or Connect
AVS › AVS Forum › Display Devices › Digital Hi-End Projectors - $3,000+ USD MSRP › Are Calibration Settings Like Braces?
New Posts  All Forums:Forum Nav:

Are Calibration Settings Like Braces?

Poll Results: How did you set up your projector settings?

 
  • 5% (2)
    I am just using a factory preset and made no other changes.
  • 5% (2)
    I started with a factory preset and made adjustments by eye with no guidance.
  • 2% (1)
    I copied settings for my model that I found online.
  • 23% (9)
    I calibrated using a disc (DVE, S&M, WoW), but no colorimeter.
  • 48% (19)
    I calibrated using a colorimeter.
  • 15% (6)
    I had a professional do the calibration.
39 Total Votes  
post #1 of 30
Thread Starter 
We all have teeth. They're white, protected by enamel, and arranged in two little rows. Pretty much the same in everyone, right? Well, almost. A lucky few might be born with perfect teeth, but most teeth are at least a little crooked. Luckily, you can find a professional who, for a modest fee, will measure your teeth and outfit you with a bespoke set of braces that will bring them into perfect alignment.

But would you ever wear braces that were fitted for someone else's teeth? Probably not.

Displays are born with pixels. Like teeth they're white, protected by glass or plastic, and arranged in neat little rows. And like teeth they rarely come from the factory perfect and need adjustment, sometimes a lot. You can also find professionals who, for a modest fee, will measure and fix your pixels.

Everyone's teeth are crooked in a unique way, and every display's settings are off in a unique way. So, it seems like copying someone else's calibration settings makes as about as much sense as wearing someone else' braces. Or does it? Here's my experience calibrating my PT-AE8000U for the first time.

Factory Defaults
Here are the RGB levels for my PT-AE8000U in REC709 mode with all settings at factory default:

Too much green, too little blue on the low end, and way too little red across the board. The bottom half of the chart should be displaying a pink line graphing the total error at each IRE, but you can't see the line because the error is too high at all points. So the factory defaults are literally "off-the-charts" bad. Color temp ranges between 7000K and 7500K.

ProjectorReviews.com Settings
When I first set up my projector, I naively copied the settings from the ProjectorReviews.com review of the PT-AE8000U. Here are the RGB levels for those settings:

Well, it looks like these settings are actually an improvement at every IRE level. There's still a lot of error, but they are substantially better than the factory defaults.

My First Calibration
Here are the RGB levels I achieved on my first try using an i1Display Pro colorimeter. I only made adjustments to RGB and overall brightness and contrast controls, and have not tried using the CMS yet.

Not too shabby. A further improvement from the ProjectorReviews.com settings at all IRE except maybe 10. And pretty close to perfect between IRE 30 and IRE 80. I'd appreciate input or tips from more experienced members.

Conclusions
Obviously, copying someone else's calibration settings is not going to get you great results. But will they be better than the factory defaults? It depends. If the defaults are really bad, and the errors all follow the same general pattern, then I think it's a reasonable assumption that someone else's calibration settings could make improvements in gross errors. In the case of the PT-AE8000U, if too little red is a general problem, that would explain why ProjectorReview.com's settings helped. To the extent that 8000's are consistently low in red, any calibrated settings that increase red are going to help.

On the other hand, the amount of additional red you need is going to vary between units, so you're not going to get really good results unless you calibrate your own unit. Plus, the overall brightness and contrast settings I ended up with were quite different from the ones I first copied.

I'd highly recommend investing in a colorimeter. I used it not only on my projector, but on an LCD TV and 3 computer monitors. They all look fantastic now, and my triple desktop monitor setup finally has consistent color across all 3.
post #2 of 30
I think copying settings is a bit like playing roulette. You might get a modest improvement, as you did with the projectorreview.com settings...or it might get worse. It gets even dicier when considerations like screen gain, screen material, and the color of surrounding walls, ceiling, and floor which are reflecting light back onto the screen. The technique of copying settings is probably somewhat less random on HDTVs than it is on projectors, because there are fewer variables in the equation.
post #3 of 30
Thread Starter 
Second Attempt

This is what I get after taking advantage of the PT-AE8000U's advanced gamma controls:

I was able to fix up IRE 10 and IRE 20, and eke out a little more accuracy out of the other points. Pretty good overall. 6300-6600K throughout.
post #4 of 30
Try reducing the main contrast control a click or two since it looks like you're running out of red at the 100%, hence the tail off of the red. This will reduce brightness slightly but improve the accuracy of the brightest whites.

