Radiance 3D LUT (5x5x5 Cube) Calibration with CalMAN 5 - Page 26 - AVS Forum
Forum Jump: 
 8Likes
Reply
 
Thread Tools
post #751 of 825 Old 05-11-2014, 07:02 PM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,640
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 175
Quote:
Originally Posted by Doug Blackburn View Post

This may be in here somewhere, but I can't find the magic search string... is CalMAN measuring 125 points for 5x5x5 LUTs & 729 points for 9x9x9 LUTS? I haven't waited around to count while doing LUT calibrations.

If each point is measured, the 30-40 minutes it takes to do the measurements for a 9x9x9 LUT with my equipment is going to turn into something like 5 hours for a 17x17x17 LUT with one of the 21XX Radiance processors. I'm guessing there's something else in play for the 17x17x17 LUTs to keep the amount of time from stretching out to 5 hours... yes? Interpolation of some of the points?

Yes, while it simply isn't possible to sub-sample a 5^3 LUT, and 9^3 is right on the edge of where sub-sampling makes sense, doing a 17^3 with CalMAN we strongly recommend using our advanced interpolation algorithms. Our latest new feature is the DLC that allows us to dynamically allocate calibration points. This gives us the ability to give better results with roughly the same time as a 9^3 with a 17^3.

We've been working with decoupled interpolated LUTs since the very beginning, it was only the Lumagen with it's sparse cube that we calibrated all the points.

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
Sponsored Links
Advertisement
 
post #752 of 825 Old 05-11-2014, 07:05 PM
AVS Special Member
 
ScottJ's Avatar
 
Join Date: Dec 2000
Location: San Francisco, CA, USA
Posts: 1,497
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 114 Post(s)
Liked: 70
Quote:
Originally Posted by Doug Blackburn View Post

If each point is measured, the 30-40 minutes it takes to do the measurements for a 9x9x9 LUT with my equipment is going to turn into something like 5 hours for a 17x17x17 LUT with one of the 21XX Radiance processors. I'm guessing there's something else in play for the 17x17x17 LUTs to keep the amount of time from stretching out to 5 hours... yes? Interpolation of some of the points?

Seems to me, if you're going to interpolate, why not just use a smaller LUT to begin with? After all, the Radiance already does interpolation within the LUT.
ScottJ is offline  
post #753 of 825 Old 05-11-2014, 07:25 PM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,640
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 175
Quote:
Originally Posted by ScottJ View Post

Seems to me, if you're going to interpolate, why not just use a smaller LUT to begin with? After all, the Radiance already does interpolation within the LUT.

Because tri-linear interpolation subject to quite a bit of artifacting and with DLC we can figure out which points are the most important to calibrate of the 4900 instead of just settling for evenly spaced 700. So between using a more advanced interpolation algorithm and DLC, using the full size of the lumagen and having CalMAN fill in the blanks is the right choice.
ScottJ likes this.

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
post #754 of 825 Old 05-11-2014, 07:26 PM
AVS Special Member
 
Doug Blackburn's Avatar
 
Join Date: May 2008
Location: San Francisco - East Bay area
Posts: 3,457
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 25 Post(s)
Liked: 228
The last 9x9x9 LUT I did for a JVC RS6710 projector has caused dark reds and dark Caucasian flesh tones and darker oranges (aka browns) to posterize and look unnatural. What do I do with a LUT cal if something like that happens? When I switch to a CMS in the Radiance with no calibration, the problem goes away. Other colors and higher lumiances of the problematic colors look fine

(Meter - Konica-Minolta CS-200, Radiance 2041 used to create that LUT, but now using a 2143 with 2041 settings copied to it - projection screen Stewart StudioTek100 in blacked-out room).

Also, most blues come out over saturated (touching the square reference box but OUTSIDE the white reference box in the CIE chart) and the dEs for those oversaturated Blues appear to be much lower than they measure manually.

Is there a document for CalMAN somewhere that explains what to do and not to do with a TV or projector when you are using CalMAN with a Radiance LUT calibration? And also what to do or not do with the Radiance processor, of course.

