Some VERY basic and novice questions!

[dweeble]

OK, this is where I am: Run a Dell Optiplex with only a basic Intel graphics card on Windows 10, monitor is an elderly Dell U2312HM. I print on a Canon Pro-100s. Objective is to calibrate such that what I see on screen is what I get when I print (yes, I do soft-proof and use correct ICC paper profiles).

Best value calibrator seems to be SpyderX Pro right now.

Is there anything in my set-up which is too old or primitive for that to be work?

If that all looks OK, might it be best to stick with idiot-proof mode on the SpyderX or is using DisplayCal going to help at my level?

Thanks in advance!

SpyderX Pro on Amazon  
Disclosure: As an Amazon Associate I earn from qualifying purchases.

[S Simeonov]

The spyder devices are not tat accurate like the x-rite ones…

[dweeble]

Perhpas not, but I’m sure either gives acceptable levels. My question is more whether my kit is capable of being “trained” effectively or whether I need a better graphics card, more tuneable monitor, ect….

[MW]

I agree with S Simeonov, the Spyder is the limiting factor. Besides lesser accuracy Spyders lack spectral corrections for your display needed for white point matching. Members of Luminous Landscape would suggest sending off a print to a custom profiler and a light booth. But it doesn’t help much if the monitor isn’t accurately characterized.

• This reply was modified 3 years, 5 months ago by MW.

[dweeble]

Can you please translate that into plain English? I did say they were VERY basic questions,  it might help if you assumed that I have little knowledge of your field 🙂

[Vincent]

Can you please translate that into plain English? I did say they were VERY basic questions,  it might help if you assumed that I have little knowledge of your field ????

Spyder X is NOT future proof (or even present day) unless you own an spectrophotometer which starts at 400 euro for the very basic ones.

I1displaypro & its variants is “software” upgradeable via CCSS corrections even if you do not own an spectrophotometer, just need that somebody measured a display with same or very similar backlight and uiploaded it to community. Also since it’s a HID device (like a “mouse”) doe snot need special drivers and will work even some day Xrite “kills it” making al i1Profiler solutions not compatible with i1d3 revA or something like that.
Cheaper variants (i1Display Studio) are not supported by HW calibration solutions if you will own a monitor with such feature, take this point seriously before buying a olwer i1d3.

Regarding your question, a U2312HM should be some kind of 6bit+dither panel. Softproofing or even comparing to a printed copy with some affordable light sources like Osram D50 CCFL or Yuji D50 VTC (awesome) implies using as calibration target for display D50 so you can use white paper color simulation (which assumes same white in display as light source.. wich leads to D50)
I’m afraid that lowering so much blye +green channels in a 6bit+dither White LED panel may be too much, even worse if your GPU does not help. But you can try, side effects may be banding but you may live with it.

Also factory/vendor profiles for that kind of printers (for canon paper, from canson, ilford…) may be not so accurate, so I’ll consider paying for a customized printer profile for your most used papers, or even consider an spectrophotometer.

These are your “limiting factors”. If we assume that they do not exist (but THEY EXIST) if you are asking for calibration target and we assume that you have some kind of good light source with more or less continuous not bumby spectral power distribution (another BIG issue) a starting recipe would be:
-Display white point matching light source (example D50 light -> D50 monitor )
-Display luminance matching  ( lux at paper from your light source DIVIDED BY pi) (example 100cd -> for a “booth” that gives ~314lux at paper)
-closest to native gammma for display, let apps deal with those conversions, not GPU if it is not so good for these tasks (no dithered LUTs)
-start with a simple matrix + single curve + black point compensation profile type.. and see how it behaves, if it can be a good match for actual display behavior. If not, try more compex ones like huge XYZLUT profiles.
-calibration speed to medium or slower, so you ensure grey calibration is as neutral as your display+GPU can do.

This way softproof on display should give a “close enough” match to a printed copy under that light source… but there are a set of requirements you must met previously, as explained above.

[dweeble]

 

Thanks for very complete reply. May I please ask some more questions based on it?

I1displaypro & its variants ………….e, take this point seriously before buying a olwer i1d3.

Got it!

Regarding your question, a U2312HM should be some kind of 6bit+dither panel. Softproofing or even comparing to a printed copy with some affordable light sources like Osram D50 CCFL or Yuji D50 VTC (awesome) implies using as calibration target for display D50 so you can use white paper color simulation (which assumes same white in display as light source.. wich leads to D50)
I’m afraid that lowering so much blye +green channels in a 6bit+dither White LED panel may be too much, even worse if your GPU does not help. But you can try, side effects may be banding but you may live with it.

It seems it is an 8 bit, does that help? Also it seems it is a backlit screen, is that as expected or does it make a difference? So the studio needs to be lit with D50 for viewing the screen and ditto the print viewing box?

Also factory/vendor profiles for that kind of printers (for canon paper, from canson, ilford…) may be not so accurate, so I’ll consider paying for a customized printer profile for your most used papers, or even consider an spectrophotometer.

I  use quite a few different papers,  will wait and see if any of the supplied profiles are weak before doing that

These are your “limiting factors”. If we assume that they do not exist (but THEY EXIST) if you are asking for calibration target and we assume that you have some kind of good light source with more or less continuous not bumby spectral power distribution (another BIG issue) a starting recipe would be:
-Display white point matching light source (example D50 light -> D50 monitor ) So if I am test viewing under D50, the software will alow me to specify a D50 white point?

-Display luminance matching  ( lux at paper from your light source DIVIDED BY pi) (example 100cd -> for a “booth” that gives ~314lux at paper) Utterly confused, sorry 🙁 Please expand!
-closest to native gammma for display, let apps deal with those conversions, not GPU if it is not so good for these tasks (no dithered LUTs) The monitor control only gives 2 gamma settings, PC or Mac, which doesn’t seem very useful. A quick Google suggests it’s actual gamma is 2.2 but tends to run rather low. Does that make a difference?
-start with a simple matrix + single curve + black point compensation profile type.. and see how it behaves, if it can be a good match for actual display behavior. If not, try more compex ones like huge XYZLUT profiles.
-calibration speed to medium or slower, so you ensure grey calibration is as neutral as your display+GPU can do.

This way softproof on display should give a “close enough” match to a printed copy under that light source… but there are a set of requirements you must met previously, as explained above.

Thanks for putting up with a novice!!!!!

[Vincent]

-It’s 6bit, it’s market segment. If you go to an Eizo with similar spec will have the same panel type (6bit) but different QC and maybe a few extra features.
White point should match lightsource, whatever it is (assumig a more ore less continuos not bumpy spectral power distribution), it was witten at the end.
But paper color and optical brighteners play ar role here. You can set display at light source to match, then rely on softproof to simulate paper white displacement (bluish, yellowish), or you can turn white paper simulation OFF (if possible) and match paper visually or numericaly (measuring reflection) or you can use both (start at D50, make minor tweaks).

-regarding “brightness” match, just google lambert cosine law and start from there. If we assume that reflection on a white paper is “close to” perfect diffiuse reflection, the amount of light reflected will be aprox. that value (same as monitor).
If we fix lux of booth high enough to be able to see tiny issues in dark printed colors (D50 500lux) you get for display D50 and 160cd, google P2 appraisal level (3664 + 12646 with TC130 procedures). See which relation holds for those two values.
In the same way if you fix display to 100cd you’ll get a stating point for getting a match on booth at ~320lux.

-gamma PC + g2.2 seem the easy choice

• This reply was modified 3 years, 5 months ago by Vincent.
• This reply was modified 3 years, 5 months ago by Vincent.

[Author]

Posts