Very Long - How to Tweak the PrintFix Scanner - (Followup to another thread)

Discussion in 'Digital Photography' started by BobS, Jan 26, 2005.

  1. BobS

    BobS Guest

    There was another thread "Help profiling monitor with Spyder2, problem in
    shadows" started on the 24th that went a bit astray after I mentioned the
    PrintFix scanner problem. Another asked about how I tweaked the scanner so
    it doesn't skew. Below are a couple of emails that I sent to ColorVison
    after receiving my scanner and what I did, with their permission to try and
    find a solution to the problem. It may or may not work on your scanner.

    Disclaimer. You do this at your own risk obviously and I'm not advising
    anyone to do this tweaking. By disassembling the scanner you will most
    likely void any warranty since you must break the seal covering one of the
    four screw holes. If your unit is still under warranty (two years), I would
    first suggest you contact Mike at ColorVision Tech Support and let him know
    you would like a replacement scanner or at least get their okay to do it so
    you don't void the warranty. He just may say okay.

    This scanner skewing is a well known problem as I've found out and I didn't
    get any hassles when I emailed them about my first scanner constantly
    skewing and jamming. I then made a suggestion for a possible fix and they
    said okay and try it. I then emailed them what I found after tweaking the
    first scanner (see below). A new scanner promptly arrived but it also has a
    skewing problem but I've done nothing to this one and it will be returned.
    I'll keep the one I tweaked so I can build some profiles and see if the
    software works.

    So with some slight modifications, the emails are posted below:

    .....................1st email describes the problem.....................

    Brand new Spyder2Pro Studio with PrintFix suite arrived this week.
    monitors just fine - looking good.

    Next I made 5 calibration prints for (Kodak, Ilford, GP, Epson,
    RoyalBrights-Matte)and I let them set overnight before attempting to build
    any printer profiles.

    Had a couple of problems that are making a believer out of me that this
    PrintFix reader is not ready for prime-time -or- I have a bad unit. I have
    uninstalled the software and reloaded it and still get the same problem.

    I can enter the B&W and CMYK calibration prints and the reader works fine.
    When I take any one of my 5 calibration prints and insert them into the
    clear plastic sleeves, I have about a 1 in 10 chance of the print making it
    through the reader before it stops, jams or starts skewing. I have cleaned
    the unit several times (remember it's brand new) using the cleaner sheet and
    isopropyl and get the same results. When I cancel the read, I have to
    gently pull the print so it will exit. It feels as though the left side
    roller wheel is looser than the right side.

    The calibration prints I made were cut from 8x10 sheets on the dotted lines
    and everything was done as per the manual for Windows. Finally after many
    attempts, I have yet to get a calibration print through (save once) that was
    not skewed in some manner or slipped, or just plain stopped (with the reader
    still trying to move the print). I have ensured the print is the same size
    as the plastic sleeve, cut squarely on the lines by using a paper cutter and
    have positioned the print within the sleeve as far as it will go, to leaving
    it back from the leading edge by about 1/2". Nothing seems to work and it
    may be that this unit is set to tight as to not allow the print and sleeve
    to go through smoothly. Print was aligned with right side as per
    instructions and there does not appear to be anything wrong with the plastic
    sleeves. Are these new sleeves you're using - maybe they're to thick?

    Finally out of desperation, I put the prints through without using the
    plastic sleeves. Finally got them to go thru within one or two attempts
    without skewing or stopping or jamming.

    I physically can't see anything wrong but as I said, I noticed that the left
    side appears not to have the same amount of pressure on the print as the
    right side - causing it to skew or jam when using the sleeve.

    Before I say to hell with this entirely, how about someone checking out a
    new reader unit there, burn it in by testing it thoroughly and then sending
    it out to me.

    ..................2nd email suggesting a potential fix...................

    Just received the shipping notification of the replacement scanner you're
    sending. Now I'm sure ColorVision has probably tried a lot of things to get
    this problem to go away but I have an idea I'd like to try if you okay it.

    I've been an electronic tech and computer maintenance tech for more years
    than I care to remember. One of the things I used to maintain were some IBM
    card readers/punches which used a double pressure roller system similar (I
    think) to what is in this scanner. The rollers may be made from a different
    material but I think the fix we used back then to prevent cards from skewing
    may also work on this. Card alignment was critical or the old punch cards
    would not be read correctly and we were constantly replacing rollers and
    making adjustments.

    Finally, one of our techs came up with a fix that involved using a fine 320
    grit wet/dry paper. He cut it to the same size as the cards and
    then glued it to a punch card. He then let this card start thru the rollers
    after it started going thru, he would pull on the card slightly to make the
    rubber rollers abrade against the sandpaper. This tended to even out the
    rollers as well as remove any glazing. It solved our problem 99% of the time
    and became a standard maintenance practice for IBM techs.

