Page 1 of 2 12 LastLast
Results 1 to 20 of 22

Thread: Any thoughts on LCD screen calibration and gamut coverage against colour temperature

  1. #1
    ajohnw's Avatar
    Join Date
    Aug 2012
    Location
    S, B'ham UK
    Posts
    3,274
    Real Name
    John

    Any thoughts on LCD screen calibration and gamut coverage against colour temperature

    As I start I recently bought an Viewsonic IPS VP2365-LED display. They do 3 which they call professional. Only one of the 27in ones has higher resolution. A bit pricey so for dpi I bought the 23in.

    I've carried out several calibrations eventually getting a respectable dynamic range, about 93% sRGB gamut coverage and more than sRGB in colour volume terms at 6500K. Looking at gamut plots I noticed that the coverage would be increased if the plot was moved down the daylight curve to warmer colour temperatures.

    I've tried recalibrating at 6000K,5500K and 5000K and the coverage does increase. There isn't much change at 6000k or between 5500K and 5000K either but at 5000K I can get a sRGB coverage of 98% plus. 5500K is just a little short of that.

    All with a gamma of 2.2.

    Then I thought to look at the tutorial on this site and noticed that they mention calibrating LCD'd to their native colour temperature, to avoid loosing out on gamut coverage I assume. The gamut that can be covered is related to the back lighting so going on the above it seems to have a colour temperature of around 5,000K looking at it simple terms. That nearly fits in with the type of LED's that are probably used for the back lighting. Visually as the tutorial suggests I can't see any differences between any of the colour temperatures as my eyes adapt as they should. Some web pages still show what I would call a painful white even at the usual monitor brightness levels. That may be down to 3,800K room lighting.

    What I can't make my mind up about is the importance of colour temperature as against gamut coverage. Also that some calibration procedure suggest that the colour temperature of the monitor should match the room lighting. 5000K is attractive in that sense as D50 lighting is readily available. For me that is just a case of changing a strip light tube. ensuring it's a decent make as the quality varies. This is the current 5000K gamut plot. I also wonder about the excess. There always seems to be some whatever the temperature.


    Any thoughts on LCD screen calibration and gamut coverage against colour temperature

    -

  2. #2
    herbert's Avatar
    Join Date
    Oct 2010
    Location
    Sussex, UK
    Posts
    471
    Real Name
    Alex

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Hi John,

    Interesting post. I imagine that the native white point of the monitor is when all three colour channels are being filtered/blocked the same amount. White is no filtering and black is full filtering. If you change the temperature then white would have at least some filtering of one or two channels to mix them in different proportions.

    Do you know where the gamut expanded with different temperature settings? Is it at the extremes of primary/secondary colours. If so then it may not matter since these are less likely to appear in your images and you are optimising for nothing.

    One thing to consider is that although the monitor will be able to cover a larger ranges of colours, any editing to make the image look pleasing will be valid on your monitor but will not translate to others that are using a standard D6500 reference. However it may be useful if you are preparing images for print since you will already be adjusted for the warmer tones of D5000.

    Alex

  3. #3

    Join Date
    Nov 2009
    Location
    Provence, France
    Posts
    910
    Real Name
    Remco

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    So the monitor can show some colours outside the gamut. Will those colours ever be used in practice, when using the sRGB working space for editing (as often advised for those that don't do their own printing)?
    And isn't the job of the profile to adjust for the difference between what the working space says and what the monitor can do (including colour temperature difference)?

  4. #4
    ajohnw's Avatar
    Join Date
    Aug 2012
    Location
    S, B'ham UK
    Posts
    3,274
    Real Name
    John

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    I think there are some miss conceptions here. Print matching for instance at D5000. The only reason that is needed is when a print is compared with an image on the screen. The only reason we might see a difference is because we are comparing one colour temperature with another. Basically it's a complicated subject but the human eye adapts to colour temperature over a huge range and has no way of knowing what the colour temperature of a view is. Have 2 views with different colour temperatures such as 2 prints or a print and the screen and we can see if they are different. What's not clear is if some one calibrates to 5000K and produces an electronic view for some one else is if it will look any different. Taking my calibrations because my eye adapts I can't see any difference in shots I calibrated on a different monitor at 5000K,5500K,6000K or 6500K. Also some web pages are painfully white at all of these temperatures. As an example of eye adaptation I can post and example. Any one who has taken photo's through a microscope will be aware of it. This particular microscope has a 30w tungsten bulb. Turned up fully it would be eye searing so it's run at 1/3 to 1/2 power for visual use. The colour temperature is extremely low well under 2000K. The bulb filament even looks orangish. The view by eye as the image is projected into the eye is the specimen in a neutral grey background. A camera shows this even set for tungsten white balance as it's too low a colour temp for it too compensate.

