Better ISF Calibration Report

Discussion in 'Displays' started by errol, Sep 13, 2002.

  1. errol

    errol Stunt Coordinator

    Joined:
    Dec 29, 1999
    Messages:
    150
    Likes Received:
    0
    Hi ya folks,
    New article from Michael TLV on a better way to present pre and post calibration results to the consumer.
    It should turn on a few light bulbs for all of us. [​IMG]
    Thanks,
    Errol
     
  2. Gregg Loewen

    Gregg Loewen Video Standards Instructor, THX Ltd.
    Insider

    Joined:
    Nov 9, 1999
    Messages:
    6,373
    Likes Received:
    32
    Location:
    New England
    Real Name:
    Gregg Loewen
    Errol,

    Thanks for sharing. Michael comes through again.

    I for one use the software that comes with the Sencore Color Analyzer. This software graphs the data points in a non linear function by placing each IRE reading within a CIE chart. This clearly shows before and after accuracy.

    Michael did a great job is showing people that 6500K is not really the correct way of describing an accurate gray scale reading. Generally, the higher the number the less red and more blue the image is. The lower the number the more red, less blue the image is. However green is not factored into to this. The amount of green does not effect the color temperature. So as with Michael's "furrypig" examples show, you can have a 6500 K image that is either has a green deficit or green gray scale push.

    When calibrating a display device a calibrator actually calibrates to the D 6500 K temperature. The "D" is a specific setting where the proper amount of green is in the image to produce a neutral gray.

    I hope my thoughts made sense.

    Michael, thanks a bunch! Your thoughts are always welcome here at HTF!!!!

    Regards

    Gregg
     
  3. Erik_C

    Erik_C Stunt Coordinator

    Joined:
    Jun 15, 2000
    Messages:
    186
    Likes Received:
    0
    That's exactly what was the case with my XBR450. Pre-calibration, it tracked D6500, but was way too green. Post calibration, it tracked D6500, but the green tint was gone. The graphs looked the same, but the tv looked much, much better.
    -Erik
     
  4. Gregg Loewen

    Gregg Loewen Video Standards Instructor, THX Ltd.
    Insider

    Joined:
    Nov 9, 1999
    Messages:
    6,373
    Likes Received:
    32
    Location:
    New England
    Real Name:
    Gregg Loewen
    Erik I think you made a typo (?). If your set precalibration tracked 6500K D (note the "D") your set was gray scale accurate. It may have been 6500K but not the "D" which indicates the correct amount of green.

    Your confusion is very common and that is why Michael is changing the way in which he reports to his clients.

    Regards

    Gregg
     
  5. Michael TLV

    Michael TLV THX Video Instructor/Calibrator

    Joined:
    Mar 16, 2000
    Messages:
    2,909
    Likes Received:
    1
    Location:
    Calgary, Alberta
    Real Name:
    Michael Chen
    Greetings Gregg

    The Sencore report uses the chroma triangle, right? I have seen an example or two of it ... where it shows the pre and the post stuff ...

    But only for the 20 and 100 ire. Can you get it to plot the information at the other spots like 30/40/50/60/70/80 ?

    Sort of like a shot gun blast scatter graph.

    I had also thought about showing a diagram like this ... so I may add something like that later.

    Regards
     
  6. Gregg Loewen

    Gregg Loewen Video Standards Instructor, THX Ltd.
    Insider

    Joined:
    Nov 9, 1999
    Messages:
    6,373
    Likes Received:
    32
    Location:
    New England
    Real Name:
    Gregg Loewen
    By default it plots only the 2 (20 and 100) points. If you select extended (or whatever it is called), it will allow you to plot all 20-100 points.
    Yeah, it looks alot like a scatter graph. It prints the before readings as o's and in red, the post readings are +'s and in black. And of course all the +'s are in the middle box which denotes D 6500K. [​IMG]
    Gregg
     
  7. Michael TLV

    Michael TLV THX Video Instructor/Calibrator

    Joined:
    Mar 16, 2000
    Messages:
    2,909
    Likes Received:
    1
    Location:
    Calgary, Alberta
    Real Name:
    Michael Chen
    Greetings
    Strangely, the scatter graph is preferable to the standard linear graph, but you often have to look at it really closely to understand what it is trying to say ...
    In the sense that the eyes do not really know what they should be focusing on so they tend to wander around the graph ...
    I guess it is also graphically boring too. [​IMG]
    Regards
     
  8. JohnnyG

    JohnnyG Screenwriter

    Joined:
    Dec 18, 2000
    Messages:
    1,522
    Likes Received:
    0
    Here's a sample of the Sencore report. I do like the format but I wish it had the option of overlaying a colour image of the chromaticity graph over the data points. Even if the image was heavily dithered, I think it would really help drive the point home.
    I certainly do like your approach Michael and I'm eager to plug some of my cal results into your spreadsheet to see how it looks!
     
  9. Mark_Hunter

    Mark_Hunter Auditioning

    Joined:
    Jul 31, 2002
    Messages:
    4
    Likes Received:
    0
    This is excellent feedback, and I am very interested in this topic.
    What type of information would be useful on such a report? Today our product, ColorFacts presents the grayscale readings directly on a CIE chart, as well as the 2D Plot that Michael is improving with his technique.
    In my personal opinion, the most useful graph for displaying improvement in the grayscale is a Red/Green/Blue error tracking chart. This displays how far the grayscale is from D65 (or any other chosen target) from black to white. It does take 2 of the charts to show changes from 'before' to 'after', however. It is still quite dramatic to see the differences, and this is an unbelievably intuitive way to show someone how you have improved their set.
    We would be happy to hear any other suggestions you would like to provide us to help improve our product, and we would be honored to work with people who are as passionate about this industry as you are.
    Thank You,
    Mark Hunter
    markh at milori dot com (re-join to contact me)
    Milori, Inc.
    www.colorfacts.com
     
  10. JohnnyG

    JohnnyG Screenwriter

    Joined:
    Dec 18, 2000
    Messages:
    1,522
    Likes Received:
    0
    Mark, I haven't heard about your product before. It looks very interesting and I'm just now dowloading the software to take a look. How long has it been on the market?
     
  11. JohnnyG

    JohnnyG Screenwriter

    Joined:
    Dec 18, 2000
    Messages:
    1,522
    Likes Received:
    0
    .
     
  12. errol

    errol Stunt Coordinator

    Joined:
    Dec 29, 1999
    Messages:
    150
    Likes Received:
    0
    Hi Mark, interesting indeed.

    r,
    Errol

    P.S., Gregg if you can get a good sample of the scatter graph, pls feel free to forward to me so I can post it.
     
  13. Mark_Hunter

    Mark_Hunter Auditioning

    Joined:
    Jul 31, 2002
    Messages:
    4
    Likes Received:
    0
    Hi Johnny G,

    Sorry about the slow response.

    ColorFacts is pretty new. It's only been on the market since fall, 2001.

    Keeping the thread on topic: We have listened to Michael and Gregg (and others) and designed a flexible, extensible reporting Engine into ColorFacts 3.0. It has the D65 "3D" graphs as well as the more traditional charts. The reports can be printed, or exported to PDF files. I think that you (Michael and Gregg) will approve.

    Mark
     
  14. JohnnyG

    JohnnyG Screenwriter

    Joined:
    Dec 18, 2000
    Messages:
    1,522
    Likes Received:
    0
    Slow?! I'm amazed you found the thread again!

    I've talked to Michael and he has plenty of nice things to say about it. I'm pretty much sold, but I have to wait until next June for my Sencore lease to expire!
     

Share This Page