AVS Forum banner

Status
Not open for further replies.
1 - 20 of 30 Posts

·
Registered
Joined
·
2,151 Posts
Discussion Starter #1
The long awaited day had come. I put the wife and kids to bed, locked the dogs out, popped off a couple of Molson Goldens and got to it.


I mounted the sensor 7 ft away on a full zoomed image, dead center. I did not leave the Panamorph in place. Looking into the glassy beast while the Pj was/is on, I can see light reflections off the prisms that I figured would only hinder the process.


My serial cable is 40 ft of Cat3 Home Depot 5 pair wire. Pretty cheapo stuff really. USB cable from sensor with a $20.00 10 ft Staples USB extension.


Radeon VE.


I'm elated to say that the precalibration and calibration wizards work perfectly flawless. Not one hitch, glitch or niche. ;)


The whole process took about an hour. That was two pre's and a final. The final showed a time of 18 minutes.


Some things I noticed during the calibrations:


1. My final contrast ratio given was 228 vs an original of 442. I saw that karos had made mention of that too. Judging by the image after calibration, I'd say that had to be a mistake somehow. Mark?


2. I did my cal. in 800 x 600, with all settings at 0. When I got to the "put the black and white boxes into their appropriate places using the sliders" section, the right ones were actually black but the left ones were green and red. No biggie. What I thought was odd was that the proper alignment of these square pegs in the round hole would change if I switched to a 1360 x 1024 resolution.


"Why on Earth would you change resolutions right in the middle of your long awaited soooo very important calibration?", you might be thinking. "Boy, what an idiot. You could have lost the whole thing."


Well, I had to. In 800 x 600, I couldn't see all of the wizard. It was larger than my desktop area, so I had to go to 1360 x 1024 to see the upper bars. I did go back to 800 x 600 before clicking OK on that screen, though.


Bottom line. I should have started in the 1360 x 1024. Mark, should there be a difference between the two?


3. The sensor casts a shadow on the image, sometimes right in a spot you really need to see, so be prepared for that. VE to the rescue again. :D All I had to do is switch to my "clone mode" via Pronto to see what I couldn't on the screen(like the sliders), make the adjustment, then switch it off. I made sure my 2nd monitor was off when any "sensing"was being done.


4. If I could have, I would have done this calibration in a black room. The light spill on my white walls around my G11 lens must have some affect on the final readings Dilard sends to the projector. I could see where doing this in a darker colored room could possibly help. Of course, this is just speculation since I have nothing to compare to.


Ok, on to the end result.


First, I have two caveats:


1. The calibration gave me banding in the image where I never had before. The wizard actually mentions that you should run the Eight Bar, Pixel Align, Tracking and Pixel Wizards after the cal., so I guess this is to be expected. The problem is I can't get rid of the banding, mostly because I can't see very well the bottom portions of the wizard. They're too dark. I've tried turning up the brightness and contrast all the way up, but it wasn't enough. Mark or anyone, is there a trick to being able to see the lower portions?


I've tried til' I was Red, Green and Blue in the face. :eek:


2. I have what looks like a very faint "hot spot" in the lower, center part of the screen. Hard to describe, but it's a small area that is slightly lighter in shade than the rest of the desktop. I figure this may have something to do with my white wall light spill or maybe the 32 point tests?


That's all the bad. Now for the good.


Here's some of the thoughts that were going thru my head while watching the only DVD I've seen a billion times(Superbit version):


"The Manuchean(sp?) ship really does have texture to it."


"Space really is the final frontier. Look at all the stars."


"WOW, my DILA finally does grey." My biggest letdown of my G11 has always been the way it did greys. Now, they really are grey. Yipeeee!!


"The Generals coat in the reproduction scene has a beautiful two tone shading to it that I've not seen before."


"Leloo's thermal bandages could've been just a weeeeeee bit thinner."


"Hmmmm, I think I'll get me another Molson Golden."


"The lapel! THERE IT IS." The lapel on the princess babes dad's dark jacket in the Diva scene has never been able to be seen distinctly without me radically turning up my brightness. At long last, I can see it!!


"Hey, who turned on my Hi Def receiver!!?"


"That hot spot on the Diva's belly isn't really a bunch of sizzling pixels after all. It's the reflection of the stage lights. Cool." :)


"The detail in the upper left area of space in the same Diva scene now shows the detail it deserves." If you look at this scene on a good CRT monitor, you'll notice a huge amount of detail in the reflection off the glass, with a great amount of stars in the background.


Those are the three crush tests I use most often for contrast and blacks; the ship, the lapel and the reflection off the glass in the Diva scene, and the calibration passed with flying colors, I'm happy to say.


If you're looking for a bunch of numbers, charts, graphs, specs., I can't be of much help. I'm a "grab a couple brews, run it through it's motions and sit back and(hopefully) enjoy" kind of guy. It's hard enough locking everyone in their rooms to do these kind of things as it is. :D


Besides, I wouldn't know an IRE from the IRS. :eek:


It was late when I finished, so I didn't have a chance to look at much more than those few scenes, but I can honestly say that the improvements were dramatic, even with my two issues mentioned above. I hope to fix those.


Bottom line, IMO:


The Dilard 2.2/Colorimeter really rocks!! Good luck to all!


Chris
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #2
Vince,


What's troubling you, my friend?
 

·
Registered
Joined
·
611 Posts
Chris,


Nothing really. If I gave offense, I apologize. I will delete the post.


Vince
 

·
Registered
Joined
·
574 Posts
Chris,


Thanks for posting your findings, I appreciate it as I await my sensor. I now need to read it one more time to pick up the details, though it looks as though you were pretty happy with the change.


Sounds stupid, but what would you rate your D-ILA before and after.


Thanks, that was a good read.


Brian
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #5
No man, don't delete it(Oops, too late).


I didn't take offense at all. I just didn't understand what you meant.


I calls 'em as I sees 'em.


I see true grey for the first time on my G11. Heck yeah, I'm excited. I'm overjoyed. I'm ecstatic. I'm gonna go me a Molson Golden. Over excited? Maybe............sorry.


Inre to the "in environment" calibration, heck, why not just put mirrors on the walls to bounce back some light at the sensor? Really, my walls are white. Bright white. "You're painting those walls grey over your divorced body" white. :D


I think the sensor should only see the light from the projector, not light being bounced back into it from my walls. Just my opinion of course.


Chris
 

·
Registered
Joined
·
574 Posts
Chris,


Also, the banding and the hot spot - are they noticable during DVD playback or only when studying grayscale images? Also, has Mark come up with a suggestion to fixing it? Please post if he has.


Thanks, Brian
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #7
Brian,


It's hard for me to put a before and after rating on stuff like this. What I see may not be at all what you see.


To say it's night and day would inaccurate. I've spent hours and hours with the previous version of Dilard to where I had my G11 the best it could be(to me) until the 2.2/Colorimeter release came, or until I sent it to someone to manually calibrate it, so for me the Dilard cal just took me to that next plateau.


The accurate(or at least more accurate) greyscale alone is huge to me. Worth the price of admission, IMO. I tell you true when I say that my greys were always either green, pink or blue tinted. Maybe other DILAs were/are better than mine. I could never get "Roy Orbison's Black and White" to look good and now, what a difference. That difference is night and day.


The hot spot is not so much a hot spot, I think, now that I've looked at it a little longer today. It's more of a pinkish spot. It has to be a still, white background to see it, so no, it's not at all visible on any moving material. I think(I hope) this may have something to do with my blue shading problem. My projector goes in to JVC Monday, so we'll see.


The banding is bad. Unfortuately, I don't have much experience adjusting this in the wizard, so I'm having a hard time with it. The banding is visible on everything. Hopefully Mark or someone can help here. I can't imagine anyone not seeing this after the calibration. I hope I'm wrong, though. for your sake. It's tough to fix.


Chris
 

·
Registered
Joined
·
4,525 Posts
Hi Chris,


I am glad that you are seeing some definite improvements with your picture. As with a couple of other folks above, I would like to see the results be even better.


I haven't seen the banding, but it doesn't sound good. Also, I am concerned about the contrast that you (and a few others) are getting. I would like to see more 5's, 6's and 7's in the first digit, and I have on our test projectors.


I am glad that you didn't have any technical issues in running the Wizards, but I think that we can begin to narrow down the environmental factors that create the best calibration. Once you get there (and back it up), you can put the system away for a while and enjoy!


I have already started work on Phil's suggestion of using the Overlay for the calibration instead of the desktop, and am considering exposing some of the basic assumptions that the Wizards make.


There is certainly a balance between making everything but the kitchen sink configurable and making the system easy to use, but if we are not getting the results that I (and others) have been seeing during testing, then maybe exposing those settings is a good alternative.


Make no mistake...I will continue to work on it until you are seeing the glorious picture that I am seeing...and banding is not part of that.
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #9
And I know you will, Mark.


Well, I just got rid of my banding. How? I had to restore using my backup of precalibration. It worked perfectly. Now I'm back to that ol' greenish image. :(


I think I'll try this again. This time in 1360 x 1024 and with some black sheets on the walls around the lens. It'll be interesting to see if the contrast ratio improves.


Hey, if the contrast ratio is looking better, but showing worse in the message box, what does that mean? Is my CR really lower now?


Chris
 

·
Registered
Joined
·
4,525 Posts
Quote:
Hey, if the contrast ratio is looking better, but showing worse in the message box, what does that mean? Is my CR really lower now?
Good question.


Since the initial contrast is taken with the precalibration wizard (and stored) and the final contrast is taken with the calibration wizard, let me make sure that they are doing it exactly the same way (same exposure times, techniques, etc.).


Trust your eyes first....:D
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #11
I've managed to get the before and after graphs uploaded to Yahoo. Very interesting. If I remember correctly, this is somewhat how it was supposed to go with the curve now sloping upward quicker than with a stock DILA.


Click here .
 

·
Registered
Joined
·
1,286 Posts
Chris:


Your yahoo link doesn't work - I get an error: "You have reached a restricted area."


So, if you were to characterize the change, would you say that the color balance is improved (better grays), the gamma is (better/worse? - I would say worse, since you have banding...), and the contrast is (better/worse? Do you trust the post calibration number?)


- Dave
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #13
The link should work now. I had the pictures in "private" mode.


Dave, the color balance and contrast is definitely better. I have the luxury of being able to directly compare the G11 image to my 17" monitor by a quick Pronto key for the VE, so I can quickly see what areas the G11 suffers in. As I mentioned above, my G11 has never done greys properly until now. As for contrast, I've always had to crank up the contrast and/or brightness on the G11 to show all the detail in a dark scene. Obviously, that ruined the picture for other standards, so I was content in not being able to see all there was to see.


The Diva scene is a perfect test for this. If you watch the upper left area, it will show the reflection of the audience and hall tiers. Behind that through the glass is the outer space loaded with stars. I could never bring all that detail into a realistic image without blowing out the brightness. On my CRT, it's very vivid. Now, after calibration, I can clearly see the distinction between the reflection in the glass and the outer stars. Much improved. Still not as good as on my CRT monitor, but I didn't expect it to be. It's 17" compared to 7 ft.


The actual blacks don't seem to be any blacker, but the detail in the blacks is vastly improved. After seeing this improvement, I reall wonder how the "after" contrast could be lower, so as of now, no, I don't trust the numbers.


The gamma charts indicate that the curve has improved by quite a bit and by my(still limited) time watching post cal., I have to agree. The banding does indeed make this hard to judge, though.


Tonight if I can clear the house ;) , I'll do another cal., this time in 1360 x 1024. I have a feeling that me being in 800 x 600 caused this banding. While in anything but 1360, I think resize is always on and I'm hoping this may be why I got the banding.


Chris
 

·
Registered
Joined
·
4,525 Posts
Yikes, Chris!

I think we've solved the banding!...are you using Dilard 2.23 ?


If you look at the low-end of your gamma, you will see a split in what should be a smooth line. The values for the first 1/16 of the gamma are exactly 1/2 what they should be.


I never thought that the version of the software with this issue was uploaded, but you seem to have gotten a copy of it. This problem is not in the current released version.
 

·
Registered
Joined
·
4,525 Posts
Also,


I verified that the precal and cal wizards take the contrast measurement identically. The numbers as reported are correct. That doesn't mean that they can't be improved, however.
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #16
I'll have to look at the version again when I get home, but it was an earlier version, not the latest


I'll load the latest and try it again tonight.


Yes, I saw the break in the graph but didn't know what to make of it. Thanks for catching this and the explanation.


I may have to take a half day off now. :D
 

·
Registered
Joined
·
574 Posts
Good eye on the graph Mark. I loaded the postcal file and was posting to ask Chris whether something was messed up with his image, because it seemed to be missing one column, until I counted them to be correct. Good thing he posted the images.


I can't wait to give this a try myself... hopefully soon.


Nice writeups Chris.


Brian
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #18
Thanks Brian.


I downloaded the most recent Dilard version today and did the calibration again tonight. This time I was in 1360 x 1024 and I put a black sheet over my screen to diffuse the reflected light.


Oddly, it didn't make bit of difference.


My first cal CRs were 442 pre and 283 post.

My second CRs are 442 pre and 275 post.


I just lost 8 points. Drats!


And the biggie...yes, the banding is still there. Double drats!!


If I could just find a way to see the Eight Bar Wizards lower IRE fields, I might be able to get rid of it. Also, in the wizard, when I move the circle over the black line to make the gap in the circle disappear, it looks like the gap is just a bit shy from being completely covered.


Any ideas?


Chris
 

·
Registered
Joined
·
806 Posts
Chris,


I had the same problem. The lower screen on the 8 bar was almost black. Entirely unusable the way it is. Somewhere I had heard that the low end adjustment should be and 20 IRE and the high end 80 IRE but I might have just been dreaming.


Mark,


Any ideas?


Phil
 

·
Registered
Joined
·
2,151 Posts
Discussion Starter #20
Thanks Phil.


I guess I'm glad to hear I'm not the only one. :)


If this is a common problem, I'm sure Mark will have a fix PDQ.


Turning up the brightness helps somewhat to see the lower portion, but not enough. I worked on it a little more last night and got it a little better, but not 100%.


My biggest question now would have to be will the Eight Bar adjustment get rid of it or do we need another software revision?


Chris
 
1 - 20 of 30 Posts
Status
Not open for further replies.
Top