Page 1 of 1


Posted: Wed Feb 11, 2009 11:28 am
by deepheat
Hi There

I have spent the last few weeks "auditioning" various Font Managers and I have been using MainType to for a few days now - so far so good!

However, there are a few items missing (or at least I was not able to find them)...
  • Duplicate Font Finder
  • Locate Corrupt Fonts
  • Rename Fonts
  • Find Font based on PANOSE attribute
While the functionality is available via a combination of various Font Managers, the layout, performance and "general feel" of MainType is much preferred. However, what I feel is missing from MainType is a comprehensive search facility as demonstrated by FontExpert2009, that allows searches via Personal Rating, Foundary and most importantly PANOSE attribute.

I notice that MainType displays the PANOSE number within the "Classification" section of the "Information" window and that appears to be able to filter on a few PANOSE "attributes", however what it doesn't do is to allow for explicit searches on (for example) "Weight=Light", "Arm Style=Non-Straight Arm/Wedge" and "Letterform=Normal/Boxed" (all legitimate PANOSE attributes). If MainType had (at least) this functionality, I would not hesitate to purchase it.

Many Thanks for your time (and for a thoroughly decent product)


Re: PANOSE Search

Posted: Wed Feb 11, 2009 12:06 pm
by Dave Crosby
Hi Dave,

So you click on the Columns icon, select Panose and move it to the head of the list, click the Panose heading to reverse the order, and everything is sorted from greatest attribute (2 Latin Text - 5 Symbol) to least (x-height etc.).

Then you open the information dialog on classification which explains what each of those numbers mean.

That tells you what the arm style etc. is.

Do you want to go beyond that and be able to sort within the Panose sort?

Re: PANOSE Search

Posted: Wed Feb 11, 2009 12:27 pm
by deepheat
Hi There Dave

Many Thanks for your speedy response.
Dave Crosby wrote: Do you want to go beyond that and be able to sort within the Panose sort?
In a nutshell - yes. That would be perfect.

Currently our font collection is around 30,000 items (currently sorted into alphabetical folders) and while your recommended workaround will indeed work, I would ideally like to be able to pick out all of the Demi Weight, Medium Contrast fonts in a quick hit so that I can see which fonts will suit a particular project/assignment/client effectively and efficiently.

Once again, many thanks for your time.


Re: PANOSE Search

Posted: Fri Feb 13, 2009 4:40 pm
by Dave Crosby
I have over 100,000 fonts and Dick Pape has over 300,000.
I think your request would be nice, but Erwin has limited time.

Perhaps Dick will explain how his system works. I just have a Panose Folder in my Fonts Folder with sub folders for the first two digits. It works well for me as I can find the style I'm looking for, then search out particulars - usually visually.

Re: PANOSE Search

Posted: Fri Feb 13, 2009 7:07 pm
by deepheat
That's an impressive number of fonts. Although it must make choosing the right one and very arduous and lengthy process!

I take it from your statement that searching within the Panose number will not make the next release of MainType - if this is indeed the case then I would be eternally grateful to Dick for sharing his cataloging methodology.



Re: PANOSE Search - Long Answer

Posted: Sat Feb 14, 2009 4:02 am
by Dick Pape
Hi Dave, Hi Dave.

I have stopped using the information within fonts. NOTHING is reliable. Too many people have had their hands and their footprints inside the body. Thieves remove dates and copyright attributions and change the names to hide their tracks. There are too many conversion programs which change from format to format so you lose track of the original font and sometimes even the original design. Fonts coming apparently directly from Adobe also seem to have traces of "fat fingers" all over (font dates of Jan 1, 1904!)- so I've given up.

My fonts are organized by internal font name (changing the external file name). Any name or family can be found in seconds using a standard directory search. Font name is the only absolute identifier easily available.

The internal font name may not adequately or correctly describe a font of course. For instance, a "Bold" font may have various weights depending. A properly calculated PANOSE weight (Weight Factor = Cap(H) / Wstem(E)) would resolve such naming problems to where a bold font is not one that is simply darker than another in the same family.

I have set up a separate classification scheme to describe a font in detail. The design is based on questions of character shapes (Identifont). A "Keywords" variable has been added to ascribe specific design traits, such as Adult, African, Airplanes, Ancient Languages, Animal Life, Architect, Art Deco, Art Nouveau, Australian and so on. I can get very close to the final font (or really group of fonts) very quickly but it's taken years to write everything down.

What's good is that the criteria are consistent and what's bad is it takes alotta time to record those variables (maybe 15 minutes a font). It's an application written in MS Access. Currently there are more than 21,000 fonts described in gory detail and over 12k more candidates in queue. This design has been sufficient for this size database. To describe a 50k data base it should be automated and new classifications should be added as you end up with too many almost duplicates of Arial or Times New Roman.

PANOSE is losing support in the industry. Perhaps only Adobe and Microsoft (The Monotype Corp.) still encode their fonts. There are too many very good fonts which don't have it, so what good does it do? Edwin can't program around the lacka data problem. (If you were to "survey the population" or something to see the percentage of fonts that contain valid PANOSE values in the very easy first position (Family Type - 2-Text, 3-Script, 4-Decorative, 5-Symbol), the result would be surprising. Possibly 25% ?? Not enough. It goes rapidly downhill thereafter).

The font user community has grown a whole lot more sophisticated and demands far finer nuances than given with PANOSE. Every designer has list of "favorite fonts" which may have very similar PANOSE numbers but offer more subtle differences in feeling and shading.

Re: PANOSE Search

Posted: Sun Feb 15, 2009 7:44 pm
by Dave Crosby
If you were to "survey the population" or something to see the percentage of fonts that contain valid PANOSE values in the very easy first position (Family Type - 2-Text, 3-Script, 4-Decorative, 5-Symbol), the result would be surprising. Possibly 25% ?? Not enough. It goes rapidly downhill thereafter
25% ?

I would guess that 80% of my fonts have no Panose information AT ALL! Probably half of the remaining 20 % have something like 0040000000 in them. I attribute this to the fact that most font makers never even knew Panose existed. There does seem to be many new fonts coming out that DO have Panose numbers (hopefully from reading Panose Your Fonts), but I understand Microsoft is thinking of abandoning Panose support in it's new Operating System.

Dick's software is good if one has the time, but I remember two totally different fonts internally named Alien.ttf that erased each other. And what good is a name like Alamo.ttf that when you finally get around to looking at it ... Awww, it is just another Arial knock off :evil:

So I continue tossing my new fonts into 0ne of five folders (2-Text, 3-Script, 4-Decorative, 5-Symbol, Special collections) and periodically move them into appropriate sub folders. At least that way I may see them again some time.

Re: PANOSE Search - Duplicates

Posted: Sun Feb 15, 2009 11:24 pm
by Dick Pape
1. Internal font coding is incomplete therefore you cannot rely on Foundry or Designer names for selecting or classifying fonts.

2. Invalid fonts often were those published before Windows XP where incomplete names were allowed. Using FC AutoNaming fixes these. Fonts that can't be read or corrected should be thrown away. Easy go!

3. Duplicate detection -- the bane pain of collecting fonts or anything. It's gotta be thorough or it really can mess you up. Most utility programs don't work thoroughly.

I go through a multi-step process trying to eliminate duplicates to reduce the volume that has to be looked at individually.

a. Fonts with same name, same file size and same file date (or name & CRC) are duplicates. Throw one away.

b. Fonts with same name and same file size are also the same font. A file date can be altered. Throw one away - I tend to keep the newer one "just because".

c. Fonts with same name could be different designs. (Same design with different sizes could be anything: fewer characters, no hinting, no kerning, older version, converted font, etc.) Gotta look at each one. My duplicate checker will display both sides of the compare so can immediately keep both if the designs are different. I will error on keeping both sides of a compare renaming one side to be (1), (2), etc.

d. In my scheme, fonts with different names are new fonts.

Every year or so I go through the files looking for fonts with the some name and verify they are different designs. (Errors do accumulate). Last count I had about 10,000 unique fonts with identical names (200mb compressed). I've even found some designers re-use their font names!

Dave C has a life with other things in it - wood working, geology, home and family. I've pared down my life to Fonts and Font Creator!

Re: PANOSE Search

Posted: Fri Jan 19, 2018 1:22 am
by Bhikkhu Pesala
MainType 8.0 now supports PANOSE data in the Font Information Panel.
Panose Information.png
Panose Information.png (9.67 KiB) Viewed 2876 times