AVS › AVS Forum › Display Devices › Display Calibration › Calibration -- What Went Right and What Went Wrong?
New Posts  All Forums:Forum Nav:

Calibration -- What Went Right and What Went Wrong?

post #1 of 6
Thread Starter 
As my calibrator of choice will not be passing through my area for a number of months I decided to try a local calibrator with the idea that (i) the picture quality would be improved over its current state and (ii) the picture quality would be further improved (i.e. optimized) when my calibrator of choice visits.

As I am less than expert in video calibration I thought that I would (i) describe the process (ii) describe why I believe/think that the key calibration components were not done and (iii) invite your feedback SO THAT I CAN INCREASZE MY LEARNING.

CALIBRATION PROCESS

Projector is an InFocus 7210
Video Processor is a Lumagen VisionHDP

1. Calibrated the projector's brightness, contrast and grayscale by DIRECTLY connecting a SENCORE unit to the projector to generate the necessary test patterns with ColorFacts software to getting readings therefrom.

The SENCORE unit fed a 720p signals to the projector as this is the projectors native rate.

This seemed to go as I thought except for the following:

-- Color was never calibrated meaning (i) primaries were not adjusted closer to their ideal values (as I was told that the projector does not have this capability) and (ii) using an SMPTE test pattern

-- Hue was never calibrated;

-- Gamma and gamut were not calibrated. Moreover, when I asked about these parameters I was told that "if the grayscale is right then the gamma and gamut are right". This does not seem right to me;

-- Grayscale was not, at this point, possible to get "spot on" as my projector has what he called a "color error" meaning that it is not possible to get BOTH the low IRE and the high IRE flat but rather 0-40 IRE flat, 40-50 IRE has a spike and 50-100 IRE is flat. This does not seem right to me; and

-- Sharpness was never calibrated.

2. Calibrated one input on the video processor by (i) calibrating grayscale by using the 11 point calibration process within the video processor (ii) calibrating the brightness, color , contrast, hue, and sharpness by feeding 720P test patterns from the SENCOR to the video processor.

The resulting settings were then copied to all other inputs on the video processor such that the settings for my two input sources (i.e. cable box and dvd player) were the same.

This seemed to go as I thought except for the following:

-- Brightness, colour, contrast, hue, ans sharpness were not calibrated sepatrately for either (i) different input sources, (ii) different input resolutions from a given source (i.e. cable box outputs 480i, 720p, 1080i) or (iii) different lighting conditions (i.e. I watch movies without light on and cable with/without light). The explanation, which does not seem right to me, was that the difference in settings are so small that there is little/no point in creating separate settings;

-- Gamma and gamut calibration request were once again answered as above.



WHAT WAS NOT DONE/WENT WRONG

1. Colour, hue and sharpness on the projector were never calibrated;

2. Brightness, colour, contrast, hue, and sharpness on the video procssor were not calibrated sepatrately for (i) different input sources (i.e. cable box and dvd) (ii) different input resolutions from a given source (i.e. cable box outputs 480i, 720p, 1080i) or (iii) different lighting conditions (i.e. I watch movies without light on and cable with/without light);

3. Gamma and gamut, on both the projector and the video processor were, for the reasons noted above, never calibrated. Should these be done for different light conditions, different sources and/or different resolutions;

4. Grayscale post calibration charts, for both the projector and the video processor, were not provided although the deltaC*, as shown in a spreasheet, was within 3.00. The post projector cailbration spike between 40-50 IRE had a deltaC* of 2.00. Should grayscale be done for different light conditions, different sources and/or different resolutions;

5. Other items that I believe shouls have been calibrated/checked but were not include:

-- Colour level (i.e. matching the colour level -- PC/Video -- with each input source);

-- Colour points, through either the projector or the video processor, were not adjusted closer to their ideal/standard values for either of REC 601 or REC 709.

A final point, at least for this section, is that the entire process took 3.5 hours which seemed long to me given that, at the end of the day, not much more that a grayscale calibration was done.