"Movies is magic..." Van Dyke Parks
THX -- ISF -- HAA
Doug Blackburn is offline  
post #755 of 825 Old 05-11-2014, 08:02 PM
AVS Special Member
 
ScottJ's Avatar
 
Join Date: Dec 2000
Location: San Francisco, CA, USA
Posts: 1,497
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 114 Post(s)
Liked: 70
Quote:
Originally Posted by sotti View Post

Because tri-linear interpolation subject to quite a bit of artifacting and with DLC we can figure out which points are the most important to calibrate of the 4900 instead of just settling for evenly spaced 700. So between using a more advanced interpolation algorithm and DLC, using the full size of the lumagen and having CalMAN fill in the blanks is the right choice.

Very interesting. I get it. (Minor point: from what I recall, the Radiance uses a more sophisticated algo. than trilinear interpolation.)
ScottJ is offline  
post #756 of 825 Old 05-12-2014, 12:49 AM - Thread Starter
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,430
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 508
Quote:
Originally Posted by sotti View Post

Quote:
Originally Posted by Doug Blackburn View Post

This may be in here somewhere, but I can't find the magic search string... is CalMAN measuring 125 points for 5x5x5 LUTs & 729 points for 9x9x9 LUTS? I haven't waited around to count while doing LUT calibrations.

If each point is measured, the 30-40 minutes it takes to do the measurements for a 9x9x9 LUT with my equipment is going to turn into something like 5 hours for a 17x17x17 LUT with one of the 21XX Radiance processors. I'm guessing there's something else in play for the 17x17x17 LUTs to keep the amount of time from stretching out to 5 hours... yes? Interpolation of some of the points?

Yes, while it simply isn't possible to sub-sample a 5^3 LUT, and 9^3 is right on the edge of where sub-sampling makes sense, doing a 17^3 with CalMAN we strongly recommend using our advanced interpolation algorithms. Our latest new feature is the DLC that allows us to dynamically allocate calibration points. This gives us the ability to give better results with roughly the same time as a 9^3 with a 17^3.

We've been working with decoupled interpolated LUTs since the very beginning, it was only the Lumagen with it's sparse cube that we calibrated all the points.

Joel, you have to measure with full 17x17x17 to have better understanding of the display; to identify where is has the most problems and where is non-linear and then correct them using your algorithm by using less corrected control points from the total 4.934 that Lumagen Radiance 21xx has.

If you don't measure the full 17x17x17 points how is possible to know where the display has problems?

Does CalMAN 5.3.5 are measuring initially the full 17x17x17 to identify these tweak points?

It's the same logic as calibrating Grayscale, if you take initially 11-Point Grayscale and calibrate it perfectly; later if you measure with 21-Point you will find spikes that were not visible with 11-Point measurements.

It's impossible with any way to make a calibration with less control points to perform better with one with the larger number of calibration points.

Interpollation is not calibration, it's guessing. You will always have better results by calibrating more control points and use less interpollation.

Consumer displays are the worst performers for linearity performance, we are not calibrating broadcasting monitors where they have better channel separation and less crosstalk.

The Lumagen's decision to move from 750 to 4.934 color points showcases that they saw the difference from their testings by comparing the 9-Point vs. 17-Point Calibration by dE reporing and by evaluating visually the results also, that's why they added 17x17x17 and not other sizes like 10x10x10 or 14x14x14.
buzzard767 and Iron Mike like this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is offline  
post #757 of 825 Old 05-12-2014, 09:16 AM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,640
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 175
Quote:
Originally Posted by ConnecTEDDD View Post

Joel, you have to measure with full 17x17x17 to have better understanding of the display; to identify where is has the most problems and where is non-linear and then correct them using your algorithm by using less corrected control points from the total 4.934 that Lumagen Radiance 21xx has.

If you don't measure the full 17x17x17 points how is possible to know where the display has problems?

