I've been perfectly satisfied with the settings I've adjusted thanks to my DVE. I have everything adjusted perfectly except I'm clueless about my contrast.

The vertical line doesn't bend and the top white box never really blooms on my direct-view Sony XBR970! I've tried the THX Optimizer, but even at "max" setting, I can distinguish among the white boxes at my highest contrast setting. The Avia disc is no help either. I know many recommend not cranking your TV to the highest setting of contrast, so I've tried to keep it around 70 (out of 100), but I notice the picture has more WOW factor when up higher.

Should I just have my setting at max? If it's a flaw in my TV and my contrast is already lower than normal, it shouldn't really do much damage, right?

More than likely your problem is that you have a digital display, where DVE and AVIA where designed primarily on some of their patterns for CRT..

A way to adjust contrast is to bring up a 100% (or 100IRE) pattern and lower it all the way and bring up, where the color may shift to another color except white (blue or pink)..reduce it until that shift disappears..another way is to bring up a 100%(white) pattern and look at it for a minute, then immediately shift to 0%(black) then right back to 100% and for a mear less than a second you might see a pink color, keep lowering it down until the colorshift goes (no longer see the pink color)..this is a little harder

I've got a CRT based display (RPTV) and I had no clue how to adjust my contrast (first with the THX Optimized and, more recently, with the DVE patterns). I can turn it up to 100% and the damned white rectangle never changed shape or grew larger. So I just kept it at 50%.

My friend had told me that Avia had a different pattern for contrast adjustment, and that it might work form -- sounds like it will not:(

Where could I find the white/black pattern?

With CRT, what you can do is bring it up all the way up, then lower it until right before it changes to grey..on 100% white

Greetings

On a CRT ... optimal contrast is determined with a 100% white windowbox pattern and a light meter. Increase contrast to get a 30 fL reading from the box.

Make sure there is no blooming.

You are set.

Test discs have contrast patterns that help you to determine where not to set the contrast ... not where to set the contrast.

Aim for the 30 fL reading for a CRT ...

Regards

Greetings

On a CRT ... optimal contrast is determined with a 100% white windowbox pattern and a light meter. Increase contrast to get a 30 fL reading from the box.

Make sure there is no blooming.

You are set.

Test discs have contrast patterns that help you to determine where not to set the contrast ... not where to set the contrast.

Aim for the 30 fL reading for a CRT ...

Regards

Michael,
What is your recommendation for RPTV's, specifically the Sony SXRD? I'm assuming that these values are for a room with moderate lighting? Thanks much for your input.

Greetings

The previous recommendations were for CRT technology. SXRD is LCOS is digital. Rules for contrast are different.

1. No white clipping.
2. When 1 achieved ... no discoloration ... (pinking of the whites)
3. Eye fatigue factors ... take down further if it still causes eye fatigue.

Regards

Greetings

The previous recommendations were for CRT technology. SXRD is LCOS is digital. Rules for contrast are different.

1. No white clipping.
2. When 1 achieved ... no discoloration ... (pinking of the whites)
3. Eye fatigue factors ... take down further if it still causes eye fatigue.

Regards

Suppose all 3 of the above can be achieved at max contrast. Are there any other factors that should be taken into account, or should max contrast be used if all 3 of the above can be achieved?

Greetings

Then MAx contrast it is for max dynamic range ... aka max contrast ratio ...

Regards

Greetings

Then MAx contrast it is for max dynamic range ... aka max contrast ratio ...

Regards

Thanks Michael. I do have one more question (I apologize for somewhat hijacking the thread, but it looks like the OP has already got an answer).

This is in the "purely theoretical" realm and is more academic than anything else, but it has been on my mind for a while now.

High contrast ratio seems to be a goal for all displays, with the limiting factors you have described thus far being the obstacles in the way, as well as the natural limiting factors inherent with a given display technology. For the sake of discussion, lets take a theoretical display that does not have any of the limiting factors you have described, and has an extremely high contrast ratio when contrast is set to Max. Lets say the light output for this theoretical set at Max contrast is 300 fL.

Now, with an 8 bit set we have 256 "shades" of gray (actually less than that using video levels) to represent every level of luminance from 0 fL to 300 fL... meaning that each integer increment from 0-255 is a relatively significant jump in light output. Our max light output (300 fL) is also very large. Now for some questions:

1) Consider the jump between digital levels (say 126 to 127) produces a relatively large jump in light output, could this cause a banding effect?

2) Let's assume the material we are viewing includes an object that the camera recorded as maximum white (235). Is it possible that this object is displayed by our TV as many times brighter than the actual object that was recorded because our contrast ratio is so high?

