LCD Monitor Calibration

[snapzz]

I have posted this on another thread but think i may get a better response with its own thread.

I have just managed to souce a Spyder from someone else. However I have a query.
When I had my old CRT calibrated it was done by adjusting the settings on the monitor. When calibrating my LCD's I had to create a profile which was done by the software. Can you not calibrate the same as the CRT or am I using a different version? I thought it was the same model although not the latest.

 

[Chaz Photos]

You should adjust BOTH first you should be setting the whitelight temp if yours has it, some have a backlight adjustment, as well the brightness and contrast PLUS then The spyder will make a profile for the colours.

 

[dalex_257]

I have posted this on another thread but think i may get a better response with its own thread.

I have just managed to souce a Spyder from someone else. However I have a query.
When I had my old CRT calibrated it was done by adjusting the settings on the monitor. When calibrating my LCD's I had to create a profile which was done by the software. Can you not calibrate the same as the CRT or am I using a different version? I thought it was the same model although not the latest.

Calibration is separate from profiling. Calibration is about getting your device (monitor in this case) as close to the standard conditions as possible (white point, black point etc). Profiling is about building a device profile that shows how it differs from the standard conditions which then allows to adjust images for this device accordingly. Ideally, you need first to calibrate your device and when it is as close to chosen standard conditions as possible, only then you profile it.

However, LCDs unlike CRTs have very little variables to deal with so their calibration is really a software gimmick. For example, the white point is not really adjustable - it is the colour of backlighting when the LCD dots are fully transparent. What you as RGB adjustments in many LCDs should really be avoided at all costs - they manipulate white point by adjusting the transparency of the RGB pixels so the white won't be truly white - this will actually reduce the monitor gamut if you do that because the maximum RGB range will be less.

The only real adjustment you can do on LCD is brighness - it controls backlight power. Anything else is always implemented via tweaking RGB levels or the image itself.

For LCDs you should always reset all their menus to the native output (no RGB adjustments) and then use profiling only and in your profiling program specify to use native white point (good profilers should have this option) - no 6500K or any other.

 

[snapzz]

Thanks

I did follow all instruction such as change monitor to factory settings first which was infact all I had to do. I could manually adjust all RGB/brightness/contrast on the monitor itself but software didnt request this.The Spyder then created a profile. As stated OP on the old CRT I did have to manually adjust the settings on the monitor to calibrate.Any further profiling was done using Adobe Gamma settings but only as a tweak.

 

[snapzz]

Calibration is separate from profiling. Calibration is about getting your device (monitor in this case) as close to the standard conditions as possible (white point, black point etc). Profiling is about building a device profile that shows how it differs from the standard conditions which then allows to adjust images for this device accordingly. Ideally, you need first to calibrate your device and when it is as close to chosen standard conditions as possible, only then you profile it.

However, LCDs unlike CRTs have very little variables to deal with so their calibration is really a software gimmick. For example, the white point is not really adjustable - it is the colour of backlighting when the LCD dots are fully transparent. What you as RGB adjustments in many LCDs should really be avoided at all costs - they manipulate white point by adjusting the transparency of the RGB pixels so the white won't be truly white - this will actually reduce the monitor gamut if you do that because the maximum RGB range will be less.

The only real adjustment you can do on LCD is brighness - it controls backlight power. Anything else is always implemented via tweaking RGB levels or the image itself.

For LCDs you should always reset all their menus to the native output (no RGB adjustments) and then use profiling only and in your profiling program specify to use native white point (good profilers should have this option) - no 6500K or any other.

So are you saying this Spyder will not callibrate my LCD only create a profile?

 

[dalex_257]

So are you saying this Spyder will not callibrate my LCD only create a profile?

Yes and that is correct - you cannot calibrate LCDs - they are just not variable (apart from the brightness).

But it will give you the best possible results so what is the problem? The LCD (if you haven't played with its RGB settings and used all native configs) will be used to its full potential this way.

 

[Chaz Photos]

Thanks

I did follow all instruction such as change monitor to factory settings first which was infact all I had to do. I could manually adjust all RGB/brightness/contrast on the monitor itself but software didnt request this.The Spyder then created a profile. As stated OP on the old CRT I did have to manually adjust the settings on the monitor to calibrate.Any further profiling was done using Adobe Gamma settings but only as a tweak.

I use the Spyder 3 and It do get you to adjust the software do as you to adjust it; it gives you 4 white blocks and you need to see al and 4 black block same again this is all done before using the spyder itself.

 

[pxl8]

I use the Spyder 3 and It do get you to adjust the software do as you to adjust it; it gives you 4 white blocks and you need to see al and 4 black block same again this is all done before using the spyder itself.

Holy Proof Read Batman!