In our process after we have gone through the first 6 steps of our AutoCube we know in theory exactly how the rest of the display should perform in theory. From there we measure an equal distribution of points and use those as the basis to begin flagging areas for DLC.
Quote:
Originally Posted by ConnecTEDDD View Post

Does CalMAN 5.3.5 are measuring initially the full 17x17x17 to identify these tweak points?
That would defeat the purpose of DLC.
Quote:
Originally Posted by ConnecTEDDD View Post

It's the same logic as calibrating Grayscale, if you take initially 11-Point Grayscale and calibrate it perfectly; later if you measure with 21-Point you will find spikes that were not visible with 11-Point measurements.
Different grayscale calibration systems work differently. Some of them certainly behave that way. For the analogy lets assume we are talking about a set of 1D LUTs with with a control mask on top. With the right interpolation algorithms, while there will certainly be more error in between the points we calibrated it's likely the amount of error is still well below the visible threshold. Also if we were to have DLC enable and one segment was outstanding in it's error we would end up doing a 13 or 15 point calibration with results more similar to a 21 point.
Quote:
Originally Posted by ConnecTEDDD View Post

It's impossible with any way to make a calibration with less control points to perform better with one with the larger number of calibration points.
That's not entirely correct. Information quality is paramount, quantity is less important. Even LightSpace out performs raw brute force with fewer more intelligently selected data points.
http://studio.spectracal.com/3d-lut-comparison-study

Having the right data is extremely important, more data is always helpful. DLC is the solution to finding out where extra data begins to have diminishing returns. DLC enables us to continue to gather more data, we can go even finer that 17x or 21x if necessary to collect information about the non-linearity areas.

Information is key, but just doing raw brute force guess work by measuring 17^3 or 21^3 isn't the right answer.
Quote:
Originally Posted by ConnecTEDDD View Post

Interpollation is not calibration, it's guessing. You will always have better results by calibrating more control points and use less interpollation.

In profile based calibrations 100% of the points are interpolated. So they must be worse? In all cases everyone is interpolating data. We are all in the same boat here. Better algorithms and better data will do a better job interpolating, but Interpolation is the basis of all 3D LUT software. Any assertment otherwise is in bad faith.

Quote:
Originally Posted by ConnecTEDDD View Post

Consumer displays are the worst performers for linearity performance, we are not calibrating broadcasting monitors where they have better channel separation and less crosstalk.

This concept has been extremely exaggerated. Certainly some consumer displays are horrible, but many are extremely similar in perfomance to broadcast monitors. In most displays the difference between broadcast monitors and standard consumer displays is more about features than quality. Some of the features of course are factory calibration and built-in 3D LUTs, which drastically improve out of the box performance, but many consumer displays have a similar performance envelope.
Quote:
Originally Posted by ConnecTEDDD View Post

The Lumagen's decision to move from 750 to 4.934 color points showcases that they saw the difference from their testings by comparing the 9-Point vs. 17-Point Calibration by dE reporing and by evaluating visually the results also, that's why they added 17x17x17 and not other sizes like 10x10x10 or 14x14x14.
Or maybe they just thought that doubling up was the right way to go. 25%, 12.5%, 6.25%, I don't really see research and innovation there, I just see doubling down on capacity.

Honestly 17^3 isn't really enough imho in order to counter the artifacts from tri-linear interpolation (the stuff done in the LUT hardware) higher density cubes are required.

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
post #758 of 825 Old 05-12-2014, 05:04 PM
AVS Special Member
 
Doug Blackburn's Avatar
 
Join Date: May 2008
Location: San Francisco - East Bay area
Posts: 3,457
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 25 Post(s)
Liked: 228
Quote:
Originally Posted by ConnecTEDDD View Post



Interpollation is not calibration, it's guessing. You will always have better results by calibrating more control points and use less interpollation.

.