    This trick may work on the scanner rollers as well. It wouldn't do any
    damage when the grit was against the top roller but without taking the
    scanner apart, I can't see how much clearance there is between the bottom
    roller and the LEDs. Obviously the sandpaper would really mess up the LEDs
    if it contacted them but if the LED strip can be easily removed or lowered
    and then run the sandpaper against the bottom roller to even it out, it may
    just work. I doubt there is any adjustment capability in this type scanner
    but I do suspect that they do have some kind of tension mechanism (leaf or
    spring) that puts pressure on the top roller.

    (Note added - there is no bottom roller. The top 4-1/2" long roller is held
    against a flat
    piece of glass that acts as the bed and tension is applied by two small
    coiled springs - one
    on each side that holds the bed up to the fixed roller.)

    If you're just going to scrap this unit anyway when you get it back, either
    let me give it a try or have someone there try it.

    ..............3rd email after getting

    Progress but not a 100% cure. When I send this unit back, I will send the
    sandpaper (320 grit) card I made up. Used a 4x6 photo blank and sprayed the
    backing with some 3M glue.

    I removed the cover and did a close inspection of the mechanicals and could
    find nothing wrong such as burrs or other deformities that could be causing
    the problem. I then measured the distance from the glass to the top of the
    rubber roller and found a variance of .007" from one end to the other.
    Rotating the roller caused that reading to vary indicating that the rubber
    roller is not as round as it should be.

    I cleaned the roller and glass and then proceeded to run the sandpaper thru
    using the Read function. The card I made has 320grit on one side and a
    glossy photo surface on the other side so as to not to cause any scratching
    or marring of the glass surface below the roller. As the sandpaper was
    being pulled thru, I would hold it back so the sandpaper would abrade the

    My USB ports certainly have enough current capability and that's one little
    tough motor in that scanner. I would let the roller make two or three
    revolutions then let the paper pull thru about a 1/2", then hold it again to
    abrade the roller. Did this about a dozen times and I could start to see an
    even wear pattern across the roller and even scuffing on the sandpaper -
    indicating the roller is now wearing more evenly. Cleaned everything and
    started some testing.

    Prior to this, you'll recall that I could not get a print thru the scanner
    that was in the plastic sleeve without it skewing, then jamming. After
    truing up the roller, I did not have a single jam - not a single one and I
    did at least 100 passes using all 5 of my calibration prints in the plastic

    All is not cured though as occasionally, the print would skew slightly and
    always to the left - meaning the left side of the scan was always lower than
    the right side of the print as viewed on the screen. There is no adjustment
    mechanism and the pressure roller is dependent on the two small springs for
    keeping an even pressure. I would seriously doubt that these are calibrated
    springs so the actual tension they apply to each side of the roller will
    vary slightly. Enough to cause the print to skew. I gently removed each
    spring and then installed it on the opposite side (right spring went to the
    left side, left spring went to the right.

    Restarted my testing and noted that the print was not skewing as much now
    that the springs were swapped. Getting close here so I again ran the
    sandpaper thru a number of iterations (as noted above) and then cleaned
    everything again.

    More testing and this is now almost perfect in comparison to what it was. I
    made up a test print with some vertical and horizontal lines so I could
    measure the skewing more easily. Again I probably ran about 100 passes and
    used the rectangular marquee tool to measure how square - or how close to
    square the print was.

    Given, that you cannot 100% of the time place a print perfectly square into
    the scanner so I needed to make a measurement from a reference point that
    would tell me if the print started squarely. Simple - measure from a center,
    vertical line. If the print skews at any point, that vertical line would no
    longer be vertical - even if off by one pixel. I ran "many" tests using my
    test print and noted a little skewing (less than .5 degrees) which is
    probably more than sufficient accuracy but then I noticed a strange skewing
    effect further on down the print.

    Even though a print would start out square (as measured on the top and down
    one side using the marquee tool) about half way down the print, the
    horizontal lines would start being skewed - about the thickness of a 3 point
    line. Found this to be a constant across all prints. It was not my test
    print, I could use your B&W pr CMYK prints and then measure the horizontal
    edges and would see that one side of an edge would always be slightly lower
    (skewed) than the other side even though the vertical lines (on my test
    prints) would be perfectly vertical.

