THE SONY SERVICE CODES - Articles, Comments, Discoveries - Page 10 - AVS Forum
Forum Jump: 
 1Likes
Reply
 
Thread Tools
post #271 of 2962 Old 07-19-2005, 04:32 PM
ADU
AVS Special Member
 
ADU's Avatar
 
Join Date: Apr 2003
Posts: 6,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 102 Post(s)
Liked: 154
After doing a bit more poking around it seems as if 6500K (or so-called "full-spectrum" or "daylight") bulbs aren't that difficult to find. The key though seems to be finding them in a convenient size/wattage, and with a high enough CRI for our purposes (namely CRT calibration). Here's a nice succinct explanation of CRI btw from Keohi.com
Quote:


The lamp should be rated at 6500 Degrees Kelvin with as high a C.R.I. rating as possible. (Color Rendering Index refers to how well a light source will "render" an object familiar. The rating is scaled from 0 to 100 with outside "daylight" being 100) The lamp used should have a rating of at least 90 CRI.

Keohi also mentions a couple other possible places to look for such bulbs. 6500k lights seem to be popular with aquarium and terrarium buffs as well, so pet & plant supply shops, in addition to home & lighting and specialty bulb shops, might be another place to try.

ADU
ADU is offline  
Sponsored Links
Advertisement
 
post #272 of 2962 Old 07-19-2005, 05:20 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Ken, you have been busy with responses todaya lot of good information.

As far as 6500K lights, if you can get close, you will probably do OK. This is just an option to get close. Just watch out for light bouncing off the walls. I did a calibration on a Sony LCD RP and the room had a creamy yellow on the walls. When the sun got in, the color analyzer jumped minus blue. It made calibration difficult.

Glen Carter
Home Theater Calibration

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
GlenC is offline  
post #273 of 2962 Old 07-20-2005, 08:00 AM
ADU
AVS Special Member
 
ADU's Avatar
 
Join Date: Apr 2003
Posts: 6,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 102 Post(s)
Liked: 154
Understood, Glen.
Quote:
Originally Posted by ADU View Post

I think there is something to be said for getting as close to ~6500k as possible for the best shadow detail.

This is pretty much exactly what I meant to say btw. What I mean/meant by this is simply that the closer the greys (and blacks) are on your TV to 6500K (or to be more precise, D65), the better the color differentiation generally should be in darker areas of the image, making details in the "shadows" easier to discern.

If you asked me why that's the case, then I'd probably be gettin a bit out of my depth, but I would take a stab and say that it may be because D65 is the color temperature/space the video content was designed for, and it may also be the native color temperature/space on the TV.

I may be wrong about this, but I suspect that arbitrarily adjusting the TV to other color temperatures would not give you the same result, and be sort of like looking at shadows through a colored/muddied haze (as billmail1's post sort of indicates). YMMV though.

Once the greyscale is adjusted to ~D65, then using backlighting or ambient light of approximately the same temperature may also help to improve the discernment of the shadow detail as well. Most 6500K bulbs are fluorescent though, and the flicker on fluorescent lights tends to give me a headache. (Dimmers have the same effect on me.) And I'm not sure if "flicker-free" fluorescents are totally free of flicker either.

ADU
ADU is offline  
post #274 of 2962 Old 07-20-2005, 02:33 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
The one key here is the color of white/gray/black is the foundation for all colors. When calibrated, R, G & B guns are set with their minimum and maximum.

A low IRE Red field would have a signal on R and none on G & B. If the black color temperature is off, at low IRE, say a little too blue, then there would also be a little additional blue in the Red, diminishing as IRE increases. Same with Magenta, there would still be too much blue until blue DRV takes over.

On the higher IRE side, if the DRV is set a little too blue, the RGB primaries would be close in color, but any color with blue in it would have just a little too much blue.

I hope this makes sense on how the calibration affects the generation of other colors. You can still have an accurate color decoder setting and an error in the color being displayed. Point being, you can calibrate to 9300K with an accurate decoder setting and Magenta will not be NTSC magenta.