However, just to point out that if you move your meter even a small distance to one side of the screen I'll bet your white balance will completely shift due to the typically poor uniformity of Panasonic LCD projectors: My AE2000 & AE3000 were like this so I never knew which reading to take as the 'correct' one.
post #5 of 30
Thread Starter 
Quote:
Originally Posted by Kelvin1965S View Post

Try reducing the main contrast control a click or two since it looks like you're running out of red at the 100%, hence the tail off of the red. This will reduce brightness slightly but improve the accuracy of the brightest whites.

Yeah I already tried that. The problem is I'm barely getting 12flL as it is, so I don't really want to lower the brightness any further.

I did discover this evening though, that if I start with Cinema 2 mode instead of REC709, I get a lot more brightness. Haven't had time to see if I can calibrate Cinema 2 as well as REC709 yet, but I may be willing will sacrifice a little on RGB accuracy and color temp for a bit more brightness.
Quote:
Originally Posted by Kelvin1965S View Post

However, just to point out that if you move your meter even a small distance to one side of the screen I'll bet your white balance will completely shift due to the typically poor uniformity of Panasonic LCD projectors: My AE2000 & AE3000 were like this so I never knew which reading to take as the 'correct' one.

Tell me about it! The poor color uniformity is plainly obvious when looking at a 100 IRE field. I try to make sure to point the colorimeter at the same spot each time so I don't drive myself crazy. But, I never notice the color shift in normal content so it's not that bad in practice.
post #6 of 30
This is what always gets me when I read about people chosing the Panasonic over other models (often the JVC) due to 'higher lumens'. You and others seem to have confirmed that once you try to calibrate them you don't get anything like the figures in the adverts. frown.gif My AE3000 was so dim after calibration that I had to compromise the settings to increase light output much as you have done. I replaced it with a JVC HD350 which was considerably brighter despite the specs saying different. My current X35 hits 15fL after calibration in low lamp with the aperture closed right down which is much brighter than the HD350 was since I had to run in high lamp and open the iris up quite a bit too.

My AE3000 and another belonging to a forum friend who brought it over to my place had the same poor uniformity: His was much worse with a noticeable pink hue on the left and green on the right. Whenever there was a snow scene or similar bright white scene, it was really obvious. However, it was considered 'within spec' when he tried to reject it.
post #7 of 30
it does look like there is a pattern of the 8000 being low on red.

http://www.hdtvtest.co.uk/news/panasonic-pt-at6000-201211052339.htm?page=Calibration

The recent 8000 I calibrated had red's in the 80% range with the best out of the box setting I could fine. This seems odd since Panasonic advertises a 'red rich' lamp. The Panasonic had the highest out of the box D65 dE's vs. any of the other models I reviewed this year. Especially in 3D, the color was a major challenge to try and calibrate behind the glasses.

In contrast, the HW50 has the best out of the box settings compared to the other models in the shoot-out thread. Out of the various samples I've seen they were relatively close with color space and gamma settings with variances in the grey scale.


I also noticed the color uniformity issue. You could move the meter a few inches in each direction and end up with different settings. As mentioned above, it doesn't show in most content but there are definitely some movies I can see it (Young Frankenstein B&W) and Art of Flight which has tons of all white fields.

re: Calibration settings - It's interesting when reviewers post their calibration settings. I don't think it's for the purpose of trying it on another model. It should be used as reference to see how far off the model being reviewed is from D65/R709. If you see massive negative gains for blue and green, you know a specific projector is weak on red. Which isn't great because you are pulling out the brightest color of green to compensate and dropping lumens.