    Any thoughts on LCD screen calibration and gamut coverage against colour temperature

    Actually it's a bit difficult to get a good white balance on subjects like this via a microscope. It's a diatom. Using a dslr mounted on the microscope it's possible to see that the colour balance is incorrect because room lighting also gets in the eye. Fit an eyecup and the orange goes to neutral grey just as it does via eyepieces.

    This is the gamut plot for sRGB 6500K with a gamma of 2.2 done without entering a value for ambient lighting. That can make a huge difference to the gamma that sRGB calibrations give. It's difficult to know what value to set if lighting varies according to where it's measured.

    Any thoughts on LCD screen calibration and gamut coverage against colour temperature

    The coloured curve is the gamut the monitor is giving. The dashed line is the actual sRGB gamut. The cross is the 6500K daylight point and the curve represents variations in white colour temperature. This gamut covers 93% sRGB. The additional colour coverage direction and the missing colour direction can be seen by the colour of the plot. I can choose the usual triangular plots but these tend to minimise the excess. The excesses are down to the calibration process testing a huge combination of modifiers to give as much of the sRGB profile as possible and at the correct gamma. There is then some possible values which may crop up that will then give colours out of the gamut. This seems to be general in respect to LCD monitors - see this for instance a review on a similar HP monitor http://www.tftcentral.co.uk/reviews/hp_zr2440w.htm. Triangular plots so the difference looks small. The ability of the HP to extend into the blue end at 6500K must be down to the back light. I suspect all led types will fall short.

    Room lighting gets complicated as well. A normal tungsten bulb at it's usual 4,800K temperature has a cri of 100%. The 100 is typical of hot filaments what ever the temperature. Other sources tend to reach 98 again at different sources. Some spikey sources such as power leds can be very low.

    Looking at my shots I can't really see any difference what ever colour temperature I calibrate too. Gamma though can make a huge difference. I can see a distinct difference between 2.2 and 2.4. The sort of difference that might be obtained by entering office lighting levels. There is also earlier calibration methods involved in calibrating to what ever the ambient lighting is. On top of that some sources point out that sRGB was cooked up for crt monitors and that lcd's have different r,g and b's. As HP and Microsoft cooked it up I suspect and over riding aspect would be that it mustn't be the same as Apple. That may answer my question. Some may have viewed Apple calibrated shots (5,500K??) on a sRGB monitor. I don't think they would see any difference.

    -

  5. #5
    ajohnw's Avatar
    Join Date
    Aug 2012
    Location
    S, B'ham UK
    Posts
    3,274
    Real Name
    John

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Well there doesn't seem to be many thoughts in this area at all. I seem to have solved my problems as follows.

    I profiled the monitor in it's native sRGB mode, It turned out that the colour temperature was around 6300K and brightness about 88 cd/m^2. I then calibrated it in it's native sRGB mode using it's native colour temperature. Dynamic range was around 330:1 which is a bit low really but it passed the test bars on here and some more difficult ones elsewhere.

    Having established that the native temperature is around 6300K I ran a number of profiles at various brightness and contrast levels in native 6500K mode. Best coverage was with contrast at 50% over a wide range of brightness values.

    I then switched to user colour and calibrated for 6300K and 100 cd/m^2 using a curves plus matrix profile. This gave me 94.5% sRGB coverage and and an attained colour temperature of 6420K presumably because I had increased the brightness over native sRGB. For the 1st time the gamut coverage was nigh on identical in perceptual and absolute. The excess coverage was also minimal at 3%.

