[Date Prev][Date Next] [Chronological] [Thread] [Top]

RE: BOD Image Scaling in GEMS



Ian, can you please clarify that statement for me.  Do you believe that adding scaling to a BOD application will make the Minolta 20/25 meet our AccuVote ballot specifications?

 Yes.   On the Minolta 20, this scaling issue seems to be the only drawback.
 
Great!  I did not realize that the Minolta was capable of meeting your various ballot tolerance requirements.  This is a great improvement over the 5SI. 
 
This news will be of particular interest to Jeff Hintz, who used the 5SI in elections last year.  I have attached two mails posted to the support list regarding his experience for reference.  I am glad to hear that the Minolta will put these problems to rest.
 
Consider the scaling feature added to GEMS -- I'll do it today (its trivial).  It will be in the 1.12.1 release.
 
Ken
 


Hi All,

I talked with HP yesterday, and they seem to feel that the duplexing is
within their tolerances with the stretching of the image on the duplexed
side of the paper.  They recommended moving up the offsets on the duplexed
side, which then gives me a little bit off at the top and also a little bit
off at the bottom.  Seems to be OK, although I have the customers still
checking the ballot when they fold on the score to see if the score goes
through an oval.

I don't have any problem with the ballot being read by the Accu-Vote, even
when the image is stretched.  My problem is that when I line up the timing
marks at the top for both the back and front, the image stretches on the
duplexed side, and sometimes the score of the ballot lands in an oval.  When
I do blank ballot testing, the Accu-Vote can sometimes read the score as a
mark.  I don't know if the Accu-Vote is actually reading the score, or some
of the toner rub off that comes from the score, or both.  I am using the 90
lb. card stock that is provided by Adkins.  It looks like the paper is cut
wrong sometimes, as well as the score lines are off sometimes too.  There
also is toner rub off.

I ran the test deck that I created for Lancaster, NE today, 307 ballots.  I
ran it once, and the totals came out fine.  I ran it a second time, and I
received some overvotes as well as some votes for a candidate.  This
happened in only two races, but if we have to do a recount, what would be
the integrity of the count???  Is there a way that we can adjust the
Accu-Vote reading of the ballot so we do not pick up hesitation marks, toner
rub off, or scored lines???

One other note, I noticed that the printing of the ovals from RX Laser seem
to be thicker than the printing of the ovals in Gems.  While doing a blank
ballot test, I picked up a udef on a race that the ovals were completely
blank, without any scores or toner.

Jeff Hintz
Global Election Systems





Hi All,

After a lot of testing and running of elections with the Ballot on Demand
system and the Accu-Vote/Accu-Feed systems, I have come to several
conclusions that I both wish to comment on and get feedback for.

It seems to be pretty much impossible to be able to print to spec from the
HP5siHM Laser Jet Printer when duplexing, whether using the RX Laser
software or using GEMS.  Although the Accu-Vote is still able to read the
ballot, the problem is that the toner is rubbing off from the ballot which
the Accu-Vote can sometimes read as marks.  Especially when the fold is
either on or close to an oval.  For example, I created a test deck for
Lancaster, NE, around 300 ballots, not folded, and used duplexing.  I first
ran all the blank ballots through the Accu-Feed with the blank test, and all
was successful.  Then I marked these ballots and ran them again, and all
totals came out fine.  But when I ran the ballots a second, third, and
fourth time, the totals never came out the same again.  In Shelby, TN, we
sorted overvotes and kicked out probably around 20 ballots due to toner rub
off.  When the customer folds the ballot, it makes the toner rub off even
worse, especially when they have a lot of text.  Also, the toner tends to
rub off onto the feed rollers of the Accu-Feeder and then the rollers become
slick and cannot pick up the ballots.  Both Accu-Feeders in Douglas &
Lancaster, NE started out picking up the ballots fine, but after several
hundred ballots they began to slip and needed help in feeding.  Is this
happening in Pima or King, or do they use ballots printed by a printing
company?  It took approximately 8 hours to scan around 5500 ballots on 2
Accu-Feeders in Douglas, NE; is that good or bad?

Another note: Douglas, NE will be doing a recount on December 1, 1998, and
all the absentees need to be run again.  I suspect that it not only will
take about 8 hours again, but the count will change due to the toner rub
off.

Here are some suggestions both from the three accounts mentioned above, and
myself.

All 3 accounts only used paper ballots for their postal absentee, and each
emphasized that it would be much easier and cost effective to use 20 lb.
bond paper, (that you can buy from an office supply store), rather than the
heavy paper we supply.  I think that this would help considerably with the
toner rub off that I experienced.  Would it be possible to enhance the
Accu-Vote & Accu-Feeder to be able to do this?
This is the first time that I created postscript files from Gems and sent
them to RX Laser to convert for their BOD system.  I did not get a very
timely turn around for these files, and they were never able to convert the
Douglas, NE files.  Again, as mentioned earlier, using the RX Laser BOD
software did not help for printing to spec while duplexing, and when the
ballots were printed shading and fonts were a lot different than Gems.
Other than the fact that RX Laser can print the ballots faster, I see no
real benefit to doing all this work for RX Laser to convert the files and
send back a CD for you to install their software.  I think that with a few
enhancements, Gems could do exactly the same thing and we wouldn't have to
be waiting or paying for RX Laser to do the job.
Another issue is the processing speed it takes to run the absentee ballots.
Douglas County, NE especially commented on this, again as mentioned above,
it took 8 hours to process their absentee ballots

Jeff Hintz
Global Election Systems