or Connect
AVS › AVS Forum › Display Devices › Display Calibration › Radiance 3D LUT (5x5x5 Cube) Calibration with CalMAN 5
New Posts  All Forums:Forum Nav:

Radiance 3D LUT (5x5x5 Cube) Calibration with CalMAN 5 - Page 12

post #331 of 747
Quote:
Originally Posted by erkq View Post

I know it's been mentioned, but I can't find it. My Lumagen/C6/CalMAN package just arrived a day early (yay!). What is the best setting to start from on my RS20? Get it as close as possible without taking away gamut that the calibration might need? This may help with banding as I remember from somewhere. Or, just give the Lumagen all the flexibility it needs by setting the RS20 at its native wide color gamut, letting the Lumagen winnow away the colors as it needs to?

Getting the projector close is better.

The less processing the lumagen has to do the better. But that's just not at the edge, if the projector already compensates for having a wide gamut it may be better to use native gamut. Probably looking at some saturation sweeps may tell you if native gamut or a calibrated gamut would be a better starting place.
post #332 of 747
Quote:
Originally Posted by sotti View Post

Getting the projector close is better.
The less processing the lumagen has to do the better. But that's just not at the edge, if the projector already compensates for having a wide gamut it may be better to use native gamut. Probably looking at some saturation sweeps may tell you if native gamut or a calibrated gamut would be a better starting place.

It seems the THX setting on the RS20 may be a good starting place, then. The gamut measure almost spot-on @ 75%. As long as the points are a little out of bounds so there's something left for the Lumagen to work with it should be OK. As I understand it, you don't want any in-bound points because the Lumagen can't create a point that falls outside the projector's range. EDIT: I forgot to ask... is this correct?
post #333 of 747
Quote:
Originally Posted by erkq View Post

It seems the THX setting on the RS20 may be a good starting place, then. The gamut measure almost spot-on @ 75%. As long as the points are a little out of bounds so there's something left for the Lumagen to work with it should be OK. As I understand it, you don't want any in-bound points because the Lumagen can't create a point that falls outside the projector's range. EDIT: I forgot to ask... is this correct?

That sounds exactly right.
post #334 of 747
Quote:
Originally Posted by sotti View Post

That sounds exactly right.

Thank you! Another question... not sure if this is the proper place for this, but as long as you're here... smile.gif

I upraded my Control to Enthusiast so I could do the color cube. But when I installed my new Enthusiast license it said, in red letters, "control excluded". But isn't that exactly what I need to control the Lumagen?
post #335 of 747
Quote:
Originally Posted by erkq View Post

Thank you! Another question... not sure if this is the proper place for this, but as long as you're here... smile.gif
I upraded my Control to Enthusiast so I could do the color cube. But when I installed my new Enthusiast license it said, in red letters, "control excluded". But isn't that exactly what I need to control the Lumagen?

That's just the way our licensing works.
The Enthusiast license is a superset of the control license, so it replaces the control license.

If you were ever to clear the licenses or re-install on a new machine, you wouldn't need to add the control license only the enthusiast.
post #336 of 747
Tried using the 3D LUT cube last night for a actual 3D calibration, all worked and it took about 1:10 minutes to run.

My one problem was with the max light out put. I could only get all 4 green lights on when I set contrast at about 77, for the little I know about actual 3D calibration the contrast should be at-least set at 90 or higher.

I used 2.2, Natural, Mid, set in my VT50 for Custom 3D mode, brightness set at 55, contrast set at 80.I use a I1pro 2, Mini .

Are there any settings I should be using other than the default 3D LUT settings?

Actually the 3D calibration came out pretty good, but I still want to get the calibration to its full potential.

ss

This is the Last 3D calibration I did, I guess CM5 RC3 is not ready for 3D 3dLUT autocal yet.

.

Update, here is a second try at a actual 3D calibration using auto 3D LUT. I set the contrast at 100, bright at 50. Once again I used my 3D glasses infront of my I1pro 2.



In the below chart it looks like there is a bug in the cube when you try to do a real 3D calibration.



Update

After beating my brains out, I think I have found the problem. It seems that the USB converter I have was messing up my 232 connection to my Mini. It seems that the USB converter was not allowing all the data to pass from laptop/CMV5E to the Mini, and that was why these charts are so bad. My old laptop has a RS 232 port, so after messing with the Mini and my laptop I got a good connection and ran a 2D 3D LUT cube, it came out great. When I get my new USB converter from Calman I will run a 3D 3D LUT cube calibration and see what happens.
Tom at Lumagen and Scott at Calman helped me out trouble shooting, thanks guys.

ss .
Edited by sillysally - 10/12/12 at 8:04pm
post #337 of 747
I did my first auto-cal and the results were extremely disappointing. I think I have basic problems that need to be solved first. It's been a while since I've been active on these forums and I may need to go back and review the RS20 calibration thread.

