Hi,

Does anyone know if there is any restriction when doing probe matching to raise temporarily the brightness of the display through physical controls, do the probe matching on the WRGB patches, then lower the brightness when doing characterization?

This can help with certain spectro who have difficulty with lower luma readings so I wonder if there is any contraindication to that.

Probably not needed, as CS does probe matching at 92 to 94% (I forget the exact number) in order to avoid color clipping. It shouldn't be an issue for ANY spectro.

This has nothing to do with clipping.
Usually if you target rec709 100 nits when doing probe matching the BLUE patch will be at around 6-7 nits and RED at 19-20 nits. Raising temporarily the physical brightness of the display can make a spectro have better precision/repeatability. After the probe matching the physical brightness should be lowered to reach your target.

I know of no Spectro that cannot read accurately above a couple of nits...
If you have found one that can't, I would suggest it will be inaccurate with any reading at any brightness level.

Steve

There is a certain spectro which is very accurate in his mft. sw. but for some reason has his measurements capped in CS at 5sec. BTW the same 5sec cap happens in another well known calibration app. Don't ask me why.

5sec max measurement time leads to not so good repeatability especially on the RED patch which is the most difficult for all spectros.

So, to work around that, I was thinking about this solution. Raise temporarily the physical brightness, do the PM, then lower the physical brightness.
Due to this artificial limitation if i raise the luma with 25 nits for example (only for PM) then the max 5 sec measurement time will generate infinitely better accuracy.

If a 5 second integration time cannot provide a valid and accurate reading above a couple of nits I would say there there is something seriously wrong with the probe...
But, just use the probe manufacturer software to take the 4 measurements.
Far preferable to altering brightness levels as you will not know if such a change alters the display's spectral response or not.

Steve

Not really, even CR-300 owners told me that they prefer to use Speed=Normal which leads to 20sec reading per patch instead of Fast or 2xFast which on their own testing is not so reliable (especially on the RED patch).

Why settle for less when the device is perfectly capable of much better precision.

Yeah, for the moment I am just doing that, taking measurements in the mft. sw. and copy/paste.

But that doesn't work for MVPM in CS and the Bodner method in another calib. sw.

Bodner uses the exact same 4 colours, so that makes no sense.
And MPVM is totally different, obviously, and I would definitely NOT recommend altering brightness for that.

Steve

The other calib sw doesn't support copy/paste at all.
For the FCMM method they have an external utility to save a PM matrix that can be later used in their sw.
For the Bodner method they don't have any external utility.

Tough luck for me.

As Bodner is not in any way related to the actual Probe Match readings, that makes no sense...
Bodner is a process/algorithm applied to a standard set of probe match values.
Just poor implementation on Calman's part.

Separately, having issues with Red is interesting too, as Blue should by rights be the more difficult.
Both because it has a lower brightness compared red, and because it can be more susceptible to issues with interference from stray light from longer wavelengths.
And blue is inherently more difficult for any spectro.
It does suggest something not really right with that probe...

Steve

No, in it's own mft. sw. the probe behaves with the same tolerances as a CR-300.
And the RED patch has the lowest repeatability on both CR-300 and that spectro. Actually the two probes compare almost the same on all 4 patches.

Far preferable to altering brightness levels as you will not know if such a change alters the display's spectral response or not.

Interesting, doesn't this also affect the spectral response when gathering data for an EDR? As far as I know there is no standard brightness when measuring spectrums for an EDR. For example I measured the spectrum on EIZO CG319X and found out that the EDR present in CS database was measured at roughly 175 nits, which is way larger brightness than this monitor is used for.

The CR spectros I have (and Jeti) are all better with red than blue as the light intensity drops.
And yes, that is also a good reason why EDRs are not that accurate.

Steve

Ok, thanks very much for the info. Have a nice weekend.

Kuky
I am aware that your issue has nothing to do with clipping. I mentioned that solely as the justification I've seen for the selection of the particular levels ColourSpace uses when probe matching.