A better DIY "by-eye" grayscale test pattern?

Discussion in 'Archived Threads 2001-2004' started by Jim A. Banville, Jun 15, 2001.

  1. Jim A. Banville

    Jim A. Banville Supporting Actor

    Joined:
    Jun 20, 1999
    Messages:
    630
    Likes Received:
    0
    Trophy Points:
    0
    I was wondering if a black screen with four 80 IRE "windows", one with "perfectly" neutral color, one with too much red, one with to much green and one with too much blue would be good, if not for roughly adjusting the high end of your grayscale by eye, at least letting you know if your current grayscale is in the ballpark on it's bright end? I suppose you could do the same with the dark end, but inverting the backgound/foreground (light gray backgound and 4 black boxes, one of the "perfect"). Of course, the boxes would be marked somehow to let you know which has too much of the color and which is supposed to be "perfect". What do ya' think?
    ------------------
    Jim
    -------
    Mancow rocks! www.mancow.com
     
  2. Jeff D.

    Jeff D. Supporting Actor

    Joined:
    Jul 10, 1999
    Messages:
    521
    Likes Received:
    0
    Trophy Points:
    0
    I don't see how this would work. First, you must consider variations in colour temperature across the screen area, a common problem with most RPTVs. Oftentimes colour temperature variations of over 1000K can be seen from the centre of the screen to the sides. Michael Chen has stated that eyeballing grayscale should never be done using the 10-step grayscale pattern, for this very reason. The test pattern you are suggesting would suffer a similar fate.
    Also, I don't see how you would get anywhere. The way I understand your explanation, you would have three "inaccurate" boxes and one "accurate" one. But as you adjust your CUTs and DRVs you will not only be affecting the "accurate" box, but all the "inaccurate" boxes as well. Certainly if you need anything (other than proper test equipment) when trying to eyeball grayscale you need a constant reference. With your proposed test pattern, one could theoretically make any of the inaccurate boxes technically accurate. Trying to adjust the accurate box by using the inaccurate boxes as reference would be fruitless - as with each adjustment you keep changing your reference.
    I have never liked nor understood people's willingness to eyeball their grayscale. When I calibrate with the Progressive Labs CA-1, I watch the screen of the TV, not just the computer monitor. As you get close to dialing in the grayscale, you make very small adjustments in the end. Watching the screen during these small adjustments, one is hard pressed to see a large difference (in, say, one "click" of a CUT or DRV parameter). Yet just one "click" can have a dramatic affect on the grayscale - sometimes making changes of over 300K.
    I would like to see someone measure their eyeballed grayscale and see if it comes to what anyone would consider close. Personally, I don't believe there is any replacement for proper measurement. That is, if you want to get the absolute best out of your TV. Eyeballing your grayscale might get you to within 1000K of your target - but why settle for that when you can get within +/- 100K??
    ____
    Jeff
    ------------------
    "You're a walking stick! It's funny! Now GO!"
     
  3. Jim A. Banville

    Jim A. Banville Supporting Actor

    Joined:
    Jun 20, 1999
    Messages:
    630
    Likes Received:
    0
    Trophy Points:
    0
    >>>I don't see how this would work. First, you must consider variations in colour temperature across the screen area, a common problem with most RPTVs. >Also, I don't see how you would get anywhere. The way I understand your explanation, you would have three "inaccurate" boxes and one "accurate" one. But as you adjust your CUTs and DRVs you will not only be affecting the "accurate" box, but all the "inaccurate" boxes as well.
     
  4. Michael TLV

    Michael TLV THX Video Instructor/Calibrator

    Joined:
    Mar 16, 2000
    Messages:
    2,909
    Likes Received:
    1
    Trophy Points:
    0
    Location:
    Calgary, Alberta
    Real Name:
    Michael Chen
    Greetings
    I think you might have better success if you just used Kodak gray reference cards instead. At least you have something to shoot for. Aside for the light in the room further skewing how you interpret what you see.
    Regards
    ------------------
    [​IMG]
    Michael @ The Laser Video Experience
     
  5. Jeff D.

    Jeff D. Supporting Actor

    Joined:
    Jul 10, 1999
    Messages:
    521
    Likes Received:
    0
    Trophy Points:
    0
    In order to eyeball grayscale (which I still do not recommend, especially with the CA-1 available for rental at an economical price), you need a "constant" reference. An optical comparator is a constant reference. You need something constant - this method does not provide it.
     