    This left me with a red channel gamma at low outputs that might cause colour casts at the dark end so I calibrated again using an XYZ LUT plus matrix. (Look Up Table). This corrected the red gamma curve and error between all channels is now low. Dynamic range is a more reasonable 560:1. That could be higher if I had used a brighter white point. but to far from native sRGB would probably cause gamut problems. I chose 100 cd/m^2 based on my room illumination. I bought a cheap used lux meter of ebay. That indicates that I have around 125 lux lighting on me when the light is on and much lower on the screen itself, about 25 Lux. I reasoned that this should make the screen look less bright. Seems to have worked out. I'm forgetting the comments about varying gama according to ambient light because from what I can ascertain photographers set it to 2.2 and iso standards basically state that the ambient light level should be less than the screen. Under 32 lux according to some sources. It is low if I turn the light out. The standards suggest D5000 lighting. That should be achievable with fluorescent multi phosphor tubes with a colour number of 950. The cri of these should be 98 but sellers seem to be miss describing tri phosphor tubes as 950 types, dropping the 0 etc and the cri's of these go down to 80 which personally I don't think would be worth using. There are some biaxial tubes about stated as being suitable for photo booths but suitable fittings seem to be none existent. It's a question of how far people want to go.

    The monitor I am using is a Viewsonic VP2365 LED which brings up another point. It seems some calibrators wont work well on LED back light panels. Some recent colorimeters have LED panels as a separate option over ordinary LCD. I seem to have succeeded with a MonacoOPTRIX gauging by the following shot. My previous monitor was an ordinary TN type with a relatively low dynamic range. The main difference I see on this monitor is that the window on the end of the church in deep shadow is slightly clearer and the shading in the clouds is more realistic. It was a little on the darker blodge side on the other monitor. Contrast stretching on the stone work is about as intended and I may have used a touch less sharpening on this monitor but probably not. Comments on these aspects would be appreciated.

    Any thoughts on LCD screen calibration and gamut coverage against colour temperature

    -

  6. #6
    herbert's Avatar
    Join Date
    Oct 2010
    Location
    Sussex, UK
    Posts
    471
    Real Name
    Alex

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Hi John,

    You have done a lot of work. I hope that your eyes are now satisfied.

    When I profiled my monitor I was just interested in seeing low error scores for the target colours. Above 3 Delta E is noticeable, less than 1 Delta E is excellent. It was low enough that I was happy. I've not tried different gammas or colour temperatures since the defaults produced results for actual verses screen colours that my eyes would not distinguish. Additionally even if I could get optimal results at gamma 3.0 and 4800K I'd probably not like it because I am used to gamma 2.2 and 6500K.

    Note that I did get a lower DeltaE using a LUT profile. However the profile could not be used by Firefox. So I changed to a matrix profile. The DeltaE is still very low and so I prefer it. I like to have colour managed web images since that is where I look at pictures most of the time.

    I've not worried about the CRI of my ambient light. Do you use a screen shield to cut down light falling on the monitor? I tried one using cardboard but saw no real difference. Perhaps my ambient was very low anyway as per your 25 lux reading. Also if you really want to minimise ambient effects then you can paint the walls around your workstation a neutral grey. These should prevent the ambient from affecting your judgement of the monitor output.

    Alex

  7. #7

    Join Date
    Nov 2009
    Location
    Provence, France
    Posts
    910
    Real Name
    Remco

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Quote Originally Posted by herbert View Post
    (...)
    Note that I did get a lower DeltaE using a LUT profile. However the profile could not be used by Firefox. So I changed to a matrix profile. The DeltaE is still very low and so I prefer it. I like to have colour managed web images since that is where I look at pictures most of the time.
    (...)
    Perhaps not too surprising: afaik, a LUT profile works directly in the graphics card, and thus works on the whole screen at OS level, and not per application, A matrix profile can work on a per application basis. That also means that your LUT profile would still adjust Firefox windows, and any others.

    Remco

  8. #8
    ajohnw's Avatar
    Join Date
    Aug 2012
    Location
    S, B'ham UK
    Posts
    3,274
    Real Name
    John

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Hi Alex - The work wasn't too bad. If something on telly, food or going out or to bed I just set up a profile or calibration. I did sit down for an hour and do a number of just quick profiles to get some idea what contrast and brightness did. My problem really was having no idea where brightness and contrast went in native sRGB mode. These and the rgb variations have a big effect on the final gamut and dynamic range.