The results were un-watchable. I think my projector isn't putting out enough light. I get 7 ft/l at 100% white and that translates to about 400 lumens in my setup. This is an RS20 with about 680 hr on the bulb and a fully open iris. My throw is 19', so that may be part of it. The C6 couldn't read the black level at all so CalMAN reported a CR in the hundreds of thousands. I guess this is normal?

So I guess what I should do is measure a known gamma curve from the RS20 with the Mini out of the signal stream and see if the result is recognizable in CalMAN as a reasonable gamma curve. I'm also considering moving the projector closer so I get better light for experimentation.

And the green bar in the 2 point white balance (I think that's what it was... the 100% and 30% thing?) was WAY out of bounds... twice red and blue. So I pulled that back on the pj. I forget how. I should definitely take notes next time!
post #338 of 747
Quote:
Originally Posted by erkq View Post

I did my first auto-cal and the results were extremely disappointing. I think I have basic problems that need to be solved first. It's been a while since I've been active on these forums and I may need to go back and review the RS20 calibration thread.
The results were un-watchable. I think my projector isn't putting out enough light. I get 7 ft/l at 100% white and that translates to about 400 lumens in my setup. This is an RS20 with about 680 hr on the bulb and a fully open iris. My throw is 19', so that may be part of it. The C6 couldn't read the black level at all so CalMAN reported a CR in the hundreds of thousands. I guess this is normal?
So I guess what I should do is measure a known gamma curve from the RS20 with the Mini out of the signal stream and see if the result is recognizable in CalMAN as a reasonable gamma curve. I'm also considering moving the projector closer so I get better light for experimentation.
And the green bar in the 2 point white balance (I think that's what it was... the 100% and 30% thing?) was WAY out of bounds... twice red and blue. So I pulled that back on the pj. I forget how. I should definitely take notes next time!

I've got an RS10 and had the same problem with light output. I was getting about 4 ft/l (!) out of it at approximately the same distance. There are a couple of threads about cleaning the internal glass between the bulb and the panels. This more than doubled my light output. Honestly, after removing the lamp, mine didn't look dirty --it's located far away from anything else in the house, but cleaning it with a good quality lense cleaner and microfiber cloth made an amazing difference. I was going to get a new bulb (got about 1250 hours on mine), but I was able to put that off a while. There are actually several glass panels of different type in there, but only the outer one facing the lamp needed cleaning. You also don't have to take apart the projector to do this, just remove the lamp.

BpM
post #339 of 747
Quote:
Originally Posted by bpmmmm View Post

I've got an RS10 and had the same problem with light output. I was getting about 4 ft/l (!) out of it at approximately the same distance. There are a couple of threads about cleaning the internal glass between the bulb and the panels. This more than doubled my light output. Honestly, after removing the lamp, mine didn't look dirty --it's located far away from anything else in the house, but cleaning it with a good quality lense cleaner and microfiber cloth made an amazing difference. I was going to get a new bulb (got about 1250 hours on mine), but I was able to put that off a while. There are actually several glass panels of different type in there, but only the outer one facing the lamp needed cleaning. You also don't have to take apart the projector to do this, just remove the lamp.
BpM

Thanks for reminding me! I've done this before but maybe I need to do it again. I forgot.
post #340 of 747
Finally had time this weekend to put Calman5 and the 5x5x5 to through the ringer. With this new 3Dlut calibration I may not want to get the new JVC I'm on the pre-order! Took me some time get the hang of using Calman5. The entire process took a bout 4 hours, but it was entirely worth the effort! I am soooooo happy that I picked up to of the Radiance XD when they were going cheap on EBAY!!! it's going to take awhile for me to view all of my blu-ray disc with the new calibration. I just watched Star Wars Episode 5 and it was like watching a new movie!!! Only one thing I did wrong.... after finishing the calibration, I did something that did generate the report.
post #341 of 747
biggrin.gif

If you haven't already, don't forget to download the Radiance Configuration utility Version 1.4 and store your setting on your HD.
http://www.lumagen.com/testindex.php?module=utilities

ss
post #342 of 747
I had a successful calibration! But the results are a little dim. I remember reading back a ways that can be solved by turning the projector's brightness up a few clicks. That makes it look a lot better but kills the black level (no surprise there). I think I need a new bulb. My RS20 is only putting out 400 lumen at high lamp. I was hoping to avoid this with less than 700 hours on it, but is seems to have faded fast. I did clean the prism again too. The colors do look better than they ever have, btw!
post #343 of 747
After experimenting with Radiance 3D LUT on both BENQ W6000 and Optoma HD8300, I found out both projector have a very good CMS. I can do a manual tuning and using the color checker, getting all colors below 3 dE. 80% of the color is below 1 dE. My conclusion is 3D LUT is great if your display CMS is not linear such as Panasonic plasma TV I own but for my projector it is not necessary.