  6. ThomasL

    ThomasL Supporting Actor

    Joined:
    Mar 13, 2001
    Messages:
    963
    Likes Received:
    0
    Trophy Points:
    0
    Michael, do you know what kind of shops/stores sell such cards? Would it be exclusively photography shops?
     
  7. Jim A. Banville

    Jim A. Banville Supporting Actor

    Joined:
    Jun 20, 1999
    Messages:
    630
    Likes Received:
    0
    Trophy Points:
    0
    >>>Red tint", "blue tint" and "green tint" are way too vague. How much of a tint? Should the box be red? Should be it light red? Grayish red? > The concept of a varying grayscale across the screen alone makes this method useless. A rotating set of squares would only further complicate things - all your references would keep changing. The point of a reference is stability. Should I adjust for proper grayscale with the neutral box in the top right or bottom left?
     
  8. Jeff D.

    Jeff D. Supporting Actor

    Joined:
    Jul 10, 1999
    Messages:
    521
    Likes Received:
    0
    Trophy Points:
    0
    It already is - the PLUGE pattern. The gray steps in this pattern will easily show whether you've got a good grayscale. I would argue that most of the grayscale errors you would see on the market today are so great that the PLUGE pattern will without a doubt highlight them. If they don't appear to be off - perhaps your grayscale is rather close.
    ____
    Jeff
    ------------------
    "You're a walking stick! It's funny! Now GO!"
     
  9. Jim A. Banville

    Jim A. Banville Supporting Actor

    Joined:
    Jun 20, 1999
    Messages:
    630
    Likes Received:
    0
    Trophy Points:
    0
    Not to beat a dying horse [​IMG] , but wouldn't the "slowly spinning wheel" setup pretty much solve the problem of screen location? I mean, you have 4 wedges (3 tinted, 1 neutral) and at the very center of the screen all 4 wedges are touching. You can press PAUSE to stop the wheel after you determine if you are getting too much color variation, but again, with the wedges touching at the center, this may not occur.
    >>>Herein lies the crux of the problem - while you are touting the advantage of 4 reference points, this would only serve to confuse matters. First, we've established these aren't references at all, since they are all changing. >I'm talking about something that could be included on soething like AVIA for free.
    ---------------------------------------------------------
    It already is - the PLUGE pattern. The gray steps in this pattern will easily show whether you've got a good grayscale.
     
  10. John_Pech

    John_Pech Extra

    Joined:
    Mar 7, 2001
    Messages:
    19
    Likes Received:
    0
    Trophy Points:
    0
    OK, call me stupid for suggesting this, for I have no clue, but what about the use of filters? Like the blue filter supplied with VE or Avia to set color and tint?
    Could you use different filters in conjunction with some sort of video pattern or patterns? Not that this would be a way of calibrating the set perfectly, but to get the set close to ideal grayscale.
    Not every set is worth the money to properly calibrate, i.e 13 to say 27", but it would be nice to get it close, at a realistic amount of $$mula$$. (Once spoiled with perfection, it seems everything you watch has to be perfect, or damn close!)
    Am I way off base here?
    ------------------
    I got two turntables and a microphone
     
  11. Jeff D.

    Jeff D. Supporting Actor

    Joined:
    Jul 10, 1999
    Messages:
    521
    Likes Received:
    0
    Trophy Points:
    0
    It will never look like one of the patterns - you will constantly be going around in circles. You're assuming that the inaccurate sections will look the same no matter what - but they are just as affected by the set's grayscale as the accurate one you are trying to calibrate.
    Colour filters would be of no use. Grayscale is about getting the proper mix of colours to achieve the proper shade of gray at 6500K. Removing any of these colours (via a filter) defeats the whole purpose - you need to see the mix of all of them.
    ____
    Jeff
    ------------------
    "You're a walking stick! It's funny! Now GO!"
     
  12. Jim A. Banville

    Jim A. Banville Supporting Actor

    Joined:
    Jun 20, 1999
    Messages:
    630
    Likes Received:
    0
    Trophy Points:
    0
    >>>wouldn't the "slowly spinning wheel" setup pretty much solve the problem of screen location?
    ------------------------------
    How is this going to help? Its bad enough that the colour temperature is going to be different all around the screen. Now you are talking about constantly moving the point you are trying to calibrate. > "Slight red tint" is vague. Everyone will have a different opinion of what a slight red tint is. We all perceive colour differently. "Slight red tint" is no accurate descriptor. What I think is a slight red tint, someone else won't. And there is no way to convince people of what they should look like.
     