 

[Chaz Photos]

So are you saying this Spyder will not callibrate my LCD only create a profile?

Any output will have a ICC Profile this is how a computer talk to any other device so you get the right details.
It like a translator when you get a computer there will be a native ICC profile for output with a spyder this only changes it for a new calibrated one.
All out puts have one windows one or manufactures one or in this case one you make with a calibration took.
Its like my printing I have one for my paper and my ink on my system with a calibrated screen and knowing how to use it I get on paper what I see on screen.

 

[Chaz Photos]

Holy Proof Read Batman!

ok I missed a letter K out

 

[dalex_257]

I did follow all instruction such as change monitor to factory settings first which was infact all I had to do. I could manually adjust all RGB/brightness/contrast on the monitor itself.

The only thing you should be adjusting is brightness. Contrast should be left to default (100%?) and all RGB levels set to the max.

As stated OP on the old CRT I did have to manually adjust the settings on the monitor to calibrate

The CRT works differently from LCD. CRT, does not use (and therefore is not affected by the colour of) the backlighting - the light comes when the beam of electrons excite phosphorus on a screen surface. Regulating analogue beam power you can create a whole spectrum of whites. The result of this is that CRTs don't have a native whitepoint as such and you can calibrate them to whatever one you need (6500K seems to be mostly preferred choice).

LCD forms the image by having liquid crystal in front of the backlight which can go from fully opaque (black) to fully transparent (white) in distinct (non analogue) set of steps (this usually dictates the number of bits per pixel - tonal levels each pixel is able to reproduce). Here I refer to pixel as one of the R, G or B components (physical pixels) on LCD matrix. This construction has one important consequence - backlight is of the same colour (mostly) throughout the whole screen (white) and it cannot be changed from pixel to pixel (unlike CRT beam intensity). This in turn results in that there is only one true white point for LCD (native) which is when all the pixels are set to be fully transparent to let all backlight through. Adjusting white point with R, G and B controls in the LCD monitors will do it by limiting maximum transparency of the R, G or B pixels which in turn means that it will eat from the useful tonal range those pixels can display.

An example to help with understanding of this. Suppose your LCD has 10 bits per pixel which means at each R, G or B pixel on a screen the transparency of LCD matrix can be varied between 0 and 1024 (so you have 1024 tonal values for each pixel, and hence per each colour channel). Suppose that your LCD backlighting has native white point around 8000K (they all are around 8000-9000K mostly). If you now will try to calibrate/set your LCD at white point 6500K, then effectively what you would be doing is reducing blue and green. A result of this "calibration" is that blue pixels will be able to only display say 750 tonal values (the maximum transparency limited at 750 to achieve 6500K white point) a little bit better for greens but still not their full range. As a result of all this, the colour gamut (all the colours you can display) have been decreased. Now the question is - is it what you really want?

 

[Martyn...]

Yes and that is correct - you cannot calibrate LCDs - they are just not variable (apart from the brightness).

Are you sure about that, I am sure that I can calibrate my Eizo Colour Edge if I wanted to, however I just let the Spyder and its software create a profile for my graphics card.

 

[dalex_257]

Are you sure about that, I am sure that I can calibrate my Eizo Colour Edge if I wanted to, however I just let the Spyder and its software create a profile for my graphics card.

Yep. I think you are confusing the availability of R,G,B controls in LCD menus - mostly all LCD screens have them but it does not mean they should be adjusted. What I meant was that you cannot truly calibrate them - at least not in the same way as CRTs.

Only a very few very expensive LCDs that use LED matrix as backlight with individual LEDs behind each pixel will allow true CRT like calibration. The rest is just software gimmick.

Calibration is about adjusting white and black points and this is just not possible on LCD (with above exception).

 

[pxl8]

Calibration is about adjusting white and black points and this is just not possible on LCD (with above exception).

Where do changes to the gfx card gamma ramp come into the equation then?

 

[dalex_257]

Where do changes to the gfx card gamma ramp come into the equation then?

In graphics card prior to the output. They are affecting the output to the monitor, just the same as if you would apply gamma to your image and then send it out.

The problem is that it does not then use the full range of the monitor - when R, G and B are adjusted in monitor control panel, it sort of applies it's own curve on top of what arrives there - to fit image RGB values to those limited by the monitor settings.

 

[Martyn...]

Yep. I think you are confusing the availability of R,G,B controls in LCD menus -

No probably the fact it is sold as a monitor that can be calibrated, not that I have ever attempted it as I am happy with the results just using the Spyder direct to the computer.

Are you familiar with the Eizo Colour Edge range?

