| Author |
Message |
akajester
Joined: 09 Jul 2008
Posts: 934
Location: Wisconsin
|
| Posted: Wed Sep 16, 2009 2:50 pm Post subject: luminance and gamma question, BG808s |
|
|
Thanks for reading this.
I have gone through Kal's excellent guide and learned more than the last dozen or so times I went through it. I'm close to getting my pj corrected. My gamma value is too low. Would lowering my luminance curve bring the gamma closer to 2.2? Or do I need a gamma boosting device? Thanks!
|
|
| Back to top |
|
 |
garyfritz
Joined: 08 Apr 2006
Posts: 12088
Location: Fort Collins, CO
|
| Posted: Wed Sep 16, 2009 2:58 pm Post subject: |
|
|
Yes, *lowering* luminance *raises* gamma. It's a bit counter-intuitive but that's the way it is.
You don't need a gamma-booster. They don't actually boost gamma -- they boost low-end luminance. The resulting gamma calculation will show a lower gamma, but that's an approximation. With a booster kicking up your low end, your luminance curve doesn't exactly fit a power curve any more so an overall gamma number is not real accurate. The HCFR gamma chart (your second picture) will show you the "gamma" at each IRE level, which is more accurate. It's basically showing you how far above/below the ideal luminance line you are at each IRE level. With a booster your low-end gamma should be low, and above 20-30IRE or so it should revert to the desired 2.2 range.
I'm drawing a blank on what it takes to lower the luminance curve to correct your gamma. If you lower Contrast you'll lower the top (100IRE) end and the whole curve will re-scale to the same place. Lowering Brightness might help, but that will tend to lower the bottom (0-30?IRE). Unless you get better advice from somebody who's not suffering from brain-fade, I'd try experimenting with the Brightness. First look at a test pattern like the 0IRE window on the GetGray or AVCHD test disks. Those patterns show 0IRE against a blacker-than-black background. If you can clearly see the 0IRE window, your Brightness is too high. Try lowering the brightness a bit and re-measure.
|
|
| Back to top |
|
|
akajester
Joined: 09 Jul 2008
Posts: 934
Location: Wisconsin
|
| Posted: Wed Sep 16, 2009 3:05 pm Post subject: |
|
|
I use the patterns from AVSHD 709. I'll check the brightness/contrast again. Thanks for the info Gary, I appreciate it.
EDIT: I can lower my green gain too. I wonder if bumping that down a few will lower the luminance.
|
|
| Back to top |
|
|
AFryia
Joined: 09 Mar 2006
Posts: 965
Location: S.E. Michigan VPH-G70Q
|
| Posted: Thu Sep 17, 2009 11:28 pm Post subject: |
|
|
| garyfritz wrote: |
I'm drawing a blank on what it takes to lower the luminance curve to correct your gamma. If you lower Contrast you'll lower the top (100IRE) end and the whole curve will re-scale to the same place. Lowering Brightness might help, but that will tend to lower the bottom (0-30?IRE). |
On a stock CRT-PJ you want to set brightness to your taste for black. Then set the the contrast/luminance. My experience is that as you lower the contrast setting, gamma will increase. Generally your blue tube is saturating too fast. When adjusting grey scale so red/green tracks a blue that saturates too fast your grey scale looks good but the gamma is low.
So you need to lower you contrast or the gain on each tube. Electronically defocus blue that will help with the saturation.
I asume you have a color meter. This is what I've done in the past.
1) adjust brightness/bias to taste.
2) set gain for white and contrast to the manufactures recomended setting (ususlly too high)
3) take grey scale reading and note average gamma
5) decrease contrast setting and repeat grey scale. Keep taking readings down to about 1/2 your max contrast setting. Any lower may kill you contrast ratio and make for a very dim picture.
When you reach your tolerance for overall brightness/contrast ratio/gamma go back and tweak you brighness/bias and gains for grey scale tracking.
I've been able to get 2.0-2.3 gamma <3 delta E on a stock PJ (no extrnal gamma processing)
The trade off is overall brightness. The image will probably not be as bright as you're used to. If you view in a light controlled room you will soon grow accustom to the better overall picture.
|
|
| Back to top |
|
|
akajester
Joined: 09 Jul 2008
Posts: 934
Location: Wisconsin
|
| Posted: Fri Sep 18, 2009 2:21 am Post subject: |
|
|
Thanks for all the info everyone. Last night I was poking around HCFR and turned on individual RGB on the graphs. I noticed that blue was much higher than normal and bringing up the averages. I EM defocused the blue a little, and dropped it's gain just a few bumps. Ran a full Gray again and Gamma was at 2.1, luminance is almost dead on! I think if I went a little lower I could get it almost perfect. It looks so good now I don't want to mess with it. 
|
|
| Back to top |
|
|
garyfritz
Joined: 08 Apr 2006
Posts: 12088
Location: Fort Collins, CO
|
| Posted: Fri Sep 18, 2009 1:48 pm Post subject: |
|
|
Congrats! Yes, defocusing blue is essential to get the color temps consistent, otherwise you end up with a "blue hump" in the middle. See this post for an explanation.
Some people can't stand the "blue fuzz" (which is usually only noticeable on white-on-black text, like credits) and prefer to go with the blue hump. If you've got a fancy enough scaler or other video processor, it's possible to custom-tweak the blue gamma curve to get proper grayscale without defocusing the blue. It'll drive your blue real hard and wear it a lot faster but it's possible.
|
|
| Back to top |
|
|
|
|