The first 2 questions pertain to our hypothetical display and really ask the question "can contrast ratio be too high?" Now we know that real world displays don't produce those kinds of levels, so my third question follows for a typical modern digital display (such as LCoS):

3) Is it at all possible or feasible to set a real world display so that the white it displays is the same luminance as the real white object that the camera captured? In other words, set the display so that if I were to look at the real object, then look at the object as displayed on screen, they would have the same luminance. Would this be desirable?

I know this is a bit academic, but I have often wondered if it is possible to set contrast too high, such that objects are displayed with a much higher luminance than they would have in real life.

Thanks for any input anyone can provide.

Greetings

In real life ... there are plenty of cases where we use our hand to blot out the surrounding light so that we can see something better.

We are effectively lowering that contrast ratio because our eyes can't handle it. This falls into the fatigue part ... so it can't happen on a display .... cuz if it does, then one didn't listen about the three items to setting contrast correctly.

I really can't answer your other stuff though ... sorry. Too academic for me. Something to ask Charles Poynton ... not a guy that teaches calibration/calibrates TVs.

Regards

Greetings

In real life ... there are plenty of cases where we use our hand to blot out the surrounding light so that we can see something better.

We are effectively lowering that contrast ratio because our eyes can't handle it. This falls into the fatigue part ... so it can't happen on a display .... cuz if it does, then one didn't listen about the three items to setting contrast correctly.

I really can't answer your other stuff though ... sorry. Too academic for me. Something to ask Charles Poynton ... not a guy that teaches calibration/calibrates TVs.

Regards

Thanks anyway Michael, I sincerely appreciate the input.

Thanks for all the suggestions, I'm doing what I can to follow them.

Greetings

Rough rule of thumb for CRT contrast without instrumentation is ... look at the same 100% windowbox and if the image hurts your eyes to look at it ... then the contrast is still too high.

Regards

Suppose all 3 of the above can be achieved at max contrast. Are there any other factors that should be taken into account, or should max contrast be used if all 3 of the above can be achieved?

I've got an XBR2 SXRD and I run contrast at MAX (and Iris at Auto1). And I've verified with a Spyder2 that I get no clipping of red, green or blue. I don't get eye fatigue and the overall brightness of the picture seems perfect to me. Only when I (rarely) watch TV in a dark room does the picture sometimes seem a little bright.

Only when I (rarely) watch TV in a dark room does the picture sometimes seem a little bright.

This is why the ISF recommends both a "day" and "night" mode be included in a display calibration service. You should consider turning down your contrast and/or using bias lighting for dark room viewing conditions. Your TV will never deliver its best picture in elevated ambient lighting. This is due to the three types of screen contamination from ambient lighting: Lambertian, specular, and haze.

You can find a wonderful graphical depiction and explanation from the National Institute of Standards and Technologies' (NIST) Flat Panel Display Laboratory. This can be downloaded from: http://www.fpdl.nist.gov/tips.html .

Best regards and beautiful pictures,
G. Alan Brown, President
CinemaQuest, Inc.

"Advancing the art and science of electronic imaging"

On a digital display that does not exibit color shift or clipping on maxed out contrast settings the image can take on a funky/digital look. What I have found to work quite well is to put up a gray ramp, turn contrast all the way down, then raise it slowley untill the white sector of the ramp is... well white compared to the next graduation of gray and also that it follows equally the luminence steps/pattern of the rest of the ramp. If the white sector jumps out at you when looking at the ramp its probably too high, if its indistinguishable from the gray next to it then its too low.


Kevin

Interesting. This seems to be a common issue with XBRs as I am struggling with contrast on my A2020 as well.

I see a bit of color shift somewhere between 90 and 94. For now I have set is in the high 80s.

Kevin. I have tried adjusting contrast between 50 and 100 (max) and find it really hard to tell when white is white. I can see a difference between the 100% bar and the first below white bar even at 50.

Good luck with your adjustments Voyeur. Its not easy.

Fred

I read a link that explained how to determine the proper luminance. It involved a piece of paper, a lighter (or any flame source, I suppose) and a dark room. The paper is held at arms length in front of the TV and will allow you to set the contrast to make the paper and TV have roughly the same luminance.

Anyway, I'm not sure if any of that makes sense, but the bottom line is that the link recommended a luminance of ~15 fL (I think that was the unit of measure) in a dark room. So I checked my TV and I'm using 19 (at a contrast of Max with good greyscale, i.e., no color clipping). Most of my viewing is in a well lit room. So is my measurement of 19 ok? I assumed that higher than 15 is ok for a lit room, but I don't know how much higher. Is there a set of recommendations for this? Also, is there perhaps recommendations for luminance across the greyscale, i.e., from 0 to 100 IRE? Or if I set the 100 IRE luminance correctly, and have the right gamma, then does that imply that my luminance is correct? It seems easier to play around with settings to hit a good luminance value rather try to hit a certain gamma curve.