Glen Carter
Home Theater Calibration

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
GlenC is offline  
post #275 of 2962 Old 07-20-2005, 06:23 PM
Member
 
loadams's Avatar
 
Join Date: Apr 2003
Location: Mount Pleasant,SC
Posts: 158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Gents, would GREATLY appreciate your imput on this procedure, tell me what I am missing.

I've decided to re-visit my grayscale settings and took a more traditional approach.

1) Put 0-100 ire ramp.
Set color and brightness to 0.
Reduce all CUTS to 0.
Reduce RDRV & BDRV to 0.
GRDV left untouched.
Set RGBS-6 (yellow) and increase RDRV until yellow. Noting where "orange"
diminishes and "lime" begins, finding a value in between.
RGBS-7, increase BDRV until high ire's look gray, white.
Record values, not writing them yet.

2) Return brightness to normal.
Return GCUT to original value.
Reduce contrast to 0.
Reduce all DRV's to 0.
Reduce RCUT & BCUT to 0.
Again, set RGBS-6 and increase RCUT to achieve yellow as described above in
step 1.
RGBS-7, increase BCUT to achieve gray. (This is for me difficult to judge)

3) Return contrast to normal.
Increase DRV parameters as recorded in step 1.
Check contrast and brightness with pluge.
Return color to normal.
Write.

Now, in my mind's eye, I have set 100 ire - 50 ire neutral, using the DRV's with no influence from the CUTS. And also, set 0-50 to neutral using the CUTS with no influence from the DRV's.

If it's a "twiddling" experiment for a rainy day, please try it. And I as I said earlier, I would greatly appreciate your imput on this procedure.

Forgive me if I'm vague, I've been riding in a car for the past 8 hours.
loadams is offline  
post #276 of 2962 Old 07-20-2005, 06:55 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by loadams View Post

Gents, would GREATLY appreciate your imput on this procedure, tell me what I am missing.

No comment!

Glen Carter
Home Theater Calibration

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
GlenC is offline  
post #277 of 2962 Old 07-20-2005, 07:14 PM
Member
 
loadams's Avatar
 
Join Date: Apr 2003
Location: Mount Pleasant,SC
Posts: 158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Gee, thanks Glen !!!
loadams is offline  
post #278 of 2962 Old 07-20-2005, 07:55 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
How do you know you have the right color of yellow? What is your reference?

We all understand that 6500K is the result of the proper balance of RGB. Cut for low IRE and DRV for High IRE and there is almost always interaction between the two.

In your method, for example, if you are a little off on yellow, say plus red, then when you try to set white, you would generally end up minus green (plus blue to compensate for plus red = minus green).

There is no way to properly set color temp without a measuring device. You can use a color analyzer or an optical comparator. An optical comparator may be direct reference to clouds, an actual 6500K optical comparator (purchased or home made) or even a B&W TV with D65 phosphor (I have a Sony PVM-96). Some B&W monitors have 5400K and 9300K phosphor, so be careful.

Glen Carter
Home Theater Calibration

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
GlenC is offline  
post #279 of 2962 Old 07-21-2005, 01:24 AM
ADU
AVS Special Member
 
ADU's Avatar
 
Join Date: Apr 2003
Posts: 6,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 102 Post(s)
Liked: 154
loadams,

I dunno. I thought I had this pretty well figured out until Glen and Ken got me all confused again so I'm gonna start over with a 6500K reference and see how that works.

What I'll be using for comparison is a 6500K fluorescent "daylight" bulb reflected off neutral white and maybe also neutral grey surfaces. The good news is that 6500K lights don't seem quite as difficult to find as I thought. The only place I had a chance to check today was Wallie-world, but they had some that seemed right in the neighborhood of what I was after-- 6500K "Sunlight" screw-in fluorescent bulbs made by Lights of America rated at 82 - 88 CRI. Those may do fine if I can't find anything better at home supply, lighting stores or elsewhere.

ADU
ADU is offline  
post #280 of 2962 Old 07-21-2005, 07:28 AM
Advanced Member
 
Reagan's Avatar
 
Join Date: Nov 2000
Location: Kentucky
Posts: 988
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Ken and Glen,

I've learned a ton from this thread. Thanks for all your contributions.