Just make sure the CMS is calibrated at 75% Saturation and 75% Luminance and recheck the saturation sweep.
Edited by rjyap - 10/15/12 at 5:30am
post #344 of 747
Quote:
Originally Posted by rjyap View Post

After experimenting with Radiance 3D LUT on both BENQ W6000 and Optoma HD8300, I found out both projector have a very good CMS. I can do a manual tuning and using the color checker, getting all colors below 3 dE. 80% of the color is below 1 dE. My conclusion is 3D LUT is great if your display CMS is not linear such as Panasonic plasma TV I own but for my projector it is not necessary.
Just make sure the CMS is calibrated at 75% Saturation and 75% Luminance and recheck the saturation sweep.

When you say that you are "getting all colors below 3 de", are you referring to the 6 primary and secondary colors (RGBCMY) or to the 125 colors that the Radiance cube can correct?

Mark
post #345 of 747
Quote:
Originally Posted by streetsmart88 View Post

When you say that you are "getting all colors below 3 de", are you referring to the 6 primary and secondary colors (RGBCMY) or to the 125 colors that the Radiance cube can correct?
Mark
I'm referring to the color checker in Calman Quick Analysis.
post #346 of 747
Quote:
Originally Posted by erkq View Post

I had a successful calibration! But the results are a little dim. I remember reading back a ways that can be solved by turning the projector's brightness up a few clicks. That makes it look a lot better but kills the black level (no surprise there). I think I need a new bulb. My RS20 is only putting out 400 lumen at high lamp. I was hoping to avoid this with less than 700 hours on it, but is seems to have faded fast. I did clean the prism again too. The colors do look better than they ever have, btw!

If it's dim, you should turn up contrast, not brightness. Make sure you run the flashing white clipping pattern and ensure that you never clip 235.
post #347 of 747
If it's too dim on an RS20 you can open the iris a few clicks. It might effect greyscale slightly though, so you should calibrate at the iris position you want to use. I agree you shouldn't increase the brightness (aka black level control) though.
post #348 of 747
I've got 2 problems, the first I can't solve until the second is solved:

1. Using a C6 and JVC RS20 projector, the calibrated pic is beautiful in higher IRE's... starting around 30% or so. Never seen anything like it. It looks like a moving photograph. But as the levels get lower it bands severely and throws in a LOT of red. My JVC's 30% and 100% grayscale could be dialed in with very small errors. Contrast Ratio came out around 22k. CalMAN liked the projector's starting place and proceeded to do it's thing. Any ideas?

2. I accidentally hit the "Video Select" button on my Pioneer BDP-51 player's remote and the screen went dead... HDMI connection broken. I had to hook it up with component directly to the projector to see the Pioneer's menu in order to turn HDMI back on. When it came back on the component pic went away (like its supposed to) and I selected the HDMI input on the RS20. Well, THERE was the Pioneer's menu, looking like everything's alright. But it disappeared in about a second. It'll wait as long as it takes you to get the HDMI input on the projector selected. But once the HDMI pic shows on screen its only got a life of about a second. An HD-DVD player plays through just fine. So I guess my next step is to plug the Pioneer directly into the RS20, bypassing the Radiance. Any ideas?
post #349 of 747
Thread Starter 
Lumagen Radiance Firmware Update (Production 101512)
Small improvement to chroma mostly noticeable in SD game sources. Some small menu fixes.
http://www.lumagen.com/testindex.php?module=updates
post #350 of 747
Thread Starter 
Lumagen Radiance Firmware Update (Production 101612)
Small improvement to chroma mostly noticeable in SD game sources. Some small menu fixes.
http://www.lumagen.com/testindex.php?module=updates
post #351 of 747
Quote:
Originally Posted by sotti View Post

Yes if you want a perfect calibration, you might loose 2-3 cd/m...

