Update or Re-Calibrate?

[Jesse]

When I install a new graphics driver (usually once a month or so) I do a clean install, which erases the old one completely and installs the new one. Does this affect the profile? When I do this, should I update the profile, re calibrate it from scratch, or does the driver not change the profile? I use a Nvidia graphics card in Windows.

[Florian Höch]

When I install a new graphics driver (usually once a month or so) I do a clean install […] Does this affect the profile?

If it’s the same monitor and a digital connection, then no.

[Jesse]

Thanks!

So I would just need to update the profile once I’ve done the initial calibration?

Side question about the number of patches, IN THEORY, I would think the most patches = the most accurate, considering there are more colors to reference. Or is there just a baseline number of patches that will yield the same results?

• This reply was modified 7 years, 10 months ago by Jesse.

[Florian Höch]

I would think the most patches = the most accurate

In theory yes, but measuring more patches also means longer measurement times, and if the instrument + display fluctuate too much during that period, you’ll be better off not turning it up to eleven in terms of the amount of patches. I would recommend against a run where the combined calibration + profiling measurement times exceed two hours, especially if your instrument is a spectrometer where you’d ideally want to keep it below an hour.

Or is there just a baseline number of patches that will yield the same results?

If you’re not overly critical, you’ll likely reach a point of diminishing returns pretty quickly (I’d say at around 1000-1500 patches). The defaults (around 425) should suffice in most situations. You can use the verification feature (with a larger verification chart) to check profile accuracy.

[Jesse]

Good to know.

So basically looking at ideal under an hour but no more than 2 (for calibration + profile) which then I can verify accuracy with a larger verification chart than the one I calibrated with for a double check.

Is this correct?

I use a colormunki display which I think is a spectrometer.

Calibrite Display SL on Amazon  
Disclosure: As an Amazon Associate I earn from qualifying purchases.

[Florian Höch]

I can verify accuracy with a larger verification chart than the one I calibrated with for a double check.

Basically yes. The verification chart hasn’t to be necessarily larger, i.e. if you’re measuring 400+ patches during profiling then the verification can have around the same. But more patches will give you a better idea about the profiling accuracy.

I use a colormunki display which I think is a spectrometer.

The ColorMunki Display is a colorimeter. ColorMunki Photo/Design are spectrometers.

[Jesse]

Yes wasn’t saying have to, but clarifying, you have made it perfectly clear now 🙂

So then can I go to 11 on patches with a colorimeter? Or is there a finite time with those as well?

[Florian Höch]

So then can I go to 11 on patches with a colorimeter? Or is there a finite time with those as well?

Point of diminishing returns is probably around 2000-2500 patches even if you’re critical.

[Jesse]

Thanks, just so I have this straight, after calibrating with 2000-2500 patches, I can then verify with a larger test chart (more patches than calibrated with) for accuracy, and from there just update the profile. I also do not need to re calibrate because of a clean install on the graphics driver. Is this all correct?

A side note question as far as settings to use for calibration. I use a LED TV as my computer monitor. I use limited range 16-235 (which was a question I asked you on the old forums and you were very helpful!) and I use the standards of a monitor for calibrating (Gamma 2.2, Whitepoint 65k, White level 120). My question is should I be calibrating it like a monitor because I’m using it as a monitor or calibrate it like a TV cause that’s what it is? Or is this a personal preference kind of thing?

As far as I know, there are not any general standards for TV’s because they are all different, so at that point I would use automatic settings on DisplayCAL. I’m just always leery about automatic settings. Any thoughts would be much appreciated 🙂

[Florian Höch]

Thanks, just so I have this straight, after calibrating with 2000-2500 patches, I can then verify with a larger test chart (more patches than calibrated with)

A few hundred for verification should suffice, unless your curiosity demands more 🙂

I use a LED TV as my computer monitor. […] I use the standards of a monitor for calibrating (Gamma 2.2, Whitepoint 65k, White level 120). My question is should I be calibrating it like a monitor because I’m using it as a monitor or calibrate it like a TV cause that’s what it is?

General (home) office / desktop and TV can use the same default parameters anyway (e.g. D65, 120 cd/m2 white).

As far as I know, there are not any general standards for TV’s

The most widely used standard for HD video playback (in a dim room) is D65, 120 cd/m2 white, Rec 709, BT. 1886 tone curve. You’ll likely want to use a 3D LUT (e.g. in conjunction with MPC-HC and madVR).

[Jesse]

So could I create two different profiles, one for desktop/games and one for video playback and swap between the two? Or would you not suggest something like that? If I remember correctly, I think in earlier versions of DisplayCAL is was not recommended.

And yep, that’s what it is, my damn curiosity! 🙂

[Florian Höch]

So could I create two different profiles, one for desktop/games and one for video playback and swap between the two?

There’s not really a need to do that: The parameters are the same.

[Jesse]

Thanks for all the answers! Keep up the great work!!!!! 🙂

[Author]

Posts