-Reagan

The truth doesn't care whether you believe it.
Reagan is offline  
post #281 of 2962 Old 07-21-2005, 10:09 AM
Member
 
loadams's Avatar
 
Join Date: Apr 2003
Location: Mount Pleasant,SC
Posts: 158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
[quote=GlenC]How do you know you have the right color of yellow? What is your reference?

We all understand that 6500K is the result of the proper balance of RGB. Cut for low IRE and DRV for High IRE and there is almost always interaction between the two.

Understand that, good point. I don't. This procedure is "eyeballing", just as all of us will do without equipment. However, in the beginning, using my homemade optical comparator, using a reference white (computer monitor @6500 at work) and Photoshop, measuring white(255-255-255), red(255-0-0), green(0-255-0), blue(0-0-255), magenta(255-0-255), and yellow(255-255-0) @ 100,90, and 20 and recording those values to transfer. Leaving GDRV alone(which the values from the monitor where VERY close) and dial in RDRV, I should have my "predicted" value for yellow. Now with RDRV set, I can use the "predicted" values for BDRV using magenta. Cycling thru RGBS and measuring for each screen and comparing their values with their "predicted" values comes close, but not perfect.

I've used the equation , written by a well known author of a well known calibration disc, to find my predicted values from reference values. AND YES, I have to give credit to UMR of this forum for dues on his eyeballing method.

The interaction is where I want to learn between the two. My whites "look" white, my dark grays "look" dark gray, it's the 40-70 ire's that have a cyan tint to them. As you have stated in an earlier thread, I want to learn where one gun cuts off and it's respective partner takes over. Probably more than I could comprehend or you would even indulge, but I should have been more clear in my first post.

Glen, you have brought some interesting points to light. I will do some more reading.

ADU, going to make the rounds today at the local depots. May go your route soon. Thanks.
loadams is offline  
post #282 of 2962 Old 07-21-2005, 11:04 AM
ADU
AVS Special Member
 
ADU's Avatar
 
Join Date: Apr 2003
Posts: 6,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 102 Post(s)
Liked: 154
If you find any other 6500K bulbs, especially with a higher CRI than the Walmart bulbs, please give a shout. Hopefully I'll have a chance to poke around a few more home and lighting shops and maybe a petstore before the weekend to see what they have. Full-spectrum 6500K lights seem to be popular with better-living/home & garden and light-therapy enthusiasts as well.

ADU
ADU is offline  
post #283 of 2962 Old 07-21-2005, 11:54 AM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by loadams View Post

The interaction is where I want to learn between the two. My whites "look" white, my dark grays "look" dark gray, it's the 40-70 ire's that have a cyan tint to them. As you have stated in an earlier thread, I want to learn where one gun cuts off and it's respective partner takes over. Probably more than I could comprehend or you would even indulge, but I should have been more clear in my first post.

OK, now think about what you just wrote (BTW, you left out Cyan (0-255-255)). If Cyan is Green and Blue and you are seeing Cyan in your gray scale, then what's the problem? ? ? ? ? ? Simple, you are minus red, right? If you add a little Red with the RCUT, you might start to see some of the 40 IRE Cyan tint go away. If you then add a little RDRV, you should see a change in the 70 IRE and down to where they interact. This error you are seeing is showing where "you" are able to see the most error in color, "you" think white is white, but the error at 70 IRE is surely carrying through to 100 IRE same with 40 IRE to 0 IRE.

You can experiment, change RDRV only and see how far down it changes the color, it may be visible at 40 IRE. If 40 IRE can be corrected with RDRV and 70 IRE doesn't go plus RED, then the interaction between CUT and DRV is at or below 40 IRE.

Glen Carter
Home Theater Calibration

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
GlenC is offline  
post #284 of 2962 Old 07-21-2005, 03:35 PM
Member
 
loadams's Avatar
 
Join Date: Apr 2003
Location: Mount Pleasant,SC
Posts: 158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Thanks Glen, I'll give it a try tonight. Guess I missed about cyan for measurement, thought once I dialed in yellow with RDRV and dialed in blue with magenta there was no need. I'm wrong on that one. Will re-visit that one also tonight. Thanks again.

ADU, will look and report back.. Today's outing was on hold, waiting on contractors.
loadams is offline  
post #285 of 2962 Old 07-21-2005, 03:36 PM - Thread Starter
Advanced Member
 
KenTech's Avatar
 
Join Date: Feb 2004
Location: Portland, OR
Posts: 714
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by ADU View Post

loadams,

What I'll be using for comparison is a 6500K fluorescent "daylight" bulb reflected off neutral white and maybe also neutral grey surfaces. The good news is that 6500K lights don't seem quite as difficult to find as I thought.

I'm skeptical of these sources and the do-it-yourself standard, until tested, because:

(1) These are consumer-lighting bulbs, and I don't trust the quaity-control on the phosphor mix. All specifications such as "6500K" have a built-in tolerance, and I'm almost willing to bet real $$ that if you buy several of these bulbs, place them side-by-side outdoors on a semi-cloudy day, and light 'em up, you will se differences. Hell, I find two bulbs from the same manufacturer bought at different times (say, Lamps of America, same form factor) don't match! If the product description claims "precision 6500K for color matching," well, there's something to be optimistic about, at least.

(2) What would you use as a reflector? All white paper of any quality has ultraviolet-conversion dyes that make them excessively blue. Chemically-untreated paper isn't white, either. I guess I would consider pigments such as titanium-dioxide or zinc oxide, or maybe some tempura whites; but tinges of yellow or some warmish color from the paint's vehicle can screw that up, too. If I were attempting to create such a product, I would test it with a colorimeter to see if I have gotten close.

(3) Any gray made up from dyes will selectively fade with time and change color; gray matte-board is like this. Only pigments are stable. A gray made from lampblack (carbon) and zinc or titanium oxides would be reasonably permanent, unless the vehicle (carrier) turns yellow with age. There are Krylon sprays that have UV-proofing qualities -- but they're very glossy.

There are so many ways that this home-brew color standard can be "contaminated" in the execution! You can always hold it up against the white-cloud standard to check for pink or green, at least.

KenTech
"We all get smart slowly."
KenTech is offline  
post #286 of 2962 Old 07-21-2005, 04:04 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Well, some may need to wait till November to see a cloud. The can be difficult at night too.

I still don't see how you can get the TV into a position where you can directly compare the clouds and the TV screen. How do you compare 20 IRE to your clouds? The room needs to have low ambient light for 20 IRE without the color of the ambient/reflected light affecting color.

As for Grey and White, here is an option http://www.kodak.com/US/en/motion/pr...4.10.8.4&lc=en
When lit with a 6500K (or close) light and depending on the amount of light, can directly be compared to different IRE levels of the screen. In a dark room, comparison to 20 or 30 IRE is possible. With brighter light, the higher IRE can be compared.

Glen Carter
Home Theater Calibration

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
GlenC is offline  
post #287 of 2962 Old 07-21-2005, 04:36 PM - Thread Starter
Advanced Member
 
KenTech's Avatar
 
Join Date: Feb 2004
Location: Portland, OR
Posts: 714
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by loadams View Post

My whites "look" white, my dark grays "look" dark gray, it's the 40-70 ire's that have a cyan tint to them.

Wait! Stop right there. You need to figuire out why that could be true, no matter *how* you arrived at that result.

If the mid-tones (40-70 IRE) have a cyan tint, and you're *sure* that the darkest tones (that actually are useful to making up a picture) are really neutral, then the intrinsic response curve (gamma) of the CRT or its drive circuitry for the RED gun does not match the curves of the other guns. The straightforward assumption made during calibration, with instruments or not, is that the curves match. For any one TV, you have no control over them.

(Please be sure first: If you have gotten away from the TV for a few moments to get another beer, say, in an evening-lighted house, the TV *will* look cyan! Your eyes have changed and will have to readjust.)

You can fudge a grayscale nonlinearity it in two ways:

(1) Ignore the almost-black color. You would tweak RCUT to make a realistic b/w picture (Color=OFF) look neutral overall, with whites and mid-tones matching. If you peer into the deepest shadows or display a really dark scene, and things look a little reddish, oh well . . . that is a good compromise because in practice your eye will not see it. If you fanatically adjust the near-blacks for perfection in a dark room, using test patterns, matching your whites perfectly, and then your mid-tones are cynaish or bluish or whatever (been there!), you *will* have screwed-up colors on your picture when you turn the color back up. The eye is very sensitive to mid-tone color! I would accept the compromise near-blacks.

(2) If the midtones are really tinted, you could try fudging one or two of the GAMx controls up one notch. My mid-tones tend a bit yellow when near-black and whites perfectly match. So I first tried knocking GAMB up one point for the GAMM=0 setting (Pro mode). That raised the mid-tone blue, leaving dark and brilliant colors alone. But on my set, it turned out to be a bit of over-correction; I have since made the GAMx settings the same, and I have reconsidered how I judge grayscale linearity, now using *real-world* video sources for judging. *That* has made me happy with the results! Selectively fudging gamma is a legitimate way to cancel out a significant nuisance-nonlinearity in one of the color-drive channels or CRT guns. I hope the need for this is rare.

************
OTHER COMMENTS: When judging b/w-image color tints by eye, several natural "bugs" get in the way. First, the eye color-fatigues quickly, and staring at the screen must be avoided -- or at least don't fixate. Move your eyes around fairly actively. Second, judging the tint of a bright white is very difficult, especially when the eye has become dim-light adapted. The dazzling white can appear reddish when it is not! One of the advantages of using midday clouds is that they are very bright themselves, and after looking at them, a glance at your screen reveals a color error pretty obviously. It's the back-and-forth that works.

My experience has now been that, when you have clean, decent whites, that's how they look watching your favorite broadcasts and DVDs, too. If the whites look *right,* check that grayscale again by turning off Color. If that looks screwed up, it is! Fix the gray scale to agree with the white. I have found consistently that, given the generous amount of bright-white that makes up many commercials on broadcast TV, especially those 4:3 pillarboxed ones on HD broadcast, you might get the impression fairly quickly that, well, they look a little pink or that they are a little "cold" relative to the rest of the picture. They probably are. So fix *that.*

A very effective method for trying out new white/grayscale settings is to mock them up first using the Warm and/or Cool settings. Remember, these are *offsets* that don't screw up your original Neutral settings. Use 2170P-1/ #14-19 to set offsets you'd like to try, remembering that 31 = no effect.

Example: You're not quite sure you've got white quite right -- looks a little blue to you. So go into service mode, set the user menu to Warm (or Cool, doesn't matter), go to 2170P-1 #14, and set everything from RDOF to BCOF to 31. (That's the same as Neutral.) Now decrease #16/BDOF from 31 to 30, and WRITE. While you're at it, you could also set the other "color temp" preset for a second experiment. Maybe decrease BDOF from 31 to 29 for a larger anti-blue change. Just set the user menu to Cool, and make the same settings as above, but BDOF will be 29. WRITE the result.

Now leave service mode, and go back to watching TV with the menu at Neutral. Still think it's too blue? Try out your new settings by switching to Warm or Cool. Screwed up? Well, your original settings for Neutral are still there, and you can try something else for Warm and Cool.

KenTech
"We all get smart slowly."
KenTech is offline  
post #288 of 2962 Old 07-21-2005, 04:57 PM - Thread Starter
Advanced Member
 
KenTech's Avatar
 
Join Date: Feb 2004
Location: Portland, OR
Posts: 714
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by GlenC View Post

Well, some may need to wait till November to see a cloud. The can be difficult at night too.

I understand this is a problem for certain areas of the US. And in LA, one can't be sure *what* color the clouds are! (Lived there once. West side.)
Quote:


I still don't see how you can get the TV into a position where you can directly compare the clouds and the TV screen.

I turn the TV about 45º, open the front door, and can easily position myself so I can look out of the front door at the sky and then back at the TV. I understand this will be very difficult for some folks. It's just that Nature gave us such a nice reference standard floating up there! So I use it.
Quote:


How do you compare 20 IRE to your clouds?

No immediate need. After getting white nearly right, I even out the grayscale on its own: It is its own standard. I can then re-check against the outdoor view. If I have moved everything toward blue, say, I can subtract one point of blue from BCUT and BDRV and re-check. Of course, there's some back and forth. "My" clouds? Oh, please! I wish.
Quote:


The room needs to have low ambient light for 20 IRE without the color of the ambient/reflected light affecting color.

So I draw the drapes and close the front door. I have to avoid the leakage of *green* from the trees and bushes just outside the windows, so I tend to tweak grayscale in the evening. I'm not under any pressure to do all of this in one go, as if I were visiting a client's home. For that, I would have a colorimeter, no question!

Glen, there an important question implicit in the above, which you can answer with your colorimeter:

Starting with a perfect grayscale and white point, does adding, say, two points to RDRV *and* two points to RCUT slide the whole red curve evenly upward? I.e. do the colors still *track* even though one of them is now displaced upward (red)? It would be good to know this because one could then make a small white-balance adjustment by adding or subtracting the *same* value from xDRV and XCUT without affecting grayscale linearity -- or at least still be reasonably close. (But maybe you don't have one of these late 34XBR960 or XS955-CRT sets.)

KenTech
"We all get smart slowly."
KenTech is offline  
post #289 of 2962 Old 07-21-2005, 05:24 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by KenTech View Post

Glen, there an important question implicit in the above, which you can answer with your colorimeter:

Starting with a perfect grayscale and white point, does adding, say, two points to RDRV *and* two points to RCUT slide the whole red curve evenly upward? I.e. do the colors still *track* even though one of them is now displaced upward (red)? It would be good to know this because one could then make a small white-balance adjustment by adding or subtracting the *same* value from xDRV and XCUT without affecting grayscale linearity -- or at least still be reasonably close. (But maybe you don't have one of these late 34XBR960 or XS955-CRT sets.)

Ken, I do not have one of these TVs. When I do a calibration, I always need to go back and forth with CUT and DRV. A change in CUT affects DRV a little and a change in DRV can affect CUT. When I have finished with the settings, I exit SM and capture readings from 10 IRE to 100 IRE and look at the x/y data to see the actual trend. I may then go back into the SM and tweak CUT or DRV by 1 or 2 if needed to get the overall gray scale more accurate. Sometimes, I have seen where a larger change in CUT (3, 4, 5, etc.) may need an opposite change in DRV by a few. Like +5 on RCUT and -2 on RDRV. I hope this helps.

Glen Carter
Home Theater Calibration

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
GlenC is offline  
post #290 of 2962 Old 07-21-2005, 05:27 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by ADU View Post

If you find any other 6500K bulbs, especially with a higher CRI than the Walmart bulbs, please give a shout. Hopefully I'll have a chance to poke around a few more home and lighting shops and maybe a petstore before the weekend to see what they have. Full-spectrum 6500K lights seem to be popular with better-living/home & garden and light-therapy enthusiasts as well.

A small pocket light like http://www.preparedness.com/pofllafl.html with a F4T5/D bulb might work fine.

Glen Carter
Home Theater Calibration

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
GlenC is offline  
post #291 of 2962 Old 07-21-2005, 05:46 PM - Thread Starter
Advanced Member
 
KenTech's Avatar
 
Join Date: Feb 2004
Location: Portland, OR
Posts: 714
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by loadams View Post

Gents, would GREATLY appreciate your imput on this procedure, tell me what I am missing.

Without actually trying it out, I would have trouble just analyzing it theoretically. And I *don't* want to mess with my existing TV's settings just now! Off the top, I think some assumptions are likely off-base:

-- That reducing anything in service mode to zero means it's actually at zero. You can't verify this. For example, you can set user Color to zero ("Min"), and it really is: you can see that it is. But those "hidden" zeroes, if they play any part in your theoretical method, may not exist.

-- "Seeing" when something is gray, yellow, white, or just tuirning green would be a nightmare of fighting the eye/mind's ability to compensate -- like trying to judge color thru the eyepiece of a video camera with auto-white-balance turned on. The eye can perceive very small differences in color when areas of color are juxtaposed, as it can differences in brightness -- we're evolved to do that! But absolutes are compensated for to an unpredictable degree -- like the color of a natural illuminant, the sun, at different times of the day.

Your method is like "starting over, clean." It seems to require more work than a good first approximation with a known standard would yield. Comparing with a cloudy day, I saw that my TV was much too blue, with a touch of green. So the first thing I did was jack red a lot and reduce blue a bit. But maybe this was easy for me, as I know my color theory from years of color-darkroom work, consulting for graphic artists, and diddling in Photoshop.

So, sorry, I don't really want to try to verify the details of your method. My bottom line would, of course, be: If you did this, and what you end up seeing is neutral whites that aren't way off-base, *and* a linear grayscale (b/w pictures are color-free), you have arrived! Precisely calibrate your white point to 6500K if you wish, but you should be able to see when results are acceptable.

I would add: If you set the xCUT parameters so you have a good black level with SBRT at 31 and the Brightness slider at 31, you are probably right in he middle of Sony's intended range. SBRT interacts with all of the xCUT settings, of course, acting like a "pedestal" for all of them, and Brightness varies *that.*

Also, the xDRV controls are uniformly modified by the Picture slider--call it a common video-gain control. My TV seemed too dim with Picture at 31, so the first thing I did was increase the *average* setting of the three xDRV controls until I was satisfied. Now HD seems perfect at 31, DVD a little higher, and the relatively dimmer SD broadcasts require Picture to be near 40. Auto-brightness limiting comes into play for those almost all-white screens of many commercials and for some bright animated shows.

Current settings for my 36XS955:
SBRT---RDRV-GDRV-BDRV---RCUT-GCUT-BCUT = 31---41-27-20---51-20-15 [rev 7.22.05]
with offsets in 2170P-3/UBOF of 0 to 4, equalizing the inputs and tuner. (The out-of-the-box RDRV was in the low 30s. Rediculous!)

KenTech
"We all get smart slowly."
KenTech is offline  
post #292 of 2962 Old 07-22-2005, 09:13 AM
Member
 
loadams's Avatar
 
Join Date: Apr 2003
Location: Mount Pleasant,SC
Posts: 158
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Okay.......

Glen, by your instructions, after diddling, I raised RCUT 3 clicks and RDRV 2 clicks. 40 & 50 ire windows look much better, as do 60 and 70 windows. Crossed gray scale ramp yields better uniformity. After I found where I wanted to go with this, I reverted back to the old values for comparison. Back and forth and such, new RDRV and RCUT values look much better. I do now see where the interaction begins and ends. It only took 2 clicks of RDRV to begin to see the cyan tint disappear from 70 ire without compromising 100 ire. Better results, and yes, whites do "look" better. I know it's not perfect, but now better. Thanks.

Ken, understand your position. I thought as long as we where "twiddling" forum hobbyists, it was something to do. Tried the GAMR curve with my old settings, will keep that in mind for future reference.

And yes, I find it helpful never to look directly at the screen when calibrating grayscale and color decoder. I rotate my eyes around the cabinet of the tv.
loadams is offline  
post #293 of 2962 Old 07-22-2005, 10:48 AM
ADU
AVS Special Member
 
ADU's Avatar
 
Join Date: Apr 2003
Posts: 6,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 102 Post(s)
Liked: 154
Quote:
Originally Posted by KenTech View Post

I'm skeptical of these sources and the do-it-yourself standard, until tested, because...

...There are so many ways that this home-brew color standard can be "contaminated" in the execution! You can always hold it up against the white-cloud standard to check for pink or green, at least.

Understood Ken. That's one reason I'm trying to find some other sample 6500K lights with higher CRIs, to see if there's any discernable difference between them. We're on the same page re color contamination as well.

A color measuring instrument is undoubtedly the most reliable approach, but this method is probably not that different than tools professional photographers and cinematographers use to judge their imagery. It's such an obvious solution in fact, I'm surprised it didn't occur to me sooner because I've had some level of awareness of the 6500K fluorescent lights for quite awhile. Never considered using them as a calibration aid though til the subject of LEDs came up. I agree with you though that until this approach is confirmed by instrument (or cloud) that it should be regarded as "experimental".

ADU
ADU is offline  
post #294 of 2962 Old 07-22-2005, 11:09 AM
ADU
AVS Special Member
 
ADU's Avatar
 
Join Date: Apr 2003
Posts: 6,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 102 Post(s)
Liked: 154
Quote:
Originally Posted by GlenC View Post

A small pocket light like http://www.preparedness.com/pofllafl.html with a F4T5/D bulb might work fine.

So basically a miniature portable 6500K flourescent light. Not bad, Glen. My only hesitation on something like this is the slightly higher price (since this is still experimental) and perhaps the 76 CRI rating on the bulb. If possible, I'd like to start with something around a CRI of 90 or higher for comparison. The size and portability certainly look convenient though. And something like that might work just dandy. I wonder if there are other compact "daylight" lights like this that professional photographers might use to look at prints?

ADU
ADU is offline  
post #295 of 2962 Old 07-22-2005, 12:09 PM
Advanced Member
 
Reagan's Avatar
 
Join Date: Nov 2000
Location: Kentucky
Posts: 988
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quick question: Is there a way to exit the service menu without turning the TV off?

Thanks,
Reagan

The truth doesn't care whether you believe it.
Reagan is offline  
post #296 of 2962 Old 07-22-2005, 12:21 PM - Thread Starter
Advanced Member
 
KenTech's Avatar
 
Join Date: Feb 2004
Location: Portland, OR
Posts: 714
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by Reagan View Post

Is there a way to exit the service menu without turning the TV off?

Nope, not that I have found. And Sony doesn't list anything in the service manual, either.

KenTech
"We all get smart slowly."
KenTech is offline  
post #297 of 2962 Old 07-23-2005, 01:03 AM
ADU
AVS Special Member
 
ADU's Avatar
 
Join Date: Apr 2003
Posts: 6,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 102 Post(s)
Liked: 154
Dunno if this'll help any, but if you're experimenting with different values and just want to recall the settings last written to memory, that can be done with 0 - Enter. So you don't have to keep turning the TV off just to get back to that stage.

This command is not recommended for butterfingers though, because the JUMP, 7 & 9 keys are all in the same vicinity and if you accidentally hit those instead, you could end up in some deep s***. You have been warned.

ADU
ADU is offline  
post #298 of 2962 Old 07-23-2005, 10:19 AM
 
CrocHunter's Avatar
 
Join Date: Jun 2004
Location: Massachusetts
Posts: 2,570
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Kentech: can you go into more detail in the MIDE section of the 17 different values.

I'm very interested and will experiment with them later on today.I will use actual images and video to see if there is any visible diference in PQ.
CrocHunter is offline  
post #299 of 2962 Old 07-23-2005, 10:30 AM - Thread Starter
Advanced Member
 
KenTech's Avatar
 
Join Date: Feb 2004
Location: Portland, OR
Posts: 714
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by CrocHunter
Kentech: can you go into more detail in the MIDE section of the 17 different values. I'm very interested and will experiment with them later on today.I will use actual images and video to see if there is any visible diference in PQ.
Sure. But I've already done that in article #12, Post #41 in this thread. Attached is the text of the article plus a chart of the various codes and how you can experiment with them. Good Luck! In the weeks since I wrote this, I haven't changed my mind on any of it -- maybe tinkered a bit with specific edge-enhancement values, but nothing earth-shaking.

 

image_processing.pdf 42.5224609375k . file

KenTech
"We all get smart slowly."
KenTech is offline  
post #300 of 2962 Old 07-23-2005, 10:37 AM
 
CrocHunter's Avatar
 
Join Date: Jun 2004
Location: Massachusetts
Posts: 2,570
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I already printed that chart a while back, just wanted to know if you knew any more discoveries that's all.

Did you use HQ images and video material to see of any differences?Off the top of your head what would you say is the most important to tinker with on that chart to give me the best PQ.
CrocHunter is offline  
Reply Direct View (single tube) CRT Displays

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off