Radiance 3D LUT (5x5x5 Cube) Calibration with CalMAN 5 - Page 18

Hi Sotti,

thanks for reply. But in THX mode You can't tweak gamma, no such setting. In the other modes You can set gamma up to 2.6, but it will not get any better than the screenshot is showing. Gamma still much too low. Think I will stuck in the THX cinema mode, the only one with acceptable brightness and accuracy, with 100 nits (only) if no one has some other ideas. EU models really suck...

I've done that too!

Hi guys, wrote it already in the Spectracal-Forum, but maybe someone can help me here.


Tried the new version yesterday. I could not make a Full 3DLUT-AutoCal (with Radiance XS):

1. It freezes sometimes while communication with the radiance, then I have to close CalMAN and start again
2. It does not get near the targets (GS Calibration), it starts to calibrate a point and then stops, but with a DeltaE over 15 on some points :shock:

My Equipment: EODIS3 profiled to an i1Pro, Panasonic Plasma 42" S10

So I tried the older version and it worked good, no freeze and the autocal does what it should do - but there I also have some issus/questions:

1. Before calibrating I have 170cd/m² on 100IRE small window, after Full 3D-LUT (5x5x5) I only have 125cd/m² :? :?: Is this normal???
2. After Greyscale I have Gamma at 2,3 with pretty small Delta E´s, but after the Color-Calibration I am at ~2,4 gamma - checked with the Quick-Analysis Workflow
3. When calibrating GS, sometimes CalMAN gets near DeltaE 0 on some points, then it changes the RGB-values again and stops calibrating with a DeltaE higher than 1,5 and moves on with the other points. Is it possible to tell CalMAN to remember the best measured value on the different points??? So if it tries to get the points better, but it is not possible - it could take the best measured combination of RGB measured before. You know what I mean??? Sorry english is not my language

Some genreal questions:

1.What is the difference between reference and adjustable patterns (CalMAN with Radiance) ???? Treid to calibrate with Reference Patterns, but the results were far away of what they should be and the picture was unwatchable
2. Recommendations to window-sizes on a Plasma??


Thank you guys! If someone needs pictures of the calibration, just tell me...

Reference patterns do not show any changes you make to the CMS in the scaler....they are always reference levels.
Adjustable patterns will show the pattern with your CMS adjustments applied. You should ALWAYS use adjustable patterns if you are using the Radiance to do the calibration.

Same answer I gave at the SpectraCal forum.

Haha yeah saw that, forgot to delete the part with the patterns Thank you anyway for the respond!

But i still need some answers on the other points, so maybe someone can help me. I know that everytime you calibrate a display you´re loosing some luminance, but isn´t it too much in my case???
But maybe the loss of luminance is the reason for the higher gamma post calibration - my display is way off in terms of color accuracy precal so the xs has to lower the luminance of white to get the luminance of the colors right....

It may not be too much. Understand that contrast could be set so high that the CMS routine cannot adjust red, blue, or green to meet their luminance targets. So it will reduce luminance of 100% White to a point where it CAN adjust the primaries to their luminance targets.

It's also common to see a difference in Gamma after the grayscale is done and Gamma after the full cube is complete. The Cube only uses a few luminance points for each color, and adjusting them will sometimes impact the overall grayscale/gamma.
Edited by Rolls-Royce - 4/9/13 at 2:44pm

How come the new Radiance 2021 with Darbee DVP does not get the same updates as the other Radiance units? Last update for the 2021 is from 3/18 while the others are up to 4/6. Anybody have any experience with the 2021?

I have 2 sources and 2 TVs. With the dual HDMI output of the 2021, is it possible for me to have 4 different 3DLUTs, one for each possible source/TV combination?

I would think the 2021 would have at least the number of memories that the Mini 3D has, and it has 8 LUT memories. The question is, how will you provide the patterns needed for a 125 pt LUT calibration from those sources? To make such a large calibration practical it has to be automated, so your source becomes the Mini 3D itself, not a BD player or a cable box. Unless, of course, you are going to hand-tweak it.

Assuming the 2021 works like the XE, yes. There are 8 CMS memories. Note that you can only have one active at a time, and both outputs will use that one.

Setting it up can be complex. The user manual describes the input memories. You would program MemA to be used for TV1 and MemB for TV2. Then you program which CMS to use:

• MemA/input1 -> CMS0
• MemA/input2 -> CMS1
• MemB/input1 -> CMS2
• MemB/input2 -> CMS3

When you switch TVs, you also need to switch the Lumagen from MemA to MemB or back, either manually or through a remote macro.

Ok, I now understand that the Radiance can only apply 1 3DLUT at any 1 time.

How about the dual HDMI outputs on the 2021? Can it send video to 2 displays at the same time using the same 3DLUT? (if I had 2 identical make/model TVs)

Yes

You'd probably want to use separate 3D LUTs.

On radiances that have separate outputs I believe you can set them up to use different CMSs and therefore have separate calibrations.

You cannot configure the outputs independently.

Here's the Radiance XE manual. I've read through it multiple times looking for this. It's not there.

Even better, here's Radiance Tech Tip 1 which explains what I was saying above. It says "The two outputs are identical and use the same calibration values."

The one big problem with the new 2021 Radiance is its outdated even before it came out. That is if you are doing a cube type of calibration.
The only reason why I would look at the 2021 is for a basic projector that has no processor or controls, but still its outdated if you want it also for a Larger LUT calibration.

ss

Spoiled by the Colorbox?

I've been following the large LUT adventures with great interest. I just remembered that someone said (from Spectracal?) that the 125 pt LUT needed the 21 pt grayscale for the math to really work. So I ran a 10% saturation sweep on my 125 pt LUT in the Mini 3D and the additional points fell into line very nicely... less than 1 dE. Maybe something would show up in luminance sweeps of greater resolution. I haven't tried that yet. Maybe the color decoder in my projector is well behaved. I don't know. But from what I see, the larger LUTs wouldn't bring me a better picture and would soak up lamp time doing it. For those starting out, $600 for a LUT box is a great deal, but I'm not sure it's worth the time and another device in the signal chain for those of us that already have a Radiance. Or, those that are buying new and want the processing features of the Radiance.

No not at all, eecolor and large LUT's are starting to come of age but there are many more problems to over come imo.
What I am saying is that the large LUT's would have been something that is the best calibration there is for home video.

Yes that is what I have been saying all along, that the Radiance (Mini 3D, in my case) does a nice job for a 21 point grayscale and a 125 point cube LUT.

In my case by adding the eecolor ($600) box in my chain I remove the Radiance Mini 3D, so now all I use the Radiance Mini 3D is for a triplet pattern generator set in reference mode .

Yes it is very true that running a 3375 point LUT takes a lot longer to run than a 125 point LUT, but the results should be better. However Calmans software needs a lot of tweaking to be able to run a large LUT well and can't depend on the color box to help out with a LUT calibration, as the (imo) Radiance Mini 3D does.
I am running a 3375 point LUT now, using Calmans latest Beta.

ss

Update; Calman ran my 3375 point LUT cube like it runs the 125 point LUT for the Radiance Mini 3D.
Edited by sillysally - 4/20/13 at 8:13pm

Last night's 1188 3D LUT took longer than previously. Have you found this?
Use to take min 30 minutes to max 37 minutes. Last night 53 minutes, and the result wasn't as good. Greyscale also took a hit in the bottom values, by elevating them.
Will retry tonight.

Turns out problem is my Bell ExpressVU Remote. When you push the 6 Key, and you're directly pointed at the XE, it messes up the XE. Had to reset to Factory, and restart. Also upped the Stimulus to 100% when calibrating my TV - much better PQ!
Edited by p5browne - 4/22/13 at 12:26pm

This might be a stupid question, but will it run on Windows 8?

Thanks.

Mark