  13. Jeff D.

    Jeff D. Supporting Actor

    Joined:
    Jul 10, 1999
    Messages:
    521
    Likes Received:
    0
    Trophy Points:
    0
     
  14. ThomasL

    ThomasL Supporting Actor

    Joined:
    Mar 13, 2001
    Messages:
    963
    Likes Received:
    0
    Trophy Points:
    0
    I think the PLUGE pattern supplied by Avia does a good job at showing you if your grayscale is out of whack but there is a range where things look generally gray but the "right" gray - that is impossible to say. Within this range, it is impossible to "eyeball" it without an outside reference. At least that is what I've learned from my personal experience. I don't think using a pattern on the tv itself as a reference will work mainly because as you tinker with the CUTs and DRVs, you'll change your reference. Any reference has to be independent of the thing you're changing. The human optical system has evolved to do an excellent job at distinguishing relative differences in colors/shades. i.e. the color we perceive is always relative to what other colors are around it and the light which is illuminating what we see. The story I like to tell is how when they were painting the inside of our new house, my wife saw the color she had picked out on a small part of a wall in relation to all the white around it. She was horrified because the more neutral off white cream color she had picked out looked very yellow. I told her wait till they're done painting and then it'll look like the color you had picked out. Sure enough, it did. It only looked blatantly yellow when our eyes were judging it against all the white around it.
    From all that I've read, for really proper eyeballing, one needs a D6500 reference such as the Kodak card and proper lighting to simulate the grayscale ramp from 10 IRE to 100 IRE. Given this, ones eyes would probably do a decent job of setting the color on the pattern on the tv screen to the same color as seen on the card. To get it as close to D6500 as possible, the only real way is to eliminate the human eyes from the equation and use a color analyzer.
    Jeff, who rents the CA-1? I have a 27 inch and at some point may be moving that to the fireplace room and upgrading to a 32 inch. When I have both, I'm seriously thinking it might be very worthwhile for me to rent this or have both sets calibrated by an ISF person.
    thanks,
    --tom
     
  15. Jim Kirkman

    Jim Kirkman Agent

    Joined:
    Feb 23, 2000
    Messages:
    31
    Likes Received:
    0
    Trophy Points:
    0
     
  16. Jeff D.

    Jeff D. Supporting Actor

    Joined:
    Jul 10, 1999
    Messages:
    521
    Likes Received:
    0
    Trophy Points:
    0
    This depends on many factors. How close is your set right now to 6500K? Will you like the results? None of this can be determined until you shell out for the calibration. As expected, there is no satisfaction guarantee - an ISF tech will get your set to spec, but there is no guarantee you will like 6500K (though I have never heard any complaints).
    If you want the absolute best from your system, however, and believe in getting as close to the technical standard as possible, proper calibration is a must.
    ____
    Jeff
    ------------------
    "You're a walking stick! It's funny! Now GO!"
     
  17. Jim A. Banville

    Jim A. Banville Supporting Actor

    Joined:
    Jun 20, 1999
    Messages:
    630
    Likes Received:
    0
    Trophy Points:
    0
    So you can't rent the CA-1 for $150?
    ------------------
    Jim
    -------
    Mancow rocks! www.mancow.com
     
  18. Michael TLV

    Michael TLV THX Video Instructor/Calibrator

    Joined:
    Mar 16, 2000
    Messages:
    2,909
    Likes Received:
    1
    Trophy Points:
    0
    Location:
    Calgary, Alberta
    Real Name:
    Michael Chen
    Greetings
    Just inquire at AV Science or TAW ... both still rent the unit. Ask them.
    Regards
    ------------------
    [​IMG]
    Michael @ The Laser Video Experience
     
  19. Jim A. Banville

    Jim A. Banville Supporting Actor

    Joined:
    Jun 20, 1999
    Messages:
    630
    Likes Received:
    0
    Trophy Points:
    0
    Is it easy to use? It is a PC based program, no? Is it just a matter of aligning some green, red and blue bar graphs? Is the only interaction with the PC via the screen probe? Do I have to move my PC to my livingroom or can I get a long extension cable for the probe? You ever hear of anyone adjusting the grayscale on a Sony KP-53HS10? It has 3 sets of menu items called CR and CB "offsets". I think they have something to do with the grayscale settings (they make sort of a "hue" change when you play with them).
    thanks!
    ------------------
    Jim
    -------
    Mancow rocks! www.mancow.com
     

Share This Page