I've spent a lot of time calibrating my XBR2 with Spyder, HCFR and Getgrey, and the picture looks great, but I only understand maybe 80% of what I'm doing - maybe not that high. I really appreciate the information shared by the experts on this forum. Thanks!

Greetings

If 100% white is standing out... then you have violated step number 2 for setting contrast. Follow the three steps.

regards

Greetings

If 100% white is standing out... then you have violated step number 2 for setting contrast. Follow the three steps.

regards

I don't think it's standing out, but I don't know if I have a properly trained eye to determine such things. If it was, would I see a spike at the end of the luminance curve I use to measure gamma?

Greetings

The Monster ISF disc has an excellent human contrast pattern to work with. Easy to see the loss of detail in the person ... and easy to see discoloration in the shirt as well.

I'd stop trusting instruments and use your eyes when it comes to setting contrast.

Regards

Interesting. This seems to be a common issue with XBRs as I am struggling with contrast on my A2020 as well.

I see a bit of color shift somewhere between 90 and 94. For now I have set is in the high 80s.

Kevin. I have tried adjusting contrast between 50 and 100 (max) and find it really hard to tell when white is white. I can see a difference between the 100% bar and the first below white bar even at 50.

Good luck with your adjustments Voyeur. Its not easy.

Fred

There will be a difference in luminence but both should be gray in look. Raise it one step at a time untill 105% looks white compared to 100%. Then look at the steps from black to white, see how each step is lighter than the next? Follow this luminence pattern and each step looks a certin percent brighter than the next, when you get to the 105% white sector it should look like it fits in with the rest of the pattern just 5% brighter than the previous sector and so on. On DVE the background above and below the ramp is shaded and not stepped, use the shading as a guide as well. As it enters the white reigon it should look very even and gradual, if there is a distinct jump or step to white back down contrast one step at a time untill it blends in.

Good luck

Kevin

Greetings

The Monster ISF disc has an excellent human contrast pattern to work with. Easy to see the loss of detail in the person ... and easy to see discoloration in the shirt as well.

I'd stop trusting instruments and use your eyes when it comes to setting contrast.

Regards

I've heard good things about the Monster disc. I'll check into acquiring one.

Thanks for your help! :-)

Isn't there another way to set the contrast on a digital display by looking at two greyscale color ramps and adjusting the contrast to the point that you can clearly dilineate the different shades in each ramp?

I believe this is mentioned here, on Step 2 (Contrast: Reverse Gray ramps and steps)

http://www.avical.com/articles/avicals_dve_user-level_video_calibration_tutorial.html

This of course assumes that the user has the DVE disc.

Greetings

Yes, but the ramp as you talk about only addresses the first point in setting contrast. What about the other two aspects?

It's a 3 step process ...

Regards

What are the other test patterns which can be used for the other 2 aspects?

Greetings

See above ... it is discussed here.

Regards

Hey Michael,

Thanks for all the help.. Can you recommend another calibration disc I can compliment my DVE disc with to help nail down the settings the best? If there is a free one that would be great but not entirely necessary.

EDIT: Nevermind, the AVS disc looks fine.

I've been perfectly satisfied with the settings I've adjusted thanks to my DVE. I have everything adjusted perfectly except I'm clueless about my contrast.

The vertical line doesn't bend and the top white box never really blooms on my direct-view Sony XBR970! I've tried the THX Optimizer, but even at "max" setting, I can distinguish among the white boxes at my highest contrast setting. The Avia disc is no help either. I know many recommend not cranking your TV to the highest setting of contrast, so I've tried to keep it around 70 (out of 100), but I notice the picture has more WOW factor when up higher.

Should I just have my setting at max? If it's a flaw in my TV and my contrast is already lower than normal, it shouldn't really do much damage, right?

I'll keep it simple: The maximum user setting for Contrast/White Level bears little to no resemblence to how things look in real life - whether viewed on Front-Proj, Rear-proj, FPD or CRT(direct view) technology. There's no reason contrast belongs maxxed - or even within 80% of max, unless the viewing location is on the surface of the SUN - Period.

Actually, there is a reason contrast is set to max at the factory: to sell TVs.

Contrast should be set with a light-output meter reading off a 100 IRE white pattern being displayed on the set in question. Video should give a reading of 30 ft Lamberts, or for a more cinema-like experience, aim for 12-16 ft Lamberts. The folks on here who actually use light meters to measure light output can provide the conversions to ftL . THAT is the surest, most fool-proof way to set Contrast/Picture/White Level. You may be surprised at just how low the control will end up being set at to achieve either level - for a direct-view/traditional TV, the setting may be no more than 20-40%! For a flat-panel-display, not much higher than that, depending on ambient lighting and what your viewing habits are.

