MTAC Workgroup 113
Address/Barcode Requirements for Flats Sequencing System
Teleconference Notes: June 14, 2007
IDEAlliance (IMB Height Update)
Jonathan Phillips from Domino presented information about the Intelligent Mail® barcode height issue. (A copy of the presentation is available on the MTAC website under WG 113 meeting notes/minutes section). Domino produces the JetArray , which is a continuous inkjet printer used in thousands of locations.
Jonathan advised caution and not to get caught up in the minutia. Jonathan said that sometimes there is a conflict between what USPS specifies and what the industry produces and we now have an opportunity to clear up the contradictions. Using the PowerPoint he walked us through the mathematical details of the height specifications.
Following is additional information based on the PowerPoint.
Slide 2: Each jet in the JetArray prints lines exactly at 0.166667”. If ink is deflected more than that it will be deflected into the line below. If the barcode is placed on the bottom line it can be taller because ink will be deflected into the clear zone, not another line. The barcode can therefore be printed up to about 0.166” tall.
Slide 3: The real restriction on barcode height is when the barcode is printed above the address block. Legacy situation dictates that every line is 0.16667” apart, which needs to be divided into human readable text, barcode and clear space. By deducting the current clear space requirement of 0.04”, the maximum practical text height is 0.1267”. Thus, if the barcode is printed higher than that, it will violate the clear space requirement.
Slide 4: If the clear space requirement is decreased to 0.028” the barcode can grow and still remain in spec. In this scenario the maximum practical text height becomes 0.13867”
In every mathematical sense, the USPS proposal to reduce the clear space requirement to 0.028” means printing with the JetArray will be as difficult as it is today but no more difficult than today.
Kodak prints a 1” address block, so the clear zone concession will make life easier for this printer too.
IDEAlliance (Carol Stream Test)
The test being conducted in Carol Stream is much more sophisticated than previous tests. It is being “set-up” the week of June 18 and testing will commence the week of June 25. The test is being set up to define boundary conditions. They will push the boundaries of decoding so a recommended technical spec can be disseminated that will be within automation compliance.
Boundaries to be tested are skew, void (gap between 2 drops), new environmentally friendly inks that have different drop characteristics, and legislation in terms of barcode compliance for automation and making sure it wholly represents what the industry can produce. Test will take two to three weeks and test results should help determine boundaries for skew, white space, void, and drop space.
We will not be testing the new 0.125” barcode height specification (only 0.134”). USPS knows from previous tests that the UFSM will fail to read when a barcode is printed below 0.130”, so the 0.125” specification will not work on the UFSM. USPS is working with our UFSM supplier to enable reading at the lower specification. There is no clear time on when the new decoders will be available, but USPS decided that in order to launch the IMB the spec needed to be loosened because the industry was not going to replace all of its existing equipment. It was asked that USPS and IDEAlliance keep in mind that if there is the ability to test on UFSM (once new decoders are available) that it be worked into the test plan.
It was mentioned that USPS may change the void specification from 0.005” to 0.01” because it is known to be an issue with the industry but does not appear to be a problem with the decoders.
A complete testing document has been produced, which can be circulated to the group. It describes the testing process and statistical sampling used. For example, the 100,000 test pieces were produced according to various specifications (different heights, skews, etc). The pieces were serialized and will be tested in batches so the same batches can be used in future tests in different facilities and on different machines.
Production of the sample images is being carefully controlled because it is recognized that different interpretations can lead to different results. For example, QUAD produced samples for USPS approximately 7 months ago. They recorded heights very carefully but when test results came back, heights recorded by USPS and by QUAD did not match. It was determined the reason was because height was being measured differently. They are trying to completely iron out these details at Carol Streambecause it can make the difference between an automatable piece and non-auto piece.
Jonathan Phillips said that looking at slides 2 and 3 of PPT, they show one example deflecting down into text and one deflecting away from text. As a manufacturer, they will have to issue best practices document for users of equipment that they will need to use so they deflect away from text.
Address Block Requirements – Draft
The address block handout attempts to show how the address block should be measured. Please look over the format to see if anyone finds it confusing (and let us know if it is).
The handout shows text point sizes but printers do not use “point” for type size. A “point” is equal to 1/72 of an inch. Therefore a 6 point font is approx 0.08”. What is the “read” difference between a 6pt font using all capital letters versus lower/upper case?
Irv Drewek stated that they print at 6 pt on paper labels and asked that we look at paper labels vs inkjet. He further said that the industry is set up for 1” labels and moving to an 8pt font minimum would have a dramatic impact on the industry because they would have to change their whole process.
The example does not show keyline or subscription line, which is essential for publishers and catalogers.
Charlie Howard stated that Harte-Hanks typically prints 2 barcodes, OEL, keyline, etc. (usually 8 lines total). The font is small butthey have had no complaints in the past, and it will be tough to jam everything into a 1” label.
Kathy Siviter asked if USPS has tested minimum point sizes. Different tests for upper vs. lowercase? Mike Amato answered “no” but that his group is meeting the week of June 18 and will discuss it then.
Other discussion items…
- What terminology does the industry use to represent type size? (we want to be consistent)
- How many lines do you use?
- B to B community uses 8 lines
- ABC auditing requires new subscription line
- Does 8pt font fit on a 1” label? (Only if a line is eliminated)
- What point size is being used on Carol Stream test mail? John Geres will be at Carol Stream and will look into that.
- The largest vendor of newspaper circulation software does 8 pt all caps – and it can be processed on AFSM and UFSM.
- Rich Bobic will poll the Fulfillment Ops group (from IDEAlliance) to see what type of info and how many lines of address group members use on labels. He will report the result back to our group. Their meeting is at the end of June, so he can get some feedback then.
- Should we specify clear space between lines of text (from saying line “should not touch” to providing a number)?
NOVA Test Analysis
Mike Amato does not have resources to continue doing analytical research on images for error analysis. Is there interest within the workgroup to conduct further analysis on the content errors identified in the test to find out what is causing it?
First there are two questions that need to be answered regarding potential release of images…
–sending images from one mail to other mailers could cause concern
–from USPS’ perspective, would release violate privacy law (because names/addresses appear on the images)
An alternative would be for mailers to investigate their own addresses. You can look at your database and compare what is printed on a piece vs. what barcode is being sprayed. CASS with DPV can be used to determine if any secondary information is missing. CASS/DPV will not provide the missing apartment information but it will flag that information is missing.
Kathy Sivitar asked if similar content error analysis tests will be conducted on the Dulles FSS. Mike responded that he is not sure because the schedule is up in the air, but he will inquire about that.
Page 1 of 2