This was the case on my old JVC RS50. The red was consistently low on that projector. It wasn't a light canon to begin with. Calibrating for D65 drop the lumens even more putting the model in RS2 brightness territory. The replacement RS55 the following year (my current JVC) made a significant improvement in this area. It was brighter and also had plenty of red. I believe they made a filter change of some kind, I don't recall.


re: meters - I think it's a good idea to get a meter than has been calibrated. The high end (10K+) meters are out of reach for most home enthusiasts. Meters like the D3 can be calibrated against the reference models with the corrections built into the software. Both Chromapure and Spectracal offer this feature. I think it's worth the extra $$ knowing the meter is as accurate as possible for a consumer product.
post #8 of 30
There are some DLP's that ship out with a new lamp at almost exactly the same color settings (I've calibrated multiple of some of the same model DLP). The problem is even here though as the lamp ages even a little bit, it starts to drift quite fast. Also, even though some DLP's ship out with nearly identical off dE's, there is always the odd one out where the calibration is way off the factory's intention. With LCOS and LCD, you will often see a larger variance even on a new lamp. That said, it depends on the DLP model, some DLP's will also not be consistent between the same model, but some are for sure. Consistent as in less than 5% to 10% difference across the entire calibration on a new lamp, again though problem is as the lamp ages even a tiny bit, the drift starts fast.
post #9 of 30
Thread Starter 
Quote:
Originally Posted by Kelvin1965S View Post

This is what always gets me when I read about people chosing the Panasonic over other models (often the JVC) due to 'higher lumens'. You and others seem to have confirmed that once you try to calibrate them you don't get anything like the figures in the adverts. frown.gif My AE3000 was so dim after calibration that I had to compromise the settings to increase light output much as you have done. I replaced it with a JVC HD350 which was considerably brighter despite the specs saying different. My current X35 hits 15fL after calibration in low lamp with the aperture closed right down which is much brighter than the HD350 was since I had to run in high lamp and open the iris up quite a bit too.

Good observation but not entirely fair. I read all the reviews before purchase, so I was fully aware that calibrated lumens would be significantly less than peak lumens. But that actually is exactly what a lot of us need for two different viewing scenarios: Bat Cave vs. Sports. In calibrated Bat Cave mode yes, I'm not getting peak brightness (12fL), but it's still plenty bright. In daytime Sports mode, where I'm not light controlled at all, I don't care about color accuracy, so I'm getting 30-40fL. My understanding is higher end pjs like the Sony HW50ES and the JVC's can't hit that peak brightness no matter how badly you screw the colors. So while I'd be happier during night time, I wouldn't be able to use it during the day.

Quote:
Originally Posted by Kelvin1965S View Post

My AE3000 and another belonging to a forum friend who brought it over to my place had the same poor uniformity: His was much worse with a noticeable pink hue on the left and green on the right. Whenever there was a snow scene or similar bright white scene, it was really obvious. However, it was considered 'within spec' when he tried to reject it.

You do realize there's been 3 new generations of the Panny since AE3000, right? smile.gif I don't know how the AE3000 looked, but I'm sure the AE8000 is a lot better. That said, it still has color uniformity problems, but they are only visible on test patterns, not normal content. I briefly owned an Epson 5020 while trying to decide, and it had perfect color uniformity as far as my eyes could tell.
post #10 of 30
I do realise it's many generations old, but having seen recent posts on the 8000 (Zombie10K's thread for one, but others too) it seems that they still haven't solved the problem and some have said it's still noticeable especially on B&W content. In the UK it's no longer the cheaper option to buy the Panasonic (it's the same price as the X35 or VW50ES) so IMHO it should be as good as them regarding uniformity. The contrast and black level is significantly poorer too, so at the same price (in the UK) I can't understand anyone buying the '8000 over the X35 or VW50ES, maybe it's different in the USA though.

I'm pretty sure than the Sony is pretty much as bright in non calibrated modes, though the X35 isn't, but I don't have any interest in watching a projector in daylight myself (though I have used my old HD350 during the day I just pulled the curtains). If that's the main use then obviously your priorities are different, but equally not much point worrying about it being calibrated which is where we came in...
post #11 of 30
Quote:
Originally Posted by Xank View Post

Second Attempt

This is what I get after taking advantage of the PT-AE8000U's advanced gamma controls:

I was able to fix up IRE 10 and IRE 20, and eke out a little more accuracy out of the other points. Pretty good overall. 6300-6600K throughout.

grey grey grey grey grey grey grey CYAN!

That "white" would drive me nuts - it is wayyyyyyyyyyyyyyyyyyy off.

I use per-color gain to start with a perfect white, and THEN calibrate the remainder of the greyscale. Yes, this will cost you a bit of brightness, but GAWD that must look horrible every time there are clouds or, well, anything white.

You've affixed quality braces to fix the teeth everywhere else, but left out the two giant buck teeth sticking right up front!!
Edited by Fat Dave - 3/18/13 at 9:55am
post #12 of 30
Thread Starter 
Quote:
Originally Posted by Fat Dave View Post

grey grey grey grey grey grey grey CYAN!

Cute! I appreciate the special effort you put into criticizing my two-day old calibration skills. smile.gif j/k.
Quote:
Originally Posted by Fat Dave View Post

Yes, this will cost you a bit of brightness, but GAWD that must look horrible every time there are clouds or, well, anything white. You've affixed quality braces to fix the teeth everywhere else, but left out the two giant buck teeth sticking right up front!!

Honestly, it wasn't so bad. I discovered while calibrating that my eyes are a lot more sensitive to red contamination of greyscale than blue contamination. This might be some kind of perceptual quirk or it could be because it seems like most factory defaults for monitors, TV's, and projectors favor blue over red, so we are just used to it.

In any case, I went for my third try last night:



Happy now? Plus, this calibration actually produced more brightness. The difference is I started with the PT-AE8000U's "Cinema 2" preset, which calibrated just as easily as "REC709", but resulted in 3-4 more ftL at 100 IRE. In this case I'm getting 16ftL from a calibrated "Cinema 2" while I was only getting 12.5ftL from calibrated "REC709".

Calibrating also had the side effect of reducing some of the blue blooming/fringing I complained about in a different thread. Hooray.

Here's a graph with a nice cyan line.

post #13 of 30
Now that's a pretty good lookin' result!
post #14 of 30
That's a much better result. smile.gif

I'm sure I read somewhere that it's best to have the red slightly lower in preference to it being higher than the blue/green. Since I can adjust each 10% (or even 5%) greyscale step in my Lumagen I always make sure that the red is always a tiny fraction lower at each step (the dEs are still well under 3 even so).

What does your gamma look like since that has a major impact on the image?
post #15 of 30
Thread Starter 
Quote:
Originally Posted by Fat Dave View Post

Now that's a pretty good lookin' result!
Quote:
Originally Posted by Kelvin1965S View Post

That's a much better result. smile.gif

Thanks guys! It's been fun so far. Very tedious at first but I'm getting the hang of it now that I have a better feel for how the controls interact and constrain each other.

I also did a bunch of benchmarking to understand the black level and peak brightness characteristics of my projector under various picture, lamp, and iris modes. A lot of things turned out to be counter-intuitive. For example, the dynamic iris' effect on increasing or decreasing brightness seems to be opposite depending on whether the lamp is ECO or NORMAL.

Quote:
Originally Posted by Kelvin1965S View Post

I'm sure I read somewhere that it's best to have the red slightly lower in preference to it being higher than the blue/green. Since I can adjust each 10% (or even 5%) greyscale step in my Lumagen I always make sure that the red is always a tiny fraction lower at each step (the dEs are still well under 3 even so).

That's a really good tip. It seems impossible to get absolutely perfect balance, so if you have to favor one ever so slightly, I agree, I'd choose blue as well.

Curious, can you plug a colorimeter into a Lumagen box and have it auto-calibrate itself? Like the way a colorimeter and software can build an ICC profile on a computer pretty much all by itself?
Quote:
Originally Posted by Kelvin1965S View Post

What does your gamma look like since that has a major impact on the image?

Heh. smile.gif Yeah that still needs a little bit of work:



Project for tonight I guess.
post #16 of 30
Quote:
Originally Posted by Xank View Post


Curious, can you plug a colorimeter into a Lumagen box and have it auto-calibrate itself? Like the way a colorimeter and software can build an ICC profile on a computer pretty much all by itself?
Heh. smile.gif Yeah that still needs a little bit of work:



Project for tonight I guess.

That gamma really needs some work, I suspect getting that right will improve the picture considerably since it's very low in the mid and high % so they will likely look a bit more washed out than they should.

You're close regarding the Lumagen, but you plug the sensor into your laptop and also connect the Lumagen to your laptop via a serial port (or serial/USB adaptor in my case). Then your calibration software (Chromapure in my case) controls the Lumagen which generates it's own test patterns while the sensor measures the result. Chromapure then adjusts that point in the Lumagen's controls (greyscale, gamma or one of 125 colour gamut points eek.gif) until the dE is within target (or as close as it can get if not). It takes about 40 minutes to run the whole thing for 11 point greyscale and gamma plus 125 point CMS, though I usually stop it after the greyscale/gamma, check a few things and manually tweak before running the 125 point advanced colour gamut part. It used to take me 4-6 hours trying to do this manually and even then I only had 8 point CMS, such is progress. cool.gif
post #17 of 30
Thread Starter 
Quote:
Originally Posted by Kelvin1965S View Post

That gamma really needs some work, I suspect getting that right will improve the picture considerably since it's very low in the mid and high % so they will likely look a bit more washed out than they should.

Ok after last night I think I have a better handle on what's happening with gamma. Here's my latest attempt at calibrating the PT-AE8000U's Cinema 2 mode:





Greyscale looks good. Gamma remains non-linear, but now averages just around the target 2.2. I tried using the advanced gamma controls to straighten out the curve, but all attempts to do that ended up throwing things completely out of whack. I can shift the entire curve up or down (that's how I got gamma to average 2.2), but I cannot change it's basic shape.

Out of curiosity, I went back to my REC709 mode-based calibration and measured its gamma:



It's almost perfectly flat but for a little dip at 90 IRE.

After a bit of experimentation, I think what's going on is that the REC709 and Cinema 2 modes simply have different gamma response curves. I'd be curious if someone can explain the optics behind how you get linear vs non-linear gamma curves. As far as I can tell, that's the principal difference between the two modes. REC709 will allow you to calibrate greyscale perfectly AND get linear gamma. Cinema 2 will also allow you to calibrate greyscale perfectly, but will give you higher overall brightness (I get 4 additional ftL's at 100 IRE) at the cost of non-linear gamma.

So I guess it's a question of priority. For me, I'd prefer the 4 additional ftL's, and frankly switching back and forth between REC709 and Cinema 2 calibrations, I could not notice very much loss of detail in either dark or bright scenes with Cinema 2's curved gamma. One thing I'm wondering is, since I can shift the entire curve up or down but not straighten it out, is it preferable to have 10 IRE (dark content) be set to 2.2 gamma, but the rest of the curve a little too bright, or is it better to shift the entire curve darker, so that the whole thing averages 2.2 gamma but some IRE's will be higher and some will be lower? Perhaps that's a matter of personal preference.
post #18 of 30
Well gamma is often the subject of fierce debate, since it is partly a subjective choice. Some aim for a totally flat response but at different levels (so 2.22, 2.3, 2.4 or even higher). Others (like me) aim for a flat response above 20% and lower the gamma as it approaches 0% to avoid black crush and improve shadow detail. Yours is almost the opposite of mine: Higher gamma low down compared to higher up, so in theory you might have more black crush/less shadow detail and slightly less 'depth' to brighter scenes.

However, you have a dynamic iris: If you took those measurements with it switched on then this could explain the results partly. Depending on whether you use small windows or full screen patterns. IMHO if you measure with DI ON then you need full screen patterns as otherwise the DI won't respond to the % that you are trying to measure since it will be partly that % in the window surrounded by a black border. It could even be a deliberate choice by Panasonic when they designed the projector since I've seen it said a few times that calibration is more accurate when based on rec709 mode (your colour gamut may well be more accurate too since I found that Cinema 2 on my old AE3000 was quite oversaturated for example).

The reason your rec709 and Cinema presets give different peak light outputs may be due to two reasons:

1. The DI opens further in Cinema mode so gives a higher peak white.
2. The colour space that is used can effect the white balance, so if you need to reduce the green in one mode to correct the white balance this will also reduce the peak white output.

When I had my old AE1000 it got damaged (and written off) but I was allowed to keep it so I took it apart and tested various things. I found that the DI opened and closed different amounts in different modes. Unfortunately even in the service menu it wasn't possible to change this (not sure if the newer models allow this in the same way that Sony projectors do).

Here is my recent gamma response. It doesn't show below 10% but I have adjusted this to drop to 2.1 at 5% which helps with shadow detail (my room currently has light walls so I suffer some washout).


Edited by Kelvin1965S - 3/19/13 at 1:19pm
post #19 of 30
As Kelvin touched on above, dynamic iris must be off to accurately measure and calibrate your gamma. You will not be measuring "Gamma" if you measure with the dynamic iris engaged, whether it be windowed or full screen. Instead, you're measuring a product of "Gamma * Iris factor". This has absolutely no relation to your actual on-screen gamma. If you calibrate with dynamic iris off, and turn dynamic iris on, it will not screw up your gamma. This can be measured and verified with stepped patterns. There are a handful on this forum that believe otherwise, but are unable to provide any evidence for some strange reason...
post #20 of 30
I have 5020. I use living room setting and just changed 7500 to 6500 color etc.
post #21 of 30
Surely you want to calibrate the gamma * iris factor, so worth measuring with the iris on? I'm not sure whether the OP measured with it on or off however as he hasn't said (that I've seen).

I think we had this argument a while ago as you didn't believe that DI adjusts gamma on the fly, but there were quite a few other members who backed me up with this. (Luckily wink.gif ) I don't have a projector with DI anymore otherwise perhaps I could produce some charts to show what happens. Basically in a dark scene the iris closes down, but the projector will boost the whiter parts of the image (up to or even into clipping) to try to counteract the effect of these parts becoming dimmer. This is changing the gamma of a scene as it is altering the difference between the white and other darker parts of the image.

Basically if you calibrate with DI OFF, then turn it ON, you are not seeing a calibrated image and visa versa. One of the reasons I prefer a projector with higher native contrast to start with. wink.gif
Edited by Kelvin1965S - 3/20/13 at 12:23pm
post #22 of 30
Thread Starter 
Quote:
Originally Posted by Fat Dave View Post

As Kelvin touched on above, dynamic iris must be off to accurately measure and calibrate your gamma.
Quote:
Originally Posted by Kelvin1965S View Post

I'm not sure whether the OP measured with it on or off however as he hasn't said (that I've seen).

Sorry to burst your bubbles, but before this thread gets too off track, I did all my calibrations with dynamic iris OFF.

So, the differences in gamma curve shape between the two modes were NOT due to the dynamic iris, but seem to be a product of the unique optical/vid processing configurations of REC709 vs Cinema 2 modes. Perhaps I will try turning ON dynamic iris and remeasuring to see if that straightens out Cinema 2's gamma curve or otherwise affects it. However, I am unsure about how to capture reliable results, as I understand dynamic iris changes based on picture content. So should I be using field patterns, windows, or APL? I've been using APL patterns to do my iris OFF calibrations. But with iris ON, I'm not even sure it will make sense as normal picture content will have different picture levels than any patter anyway.

I could be convinced otherwise by wiser forum members, but right now my plan is to never engage dynamic iris during normal viewing.

Kelvin, you made a good point earlier, where if a flat curve is not possible, lower gamma at lower IRE's would preferable, as would higher gamma at higher IRE's. Unfortunately, as you pointed out my Cinema 2 curve is the opposite of that.
post #23 of 30
No bubbles burst here. wink.gif

As you aren't using DI then carry on using which ever pattern you prefer since there isn't any issue like with a Plasma screen limiting the light output. The difference in light output between your calibrations could well be due to the colour space that these settings apply which might also effect the white balance as I mentioned previously. The gamma response may well have been set up differently, though I'm not really sure what it is trying to achieve in Cinema 2 mode since it's the opposite of what many calibrators do.

Technically not using the DI gives a more accurate result, but unfortunately you're then left with around 6000:1 native contrast which is why many owners tend to use it so that they get more apparent contrast (and deeper fade to black). I know that Phil Hinton on AVF prefers not to use it when he has reviewed Panasonic AExxxx models in the past but he got tongues wagging on here because of it, but of course it's entirely your choice. Accuracy verses contrast put simply.