COMMENTS/FEEDBACK

At this point -- BECAUSE I AM INTERESTED IN LEARNING -- I would appreciate your comments/feedback on the above. I ammost interested in how the process should have been improved, which settings need to be done for separate input sources, input resolutions and light conditions, and finally, what should have been done that was not done.


LUMAGEN COMMENTS/QUESTIONS

In addition I noticed the following with the respect to the operation of the Lumagen which I would apprecaite feedback on:

1. Although the grayscale settings from input 1 to input 2 we properly copied I noticed that the settings fro input 2 had to be adjusted/tweaked. I do not understand why this would be the case given that the test patters are internally generated and all parameters/settings were the same.

2. Although the copy feature worked well it would be convenient/helpful to have the parameters copied to all input resolutions for a givgen input (i.e. brightness, color, contrast, etc.) parameters were not copied uniformally for ALL input resultions within a given input? Is there a reason for this? Would it be possible to add this feature?

3. After upograding to firmware version 122206 depressing the ENTER/OK key when watching a DVD whose signal comes into the video processor as 480p (i.e. can not do 480i over HDMI) the message "IN2A 16:9 Z1 480/576P-60 AUTO1" is displayed. What does the Z1 mean as I have never seen this before?

4. Appear to have an overscan problem in that there is information/noise at the top of the signal from the cable box. Am I correct in assuming/understandintg that the best way to adjust this is MENU->OUT->RES->SIZE?

5. Suggestion...it would be GREAT were Lumagen to provide a list of parameters/variables that are (i) gloab (ii) input specific and (iii) input/resolution specific.

And finally, thanks in advance to everyone who will respond to this thread!!
post #2 of 6
I'm not a professional calibrator, but I do dabble a bit. Krasmuzik will be able to give you more definite answers, but I'll have a go and if I'm wrong I'll learn from any corrections.

It is true that on many projectors you can't alter the primaries or secondaries, so in that respect the gamut cannot be altered. You might be able to alter them with a colour correction filter or by reducing the color saturation to reduce the gamut of it exceeds the target gamut. but that's a bit of a kludge that may or may not work. Quite often only one primary may exceed the target so lowering the color saturation will make the other two under saturated..

Color error may be the decoder, which is not unheard of, so if you have a red push for example (used to be quite common on tvs), then you can't do much about it.

Many calibrators use signal generators, and although I feel it's better to use the actual sources, apparently there is rarely little deviation (around 1% IIRC) between sources so little to be gained. A generator makes the process quicker and more accurate I believe.

Gamma is often limited to the ones you can select on the projector, so you cannot customise them directly. Some have Look Up Tables that can be altered, or you can iron out humps and dips using an external scaler if it has the ability.

Brightness and contrast can only be set provided the source has a test facility sending the correct levels to the display. TV used to transmit test cards years ago but that rarely happens now, so it can't easily be done. DVDs can use a test disk (DVE, Avia) though so that's easy. Off air tv transmissions can vary too I think (maybe due to the tuners we use), so that might be why it's not easy to set them without a transmitted test card, though I'm not 100% on that. Sometimes if a DVD player has an RF out, you can use that, or use it via a VCR which does. It can involve a lot of cable swapping though.

Sharpness AFAIK doesn't alter anything except haloing and ringing, so use it to your personal preference. You can use a test pattern to see how much ringing you have around lines and text, but quite often lower is better as it introduces less artefacts (on some pj's too low can cause visible problems too, so going below a center position on the adjuster may do that).

A dE of under 3 is pretty good. Above 10 is noticeable, but obviously the closer to zero the better. It's a value that represents perceivable deviation IIRC.

3.5 hours doesn't sound bad to me to be honest.

Gary
post #3 of 6
Well it only takes me an hour to calibrate an Infocus - but that is because I sell them so am very familiar with their calibration nature and can get right down to business when I was the one that installed it. The SP7210 put me out of business on this though because they are 0dE at factory - only after significant lamp hours does it need a videophile tune up - unless it is used as a daily TV that could be a year or more!

