Datacolor User to User Support Group.

On Jan 5, 2013, at 8:41 AM, "gerard_culemborg" <yahoo@...> wrote:

> 
> My interest in calibrating (or profiling) displays is not just to get visually acceptable colors, but I also want to use the colors or gray levels for scientific experiments. Therefore, I am interested in how well the Spyder 4 calibrates a display.
> 
> I was hoping that with the advanced analysis tool of the Elite version it would be possible to analyze the calibration. Since this tool is meant to look at the hardware only without calibration, I did the following work-around, which I was hoping anyone can tell me if this works: I have two displays and calibrated the second one. On the first I open the color management tool of Windows 7 (x64) and see that indeed the calibrated icm file is the default one. On the second monitor I start the advanced analysis tool, for example the tone response. Right after it started I select on the first monitor the calibration icm file and default it again. This seems to work, because the colors slightly change on the second monitor. So, I think this way I can use the advanced analysis to measure the calibrated state.
> 
> I followed the above procedure for my Fujitsu P23T-6 LED IPS monitor in photo mode before and after calibration. I posted the tone response curves before and after calibration in the photos section of this message board in the album 'display calibration verification'. It is clear that before calibration the curve was not nice, but after calibration it still does not look anything like the 2.2 gamma curve (even with iterative gray balance calibration), although it does look different now. The strange thing is that when doing a CheckCal procedure, the software mentions perfect gamma and near perfect color temperature, which does not seem to match with the advanced analysis results with work around.
> 
> So my questions are if anyone knows
> - if my workaround is indeed valid?
> - if indeed valid, why the gray scale calibration seems so bad, while the CheckCal says it is ok (is there an inherent limitation for the calibrating LUT or is it a Spyder limitation and would maybe other hardware perform better)?
> - if my workaround is not valid, does anyone know other ways to verify the calibration?

>
> >>My interest in calibrating (or profiling) displays is not just to get visually acceptable colors, but I also want to use the colors or gray levels for scientific experiments. Therefore, I am interested in how well the Spyder 4 calibrates a display.
> 
> For scientific experiments a laboratory-grade gun-type meter such as the Minolta CS2000 is recommended. Unfortunately it costs about as much as a new car...
> 
> >>I was hoping that with the advanced analysis tool of the Elite version it would be possible to analyze the calibration. 
> 
> Not in the current generation of Spyder software. The current analysis toos are analyzing the display, not the calibration. With the exception of the color accuracy test, none of them measure the display's calibrated/profiled state. 
> 
> >>Since this tool is meant to look at the hardware only without calibration, 
> 
> Sorry, I guess the last comment was unnecessary...
> 
> >>I did the following work-around, which I was hoping anyone can tell me if this works: I have two displays and calibrated the second one. On the first I open the color management tool of Windows 7 (x64) and see that indeed the calibrated icm file is the default one. 
> 
> Caution must be used under Windows, as profile selection at the OS level does not flash the video LUTs 
> 
> >>On the second monitor I start the advanced analysis tool, for example the tone response. Right after it started I select on the first monitor the calibration icm file and default it again. This seems to work, because the colors slightly change on the second monitor.
> 
> This would imply a failure of your Windows system/video card to some rate the two display profiles/calibrations. Flashing one should NOT effect the other. Try adding a second video card. 
> 
> >> So, I think this way I can use the advanced analysis to measure the calibrated state.
> 
>  I don't see anything in your description that convinces me that the correct calibration for the second display has been applied by applying a profile on the other display. 
> 
> >>I followed the above procedure for my Fujitsu P23T-6 LED IPS monitor in photo mode before and after calibration. I posted the tone response curves before and after calibration in the photos section of this message board in the album 'display calibration verification'. It is clear that before calibration the curve was not nice, but after calibration it still does not look anything like the 2.2 gamma curve (even with iterative gray balance calibration), although it does look different now. The strange thing is that when doing a CheckCal procedure, the software mentions perfect gamma and near perfect color temperature, which does not seem to match with the advanced analysis results with work around.
> 
> Hence implying that the workaround isn't working around. 
> 
> >>So my questions are if anyone knows
> - if my workaround is indeed valid?
> - if indeed valid, why the gray scale calibration seems so bad, while the CheckCal says it is ok (is there an inherent limitation for the calibrating LUT or is it a Spyder limitation and would maybe other hardware perform better)?
> - if my workaround is not valid, does anyone know other ways to verify the calibration?
> 
> I've responded to all but the last above. The last may have to wait for Spyder software with extended functionality. Verification occurs in the current software, but not with the "do it yourself" toolbox you are looking for, just simple confirmation as you note. 
> 
> C. David Tobie
> Global Product Technology Manager
> Imaging Color Solutions
> Datacolor inc. 
> cdtobie@...
> www.datacolor.com
> 
> On Jan 5, 2013, at 8:41 AM, "gerard_culemborg"  wrote:
> 
> > 
> > My interest in calibrating (or profiling) displays is not just to get visually acceptable colors, but I also want to use the colors or gray levels for scientific experiments. Therefore, I am interested in how well the Spyder 4 calibrates a display.
> > 
> > I was hoping that with the advanced analysis tool of the Elite version it would be possible to analyze the calibration. Since this tool is meant to look at the hardware only without calibration, I did the following work-around, which I was hoping anyone can tell me if this works: I have two displays and calibrated the second one. On the first I open the color management tool of Windows 7 (x64) and see that indeed the calibrated icm file is the default one. On the second monitor I start the advanced analysis tool, for example the tone response. Right after it started I select on the first monitor the calibration icm file and default it again. This seems to work, because the colors slightly change on the second monitor. So, I think this way I can use the advanced analysis to measure the calibrated state.
> > 
> > I followed the above procedure for my Fujitsu P23T-6 LED IPS monitor in photo mode before and after calibration. I posted the tone response curves before and after calibration in the photos section of this message board in the album 'display calibration verification'. It is clear that before calibration the curve was not nice, but after calibration it still does not look anything like the 2.2 gamma curve (even with iterative gray balance calibration), although it does look different now. The strange thing is that when doing a CheckCal procedure, the software mentions perfect gamma and near perfect color temperature, which does not seem to match with the advanced analysis results with work around.
> > 
> > So my questions are if anyone knows
> > - if my workaround is indeed valid?
> > - if indeed valid, why the gray scale calibration seems so bad, while the CheckCal says it is ok (is there an inherent limitation for the calibrating LUT or is it a Spyder limitation and would maybe other hardware perform better)?
> > - if my workaround is not valid, does anyone know other ways to verify the calibration?
>

On Jan 5, 2013, at 6:07 PM, "gerard_culemborg" <yahoo@...> wrote:

> I wasn't quite clear with the workaround explanation. I actually do change the profile of the second display (the one I calibrated) and not the first display. I just mentioned that I do this on the first display to note that I am not switching applications on a single monitor.
> 
> So I am still confused why the calibration did not go right. I also tried making a profile only and based on this trying to set certain luminance values on the screen, measured with a Minolta luminance meter, and these did not give the correct tone response curve either. Is there a reason why the hardware tone response curve as shown in the file 'before_calibration.jpg' in the photos section is too difficult to calibrate or create a good profile of that can be used in a CMS?

>
> >>I wasn't quite clear with the workaround explanation. I actually do change the profile of the second display (the one I calibrated) and not the first display. I just mentioned that I do this on the first display to note that I am not switching applications on a single monitor.
> 
> Still, applying a profile under Windows at the OS level will not flash the video calibration data to the Video LUTs, as it will on a Mac. So the idea that this will allow you to measure the "calibrated' state would be incorrect, not to mention that our current software would remove the LUT corrections before measuring anyways, as that is the way it was designed. 
> 
> >>So I am still confused why the calibration did not go right. I also tried making a profile only and based on this trying to set certain luminance values on the screen, measured with a Minolta luminance meter, and these did not give the correct tone response curve either.
> 
> Why would a profile built to just define the current state of the display, not calibrate it, reduce a correct tone response curve? You told it not to. 
> 
> >> Is there a reason why the hardware tone response curve as shown in the file 'before_calibration.jpg' in the photos section is too difficult to calibrate or create a good profile of that can be used in a CMS?
> 
> I don't quite understand the question. Correcting the TRC, per channel, is a calibration function. Not a profile function. Where the "too difficult" part comes in is not clear to me.
> 
> C. David Tobie
> Global Product Technology Manager
> Imaging Color Solutions
> Datacolor inc. 
> cdtobie@...
> www.datacolor.com
> 
> On Jan 5, 2013, at 6:07 PM, "gerard_culemborg"  wrote:
> 
> > I wasn't quite clear with the workaround explanation. I actually do change the profile of the second display (the one I calibrated) and not the first display. I just mentioned that I do this on the first display to note that I am not switching applications on a single monitor.
> > 
> > So I am still confused why the calibration did not go right. I also tried making a profile only and based on this trying to set certain luminance values on the screen, measured with a Minolta luminance meter, and these did not give the correct tone response curve either. Is there a reason why the hardware tone response curve as shown in the file 'before_calibration.jpg' in the photos section is too difficult to calibrate or create a good profile of that can be used in a CMS?
>