Using a light meter whilst setting white and black levels

Before I took my H77 in for it's upgrade to H78, I thought I'd measure the contrast ratio so that I could see if the H78 mod increased the CR greatly. I found that when setting WinDVD6 by eye, I was only getting around 2200:1 CR. This is a newly set-up PC so I was surprised at this figure as I was getting aroyund 2600:1 before using the same set-up.

I had my light meter set-up and taking readings, and while I was using Avia to set the black level from my HTPC (now using PowerDVD 6 to compare), I found the black level would read 0.36 lumens. Setting the white level would read 958 lumens.

Setting the black and white levels by eye sometimes would not achieve these figures, especialy the white level, so I relied on the light meter to ensure that when properly set-up, I had the best black level and the best white level whilst still keeping both moving bars visible. Final calibrated contrast reading was as the above figures suggest: 2661:1. That was higher than my previous 'bye eye' setting of around 2600:1.

I then tried setting up WinDVD6 and found that although I could achieve the same figures, I couldn't get them both at the same time when trying to get both white and black level settings correct. It seems that as brightness and contrast effect each other when adjusting, the adjustments on WinDVD6 were not allowing me to set the levels without some compromise. The final CR with WinDVD6 was around 2550:1.

I had to use the software player adjustments as well as the projector adjustments because with the player settings at zero or midpoint, the moving bars were not being made visible, so I had to use the software player controls to display them.

I'm not sure if the differences with PDVD and WDVD were due to the overlays using PC levels differently - maybe WDVD was using the correct range of 16 to 235 within the PC levels, and PDVD was using PC levels and expanding 16 to 235 to match, so my next test would be to see if below black and above white levels could be made visible with both, so that's something to try later.

I was using the Extech equivalent of the Robin RT-24 which can be bought on ebay for around £20. If you want to measure your own contrast ratio at home, a lightmeter with that kind of ratio will be fine. It doesn't have to be calibrated as you're just measureing relative to your own meter, and not against someone elses, so although it might not be 100% accurate across the range, you can still get the brightests white and blackest black with it. If you want to get the best contrast out of your pj, then a light meter seems a good way to help you achieve it.

Gary.

 

Useful stuff Gary.

Also, a light meter is pretty much essential for adjusting Iris's like on the Z3. Because your eye will not detect the point at which the Iris offers optimum light reduction.

For instance, -63 on the Z3 is the lowest it will go (from memory!) but -53 yields the lowest actual brightness before it starts to increase. This is optimum for the darkest black (not contrast ratio - which will stay the same irrespective of iris) if you have a black-out room.

 

There are many cheapo lux meters available which only have a 1 lux resolution - pretty useless really for this application if you want a realistic measurement - you really need one with 0.01 lux resolution for the black measurement, so be careful on eBay! (or just make sure you buy the same one Gary uses)

 

Good point Mike,

The Robin/Fluke RT-24 is identical to the Extech meter that comes with SMART III, and works just as well. I can't find any on ebay right now, but tehy do keep coming up. They often go for around £20 which is cheaper than the £79 it seems to retail for.

For black level you need the meter to read into decimal places, even when you place it close to the pj lens, just to make sure you're getting a reading that's well within the range of the meter - anything that's reading near the extreme could be innacurate. For example, I was measuring teh NEC HT1000 at the screen, and the readings for black level were so close to zero that the result suggested in excess of 3000:1 contrats ratio. Putting it much closer revealed a more accurate figure of 2000:1.

A similar thing happened with teh Colorfacts Trichromat sensor - it would read it's minimum reading if too far away so you weren't reading the black level at all.

Gary.