Hi Joel and Derek,

I am interested in purchasing Calman 5 Enthusiast and have many detailed questions I'm hoping you can help with:

1) Regarding Joel's comments about brightness quoted above - this is a great concern for me, as it is other members who have voiced opinions about it. I want the 3D color cube, but do not want to sacrifice much light in the process. I read some of your later posts that said you anticipated making some changes that would lower your overhead/buffer to eek out more brightness - was that successful and how much more brightness?

In another post you mentioned if we really really don't want to lose brightness would could change some targets, but I did not understand that explanation. Can you elaborate?

Also - I have a VW95 and have previously measured its Wide color space. In this mode there is enough Y for all 6 colors. Does this mean you would not lower brightness? Or does it just mean that you will not have to lower it as much (lowering it to give you a buffer etc)?

2) I tried the software in evaluation mode with the simulated meter. I realize from other posts that to get the 3D color cube going you must choose it instead of regular CMS. However I searched high and low and couldn't find any option for the color cube. Using the Enthusiast workflow it did not offer a 3D cube either, just regular CMS. Am I missing something here, or are these options not presented because I am in evaluation mode and need to be connected to a Radiance for those options to even be shown?

3) I saw in a post where one of you recommended trying the auto-calibration in simulation mode to get a feel for how it works. But when I try it says its not available in evaluation mode?

4) I have an Eye One Pro that's about 9 years old and hasn't been calibrated/tuned from X-Rite, and an i1D3. Given that I am not going to calibrate the Eye One Pro, would you say that the i1D3 is likely to be more accurate and I should use that? Or would you still recommend training my i1D3 to the Eye One Pro?

5) I'm confused about the meter training option. I want the reference meter to be the i1D3 off the screen with diffuser off, and the target meter (the one I will do the calibration with) to be the i1D3 facing the projector lens with diffuser on. On the Create Profile would I just have the i1D3 selected as both the reference and target meter at the top, and then at the bottom for Display Type choose LCoS Front Projector (facing projector)? Is this all that is needed?

One thing that's confusing is that at the top of that screen - if I choose Advanced options, under the Reference Meter it gives me a Meter Mode where you can choose front or screen facing. But for the Target Meter selection it does not. Why is that?

Once the profile is created/measured you just select it from the Meter screen and it is automatically used? There does not appear to be any indication in the GUI besides that selector that a trained meter is in use?

6) For the i1D3 is there an option in the software where we can tell Calman whether the diffuser is on or not? Is this necessary?

7) A trick I use with the Radiance and my VW95 is to use a small bump of 0.3 IRE at 0% in its 21pt calibration. This makes it so my level 17 is barely visible (brightness control on the pj is too coarse). If I make the change manually in the Radiance, will the auto-calibration clobber it? Or is there a way I can tell Calman to set it at that and not touch it?

8) The VW95 does not provide the ability to put the pj into 0-255 vs 16-235 mode (like the JVC line does). Instead it auto detects the levels based on the Radiance. I have the Radiance set to output YCrCb 4:2:2 Video. Is this setting ok when working with Calman?

9) Talk about measuring and calibrating with headroom up to 255 has me nervous. It may be the gold-standard recommendation, but in my case I want to clip at 235. IMO the headroom up to 255 not worth the tradeoff in brightness. I don't want to debate this with folks because I understand their reasoning and they are entitled to their view as am I. I only bring it up because I want to make sure that if I use auto cal for grayscale/gamma and/or CMS/cubed its not going to dim me for this headroom. Is there a setting where I tell it this is my preference?

10) I'm not familiar with dE2000, but am a bit familiar with dE94. Am I correct to assume that dE2000 takes Y into account? For example, as I understand it with dE94 your color may be a fair amount off at 5% yet it is so relatively dark that the color error is not really perceivable, and therefore you can still wind up with an excellent dE value. dE 2000 works the same way? With dE94 I had always heard to target 1.5 or lower - that anything at those levels is indistinguishable by eye to the standard. Is 1.5 the same threshold for dE2000? I ask because the charts seem to be forgiving (green) up to 3.0.

Thanks and I look forward to your responses!
post #352 of 747
Quote:
Originally Posted by lovingdvd View Post

...
9) Talk about measuring and calibrating with headroom up to 255 has me nervous. It may be the gold-standard recommendation, but in my case I want to clip at 235. IMO the headroom up to 255 not worth the tradeoff in brightness. I don't want to debate this with folks because I understand their reasoning and they are entitled to their view as am I. I only bring it up because I want to make sure that if I use auto cal for grayscale/gamma and/or CMS/cubed its not going to dim me for this headroom. Is there a setting where I tell it this is my preference?
...

I'm a newb with this calibration stuff and have opinions on a lot of what you asked, but I'll spare you because of my noobish status. But this question... #9... I agree with you wholeheartedly. I run my video chain at 16-235 and am interested in brightness much more than WTW.

CalMAN has a display evaluation sequence and part of that is a clipping measurement series. It's very useful and mine follows the slope properly right up to 1.00 (100%) and then levels off immediately and looks like a table. Perfect! IMHO. I believe if your projector is pushed right up against the 235 clipping level and the chain is a "video" level HDMI then CalMAN wil calibrate with that clipped 1.00 point as 100% on the gamma curve.

BTW, I solved all my problems in my previous posts and am really loving this thing!
post #353 of 747
Quote:
Originally Posted by lovingdvd View Post

Hi Joel and Derek,
I am interested in purchasing Calman 5 Enthusiast and have many detailed questions I'm hoping you can help with:
1) Regarding Joel's comments about brightness quoted above - this is a great concern for me, as it is other members who have voiced opinions about it. I want the 3D color cube, but do not want to sacrifice much light in the process. I read some of your later posts that said you anticipated making some changes that would lower your overhead/buffer to eek out more brightness - was that successful and how much more brightness?
In another post you mentioned if we really really don't want to lose brightness would could change some targets, but I did not understand that explanation. Can you elaborate?
Also - I have a VW95 and have previously measured its Wide color space. In this mode there is enough Y for all 6 colors. Does this mean you would not lower brightness? Or does it just mean that you will not have to lower it as much (lowering it to give you a buffer etc)?
Correct, if you have enough luminance in your primaries we don't touch 100% white any more than we need to get it to D65.
Quote:
Originally Posted by lovingdvd View Post

2) I tried the software in evaluation mode with the simulated meter. I realize from other posts that to get the 3D color cube going you must choose it instead of regular CMS. However I searched high and low and couldn't find any option for the color cube. Using the Enthusiast workflow it did not offer a 3D cube either, just regular CMS. Am I missing something here, or are these options not presented because I am in evaluation mode and need to be connected to a Radiance for those options to even be shown?
You need to use the Color Cube 3D Lut workflow, it's enabled with the enthusiast license. We only show the workflows you're licensed to use. I don't believe the Color Cube workflow is available in the evaluation license.
Quote:
Originally Posted by lovingdvd View Post

3) I saw in a post where one of you recommended trying the auto-calibration in simulation mode to get a feel for how it works. But when I try it says its not available in evaluation mode?
I'm not sure when I would have said that, but yes we disabled the ability to run it without real hardware as it was causing more support calls than anything.
Quote:
Originally Posted by lovingdvd View Post

4) I have an Eye One Pro that's about 9 years old and hasn't been calibrated/tuned from X-Rite, and an i1D3. Given that I am not going to calibrate the Eye One Pro, would you say that the i1D3 is likely to be more accurate and I should use that? Or would you still recommend training my i1D3 to the Eye One Pro?
That's actually a good question. I'm not sure I have a good answer for that. Nearly all of the i1Pro's we take in for recertification meet their factory spec, so in all likely hood it will probably be slightly more accurate at high light levels than the i1 Display Pro(3).
Quote:
Originally Posted by lovingdvd View Post

5) I'm confused about the meter training option. I want the reference meter to be the i1D3 off the screen with diffuser off, and the target meter (the one I will do the calibration with) to be the i1D3 facing the projector lens with diffuser on. On the Create Profile would I just have the i1D3 selected as both the reference and target meter at the top, and then at the bottom for Display Type choose LCoS Front Projector (facing projector)? Is this all that is needed?
One thing that's confusing is that at the top of that screen - if I choose Advanced options, under the Reference Meter it gives me a Meter Mode where you can choose front or screen facing. But for the Target Meter selection it does not. Why is that?
Once the profile is created/measured you just select it from the Meter screen and it is automatically used? There does not appear to be any indication in the GUI besides that selector that a trained meter is in use?

The i1 Display 3 will have different options, if you use just the display 3, you'll want UHP facing screen. If you decide to profile the i1 Display 3 against the i1 Pro, you'd still do basically the same thing, but the profile effectively overrides meter mode.

Since the i1 Pro is a spectro it only has three modes, emissive, illuminant (With it's little cap on), and reflective, but we make a big list of lots of options. Basically they are all emissive except facing projector or reflective.

Quote:
Originally Posted by lovingdvd View Post

6) For the i1D3 is there an option in the software where we can tell Calman whether the diffuser is on or not? Is this necessary?
Yes if you want to measure from the projector and use the diffuser you can do that.
Quote:
Originally Posted by lovingdvd View Post

7) A trick I use with the Radiance and my VW95 is to use a small bump of 0.3 IRE at 0% in its 21pt calibration. This makes it so my level 17 is barely visible (brightness control on the pj is too coarse). If I make the change manually in the Radiance, will the auto-calibration clobber it? Or is there a way I can tell Calman to set it at that and not touch it?
CalMAN will overwrite that change, if you want to add it back after the auto cal, you certainly can. It will raise your dE's a little, but not much. Also to note, 17 isn't necessarily suppose to be visible. Depending on your gamma formula 17 may be below a perceptible difference, but on target. We do have other gamma formula's like BT.1886 that take black level into account and do a better job with shadow detail.
Quote:
Originally Posted by lovingdvd View Post

8) The VW95 does not provide the ability to put the pj into 0-255 vs 16-235 mode (like the JVC line does). Instead it auto detects the levels based on the Radiance. I have the Radiance set to output YCrCb 4:2:2 Video. Is this setting ok when working with Calman?
Yup YCrCb 4:2:2 output is what most people use.
Quote:
Originally Posted by lovingdvd View Post

9) Talk about measuring and calibrating with headroom up to 255 has me nervous. It may be the gold-standard recommendation, but in my case I want to clip at 235. IMO the headroom up to 255 not worth the tradeoff in brightness. I don't want to debate this with folks because I understand their reasoning and they are entitled to their view as am I. I only bring it up because I want to make sure that if I use auto cal for grayscale/gamma and/or CMS/cubed its not going to dim me for this headroom. Is there a setting where I tell it this is my preference?
The cube calibration doesn't go over 235 for the radiance, so we aren't going to measure or change anything as part of the AutoCal. If you set it up that 235 is perfect D65 right at clipping, it's likely we won't even touch your light output.
Quote:
Originally Posted by lovingdvd View Post

10) I'm not familiar with dE2000, but am a bit familiar with dE94. Am I correct to assume that dE2000 takes Y into account? For example, as I understand it with dE94 your color may be a fair amount off at 5% yet it is so relatively dark that the color error is not really perceivable, and therefore you can still wind up with an excellent dE value. dE 2000 works the same way? With dE94 I had always heard to target 1.5 or lower - that anything at those levels is indistinguishable by eye to the standard. Is 1.5 the same threshold for dE2000? I ask because the charts seem to be forgiving (green) up to 3.0.
Thanks and I look forward to your responses!

The only dE formula in our product that doesn't use luminance at all is dE UV. All the others do include luminance, although it can be factored out on a chart by chart basis as appropriate.

The dE2000 numbers are typically very similar to dE94, they made some tweaks to the formula to compensate for known issues with L*a*b* colorspace so that the error is more uniform across different hues. So an error in blue of 1.5 looks similar in it's difference to an error in red of 1.5.

We use 3 as line because in general for moving content that's where thing start to become imperceptible. On the other hand you'll notice our cube AutoCal threshold is at 1.5 and our Greyscale AutoCal is at 0.5. Everyone's ability to perceive color is a little bit different, but even at a dE of 1, if the colors are butt fit next to each other with no separation you may be able to perceive a difference. You can trade your time for accuracy as well, in our option panel you turn the dE threshold for both down to 0, then we will go until we've measured a bit on either side of the target and chosen what we think is the absolute best value for your display.

On a grayscale calibration with the radiance set to 0 for AutoCal we usually see dE values in the 0.2 range.
post #354 of 747
Quote:
Originally Posted by sotti View Post

Correct, if you have enough luminance in your primaries we don't touch 100% white any more than we need to get it to D65.
You need to use the Color Cube 3D Lut workflow, it's enabled with the enthusiast license. We only show the workflows you're licensed to use. I don't believe the Color Cube workflow is available in the evaluation license.
I'm not sure when I would have said that, but yes we disabled the ability to run it without real hardware as it was causing more support calls than anything.
That's actually a good question. I'm not sure I have a good answer for that. Nearly all of the i1Pro's we take in for recertification meet their factory spec, so in all likely hood it will probably be slightly more accurate at high light levels than the i1 Display Pro(3).
The i1 Display 3 will have different options, if you use just the display 3, you'll want UHP facing screen. If you decide to profile the i1 Display 3 against the i1 Pro, you'd still do basically the same thing, but the profile effectively overrides meter mode.
Since the i1 Pro is a spectro it only has three modes, emissive, illuminant (With it's little cap on), and reflective, but we make a big list of lots of options. Basically they are all emissive except facing projector or reflective.
Yes if you want to measure from the projector and use the diffuser you can do that.

CalMAN will overwrite that change, if you want to add it back after the auto cal, you certainly can. It will raise your dE's a little, but not much. Also to note, 17 isn't necessarily suppose to be visible. Depending on your gamma formula 17 may be below a perceptible difference, but on target. We do have other gamma formula's like BT.1886 that take black level into account and do a better job with shadow detail.
Yup YCrCb 4:2:2 output is what most people use.
The cube calibration doesn't go over 235 for the radiance, so we aren't going to measure or change anything as part of the AutoCal. If you set it up that 235 is perfect D65 right at clipping, it's likely we won't even touch your light output.
The only dE formula in our product that doesn't use luminance at all is dE UV. All the others do include luminance, although it can be factored out on a chart by chart basis as appropriate.
The dE2000 numbers are typically very similar to dE94, they made some tweaks to the formula to compensate for known issues with L*a*b* colorspace so that the error is more uniform across different hues. So an error in blue of 1.5 looks similar in it's difference to an error in red of 1.5.
We use 3 as line because in general for moving content that's where thing start to become imperceptible. On the other hand you'll notice our cube AutoCal threshold is at 1.5 and our Greyscale AutoCal is at 0.5. Everyone's ability to perceive color is a little bit different, but even at a dE of 1, if the colors are butt fit next to each other with no separation you may be able to perceive a difference. You can trade your time for accuracy as well, in our option panel you turn the dE threshold for both down to 0, then we will go until we've measured a bit on either side of the target and chosen what we think is the absolute best value for your display.
On a grayscale calibration with the radiance set to 0 for AutoCal we usually see dE values in the 0.2 range.

Thanks Joel. A couple quick follow ups for now please:

1) So if I do the i1D3 only, do you think it will read down accurately (with good repeat-ability) at 5% from off the screen?

2) If I want to train it to itself - reference of off the screen and trained meter facing the lens - what is the exact procedure to follow? Just want to have my ducks in a row so I am not confused when I go to do it. You mention there is an option to tell it whether the diffuser is on or not. Is this in the meter properties? I would think that if you told it it was facing the pj it would automatically assume the diffuser is on?

3) I read that the Radiance now supports increments of 0.1 (finer "resolution"). Does the latest Calman 5 take advantage of this? This is for grayscale/gamma yes?

Thanks again!
post #355 of 747
Quote:
Originally Posted by lovingdvd View Post

Thanks Joel. A couple quick follow ups for now please:
1) So if I do the i1D3 only, do you think it will read down accurately (with good repeat-ability) at 5% from off the screen?
Most likely yes, but without knowing the gamma formula and 100% light output it would be hard to know for sure. The i1 Display Pro reads well down to 0.05 cd/m or so.
Quote:
Originally Posted by lovingdvd View Post

2) If I want to train it to itself - reference of off the screen and trained meter facing the lens - what is the exact procedure to follow? Just want to have my ducks in a row so I am not confused when I go to do it. You mention there is an option to tell it whether the diffuser is on or not. Is this in the meter properties? I would think that if you told it it was facing the pj it would automatically assume the diffuser is on?
We don't support profiling to the same meter, if you wanted to do that you can use the screen offset (x,y in absolute, Y as a percentage 0-1).
Quote:
Originally Posted by lovingdvd View Post

3) I read that the Radiance now supports increments of 0.1 (finer "resolution"). Does the latest Calman 5 take advantage of this? This is for grayscale/gamma yes?
Thanks again!

The radiance has always supported 0.1 steps internally and you've always been able to enter a .1 step into our DDC controls.

Our grayscale autocal gets to dE's of 0.2 using the equivalent of one step on 8bit increments. We don't find any visual improvement by using the smaller step size for AutoCal, some panels don't work at all since they only respond to 8bit steps anyway, and the whole procedure takes incrementally longer with the smaller step size. Not to mention for most people who simply leave the dE threshold at 0.5 it never even gets to where it would be making 0.1 adjustments. If we exposed DDC step size as an option, it's something people could change, break the AutoCal procedure, generate support calls, ask about what it does on the forums, ect... So if you want to go back over the grayscale and fine tune it with .1 steps because you know your display will benefit from it, you can use the DDC controls to do that.
post #356 of 747
Quote:
Originally Posted by sotti View Post

Most likely yes, but without knowing the gamma formula and 100% light output it would be hard to know for sure. The i1 Display Pro reads well down to 0.05 cd/m or so.