Only the greyscale can be calibrated on it - everything else is fixed in the PJ design. I usually use gamma 'film' though you might prefer gamma 'CRT' - just make sure white peaking was left at 0. The key with the grayscale on it to know is that the presets are optically optimized - if your lamp has drifted you only want to cut the RGB gains to compensate - not increase. As far as video or PC levels over DVI - Bob Williams published the gain/offset numbers to use to maximize contrast (66/39?) as your calibration starting point for Video DVI sources - while you could use brightness/contrast to compensate instead - the gain/offset adjustment has more bits. You can also take colorspace out of auto if you have a REC601 vs. REC709 decoding issue on upconversions - but you cannot calibrate the decoding or gamut.

I use the set top box black bars to find black levels of TV sources - just set the box and PJ up to be 4:3 and match up the black bars. White level is harder since you need a white scene and DVR to freeze it on - and tint/color varies with each station even if you record the early AM test pattern. Since it varies per broadcast - a generator is guaranteed not to match the station. I used to do the DVD with RF out over CH3 trick for TV tuners - but that is no guarantee to match the broadcast either.

But I think it is silly to do basic video adjustments for a DVD player with a reference test generator - the adjustments are there because sources are not reference. I use the reference generator to find the the reference adjustment on the projector (which should be default on SP7210) - and then adjust either the source or display - which ever causes least banding/dithering issues. Had one phillips player that was way out of wack on tint - but any adjustment caused various colors to sizzle with digital noise. I made the customer go to circuit city with me and buy a new player.

On Infocus you only get the full set of video adjustments when the Faroudja is engaged which means 480i. They use PixelWorks without adjustments for the other rates. Also there are three calibration memories - the manual is misleading indicating per input - but they are global. So a scaler with basic adjustments and more calibration memories is a good idea, but I am not conviced on using scalers for display calibration - only source adjustments.

For day mode I set up for 7500K with white peaking full and bright room gamma- this makes it brighter than spec - and ambient lite pollutes any attempt at D65 anyways. Some calibrators like to raise brightness for day use for elevated black levels - but I don't - I guarantee in the NW (as in canada) - the room brightness varies from day to day.


2dE is good since 3dE is considered the difference limen. While you could tweak gamma/grayscale with an external scaler - I don't think that is a good idea on an 8bit DVI channel.

Lumagen questions will probably get a better response in the VP forum.
post #4 of 6
Hi Kevin,

Some good tips in there about setting the black level using black bars. I've done that with scope movies on a DVD but didn't think about that for tv by forcing it to 4:3.


Quote:


I made the customer go to circuit city with me and buy a new player.

Damn you're good.

Gary
post #5 of 6
Quote:
Originally Posted by Joelc View Post

As my calibrator of choice will not be passing through my area for a number of months I decided to try a local calibrator with the idea that (i) the picture quality would be improved over its current state and (ii) the picture quality would be further improved (i.e. optimized) when my calibrator of choice visits.

As I am less than expert in video calibration I thought that I would (i) describe the process (ii) describe why I believe/think that the key calibration components were not done and (iii) invite your feedback SO THAT I CAN INCREASZE MY LEARNING.

CALIBRATION PROCESS

Projector is an InFocus 7210
Video Processor is a Lumagen VisionHDP

1. Calibrated the projector's brightness, contrast and grayscale by DIRECTLY connecting a SENCORE unit to the projector to generate the necessary test patterns with ColorFacts software to getting readings therefrom.

The SENCORE unit fed a 720p signals to the projector as this is the projectors native rate.

This seemed to go as I thought except for the following:

-- Color was never calibrated meaning (i) primaries were not adjusted closer to their ideal values (as I was told that the projector does not have this capability) and (ii) using an SMPTE test pattern

-- Hue was never calibrated;

-- Gamma and gamut were not calibrated. Moreover, when I asked about these parameters I was told that "if the grayscale is right then the gamma and gamut are right". This does not seem right to me;

-- Grayscale was not, at this point, possible to get "spot on" as my projector has what he called a "color error" meaning that it is not possible to get BOTH the low IRE and the high IRE flat but rather 0-40 IRE flat, 40-50 IRE has a spike and 50-100 IRE is flat. This does not seem right to me; and

-- Sharpness was never calibrated.