The upshot being your choice as to whether you need the extra light output of Cinema 2 or the better gamma response of rec709. No reason why you can't calibrate both modes and chose according to the content and/or viewing conditions. When you start measuring the colour gamut you might find that the rec709 mode is less over saturated than Cinema 2 so again will be easier to calibrate accurately. Whether you prefer this is another mater, but again you have the choice which to use.
post #24 of 30
Quote:
Originally Posted by Kelvin1965S View Post

...I think we had this argument a while ago as you didn't believe that DI adjusts gamma on the fly, but there were quite a few other members who backed me up with this. (Luckily wink.gif ) I don't have a projector with DI anymore otherwise perhaps I could produce some charts to show what happens. Basically in a dark scene the iris closes down, but the projector will boost the whiter parts of the image (up to or even into clipping) to try to counteract the effect of these parts becoming dimmer. This is changing the gamma of a scene as it is altering the difference between the white and other darker parts of the image.

Basically if you calibrate with DI OFF, then turn it ON, you are not seeing a calibrated image and visa versa. One of the reasons I prefer a projector with higher native contrast to start with. wink.gif

You can't calibrate with the DI on at all. It will NOT be accurate. This has been well established and is backed by every manufacturer.
post #25 of 30
Thread Starter 
Quote:
Originally Posted by Fat Dave View Post

You can't calibrate with the DI on at all. It will NOT be accurate. This has been well established and is backed by every manufacturer.

Didn't think so. Good to hear I have the backing of every manufacturer.

Plus, I'm pretty sure Panasonic's DI alters more than just iris aperture. It appears to make video processing/color balance changes on the fly. Trying to calibrate against that sounds like a handful of buttered worms.
post #26 of 30
Ergo, when you watch with DI on it can't be accurate either. wink.gif

Glad we got there in the end. smile.gif
post #27 of 30
Thread Starter 
Quote:
Originally Posted by Kelvin1965S View Post

Ergo, when you watch with DI on it can't be accurate either. wink.gif

Glad we got there in the end. smile.gif

I wasn't watching with DI either. smile.gif
post #28 of 30
Glad to hear it. smile.gif
post #29 of 30
Quote:
Originally Posted by Kelvin1965S View Post

Ergo, when you watch with DI on it can't be accurate either. wink.gif

Glad we got there in the end. smile.gif

brutal.

When you're displaying a single shade of grey ONLY, you cannot determine the "actual" gamma on a DI setup, because there is no reference white.

If you display for example 10% bars with a properly gamma-calibrated DI system with DI engaged, you will see that the gamma displayed is exactly what you've calibrated with DI off.

So, yes it does work. At least on the Sony line, as that's the extent of my measurements with DI.
post #30 of 30
Quote:
Originally Posted by Xank View Post

Greyscale looks good. Gamma remains non-linear, but now averages just around the target 2.2. I tried using the advanced gamma controls to straighten out the curve, but all attempts to do that ended up throwing things completely out of whack. I can shift the entire curve up or down (that's how I got gamma to average 2.2), but I cannot change it's basic shape.

I'm not sure, but you may be able to flatten your gamma by adjusting the gamma correction for overall luminance (Y). Essentially, by adjusting the individual gamma for R-G-B, you have achieved nearly perfect color balance and color temperature from 10 IRE - 100 IRE. However, that does not mean that the total light output level is increasing linearly from 10 IRE - 100 IRE.

Now that you have R-G-B dialed in for an exceptional grayscale, you may be able to tweak the 15 point gamma correction for Y to flatten the overall gamma. If successful, you will get better black detail and more contrast (less washed-out-looking) across all IRE levels. I don't know how interactive the Y gamma control is with the R-G-B gamma controls...hopefully not too interactive. If it is, you might have to sneak up on a flat gamma through an iterative process of flattening overall gamma using Y gamma correction, re-balancing R-G-B grayscale with the individual R-G-B gamma correction, and finally re-visiting overall gamma with the Y gamma correction.
New Posts  All Forums:Forum Nav:
  Return Home
AVS › AVS Forum › Display Devices › Digital Hi-End Projectors - $3,000+ USD MSRP › Are Calibration Settings Like Braces?