(Gosh, I'm beginning to sound like Imus does on autism or cattle futures!!) :D

regards,

I'll keep it simple: The maximum user setting for Contrast/White Level bears little to no resemblence to how things look in real life - whether viewed on Front-Proj, Rear-proj, FPD or CRT(direct view) technology. There's no reason contrast belongs maxxed - or even within 80% of max, unless the viewing location is on the surface of the SUN - Period. ... Video should give a reading of 30 ft Lamberts, or for a more cinema-like experience, aim for 12-16 ft Lamberts.

How can you possibly say that the contrast should be no more than 80% of max? Are you familiar with the light output of all TVs? At max my TV is 19 ft Lamberts. You recommend as high as 30 ft Lamberts.

How can you possibly say that the contrast should be no more than 80% of max? Are you familiar with the light output of all TVs? At max my TV is 19 ft Lamberts. You recommend as high as 30 ft Lamberts.

What type of TV/brand/age is your TV? Actually, I didn't recommend 30 ftL for video, SMPTE recommends it! :)

I don't recall where I read it, but it was stated somewhere that many direct-view and flat panel displays, at factory default, were capable of up to 100ftL with those factory defaults left untouched. That's enough light output to read or play cards by, or to turn off what most people call lights, and let the set illuminate the room! Not to mention that some flat-panels have their backlights set to 100%(either in the user menu(you can lower it), or internally: in which case you're SOL). Combine all that with color temps 10,000 kelvin or up, and you know what I mean.

For a TV to seem "dim" or not quite bright enough, even with contrast up all the way, it's probably at lest 10 years old and/or nobody ever bothered to lower the contrast. It's burned out! Using AVIA and DVE, my tube sets contrast is about 35-40% - comfortably bright enough to view even at 12 noon(indirect sunlight of course!)

What type of TV/brand/age is your TV? Actually, I didn't recommend 30 ftL for video, SMPTE recommends it! :)

I've got a 70" Sony XBR2, a rear projection LCOS TV. It's about 14 months old, and still has the original bulb. So it's possible that I've lost a few ftL due to bulb age.

I don't recall where I read it, but it was stated somewhere that many direct-view and flat panel displays, at factory default, were capable of up to 100ftL with those factory defaults left untouched. That's enough light output to read or play cards by, or to turn off what most people call lights, and let the set illuminate the room! Not to mention that some flat-panels have their backlights set to 100%(either in the user menu(you can lower it), or internally: in which case you're SOL). Combine all that with color temps 10,000 kelvin or up, and you know what I mean.

100ftL?! Holy crap! That would burn my retinas. :-)

For a TV to seem "dim" or not quite bright enough, even with contrast up all the way, it's probably at lest 10 years old and/or nobody ever bothered to lower the contrast. It's burned out! Using AVIA and DVE, my tube sets contrast is about 35-40% - comfortably bright enough to view even at 12 noon(indirect sunlight of course!)[/QUOTE]

Yeah, tube sets are still considered one of the brightest technologies, aren't they? Aren't they brigher than a plasma? It seems that rear projection is one of the dimmer technologies. I could be wrong.

I've got a 70" Sony XBR2, a rear projection LCOS TV. It's about 14 months old, and still has the original bulb. So it's possible that I've lost a few ftL due to bulb age.



100ftL?! Holy crap! That would burn my retinas. :-)

For a TV to seem "dim" or not quite bright enough, even with contrast up all the way, it's probably at lest 10 years old and/or nobody ever bothered to lower the contrast. It's burned out! Using AVIA and DVE, my tube sets contrast is about 35-40% - comfortably bright enough to view even at 12 noon(indirect sunlight of course!)

Yeah, tube sets are still considered one of the brightest technologies, aren't they? Aren't they brigher than a plasma? It seems that rear projection is one of the dimmer technologies. I could be wrong.[/QUOTE]

Yep - good old direct-view tube TVs will always be the brightest thing going - until SED finally gets off the ground(the mfgs have been promising this technology would definitely be out by next fall - for the past 10 or so falls!)

Rear-pro, particularly models going back more than 5 years, weren't as bright as ones made less than 5 years ago. They also tend to have directional issues - you need to sit directly in front to get full brightness of image/color range/focus.

Again, Boat, 100ftL light output is useless. Its only purpose is to sell the display, to move it off the showroom floor and into your horse. After that, it's up to you: joe-consumer, to figure out how to set the picture controls, or do a self-calibration, or ISF-level calibration.

Although judging by your handle, 100ftL could come in handy if that's the measurement of your bow-light(!), if you go out after dark.

With a LCOS as you described, the first thing to do when you get it home - even before you get signal to it - is to power it up, get into the user menus, and lower the contrast to 50%, and select a Warm1 or Warm2 color temp.

Your bulb life at the factory defaults will be like a 9 iron or a pitching wedge - all height(very bright) and little distance(bulb life longevity). You want the bulb life curve to be more like a 1 or 2 iron - modest height(more realistic reproduction of image) with great distance(lonnnnnger bulb life!):cool: