QUESTIONS pertaining to the code and symbol technology
---------------------------------------------------------------------------------------------
Return to Questions & Answers page.
Section 3 Question #1 - Why was it needed to create the middle guard bars (segment): 01010, as the digits are coded in such a way that scanner would not confuse what is from the left side and what is from the right side of the bar-code? Why not 0010100 that has the same width as the other segments? 9/29/98
Answer- The center pattern, which is shared by both the right and left halves of the symbol, serves to tell the logic which direction the beam is crossing the symbol (note that scanners need not read both halves of the symbol on the same pass nor in the same direction). The distance from the starting edge to the end of the second bar is 4 modules. The logic determines that this "character" is much more narrow than the others, thus it is the center. If 0010100 were used the character width would be 5 modules, still quite narrow, but less distinctive. Another reason, perhaps more important, is it reduces the overall width by 2 modules allowing the printing tolerances to be increased ever so slightly. Nothing would be gained by the wider center character, 0010100.
---------------------------------------------------------------------------------------------Section 3 Question #2 - Why are there two ways to encode each digital value and when is each used? 12/6/98
Answer- The U.P.C. symbol is designed so that even a simple system can read it in halves. The scanner may read either the left or right half first. Also a half may be read from --left to right--, or --right to left--. It is necessary for the logic to be able to determine which data is from the left half and which data is from the right half, of the symbol. The two halves are mirror images of each other, in that from the margin toward the center, there are two guard bars, followed by six characters, and the center bars. The characters (from the guard bars) start with a space and each character is composed of 2 spaces and 2 bars. There are two sets of 10 characters. One set is called EVEN parity and the other is called ODD parity. Those with an even number of black modules, such as "6" coded "WWWWBWB" (0000101) are from the even set. While those with an odd number of black modules, such as a "6" coded "WBWBBBB" (0101111) are from the odd set. A symbol half composed of six characters from the ODD character set belongs on the left (and is considered country "Flag 0"). A symbol half composed of six characters from the EVEN set belongs on the right. There are 20 positional arrangements of 3 ODD and 3 EVEN characters in a 6 character group. Ten of these arrangements are used to indicate Version E. Nine arrangements are used for left half of the EAN-13 symbol (they also represent value of the country "Flag", 1 through 9). The 20th arrangement is used to indicate the left half of Version D symbol. Examples:
G OOOOOO C EEEEEE G = UPC-A, Flag 0
G EOEEOO Ep = UPC-E with a check digit value 6
G OEEOEO C EEEEEE G = EAN-13, Flag 9
Where:
G = guard pattern
C = center pattern
Ep = end pattern
E = a digit encoded from the "even" set
O = a digit encoded from the "odd" set. By using only the three patterns:
1- 3 ODD & 3 EVEN
2- all EVEN
3- all ODD
a considerable amount of error checking is obtained.
---------------------------------------------------------------------------------------------------Section 3 Question #3 - Why did you choose modulo 10 and the weighting system of 1-3-1-3...? 10/8/99Answer- Any modulo equal to, or greater than, the number of characters in the data set will detect single digit errors*. A modulo greater than that will also detect some double digit errors. In the case of the U.P.C which is numeric only, a modulo 10 was the largest that one could represent in the check digit and was sufficient to detect all single digit errors.
*Example of a modulo less than the size of the character set:
Assume a modulo 7 were used with a numeric character set of 10 characters. If the numbers 1 & 8 were included in positions weighted 1-3 respectively, the pair could be replaced with 1 & 1 and the single digit error would go undetected.The weighting system is used to detect number transpositions which are the most common form of human keying errors (not so with scanning errors). Examples of this are keying 89 when 98 was intended or 76 when 67 was intended. The weighting used was actually proposed by a colleague of mine who was convinced it would also help detect certain two bit scanning errors. However it is not the best choice for detecting transposition errors. The weighting system of 1-3-1-3-1… will only detect 89% of the transposition errors. The weighting of 1-2-1-2-1-2… (used on credit cards) will detect 100% of this type of error. For the first ten or twenty years this fact was a mute point because in 1973 the UCC elected to print only eleven digits in human readable form, leaving off the check digit. They were concerned about the time required by the check out clerk to key an extra digit. Since not everyone was convinced scanning would replace keying, this trade off of productivity gain vs. error detection appeared prudent.
Section 3 Question #4 - Recently I have had several questions pertaining to the ISBN numbers and the U.P.C. used on books. I believe the following answer covers all of them.
Answer-
The U.P.C. symbol on books staring with the digits 978 is called a Bookland EAN code. To convert the ISBN to the Bookland EAN, begin with the Bookland Prefix 978. The next 9 digits are the first 9 digits of the ISBN, with the hyphens deleted. The 13th digit is the check digit, calculated mathematically using the preceding digits.Quite often there is 5-digit add-on to encode the retail price of the publication the right. The first digit of the add-on is the currency indicator. Five is the designation for U.S. dollars.
I recommend the folowwing web sites for more complete information:
Bar Code Graphics Incorporated and Morris Publishing
These sites include excellent descriptions of the bar codes used by the publishing industry.R.R. Broker assigns ISBN numbers – address:
U.S. ISBN Agency
121 Chanlon Road,
New Providence, NJ 07974
Tel: 877-310-7333
Fax: 908-665-2895
isbn-san@bowker.com
---------------------------------------------------------------------------------------------------||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
If anyone sends me a question pertaining to the U.P.C. symbol, history, or the like, which can be answered concisely via e-mail, I will send the answer to their mail box and post both question and answer of some of them here.
I truly enjoy helping people, however, it would be nice if you send a "thank you"
e-mail if I am of help!
Please put "U.P.C. Question" in the subject to avoid the possibility of your
question being thrown out with the trash mail.====================================================
![[symbol]](isbn-sym.jpg)
For contact information please go to my Contact Page.
Revised: October 2006