We don't support profiling to the same meter, if you wanted to do that you can use the screen offset (x,y in absolute, Y as a percentage 0-1).
The radiance has always supported 0.1 steps internally and you've always been able to enter a .1 step into our DDC controls.
Our grayscale autocal gets to dE's of 0.2 using the equivalent of one step on 8bit increments. We don't find any visual improvement by using the smaller step size for AutoCal, some panels don't work at all since they only respond to 8bit steps anyway, and the whole procedure takes incrementally longer with the smaller step size. Not to mention for most people who simply leave the dE threshold at 0.5 it never even gets to where it would be making 0.1 adjustments. If we exposed DDC step size as an option, it's something people could change, break the AutoCal procedure, generate support calls, ask about what it does on the forums, ect... So if you want to go back over the grayscale and fine tune it with .1 steps because you know your display will benefit from it, you can use the DDC controls to do that.

How hard is it to add support for profiling the same meter? Something you can add? In my case it would be more than just the difference in Y because the screen is involved in the reference but not field. Also the diffuser is involved in the field but not reference measure.

I guess in my case it'll be best to train the i1d3 as the field pointed into the pj lens, and the EyeOne Pro off the screen as the reference. Sound good?

Also where will I find the option to tell Calman whether the i1D3 has the diffuser in place or not. Thanks!
post #357 of 747
Quote:
Originally Posted by lovingdvd View Post

How hard is it to add support for profiling the same meter? Something you can add? In my case it would be more than just the difference in Y because the screen is involved in the reference but not field. Also the diffuser is involved in the field but not reference measure.
I guess in my case it'll be best to train the i1d3 as the field pointed into the pj lens, and the EyeOne Pro off the screen as the reference. Sound good?
Also where will I find the option to tell Calman whether the i1D3 has the diffuser in place or not. Thanks!

Profiling with the i1Pro off the screen would be the way to go.

The diffuser shows up in the meter mode.
post #358 of 747
Quote:
Originally Posted by sotti View Post

Profiling with the i1Pro off the screen would be the way to go.
The diffuser shows up in the meter mode.

Thanks. What type of support does Calman have for 3D calibration? Given my two meters, EyeOne Pro and i1D3, and given that the i1D3 likely will not be able to read low enough off the screen - what would you recommend as the procedure to follow for profiling the meters and performing a full calibration in 3D mode (greyscale, gamma, and gamut/cube). The VW95 puts out a lot of extra blue in 3D mode and the gamma is off, so a calibration is important in this mode.

Ideally I would like to put 3D glasses over the EyeOne Pro and read off the screen as the reference, and then use an i1D3 with NO glasses and read out of the lens as the trained meter and save a profile of this. This way for subsequent calibrations I could just reuse the profile with the i1D3 facing the screen and no glasses - so the nice thing about this approach is I only have to go through the trouble of taping glasses to a meter once, and then for all subsequent touch ups or recalibrations down the road I don't have to go through the trouble of taping on glasses. Will this approach work?

Also does Calman support DDC for Samsung plasmas such as the PN63C8000 or its new line of LED displays, or LG LED displays? I found a list somewhere that had some supported devices but it had mostly Panasonic and JVC pjs so I wasn't sure if that was the complete or current list. If these devices are not supported might they be soon? Thanks!
post #359 of 747
Over in the other thread I did give a quick description of it here.
http://www.avsforum.com/t/1423111/calman-5-release-notes-and-discussion/360#post_22505598

Basically you can read from the screen with the glasses for the i1Pro, and still read from the projector with the i1 Display Pro(3). That way you'll still have plenty of light sensitivity, but it will be corrected for the glasses.


As to Samsung DDC and AutoCal, we'd love to support their products, but we haven't had that opportunity yet. Here is the list of products we do currently support:
http://store.spectracal.com/support/hardware-support.html
post #360 of 747
I finally figured this thing out, or rather I figured out what not to do to my projector for the initial setup. The results are spectacular. My RS20 looks awesome (with a new bulb). Re-watched Seabiscuit on HD-DVD. Wow... the pop a nice 2.2 gamma gives and all the different colors in the carpet of the high desert in the opening scenes is amazing. And the horses look so real... the rich color content of their coats... even my wife was impressed! smile.gif Of course, I've never had my theater calibrated before.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › Radiance 3D LUT (5x5x5 Cube) Calibration with CalMAN 5