2. Calibrated one input on the video processor by (i) calibrating grayscale by using the 11 point calibration process within the video processor (ii) calibrating the brightness, color , contrast, hue, and sharpness by feeding 720P test patterns from the SENCOR to the video processor.

The resulting settings were then copied to all other inputs on the video processor such that the settings for my two input sources (i.e. cable box and dvd player) were the same.

This seemed to go as I thought except for the following:

-- Brightness, colour, contrast, hue, ans sharpness were not calibrated sepatrately for either (i) different input sources, (ii) different input resolutions from a given source (i.e. cable box outputs 480i, 720p, 1080i) or (iii) different lighting conditions (i.e. I watch movies without light on and cable with/without light). The explanation, which does not seem right to me, was that the difference in settings are so small that there is little/no point in creating separate settings;

-- Gamma and gamut calibration request were once again answered as above.



WHAT WAS NOT DONE/WENT WRONG

1. Colour, hue and sharpness on the projector were never calibrated;

2. Brightness, colour, contrast, hue, and sharpness on the video procssor were not calibrated sepatrately for (i) different input sources (i.e. cable box and dvd) (ii) different input resolutions from a given source (i.e. cable box outputs 480i, 720p, 1080i) or (iii) different lighting conditions (i.e. I watch movies without light on and cable with/without light);

3. Gamma and gamut, on both the projector and the video processor were, for the reasons noted above, never calibrated. Should these be done for different light conditions, different sources and/or different resolutions;

4. Grayscale post calibration charts, for both the projector and the video processor, were not provided although the deltaC*, as shown in a spreasheet, was within 3.00. The post projector cailbration spike between 40-50 IRE had a deltaC* of 2.00. Should grayscale be done for different light conditions, different sources and/or different resolutions;

5. Other items that I believe shouls have been calibrated/checked but were not include:

-- Colour level (i.e. matching the colour level -- PC/Video -- with each input source);

-- Colour points, through either the projector or the video processor, were not adjusted closer to their ideal/standard values for either of REC 601 or REC 709.

A final point, at least for this section, is that the entire process took 3.5 hours which seemed long to me given that, at the end of the day, not much more that a grayscale calibration was done.



COMMENTS/FEEDBACK

At this point -- BECAUSE I AM INTERESTED IN LEARNING -- I would appreciate your comments/feedback on the above. I ammost interested in how the process should have been improved, which settings need to be done for separate input sources, input resolutions and light conditions, and finally, what should have been done that was not done.


LUMAGEN COMMENTS/QUESTIONS

In addition I noticed the following with the respect to the operation of the Lumagen which I would apprecaite feedback on:

1. Although the grayscale settings from input 1 to input 2 we properly copied I noticed that the settings fro input 2 had to be adjusted/tweaked. I do not understand why this would be the case given that the test patters are internally generated and all parameters/settings were the same.

2. Although the copy feature worked well it would be convenient/helpful to have the parameters copied to all input resolutions for a givgen input (i.e. brightness, color, contrast, etc.) parameters were not copied uniformally for ALL input resultions within a given input? Is there a reason for this? Would it be possible to add this feature?

3. After upograding to firmware version 122206 depressing the ENTER/OK key when watching a DVD whose signal comes into the video processor as 480p (i.e. can not do 480i over HDMI) the message "IN2A 16:9 Z1 480/576P-60 AUTO1" is displayed. What does the Z1 mean as I have never seen this before?

4. Appear to have an overscan problem in that there is information/noise at the top of the signal from the cable box. Am I correct in assuming/understandintg that the best way to adjust this is MENU->OUT->RES->SIZE?

5. Suggestion...it would be GREAT were Lumagen to provide a list of parameters/variables that are (i) gloab (ii) input specific and (iii) input/resolution specific.

And finally, thanks in advance to everyone who will respond to this thread!!


Thanks for the effort. Great help to me.
post #6 of 6
Thread Starter 
Gary and kevin, many thanls for the feedback..much appreciated...will see what happens my the other calibrator gets his hands on things...
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › Calibration -- What Went Right and What Went Wrong?