This is the one I have LINK

 

[pxl8]

I have to say I'm not convinced either. I've adjusted the RGB values on my monitor as part of the Sypder 3 calibration to get the colour temp close to 6500 and the cd/m2 to around 130. The results give incredibly good screen to print matching for colour and brightness. As for losing bandwidth I can't say I've noticed anything as a result.

 

[dalex_257]

No probably the fact it is sold as a monitor that can be calibrated, not that I have ever attempted it as I am happy with the results just using the Spyder direct to the computer.

Are you familiar with the Eizo Colour Edge range?

This is the one I have LINK

Yep. The problem is that term "calibration" is often used to refer to the whole process - a.k.a. calibration and profiling. In case of Eizo they just refer to hardware LUt and presence of the colourimeter. Eizo are better than the average DELL or Apple however - they have hardware LUT table which means you don't store it at the graphics card and your pixels changed in least erroneous way. What I said though about LCDs are still applicable to them.

Any good colour management book or even calibration and profiling program (like basICColor Display) will recommend you to leave LCD WP as native and don't tweak it. It is simple really - the less you tweak R,G,B and contrast levels on LCD the more of the available native monitor gamut you will have to display your images.

 

[dalex_257]

I have to say I'm not convinced either. I've adjusted the RGB values on my monitor as part of the Sypder 3 calibration to get the colour temp close to 6500 and the cd/m2 to around 130. The results give incredibly good screen to print matching for colour and brightness. As for losing bandwidth I can't say I've noticed anything as a result.

I was not saying it would be glaringly obvious on every single image. The point here is whether you are after the best possible quality and consistency or not. The whole effort of using calibration/profiling to me personally means that I am - if I started to care about calibration then I'd do it properly.

It is the same really as to argue whether JPEG is the best format in the world or not for high quality images... For some people it will be sufficient and for some others it will not (me included - I do not like the lossy compression). Same here. If you think that limiting (compressing) your LCD gamut is fine as soon as it is not visible to you then for you it is ok. For me it is not

If you want, you could experiment and generate a gradient image which could be displayed on the "uncalibrated" LCD smoothly and yet have some banding when you "calibrate" it to non-native WP (for 6500K WP calibration I'd use pure blue channel to see that). I would not do it since knowledge of how and what LCDs do for displaying is more then enough for me.

Edit: have a look at this thread where Andrew Rodney comments on this very point.

 

[Martyn...]

Any good colour management book or even calibration and profiling program (like basICColor Display) will recommend you to leave LCD WP as native and don't tweak it. It is simple really - the less you tweak R,G,B and contrast levels on LCD the more of the available native monitor gamut you will have to display your images.

Thanks for that, that is what I do, the print to screen accuracy is superb, so I had no intention of messing with the monitor, plus to do so involves running upstream and downstream connections, much simpler just to plug the spyder into the usb on the mac, and let it sort it out.

 

[pxl8]

If you think that limiting (compressing) your LCD gamut is fine as soon as it is not visible to you then for you it is ok. For me it is not

But what purpose does all this extra gamut serve if it's beyond reach of your working space or output device? If your monitor profile shows it's capable of extended saturation that will lead to banding when working with colour managed apps when using a working or proofing to a space that is smaller than the monitor. It's not different to the sRGB or AdobeRGB argument, more isn't always better :nono:

 

[dalex_257]

But what purpose does all this extra gamut serve if it's beyond reach of your working space or output device? If your monitor profile shows it's capable of extended saturation that will lead to banding when working with colour managed apps when using a working or proofing to a space that is smaller than the monitor. It's not different to the sRGB or AdobeRGB argument, more isn't always better :nono:

Your monitor is your output device and you are limiting its gamut colour space deliberately by tweaking its white point.

I am editing all images in LR with ProPhotoRGB which is significantly larger then any LCD can display so your point is mute here. Only and only if you are doing something in sRGB and with 8bits - only then what you are saying is true.

 

[pxl8]

My monitor isn't my output device, it's my editing device. My final output is a printer in sRGB with jpegs so, yes what I'm saying is true and that's the point.

Calibrating, for example, a wide gamut monitor such as the Dell 24" using native white point will result in completely unrealistic saturation levels - it might look super punchy on screen but won't come close to matching a print - been there done that, avoid wide gamut like the plague as a result

 

[Chappers]

I have to agree with Pxl8.

I use a GM Eye 1 Display 2, with ColourEyes Display Pro.

The first part of the procedure is to set white point ,brightness and contrast ( I use the recommended L* for contrast) These are all hardware set by using the OSD on the monitor. Once set the software then runs it's calibration and profiling procedure .

 

[dalex_257]

My monitor isn't my output device, it's my editing device. My final output is a printer in sRGB with jpegs so, yes what I'm saying is true and that's the point.