    After thinking about this for awhile, I think it must be buffer latency
    causing this "electronic skewing". The data is read serially across the
    scanner and what I'm seeing is the latency of the data transmission I
    believe. I would think that this could be taken care of in the software.
    Again, it's consistent and although barely noticeable unless you're looking
    at things very closely as I've been doing today. But couple that electronic
    skewing with any other mechanical skewing and you end up with significant

    One other factor that may be contributing to the skewing is the default
    settings being used on the Read function. The default of 400dpi is fine but
    the software initially starts the scanner at some lower resolution (faster
    speed) for about the first 1/2" or until it reads something of a particular
    color value, then it slows to the 400dpi speed. That initial burst of speed
    may not seem like much but looking at the forces involved for the size of
    the objects we're dealing with and that is about the equivalent of a car
    peeling out from a red light - quite a jerk start for a small scanner. I
    started using "User Settings" and manually adjusting the window so the
    scanner would start reading immediately and made the size 4.3" (max width
    you can set x 5") and then running prints thru with and without the plastic
    sleeve. Results were consistent in that by not allowing the scanner to do
    the fast initial feeding, then slowing, the prints had less tendency to skew
    at the start. I got significantly more good (square) starts using this
    method than by using the defaults.

    So to sum it up. From my point of view there are three problems:

    1. Pressure roller not perfectly round.
    2. Spring tension not even.
    3. Software needs some tweaks to slow down the start and to optimize the
    data by buffering it. Perhaps adding some markers to the calibration prints
    would help so the software knows how far skewed it is and then determine to
    allow or fail the calibration read.

    While I was writing this, the other scanner arrived. After I check it out,
    I'll send one of them back. Then I can get onto actually making profiles
    for the i9900 and seeing how good this software really is......

    ....................end of emails................................

    There were several emails in-between these but the ones above describe the
    process of tweaking the scanner. You'll note that you do not need to take
    the scanner apart to use the wet/dry 320 grit paper hence leaving the
    warranty intact. Be absolutely sure the grit of the sandpaper is facing
    up - never down or it will ruin the glass surface that protects the LEDs
    below. Also be sure you only use a good quality wet/dry paper so the grit
    does not come off. You can find this paper at automotive stores and it's
    about $1 per 8x10 sheet.

    You'll find that the rubber roller is very tough and it will take several
    tries to notice a difference. Be sure to clean everything before running
    your cal prints thru. It's a shame they went with using this type of
    scanner. There are small (5x7) and far better flat bed scanners on the
    market that could be used. Perhaps we can convince them to modify their
    software so it will work with other scanners - like the Monaco (from what
    I've read).

    Anyhow, I hope this has enlightened anyone that was interested and not
    confused. Guess I could take some pics of my scanner (cover removed) if
    needed and post them in the abpo newsgroup if you need further
    clarification. As it is, I'm sure someone will start complaining about how
    long this post is...........

    Be good,

    Bob S.
    BobS, Jan 26, 2005
    1. Advertisements

  2. BobS

    bmoag Guest

    I was going to get one of these. Thnaks for the warning.
    bmoag, Jan 26, 2005
    1. Advertisements

  3. BobS

    Biggles Guest

    Many thanks for taking the time to post this information.
    I'll give it a try.

    Dave Stewart
    Biggles, Jan 26, 2005
  4. BobS

    BobS Guest


    Just to insure there is no misunderstanding, I'm not telling anyone to stay
    away from this product because of a cheapo scanner problem that can be
    easily tweaked. Quite the contrary. Since having this problem, I've been
    looking for reviews of other products such as the Monaco and G-M's and this
    site did some comparison testing of the various
    systems you should read. I think you'll find they all have some deficiencies
    and that PrintFix is as good as any - save the scanner skewing problem.

    I have yet to make some "real" calibration prints and start using the
    software to make profiles and tweak those. The ones I did make are suspect
    because of the slewing problems. So now that I have that little problem
    partially solved, I'll go ahead and make some new calibration charts and
    then evaluate the software. I know the profiles will not be perfect
    renditions of what I see on the screen because of the gamut differences but
    should be close so they can then be tweaked manually and then saved for each
    type of paper.

    At this price range, I doubt anyone has a perfect solution and the systems
    that probably are near perfect cost big bucks and have a big learning curve
    from what I've been reading. If you go with the Monaco, you'll need your
    own scanner and I haven't checked what kind of specs it needs to meet but
    that's an added cost if you don't already have a scanner.

    I may in-fact return this whole package for a refund but it won't be because
    of the scanner alone. If the software doesn't work well enough or is buggy,
    then I'll move on to probably the Monaco Pro and give it a whirl.

    The Spyder2Pro (the new model) does a very decent job of calibrating
    monitors based on my own experiences with it here and what others are saying
    about it. Granted, you have to look hard to find some reviews and they're
    few and far between but if you can believe what others are saying -
    ColorVisions Spyder2Pro and PrintFix suite are not perfect but have been
    given some points over the others in it's class.

