after the 'z'.
I looked into some old TT symbolfonts and most only go up to the 'z'. Only exception is ZapfDingbats. I have no idea if that was done to make typing the symbols easy or if there was a technical reason for it.

erwindenissen's picture

When you type (or paste) text in Word while you've selected a Symbol font, it automatically tries to convert the Unicode codepoints to Symbol codepoints. That is why copy/paste from Word does make it work in Photoshop.

Symbol fonts are considered legacy nowadays. Besides it is plain wrong to have a Symbol font which contains letters.

Ultimate Hammer Bro's picture

Sorry for taking this long to reply, but I'll add that I'm not using Mac, but Windows (7).

Syndicate content Syndicate content