    I'm on linux and have installed the profile system wide. If I view the shot I posted now looks the same in anything including firefox even with it's colour management disabled. Also Opera which I mostly use. That is criticised for having no colour management at all. I don't understand a lot of the colour management jargon but asking some one who does - the author of dispcalgui - it seems that he adds a messed up rgb matrix option to the LUT's to check that applications are choosing to default to the LUT as they should. It seems some might use the matrix. I can visualise what a LUT is and a curve but pass on the matrix. In fact I'm surprised anything other than a LUT is used. Large file though. It took several mins to put it in place where ever it goes.

    Some sites show the same photo in sRGB,AdobeRGB and another (print?). The sRGB and Adobe ones do look the same The other without any colour space info appears to be different especially at the blue end. I understand these should default to sRGB so I may have something else to sort out..

    I've yet to find a way of getting delta E plots out of the software. I may try "verify profile" shortly and see what that does. It will probably take an hour. All I have viewed so far is gamma curves. Turning the monitor brightness up cause the red output to rise more quickly than green and blue. The LUT cured that. They now track apart from a much smaller error in red. I can't measure that on the graph. I will complain to Argll if I can't get a plot of actual colour errors as this seems to be the main metric.

    Ambient lighting is a pain. I did toy with the idea of trying to sort it out properly. It looks to be very difficult to get the correct lighting. Even the gas works type fluorescent photo flood bulbs aren't multi phosphor. They probably show poor reds. The iso standards seem to indicate that sensible 6500K lighting isn't feasible so use 5000K which can be achieved with fluorescent types and get round the miss match with the screen by setting lowish ambient levels. It seems less than 32 Lux is the latest idea. When I turn the light out I get about 7 as there is little natural light where I work. If that proves to give me head aches due to squinting I will probably add some halogen lighting. The rooms main lighting is behind the screen. With it on I get about 23 lux at the screen all reflected and about 120 lux on me. At zero distance Lux seems to equal Cd's so the screen doesn't look too bright with the light on. I'm guessing on that equality. There are some interesting test bars etc on this site and also info on iso standards and colour management http://www.northlight-images.co.uk/articles.php.

    It's all been interesting though as it has mostly improved one of my favourite photo's. The church. It won a mini comp by a long head so I can tell myself that those that didn't vote for it must need their monitors setting up. I'm not sure strict colour realism is needed in my case. A commercial photographer has different needs with product shots etc. Also as mentioned in the tutorial some how or the other our eyes adjust to colour temperature as soon as the object what ever it is lies in the centre of our vision.

    I can't tie down the gamma aspects against light levels. It sounds like it has lost it's way over time as all monitors seem to come out with default gammas of around 2.2. I have noticed that accounting for light levels in the way suggested can push the gamma up a lot and that at say 2.4 my shot to me looks over processed. Other than that all I would notice looking carefully now is that maybe the grass in the lower left nearest the camera may be a bit over sharpened. I assume the clouds don't have a colour cast. That's straight camera white balance. As usual it's often ok apart from the usual problems with some warmish artificial lighting.

    -

  9. #9
    ajohnw's Avatar
    Join Date
    Aug 2012
    Location
    S, B'ham UK
    Posts
    3,274
    Real Name
    John

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    I have managed to get a delta E plot out of the software but am not sure if I have used the correct test chart. I think so as the software seems to adjust for the measured colour temperature. I'm inclined to think stinking fish on the odd ball. There is a similar error in a 100 point grey scale test but I can only get the average and max for that an no value on the face of it. If I look at the test images here http://www.northlight-images.co.uk/a...hite_test.html other than some loss at the extreme black end pf the circular tests that is a bit maybe I can't see any problems. The chart used to measure the errors is labelled ColorChecker.cie and evaluation mode listed as rgb in the results. Extra statistics show 3 readings of the same colour with delta E of 8.1 , 4.8 and 4.64. The other odd thing is that the problem colour swatch is 2 tone in the results

    Any thoughts on LCD screen calibration and gamut coverage against colour temperature

    -

  10. #10

    Join Date
    Nov 2009
    Location
    Provence, France
    Posts
    910
    Real Name
    Remco

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Not sure that that extreme ΔE* is an outlier. The nominal values and device values are rather extreme, in that there's no red at all, and the L*a*b* is also rather extreme compared to the others: there's no other patch with with both a* and b* <-20. It looks like the profile couldn't lower the red enough to get a decent correction (no negative values allowed).