    You'll note that I have not slammed this product or otherwise tried to
    dissuade anyone from purchasing it. Just making others aware of what I did
    to try to fix a scanner problem. If you get it, and the scanner doesn't
    work, call or email them and I'm sure they'll send you another scanner right
    away as they did me - no hassle. The Tech Support guy's are helpful but
    they didn't make the scanner and have to live with it too. ColorVision
    should be addressing this problem and getting it solved - they've known
    about it for a long time now.

    Wish someone that has the Monaco Pro would jump on in here and give their
    "real-world" opinion of that product so we could learn more about it. Maybe
    what I need to do is just order one and try it out - providing they have a
    return policy that is.

    Bob S.
    BobS, Jan 26, 2005
  5. I am telling you to stay away from this product. Even with a
    non skewing scanner there are huge problems. First - dust!
    Second - newton rings. Third - strange patch colors - only very
    light colors and no greys. This thingie does not work IMHO.
    I did not find any reviews at this site. But if I am understanding you
    correctly you are saying that Monaco and G-M are just as bad as PrintFIX.
    That I don't think is a good recommendation for PrintFIX.

    Roland Karlsson, Jan 26, 2005
  6. BobS

    BobS Guest


    I stand corrected, the site did not do an evaluation
    of PrintFix, they did the Spyder2Pro which was part of the suite I
    purchased. Thanks for noting my error.

    No I didn't say that Monaco and G-M are just as bad - no where near those
    words. I did say, from what I've read, for the price range, they all appear
    to have problems. Since I only have the PrintFix I cannot do a comparison
    against the other two.

    On the other hand, you apparently do have some experience with the PrintFix
    profile software. Is this with the latest version or their first version?
    Can you elaborate on the huge problems so I can verify them when it test
    this for myself? Appreciate your input. Any comments on the DoctorPro
    plug-in that is used to tweak the profiles?


    Bob S.
    BobS, Jan 26, 2005
  7. Yepp - I bought it - and I found so many problems that
    I returned it and got my money back. The DoctorPro plug-in
    I never used as I never got any profile to doctor.

    The problems were:
    1. The scanner could not feed the sleeves
    2. The sleeves are extreme dust magnets
    3. The sleeves were badly scratched
    4. There was strong newton patterns, i.e
    interference pattern caused by tight
    glossy surfaces.
    5. The scanner produced a low quality and
    very grainy result with heavy stripes.
    6. I got very different callibrations when using
    the callibration patches.

    Those problems above made it impossible to get any
    useful scans. So - I gave up.

    There are also other problems that are of a more
    theoretic nature that I could not really test:
    1. The printed patch is very light and contains only
    rather clear colors. I cannot see how those patches
    could callibrate dark and low saturated parts, e.g.
    2. The sleeve attenuates the light rather strong. This
    have to be a strong disadvantage. It would be better
    to scan without any sleeve.

    There was also a last problem that irritated me strongly.
    I have a Canon printer, but the manual did only tell how
    to set up Epson printers. The interface is very different
    for Canon printers and it was impossible to know how to
    do it. So - I contacted Color Vision support. They were very
    helpful and gave me totally incorrect information. I am not
    sure that they even have seen a Canon printer - at least
    not the one I was talking to. After five or six mails I gave
    up. Then, in the PrintFIX 1.2 software, they eventually added
    a correct manual for Canon printers.

    And then - of course - you know the rest if the story. Even
    with correct information - it did not work.

    Roland Karlsson, Jan 27, 2005
  8. BobS

    BobS Guest


    It's a mute point now since I just sent the whole PrintFix suite back but
    for those that will read the archives, some clarifications.

    Your version was earlier than the one I just received since the manual did
    have the Canon printer setup and it also had the new Spyder2Pro puck and
    software that was released fairly recently.

    I did some more research and had several email exchanges with someone that
    has reviewed the PrintFix software, DoctorPro plug-in and has compared them
    to other products. Since I don't have his permission to quote him, I'll
    just say I followed his advice and sent it back.

    The problems you noted below and the ones I commented on are problems that
    Pantone / ColorVison should be addressing but seem to be simply ignoring.
    This product came out in May of 2003 and it still has the same problems that
    were noted right after it came on the market.

    It doesn't appear that there is another product available in this price
    range ($400 USD) that does both monitor calibration and printer profiles -
    reliably. One would tend to think that one of the three manufacturers
    (ColorVison, Monaco, G-M) would be able to get it right and be able to
    market it in this price range.

    Thanks for your comments. Oh yeah... what did you end up with or are you
    still looking?

    Bob S.
    BobS, Jan 27, 2005
  9. Hmmm .. my Canon seems to be rather good callibrated from factory
    and it seems to be rather independent with respect to paper. So -
    I live with it as it is.

    Roland Karlsson, Jan 27, 2005
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.