This is not exactly true. Take the case of frame interpolation. You can take 2 frames where there is a moving object, let's say a race car... and let's say the car is moving left to right. Frame 1 has a blurred object in the left side of the frame and frame 2 has an equally blurred object on the right side of the frame. If you look at those 2 frames, let's say it is blurred so much that you cannot tell if it is a car or some other object. Using frame interpolation, you can create a new frame between 1 and 2 that is sharper than either frame 1 or frame 2. It can be so much sharper that you may be able to tell the type/model of car when that would be impossible using only a single blurred frame. There are different levels of interpolation... certainly low level interpolation might only be a little better than guessing. But if you use multiple 3D points for interpolation, you can get much better results. Back to the frame interpolation comparison... if you have multiple frames showing the blurred object in sequence, you can get increasingly sharp interpolated frames...2 frames allows a somewhat sharper intermediate image. If you have 3 "original" frames, you can get 2 intermediate frames that are sharper than the interpolated frame you get from 2 frames. If you have 4 frames to work with, you can get sharper yet...and so forth. Similarly, if you are using 3D measurements, you can use multiple 3D measurements in the same region to improve your estimate of how an intermediate point should be corrected. You can even use multiple 3D measurements where there is a matching coordinate to help you predict a correction.

That said, there's nothing like real measurements to create a perfect LUT, but if it would take 5 hours of measurements to complete 4913 measurements... who is going to wait around that long? Maybe a hobbyist who can start the process before bedtime and find it complete in the morning. If you can interpolate to dE= 0.5 or less, the faster time to complete measurements is a pretty good tradeoff to get the measurements down to a reasonable elapsed time. But that means whoever is using the software ought to be able to select 4913 measurements and epect to wait a long long time for results, or they should be able to select some smaller number of measurements plus "math" to create other reference points.

"Movies is magic..." Van Dyke Parks
THX -- ISF -- HAA
Doug Blackburn is offline  
post #759 of 825 Old 05-12-2014, 10:28 PM
AVS Special Member
 
p5browne's Avatar
 
Join Date: Jun 2008
Location: Sudbury, Ontario, CANADA
Posts: 4,014
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 183 Post(s)
Liked: 204
New Calman 5.3.5 RC1 now available to download:

http://color.spectracal.com/downloads/

Note: Use Internet Explorer since Google Chrome appears not to work for the download.
p5browne is offline  
post #760 of 825 Old 05-13-2014, 12:37 PM
AVS Special Member
 
buzzard767's Avatar
 
Join Date: Dec 2005
Location: Naples, FL & Wausau, WI
Posts: 4,106
Mentioned: 5 Post(s)
Tagged: 1 Thread(s)
Quoted: 445 Post(s)
Liked: 973
Quote:
Originally Posted by Doug Blackburn View Post

Quote:
Originally Posted by ConnecTEDDD View Post



Interpollation is not calibration, it's guessing. You will always have better results by calibrating more control points and use less interpollation.

.

This is not exactly true. Take the case of frame interpolation. You can take 2 frames where there is a moving object, let's say a race car... and let's say the car is moving left to right. Frame 1 has a blurred object in the left side of the frame and frame 2 has an equally blurred object on the right side of the frame. If you look at those 2 frames, let's say it is blurred so much that you cannot tell if it is a car or some other object. Using frame interpolation, you can create a new frame between 1 and 2 that is sharper than either frame 1 or frame 2. It can be so much sharper that you may be able to tell the type/model of car when that would be impossible using only a single blurred frame. There are different levels of interpolation... certainly low level interpolation might only be a little better than guessing. But if you use multiple 3D points for interpolation, you can get much better results. Back to the frame interpolation comparison... if you have multiple frames showing the blurred object in sequence, you can get increasingly sharp interpolated frames...2 frames allows a somewhat sharper intermediate image. If you have 3 "original" frames, you can get 2 intermediate frames that are sharper than the interpolated frame you get from 2 frames. If you have 4 frames to work with, you can get sharper yet...and so forth. Similarly, if you are using 3D measurements, you can use multiple 3D measurements in the same region to improve your estimate of how an intermediate point should be corrected. You can even use multiple 3D measurements where there is a matching coordinate to help you predict a correction.

That said, there's nothing like real measurements to create a perfect LUT, but if it would take 5 hours of measurements to complete 4913 measurements... who is going to wait around that long? Maybe a hobbyist who can start the process before bedtime and find it complete in the morning. If you can interpolate to dE= 0.5 or less, the faster time to complete measurements is a pretty good tradeoff to get the measurements down to a reasonable elapsed time. But that means whoever is using the software ought to be able to select 4913 measurements and epect to wait a long long time for results, or they should be able to select some smaller number of measurements plus "math" to create other reference points.