    And the two toned colour patch could be the requested colour and obtained colour, both passed through the profile to give you an idea of the difference. The difference is only visible in that patch, as it has a large error.

  11. #11
    ajohnw's Avatar
    Join Date
    Aug 2012
    Location
    S, B'ham UK
    Posts
    3,274
    Real Name
    John

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    I wonder if the colour is out of the gamut as I am not sure which chart to use. Hoping to get an answer on that shortly. I have run the software again using a chart labelled verify extended and this one offers the ability to check at 6500K or the measured colour temperature. As effectively native I set that option. These charts are using the LUT profile. They are generally better in rgb than the curves plus matrix. The worst colour error is dropped from 2.8

    One fail on grey scale metrics but I wonder if an lcd monitor can pass it. More work as there are some black point compensation features I haven't enabled,

    Any thoughts on LCD screen calibration and gamut coverage against colour temperature

    This is the entire colour chart results. I suspect these aren't too bad for an 180 monitor and my meandering about as I am not familiar with doing this sort of thing.

    Any thoughts on LCD screen calibration and gamut coverage against colour temperature

    Opinions welcomed. It's an interesting area. Seems to me that it isn't possible to evaluate a photo without a fair amount of playing around in this area. I would guess shots would have to be examined with combinations of 6250k - 6760K, dynamic ranges from 300-800 : 1 and gammas from 2 to 2.5 to see it as others might. The worst aspect seems to be gamma variations.

    The red channel does seem to have a problem but I wonder if it's response time.
    -

  12. #12
    herbert's Avatar
    Join Date
    Oct 2010
    Location
    Sussex, UK
    Posts
    471
    Real Name
    Alex

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Quote Originally Posted by revi View Post
    Perhaps not too surprising: afaik, a LUT profile works directly in the graphics card, and thus works on the whole screen at OS level, and not per application, A matrix profile can work on a per application basis. That also means that your LUT profile would still adjust Firefox windows, and any others.

    Remco
    Hi Remco,

    Hardware calibration is expensive and so quite rare. My monitor set-up does not support it. If it did I am sure it would have been all over the monitor specifications. So we are left with software calibration:

    A LUT is a list of input colours (RGB triples) and then the converted output colours. However since there are billions of input colours the list does not contain them all. When the input triple is missing the colour management engine has to do some interpolation using the surrounding points it does have. This is probably where there are differences between colour management engines. For example Photoshop allows you to choose between the Adobe engine or a different one (but I can't remember its name).

    A matrix profile is a simple 3x3 matrix. The input RGB triple is multiplied by the 3x3 matrix to get the output.

    As such not only is a matrix profile much smaller than a LUT profile (so it can be embedded into images), it is also very easy to implement. In both cases it is possible to draw a curve that shows how each channel (R, G or B) is being mapped to the output. This is what John must be using for his profiling analysis.

    I don't think that the profile produced by profiling works at the graphics card unless supported by your system. In most cases this would not be true and the profile is then applied on an application level using software. For my profiling set-up I can only save profiles to the system as a file. I have no option to write it to the graphics card.

    ---

    Edit: I made a mistake. The LUT profile is loaded into the graphics card by the operating system at start-up. This happens automatically on Windows 7 but requires a boot utility on Windows XP.

    ---


    When Firefox is performing colour management it has two options:

    1. It can request help from the operating system to do a conversion, e.g. please convert this AdobeRGB image to the monitor profile. In this case it would be the operating system (Windows in my case) that cannot handle the LUT profile which I find surprising.

    2. It requests the monitor profile from the operating system, does the conversion itself and then sends the image to the monitor requesting that no management is performed by the operating system on the graphics. In this case it is Firefox that cannot handle the LUT. I think this is more likely. It would make sense that the browser would want to perform all graphics operations internally and then send the entire layout (text, images, etc.) to the screen. Later versions of Firefox are hardware accelerated which indicates to me that the browser is doing the entire graphical workflow.

    It may be that Firefox has made a choice to not ask for conversion if a LUT is involved. Or it may be that Windows refuses to do the conversion if it is a LUT profile. Or maybe Firefox cannot convert using a LUT profile. In either case I do not know the exact details of why it doesn't work and can only speculate. The point is that I noticed the difference and prefer using the matrix based profile since web browsing is important to me.

