Monday, March 23, 2015

Poor RGB separation in Plasma TVs and poor Barco broadcast monitors!

I spent a day over the weekend at a customer's facility; they are a large production house with a decent number of edit, grade and audio rooms. An old industry pal has recently become the tech manager there and he's trying to get them up to standard. So - I've been in calibrating monitors and he also asked me to give him an assessment of how easily he could LUT their plasmas to get them to Rec.709.
The first observation is how bad Barco broadcast displays have become! I was very used to them in the eighties at the BBC but the RHDM-2301 is a sorry excuse for a TV monitor.  The marketing material say; 
The RHDM-2301P is the perfect reference monitor for Directors of Photography (DoP) on set during film acquisition, as well as for dailies processing. The Grade-1 color accuracy and stability means that two RHDM-2310P monitors will show identical pictures even on two distant sets.
Which might be the case, but the problem is that they only give you access to adjust the white-point of the monitor (which on these two were both wrong AND different; one was a tad blue, the other a tad red) - the menu tells you what the CIE 1931 color space chromaticity coordinates of D65 are, namely x=0.3127, y=0.3290 but the monitor is not kicking out that colour - it was a bit blue in the white; around x=0.3045 and y=0.3315! This is why monitors tell you what RGB values they are driving and you measure and make the display correct using the x,y,Y values off you photometer or spectralradiometer. For them to tell you what the white point is in terms of x & y and them be wrong is monstrous! They do allow you to tweak the x & y but only in the whites; the monitor is also incorrect in the blacks; greyscale tracking is wrong!
It's also kinda pointless for them to tell you the values of the primaries as well! They can't be adjusted.
So - rant about Barcos over, here are some test results for the cheaper Panasonic 32" plasmas they also had;

Using LightSpace I profiled the display. You can see the gamut of the TV is bigger than Rec.709. The other worry is how bad the RGB separation is. However; after taking a 17-point profile (around 5000 colours) the software reports better than 100% compliance. Good news!

This is the resulting LUT cube and you can see 709 is entirely contained. There are no tightly packed points anywhere indicating a LUT will make things better.

So, loading the cube into our ISmini LUT box and re-running;

1 comment:

Liz Baron said...

I can only agree about the Barco 2301. They were/are a very expensive monitor with limited control. Apparently there is a separate licence for 'Post Production' use which gives access to more parameters, but all the ones I've seen have only had this white point adjustment. As they age the backlights drift too, resulting in some which are impossible to white balance without major surgery.