AVS Forum banner
1 - 10 of 10 Posts

·
Banned
Joined
·
5,511 Posts
Discussion Starter · #1 ·
As you can see from the chart, everything looks pretty darn good, except for that green spike at 10. With this Regza, there is a pretty large green deficiency in the mid range (higher blue), so when you raise it to get the mid right, green takes over the 10-20 range.


To lower the Green, my adjustments would mean moving from everything under deltaE 3 to under 5. With the current scatter I have, and the green spike, would you sacrifice the lower deltaE to reign in the green, or would you leave it alone as is?


Thanks.

 

Regza 0227.pdf 100.53515625k . file
 

·
Registered
Joined
·
12,694 Posts

Quote:
Originally Posted by ElwayLite /forum/post/18307322


As you can see from the chart, everything looks pretty darn good, except for that green spike at 10. With this Regza, there is a pretty large green deficiency in the mid range (higher blue), so when you raise it to get the mid right, green takes over the 10-20 range.


To lower the Green, my adjustments would mean moving from everything under deltaE 3 to under 5. With the current scatter I have, and the green spike, would you sacrifice the lower deltaE to reign in the green, or would you leave it alone as is?


Thanks.

Can you split the difference? Reduce the green spike at 10% without compromising 40%-70% too much.
 

·
Banned
Joined
·
5,511 Posts
Discussion Starter · #3 ·
One or two clicks on the low end always put 50 over delta 3, which is prob not huge because its blue that is high and green that is low.
 

·
Registered
Joined
·
2,760 Posts
I would sacrifice some midrange to reduce the low end green. Too much green is the worst subjective direction for the white balance. Too much blue isn't that big of a deal, comparatively.

When you have a large green emphasis like that at the low end, dark suits and even beards/hair or a 5 o'clock shadow on someone's face can have a nasty greenish tinge. I learned that on my own TV when I was first starting out. My wife and I liked the X-Files. If I had green emphasis at the low end, Moulder's stubble would look greenish. Very annoying.
 

·
Banned
Joined
·
5,511 Posts
Discussion Starter · #5 ·
Thanks Chad. Thats what I thought I was seeing, suits and afro's would have this weird tinge to it and it was bugging me. Of course the wife gave me an evil look when I said I was gonna stay up late again messing with it
, because she thinks it looks great. Problem with paying for some calibrations, and then buying your own meter, you start looking at every other thing and become anal about it.


Here is the new chart and a couple of comments:


*) I went in with the attitude, if Im gonna give up some middle, I'll give it up to blue (learned that from Gregg Loewen)


*) Green at 10% was around 120 and Red was about 95, now they are both sitting on top of each other around 105, and blue is higher but still deficient. Do not think I can get it any better than this.


*) My only moves on the Gains and Offsets were to take 3 Greens out of the low end and 1 Red out of the high end


*) The 20% windows went from 107 Green and 95 or so Red, to 105 Red and about 100 Green. Blue comes up faster and is almost 100 at this point.


*) From 30 on its either a little Blue heavy, or Red and Blue are right on top of each other; Green is deficient thought


*) Gamma was interesting. Originally, on the flat line chart, it started out a little over 2.35 and then dropped below 2.35 in the higher ranger, for an avg of 2.32. Once I made the above adjustments, the line became the flattest I've ever seen messing with this tv, and its darn near a perfect 2.35.


I figure equaling Red and Green near 105 at 10% and having a little red on top at 20%, as well as blue higher on both, should fix most of my problem.

 

Regza 3-13.pdf 101.1103515625k . file
 

Attachments

·
Registered
Joined
·
12,694 Posts

Quote:
Originally Posted by ElwayLite /forum/post/18309133


Thanks Chad. Thats what I thought I was seeing, suits and afro's would have this weird tinge to it and it was bugging me. Of course the wife gave me an evil look when I said I was gonna stay up late again messing with it
, because she thinks it looks great. Problem with paying for some calibrations, and then buying your own meter, you start looking at every other thing and become anal about it.


Here is the new chart and a couple of comments:


*) I went in with the attitude, if Im gonna give up some middle, I'll give it up to blue (learned that from Gregg Loewen)


*) Green at 10% was around 120 and Red was about 95, now they are both sitting on top of each other around 105, and blue is higher but still deficient. Do not think I can get it any better than this.


*) My only moves on the Gains and Offsets were to take 3 Greens out of the low end and 1 Red out of the high end


*) The 20% windows went from 107 Green and 95 or so Red, to 105 Red and about 100 Green. Blue comes up faster and is almost 100 at this point.


*) From 30 on its either a little Blue heavy, or Red and Blue are right on top of each other; Green is deficient thought


*) Gamma was interesting. Originally, on the flat line chart, it started out a little over 2.35 and then dropped below 2.35 in the higher ranger, for an avg of 2.32. Once I made the above adjustments, the line became the flattest I've ever seen messing with this tv, and its darn near a perfect 2.35.


I figure equaling Red and Green near 105 at 10% and having a little red on top at 20%, as well as blue higher on both, should fix most of my problem.

Does CalMAN ignore the dE's below 20%? 0% has no dE and 10% looks lower than 20%, which doesn't seem right.
 

·
Banned
Joined
·
5,511 Posts
Discussion Starter · #7 ·
0% has nothing because this is a LD set and the whole screen is black, at least that was my thought. Im not sure why 10% is better than 20%, should be opposite, I would think.
 

·
Registered
Joined
·
6,224 Posts

Quote:
Originally Posted by PlasmaPZ80U /forum/post/18309624


Does CalMAN ignore the dE's below 20%? 0% has no dE and 10% looks lower than 20%, which doesn't seem right.

Calman doesn't ignore those dEs. As the OP alluded to, 0% has no dE because it's luminance is so low. As luminance climbs at 0%, Calman does start displaying dE numbers for it. Although I'm not 100% sure, I think that 10% has a marginally lower dE than 20% because its red and green are somewhat closer to their target levels than at 20%, even though blue is slightly farther out at 10%. Grayscale dE also takes gamma into account, so even though the target luminance for each point isn't shown, if luminance is closer at 10% to its target point than at 20%, this will improve dE there.
 

·
Registered
Joined
·
12,694 Posts

Quote:
Originally Posted by Rolls-Royce /forum/post/18309804


Calman doesn't ignore those dEs. As the OP alluded to, 0% has no dE because it's luminance is so low. As luminance climbs at 0%, Calman does start displaying dE numbers for it. Although I'm not 100% sure, I think that 10% has a marginally lower dE than 20% because its red and green are somewhat closer to their target levels than at 20%, even though blue is slightly farther out at 10%. Grayscale dE also takes gamma into account, so even though the target luminance for each point isn't shown, if luminance is closer at 10% to its target point than at 20%, this will improve dE there.

So, the dE's are for 3-D grayscale (xyY)? I ask because in the initial post the .pdf file showed a much larger error at 10% than 20% in RGB levels yet the dE was quite smaller at 10% compared to 20%.
 

·
Registered
Joined
·
6,224 Posts
Yes, CalMAN uses gamma in the computation of grayscale dE by default. So if you're off your target luminance, your dE will suffer. Color accuracy and luminance are weighted to varying degrees depending on the formula used, so I'm not sure at which point one becomes more crucial than the other.
 
1 - 10 of 10 Posts
Top