    I would add that this is because I have a wide-gamut monitor. If I had an sRGB monitor then I would not notice. But colours do look out-of-whack on a wide gamut monitor if not passed through a profile.

    Alex
    Last edited by herbert; 31st October 2012 at 04:44 PM. Reason: The LUT profile is loaded into the graphics card by the operating system at start-up

  13. #13
    herbert's Avatar
    Join Date
    Oct 2010
    Location
    Sussex, UK
    Posts
    471
    Real Name
    Alex

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Quote Originally Posted by ajohnw View Post
    Hi Alex -
    I'm on linux and have installed the profile system wide.
    I don't know how profiles are implemented on Linux. But you have a lot of tools to test things out. One of the advantages of the system. I am glad you have DeltaE values. They are the standard to work with.

    Just keep in mind the bottom line is to make sure that the colour actually emitted from your monitor is close to (i.e. indistinguishable from) the colour it is meant to be emitting. Then make sure that this is true of all the colours you are likely to encounter when working with images.

    Alex

  14. #14
    ajohnw's Avatar
    Join Date
    Aug 2012
    Location
    S, B'ham UK
    Posts
    3,274
    Real Name
    John

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    I'm more convinced than ever that the errors are down to step changes on a colour channel causing errors on the next patch. One other advantage of linux is that I can query this with the authors. The brown red for instance in the last jpg I posted is mid red and very definitely look mid red in a red test bar. Linux comes with a full set of bars. The brown red is immediately after the full house white and the earlier jpg worst error also has a step change before it. In fact all error around 1 or so have stepish changes before them as well. I suspect this is down to my monitor being a LED type and trying to be green.

    As to profiles and Linux it's the same as other systems - in the video card generally but there is also a communication channel to the monitor. I suspect that is part of EDID. Last time I upgraded my graphics card I was amazed how much memory it had. I suspect I know why now but the LUT could also reside in main memory and be handled by the graphics card driver. The tutorial on here states this about where corrections lie

    The Look-Up Table (LUT) is either controlled by your video card or by your monitor itself, so it will be used regardless of whether your software program is color managed unlike with the color profile. The LUT is usually loaded immediately after booting up into your operating system, and is used identically regardless of what your monitor is displaying.

    Putting a correction into an image doesn't make sense to me other than for translating adobe rgb to sRGB. I have always checked that adobe rgb and srgb look the same on my monitors.

    I understand Windows 7 and earlier can use profiles system wide but that there can be problems it reloading next time the system is restarted. There are odd bits about on the web about this. Monitor loading is more dubious but there are colorimeters about that will perform all of the settings on a monitor automatically. Windows does make use of EDID and has done for some time and I suspect brightness and contrast control etc are an extension of that. There is some information about EDID use on the web. It seems that the main problem with using it in past forms is manufacturers not really taking the time needed for it to give truly accurate colour information. When running a locally colour managed application these load the profiles as and when required so installing system wide shouldn't interfere with this. My colour managed applications are still doing their own thing other than the Gimp. It's decided to use the profile I have installed all on it's own. I have 3 different colour management systems installed on my system. No sign of any fights or other problems. Firefox uses yet another but I have disabled that for now.

    -

  15. #15
    herbert's Avatar
    Join Date
    Oct 2010
    Location
    Sussex, UK
    Posts
    471
    Real Name
    Alex

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Hi John,

    Thanks for the extra info. I did some more reading and I stand corrected, the LUT profile is loaded into the graphics card by the operating system at start-up. This happens automatically on Windows 7 but requires a boot utility on Windows XP.

    So why I had problems with Firefox I don't know. Perhaps a spate of switching around profiles during one session is not recommended and a reboot should be done to make sure all things are OK. Maybe I experienced some of the problems that you mentioned about the Windows 7 LUT boot process being sporadic.

    Alternatively it may be that Firefox is detecting that there is a LUT profile in place and following a different process to that used with a matrix profile on the assumption that the graphics card is now involved.

    On a different note it is interesting that your profiling errors are greatest when your patches switch dramatically in channel output. Do you have an option to add a delay to the profiling software for the monitor to stabilise on the new colour before reading the patch? Alternatively can you alter the input patch order to avoid big steps?

