As I'm thinking about +SpectraStudy #SensorMaster next feature instead of doing proper Saturday night activities, I got curious about how the spectrum of the displays right in front of me look

So there's one each for the Dell 2407 WFP, Galaxy S4, Nexus 5 and Nexus 10.

If the AMOLED and mobile LCDs spectrum are rather typical, the CCFL-backlight desktop monitor spectrum is unlike anything I've seen before.

Now I understand why no software nor sensor I used so far (based on the standard but old CIE 1931 2° observer model) has been capable of color-matching it with others, not even remotely.

This is great for me because that'll be the perfect benchmark for the physiologically relevance of the color matching function I'm integrating in my program.

#supercurioBlog #color #calibration #measurements #display #development


In Album Some white spectrum

Source post on Google+

Published by

François Simond

Mobile engineer & analyst specialized in, display, camera color calibration, audio tuning

16 thoughts on “As I'm thinking about +SpectraStudy #SensorMaster next feature instead of doing proper Saturday night activities, I got curious about how the spectrum of the displays right in front of me look”

  1. +Friedrich Sinofzik thanks. I know it could pay off as contracts with manufacturers to help them calibrate their displays using the skills acquired and software developed (I have a proposal already)
    The only problem is that then it will compromise my ability to review devices because of potential conflict of interest.

  2. I understand, because this my write about "crowdfund" or similar project..

    Maybe not work for this manufacturers but create your (or with someone more) own business for software or hardware or both..

    A simple hardware but effective and a good analytic software easily can get some kind of support ..

    Who knows? 😉
    I really wish you the best..

  3. A white LED is basically engineered to have the most power efficient white light spectrum, so its 3 peaks close the center of each primary wavelength. The spectrum you see for the CFL is pretty typical, the output is a bunch of spectral lines corresponding to electron transitions in the gas. They're basically randomly positioned, and the lamp is made white by mixing gasses or coatings to balance out the spikes.

  4. Is your software based on a spectrometer for read out, for a tricolor device like an Xrite profilers? If the latter, one way to commercialize it would be to sell software that worked with existing commercial hardware. FWIW I've found that most are ok for calibration but pretty bad for actually comparing devices or assessing calibration after the fact.

  5. Building a decent spectro is pretty easy; building a spectro that's price competitive with the closed-hardware companies is hard. I've got a prototype on my desk now, but it's just too expensive with too a little margin for manufacture, even in batches of 500. The price of the sensor is coming down really fast, and I'm hoping to announce something either late this year or early next year.

  6. +Richard Hughes I look forward to that very much, and will gladly help any way I can.
    So far, the mobile press and reviewers tend to not be equipped in display measurement equipment due to the cost first, then they don't know what to do with the data they're presented.
    As there's a lot of variety in the display technologies used in mobile than ever (type of backlight, blue led + phosphors, OLED, quantum dots) spectrophotometers are more useful than colorimeters.
    Later I hope to deliver calibration solutions for mobile devices (working prototype stage ATM), and affordable sensors will be key as well.

  7. +Richard Hughes Yes it's always gonna be a tough sale regardless of the price, so that's a decent price point.
    Do you think the performance will be comparable to a used i1pro? Particularly when reading dark samples, which is useful to measure the near black response of AMOLED and new LCDs with contrast ratio increasing beyond 1400:1 like I've seen recently.

  8. +Richard Hughes​ in my own software it would be no trouble at all to have the spectrum only for some readings.

    An instrument containing both sensors would be a winning combo: reading the spectrum to correct XYZ values according to the chosen color matching function and using only the colorimeter sensor for speed and sensitivity the rest of the time.

    What you mention makes me wonder if the spectrum shape changes depending on RGB values, either on LCDs and OLED.
    For the later, maybe the materials emit something a bit different lower intensities? I'm gonna test that!

    With this dual sensor instruments, in case the spectrum varies enough to make a difference, I wonder if an interpolation between several correction matrices depending on RGB levels could be used.

Leave a Reply