Originally Posted by mikela
I have a relatively new RS2000/DCR and a Lumagen. Is it best practice to use Autocal first to set gray scale prior to profiling? I ask because I just purchased a Datacolor SpyderX only subsequently to learn that it has no diffuser. This apparently makes it not suitable for projectors. If that is true, is it correct to assume that it is not really necessary to use Autocal and that LS will make the necessary corrections using the methodology above?
Yes it is also my understanding that the Spyder X has to face the screen, unlike the Spyder 4/5 that had to be used facing the lens with the Autocal. Be carefull, the Spyder X isn't suppoted yet by any of the JVCs, and will most likely only be supported on the current models. A new version of the software has been announced that will support it. It will need a new f/w that will be available on the new 4K models in October/November. No idea about the other models. I have bought a Spyder X and will test it as soon as the new version of the Autocal supporting it lands. I'll post rewsults in the JVC calibration thread.
Originally Posted by Dominic Chan
If you look at the JVC Autocal instructions they specifically say “With the light receiver of the optical sensor facing the projector, install the optical sensor between the projector and the screen.”
Yes, that was for the Spyder 4 and 5 and for older models. At least that's my understanding. It changes with the Spyder X, which is nice provided it can read accurately from off the screen.
The improvements between the 5 and the X are more significant apparently than between the 4 and the 5, which were minimal. I've read 2X faster and 10X better low light sensitivity for the X compared to the 5. However, low light sensitivity would still be inferior to the i1displayPro, which isn't good news because the i1displayPro can't read black accurately on high contrast projectors such as most JVCs. I sent back two of them for that reason before buying my Discus.
Originally Posted by Light Illusion
Our preference is to always take readings off the screen, as for colour/grey scale accuracy you need to include the screen in the measurements.
Additionally, various tests we have performed on the i1D3 diffuser have shown it is not really 'pure' enough for accurate profiling.
(Someone really should run full tests on that, as we have just done basic ones, and then given up based on the initial tests.)
I agree that reading off the screen is preferable, and I agree that from an accuracy point of view, getting a moderately accurate Spyder is a lottery as far as colour accuracy is concerned. Most of them are fairly off.
However, this isn't the most important factor. Their main use with the JVC Autocal is for gamma autocal, to correct the gamma drift which can be quite severe on the JVCs. They do a great job for that, as color accuracy is mostly moot for that use. Then if you want to use them for color calibration, you can either correct them with a custom profile using a more accurate method (time-consuming, complex but works very well) or you can use an i1pro/i1pro2 for color calibration, and that works very, very well.
The JVC Autocal makes it possible to get a very good baseline, and for many that will be good enough. I prefer to apply a small 3D LUT over that just to get near reference results, of course with the limitation than my spectro is only an i1pro2, so it's close to "near reference".
By the way, I appreciate that you seem to have improved your madVR implementation recently. I am snowed under with work at the moment so I haven't had the time to take a look, but I will when I get a chance. Thanks for keeping madVR users in mind, I'm sure that it will be handy when you will need to support the Envy