Calibrating, for example, a wide gamut monitor such as the Dell 24" using native white point will result in completely unrealistic saturation levels - it might look super punchy on screen but won't come close to matching a print - been there done that, avoid wide gamut like the plague as a result

Even considering recent (supposedly near aRGB) 16 bit printers like Canon 9500 Mk II??

FYI, printers are never in sRGB and are not even coming close - that was a unified space for what monitors are supposedly capable of displaying. Inkjets colour spaces (in current printers) do intersect with sRGB but are not contained within it.

IMO what you are doing is no better than just edit in printing space - what's the point in using all wide gamut spaces like aRGB and ProPhotoRGB then?

 

[pxl8]

IMO what you are doing is no better than just edit in printing space - what's the point in using all wide gamut spaces like aRGB and ProPhotoRGB then?

The reality is there is no point in editing in a space larger than the output device can handle, because the output device can't handle it.

The golden rule is always, always use a working space that's only big enough for the colours in the image that the output device can render - anything bigger and you'll either clip the colours or introduce banding. Because, yep, the output device can't handle it.

For fine art stuff and/or low print run wide/large gamut is fine provided you can work and print in 16bit.

Back in the real world of a working photographer getting prints from labs such as DSCL it's 8bit jpeg with sRGB as a working space.

Read the thread about sRGB v AdobeRGB in my sig, it's got some examples that clearly show the utter waste of tonality for the sake of saturation that never used.

 

[dalex_257]

The reality is there is no point in editing in a space larger than the output device can handle, because the output device can't handle it.

The golden rule is always, always use a working space that's only big enough for the colours in the image that the output device can render - anything bigger and you'll either clip the colours or introduce banding. Because, yep, the output device can't handle it.

This rule was and still is only applicable to low resolution spaces (like 8bits files). Not so if you edit in 16 or more bits which is where more realistic display come into count. The end mapping to limited gamut like printer's should relly be left down to ICC to handle - after all this is why there is a lot of tals about printer profile quality and it's exactly the reason why it is not easy to build a good printer profile.

I will still remain convinced that the less loss you encounter during editing - the better the final result will be and to that matter the ultimate editing space for me willbe the one that encodes colour with floating point (32 bits) when PS and LR will finally arrive there.

I guess it is pointless to argue really - you have your opinion and I have mine both based on either experience or mathematics or mixture of those. So I will try to stop here ;-)

My main point of posting was to clarify to OP what calibration really is and what is the underlying difference between LCD and CRT - the choice I'll leave of what is better and what is not I'll leave to others...

 

[snapzz]

Wow loads of techy stuff flying around here now!

I think I get the gist of it now. The main reason for asking is I have a multi monitor setup for showing images at events. The monitors are identical in make and model and are connected via network switch to my laptop. What I wanted to do was to calibrate all three monitors. Its plain that that is not possible so I ended up by setting all three to default and create a profile from one. Taking into account wear & tear may mean each may be slightly different it is not critical as they are only viewing aids.

Many thanks to everyone for your help.

 

[Martyn...]

Wow loads of techy stuff flying around here now!

I think I get the gist of it now. The main reason for asking is I have a multi monitor setup for showing images at events. The monitors are identical in make and model and are connected via network switch to my laptop. What I wanted to do was to calibrate all three monitors. Its plain that that is not possible so I ended up by setting all three to default and create a profile from one. Taking into account wear & tear may mean each may be slightly different it is not critical as they are only viewing aids.

Many thanks to everyone for your help.

Spyder 2 would only allow one monitor to be calibrated, however I think that Spyder 3 allows multiple monitor calibration, worth checking out.

 

[Chaz Photos]

The reality is there is no point in editing in a space larger than the output device can handle, because the output device can't handle it.

The golden rule is always, always use a working space that's only big enough for the colours in the image that the output device can render - anything bigger and you'll either clip the colours or introduce banding. Because, yep, the output device can't handle it.

For fine art stuff and/or low print run wide/large gamut is fine provided you can work and print in 16bit.

Back in the real world of a working photographer getting prints from labs such as DSCL it's 8bit jpeg with sRGB as a working space.

Read the thread about sRGB v AdobeRGB in my sig, it's got some examples that clearly show the utter waste of tonality for the sake of saturation that never used.

It is always better to Start out with more then you need at the end as you edit any image you are degrading it on one way or another
AS you output in a 8bit Jpg, go by what you are saying you might as well shoot in Jpg mode.
Shooting in RAW and editing in 16bit Tiff you will have MAX info to start with only then reducing it to the colour space you need and 8bits.
Also so many do not know How to print from Photoshop correctly with the right profile and printer settings; I have to keep on teaching this over and over in my club.
One last thing you have to remember that a screen is light produce where as a print is reflected light you need to set up Prof View correctly