    Alex

  16. #16
    ajohnw's Avatar
    Join Date
    Aug 2012
    Location
    S, B'ham UK
    Posts
    3,274
    Real Name
    John

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Quote Originally Posted by herbert View Post
    Hi John,
    On a different note it is interesting that your profiling errors are greatest when your patches switch dramatically in channel output. Do you have an option to add a delay to the profiling software for the monitor to stabilise on the new colour before reading the patch? Alternatively can you alter the input patch order to avoid big steps?

    Alex
    Hi Alex
    I've ruled out the delay being a problem by editing the patch file. I changed the swatches before the problem 127,0,0 red so that it repeated and also tried grey and black before it. The delta E remains high what ever. I have managed to get the rest under one by "hacking" the monitor. Viewsonics can be powered up in a factory mode that allows various changes on how the channel behave. Best I could do on the problem red was a delta E of about 2.2 or a bit lower. It's odd. I have seen comments about colorimeters not profiling LED back lit monitors and am beginning to wonder if that is the problem. I have seen comments that certain Huey's will not calibrate one well, that the latest one will - but - and one seller on ebay mentions that a used one of another make wont either.

    I have added 127,0,0 to graphics and see the same brownish red as the swatch in the test reports so the test bars I have used must be misleading me. I'm not sure how unless they sweep through 127 rather quickly.

    I can have a monitor break now as I am away 4 or 5 days. I'm reluctant to just buy a new colorimeter as it might turn out to be a monitor problem. If I do I would favour the Colormunki over the Syder4 going on a none commercial review site but it seems it will be a while before linux supports it.. The reviews suggest both of these for led back lit displays.

    -

  17. #17
    Cantab's Avatar
    Join Date
    May 2012
    Location
    Canada (west coast)
    Posts
    1,238
    Real Name
    Bruce

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Quote Originally Posted by herbert View Post
    The LUT profile is loaded into the graphics card by the operating system at start-up. This happens automatically on Windows 7 but requires a boot utility on Windows XP.
    Alex, I'm jumping into this discussion with considerable trepidation but after I profiled my CRT monitor, my XP desktop computer immediately started to load the profile at start-up. I didn't need to install a special boot utility, but perhaps it came with XP Professional? My comment is probably of historic interest for most folk and, yes, I do have a computer with Windows 7 but it's an LCD laptop and I prefer the larger screen on my still functioning CRT.

  18. #18
    herbert's Avatar
    Join Date
    Oct 2010
    Location
    Sussex, UK
    Posts
    471
    Real Name
    Alex

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Quote Originally Posted by Cantab View Post
    Alex, I'm jumping into this discussion with considerable trepidation but after I profiled my CRT monitor, my XP desktop computer immediately started to load the profile at start-up. I didn't need to install a special boot utility, but perhaps it came with XP Professional? My comment is probably of historic interest for most folk and, yes, I do have a computer with Windows 7 but it's an LCD laptop and I prefer the larger screen on my still functioning CRT.
    Hi Bruce,

    Thanks for the info. Maybe the profiling software also installed a utility for you? Or maybe XP is doing it for you.

    You could have a look in your start-up programs by downloading the extremely useful Autoruns software:

    http://technet.microsoft.com/en-us/s.../bb963902.aspx

    It was written by Microsoft developers to summarise all that is started when your computer boots. You can then disable them. Even if you don't find a colour profile utility it will will help you get rid of a whole bunch of start-up stuff that slows your computer down.

    Alex

  19. #19
    herbert's Avatar
    Join Date
    Oct 2010
    Location
    Sussex, UK
    Posts
    471
    Real Name
    Alex

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    @John,

    A DeltaE of 2.2 is still below the general threshold of 3 that is the level where you can see a difference. But if you can see it and it really bothers you then it may be that your monitor is the culprit. Fixing it will then become expensive.

    Alex

  20. #20

    Join Date
    Jul 2011
    Location
    A Pacific Island
    Posts
    925
    Real Name
    Andrew

    Re: Any thoughts on LCD screen calibration and gamut coverage against colour temperat

    Bruce, your profiling software runs because it has been added to the start-up software in XP. Others such as iTunes, 3rd party media drivers, blue tooth drivers, antivirus, etc may also there depending on the load instructions included in the software. For a list select START>RUN type MSCONFIG select OK. A panel will come up. Select the STARTUP tab and there's your list. Look but don't touch unless you know how to proceed.

Page 1 of 2 12 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •