Copyright ©1991, 1992, 1995-97 Hewlett-Packard Corporation.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, recording, or otherwise, without the prior written permission of Hewlett-Packard Corporation.
The information in this guide is furnished for informational use only, is subject to change without notice, and should not be construed as a commitment by Hewlett-Packard Corporation. Hewlett-Packard Corporation assumes no responsibility or liability for any errors or inaccuracies that may appear in this book. The classification system described in this guide is furnished under license and may only be used or copied in accordance with the terms of such license.
Hewlett-Packard, PANOSE, PANOSE Classification Numbers, and PANOSE digits are trademarks of Hewlett-Packard Corporation. Other brand or product names are the trademarks or registered trademarks of their respective holders.
For additional information, please contact us at firstname.lastname@example.org.
Printed in the United States of America
Printed: February 14, 1997
1.0 Introduction | 2.0 Latin Text and Display | 3.0 Latin Hand Written
4.0 Latin Decorative | 5.0 Latin Pictoral | 6.0 Summary of Variables
7.0 Calculated Variables | 8.0 PANOSE Submission Form
9.0 PANOSE Classification Sheet
1.1 PANOSE Classification
1.1.1 Metrics Guide
The PANOSE Typeface Matching System was developed by Benjamin Bauermeister and is exclusively licensed to Hewlett-Packard Corporation in Seattle, Washington. Under copyright this document is being provided to third party vendors of typeface products and type related utilities to aid in understanding the details of the PANOSE Typeface Matching System and to describe the process of assigning PANOSE Classification Numbers to typefaces.
Hewlett-Packard does not restrict the use of PANOSE Classification Numbers in typeface products developed by third parties. Hewlett-Packard does however strongly urge you to submit written notice and samples of typefaces developed with the PANOSE Typeface Matching System to Hewlett-Packard Corporation. Submission forms are available in Section 8 of this document This provides Hewlett-Packard with the information necessary to enhance, expand, and solidify the PANOSE Typeface Matching System as it is applied to an increasing variety of font designs. We welcome your feedback.
In addition Hewlett-Packard reserves the use of the PANOSE trademark exclusively for typefaces that have been officially verified and registered with Hewlett-Packard. PANOSE licensing information can be obtained by contacting us. Typeface classification and verification services are also available from Hewlett-Packard. Fonts classified by Hewlett-Packard are allowed full use of the PANOSE trademark. Sample classification submission forms are found in Section 9 of this document. Separate licensing is available for use of the mapping algorithms for cross platform and web font solutions.
This document provides a detailed record of the process required to classify a Latin text, display, handwritten, or decorative face or symbolic fonts with the PANOSE Typeface Matching System. Great care has gone into making this document as complete as possible in order to avoid any conflicts or confusion that may arise in the classification of typefaces. However, with the varied world of type design, this document is subject to change as more data regarding lesser known faces become available. Please feel free to send your additions or clarifications to by contacting us
The process of determining a PANOSE Classification Number starts with a series of measurements. While the total number of measurements is nearing sixty-five for a text font, far fewer are required to classify most typeface designs. Still, until the classification parameters are familiar, diligence should win out over expediency in the pursuit of creating PANOSE Classification Numbers that are consistent and correct.
Ratios are computed using the measured values. The classification criteria for PANOSE is based on these inter-related ratios. Because of this, PANOSE measurements can be taken on samples of type at any size, as long as the measurement system is not changed during the classification process of a given face.
The rules for proper measurement, laid out below, are being refined so that minimal human intervention will be required to classify a typeface. At this time, however, there are no approved tools for the automation of PANOSE Classification Number assignment. This document tries to provide detailed descriptions about the mechanics of measuring every attribute, while presenting methods to quickly glean the same information using visual feedback only.
PANOSE is a classification system for visual attributes of type only. There is no information contained within a PANOSE Classification Number that pertains to the character widths, spacing metrics, or advance widths. This simplifies the process of classification because we only describe attributes that can be seen and measured.
Figure 1 - PANOSE Universe
PANOSE Classification Numbers used in static outline fonts consist of ten digits. The first digit defines what type of font is being classified (Latin Text, Symbolic, Japanese Decorative, etc.) (Figure 1) and the following 9 provide classification within that type. Thus digits 2 through 10 change meaning depending on what the first digit is. There is always a digit that expresses weight, one that expresses aspect ratio, and a monospace / proportional flag but their position may change depending on the type of font. The order of the digits is roughly the order in which they are useful in maximizing the separation of closely related faces. Thus the position of digits expressing the same quantity (for example aspect ratio) will not always be the same (it is digit 4 in Latin Text and 5 in Latin Handwritten). Below is a summary of all the digits in all the presently defined font types:
1. Family Kind (= 2 for Latin Text)
2. Serif Style
6. Stroke Variation
7. Arm Style
Latin Hand Written:
1. Family Kind (= 3 for Latin Hand Written)
2. Tool Kind
5. Aspect Ratio
1. Family Kind (= 4 for Latin Decorative)
6. Serif Variant
10. Range of Characters
1. Family Kind (= 5 for Latin Symbol)
5. Aspect Ratio & Contrast
6. Aspect Ratio of Character 94
7. Aspect Ratio of Character 119
8. Aspect Ratio of Character 157
9. Aspect Ratio of Character 163
10. Aspect Ratio of Character 211
PANOSE uses an expanded form for other applications such as distortable type. Contact Hewlett-Packard Corporation for more information about the expanded form of PANOSE.
While measuring typefaces for the PANOSE Typeface Measuring System is generally straightforward, be aware of the following special considerations when classifying a typeface to assign a PANOSE Classification Number:
1.1.3 Italic Character Forms
This version of the PANOSE Classification Metrics Guide not only contains specific weight settings for fonts, but also contains classification parameters to isolated italic fonts based on their character skew. There are several measurements in this classification document that require special treatment if an italic font is being classified. Sometimes the horizontal distance between two points is required and the two points are not on the same horizontal plane. A simple measurement of the horizontal distance between these points is not going to yield accurate results. Compensating for the character skew when the measurement is taken will result in proper compliance with the PANOSE Typeface Matching System scheme. Whenever possible, the special cases of compensation for italic fonts have been noted in this document.
1.2 Classification Samples
It is important to work from a properly printed character sample when measuring a typeface to determine its PANOSE Classification Number. Unlike other systems of classification, a large type sample does not necessarily benefit the process of assigning PANOSE Classification Numbers, in fact, it often hinders the classification process. Keep the following concerns in mind when generating printed samples for PANOSE measurement. A sample classification sheet is available in section 9 of this document.
The printed sample characters should be large enough to view the details of the characters clearly. More importantly, make certain that all the rasterization hints applied to the character form are diminished and no longer affect the glyph outline. Characters of 200 points in size are sufficient for most PANOSE classifications. The exception to this is the uppercase O, which is used to determine the Letterform and Stroke Variation digits, and the uppercase I, which is used to determine the Serif classification parameter. The measurements for these digits are quite subtle and require a significantly larger character sample; 400 points in size is usually sufficient.
The use of low resolution output can be of assistance in determining a PANOSE Classification Number. The minute stair-steps that occur in the curves and serifs of a laser printed sample at 300 dots per inch (dpi) often indicates the inflection point of a serif from the stem, the bow of a stem, or the extreme boundary of a rounded character form. For this reason, 300-600 dpi output is recommended. Additionally, hard paper that presents each laser dot cleanly is of benefit. Resolution enhancement technology diminishes the advantages of lower resolution printing, yet will not adversely effect the correctness of the measurement attained from the sample.
1.2.3 On-Screen Measurement
It is quite possible to take the PANOSE measurements in a font editor software package. General illustration and drawing packages are not recommended for this process. The advantages of on-screen measurements are two fold: the measurements can be taken at any resolution and the true points of extremities and tangency can be located more precisely. The only drawback in using a font editor for PANOSE measurements is that most current tools lack elegance in the methods for measuring the distance between two random points on an outline. Undoubtedly these tools will change and improve over time, at which point on-screen measurement will provide a superior alternative to printed samples.
1.3 Extents and Theoretical Edges
Two general measurement terms frequently used in this documentation are “extents” and “theoretical edges.”
Extents define the upper, lower, left and right bounds of the object being described or measured. For example, the left extent of a glyph indicates the horizontal location where a vertical line first contacts the shape of the glyph. Similarly, the upper extent of the uppercase O glyph indicates the vertical location where a horizontal line first comes in contact with the glyph shape. Extents for italics are often determined on the skew.
Theoretical edges are used to indicate a straight line where a very subtle curve exists. Stem edges are often at theoretical edges because stem designs contain slightly bowed lines-not straight-lines. A theoretical edge is a line halfway between the right-most and left-most extents of the true edge of a stem.
1.4 Baseline location
All of the measurements necessary to determine a PANOSE Classification Number are based on the visual and physical properties of the glyph shapes, not on assumed locations of shapes, etc. The one exception to this is the baseline location. Whenever possible, use the information available from the system that is printing the character samples to indicate the correct location of the baseline. If a true baseline is not provided, the baseline is assumed to be the location of the theoretical edge of the lower edge of the lowest horizontal arm of the uppercase E. If this does not result in a horizontal line, the baseline is placed at the midpoint of the aforementioned theoretical edge.
1.5 Digit values of 0 and 1
The reader will notice that the value 0 and 1 are defined as Any and No Fit for every digit in the PANOSE system. These have specific meanings to the mapper. 0 means match that digit with any available digit. This allows the mapper to handle distortable typefaces such as multiple master fonts in which, for example, weights may be variable or serifs may change. 1 means that the item being classified does not fit within the present system. There are two possible causes of this. First is that there has been no work done on that family of faces, for example at the present time an Arabic cursive font would have the PANOSE number 1 1 1 1 1 1 1 1 1 1 as there has as yet been no work done on Arabic fonts. The mapper will recognize the font as having a valid PANOSE number and accept it but will only do name matching, not font substitution. The second possibility is that within the classification scheme there is nothing that fits the particular case that is being classified, for example a completely new shape of serif in a Latin Text font that does not fit the existing design space. A 1 would indicate that the serif doesn't fit but would still allow the mapper to do substitutions. If, in the process of classification, you find an example of something that does not fit within the present classification scheme, please contact us so we can evaluate it for possible expansion of the PANOSE system. We are well aware we have not fully described the typographic universe and are prepared to extend PANOSE as the need arises.