A K10-A in the LightSpace closed loop mode measures 17^3 in about an hour and a quarter.

Buzz
THX Certified Video Calibrator

 

buzzard767 is offline  
post #761 of 825 Old 05-13-2014, 01:06 PM - Thread Starter
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,430
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 508
Quote:
Originally Posted by Doug Blackburn View Post

but if it would take 5 hours of measurements to complete 4913 measurements
Quote:
Originally Posted by buzzard767 View Post

A K10-A in the LightSpace closed loop mode measures 17^3 in about an hour and a quarter.

Calibration Reports using CalMAN for Verification, LS with 17-Point Cube (4.913 Color Points) can be found here (It took 1H 22M).

and with 21-Point Cube (9.261 Color Points) can be found here. (It took 2H 34M).

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is offline  
post #762 of 825 Old 05-13-2014, 10:20 PM
AVS Special Member
 
p5browne's Avatar
 
Join Date: Jun 2008
Location: Sudbury, Ontario, CANADA
Posts: 4,014
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 183 Post(s)
Liked: 204
Lumagen FW Updates for XE XS XD and Mini # 050314

http://www.lumagen.com/testindex.php?module=updates
p5browne is offline  
post #763 of 825 Old 05-17-2014, 05:37 PM
Member
 
tomhahn's Avatar
 
Join Date: Mar 2002
Location: San Francisco Bay Area, East Bay
Posts: 147
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Doug Blackburn View Post

The last 9x9x9 LUT I did for a JVC RS6710 projector has caused dark reds and dark Caucasian flesh tones and darker oranges (aka browns) to posterize and look unnatural. What do I do with a LUT cal if something like that happens? When I switch to a CMS in the Radiance with no calibration, the problem goes away. Other colors and higher lumiances of the problematic colors look fine

(Meter - Konica-Minolta CS-200, Radiance 2041 used to create that LUT, but now using a 2143 with 2041 settings copied to it - projection screen Stewart StudioTek100 in blacked-out room).

Also, most blues come out over saturated (touching the square reference box but OUTSIDE the white reference box in the CIE chart) and the dEs for those oversaturated Blues appear to be much lower than they measure manually.

Is there a document for CalMAN somewhere that explains what to do and not to do with a TV or projector when you are using CalMAN with a Radiance LUT calibration? And also what to do or not do with the Radiance processor, of course.

Was there ever a reply to this? I have a 6710 , and am contemplating a CM bundle, but this is discouraging, I have to say.
tomhahn is offline  
post #764 of 825 Old 05-17-2014, 06:10 PM
Advanced Member
 
Iron Mike's Avatar
 
Join Date: Oct 2006
Posts: 795
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 25 Post(s)
Liked: 361
Quote:
Originally Posted by tomhahn View Post

Was there ever a reply to this? I have a 6710 , and am contemplating a CM bundle, but this is discouraging, I have to say.

if u are concerned about the investment, maybe best to get an evaluation version of CM and other applicable solutions and see what fits best your needs and desired results...

calibration & profiling solutions: Lightspace, Spaceman, Calman, Argyll, ColorNavigator, basICColor
profiling & calibration workflow tools: Display Calibration Tools
meter: Klein K-10 A, i1Pro, i1D3
AVS thread: Lightspace & Custom Color Patch Set & Gamma Calibration on Panasonic 65VT60
Iron Mike is offline  
post #765 of 825 Old 05-17-2014, 07:48 PM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,640
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 175
Quote:
Originally Posted by tomhahn View Post

Was there ever a reply to this? I have a 6710 , and am contemplating a CM bundle, but this is discouraging, I have to say.

My guess here is that he started with the projector in it's widest mode, which means that it took quite a bit of correction to get it back to rec.709. That would likely be the cause of the posterization. Or perhaps there was some setting in the projector that was enabled either flesh tone correction or and adjustment to the orange axis that meant we needed to apply more correction than the display could hold.


As to the comment about the blues, there really isn't enough context there to know how much of a problem there was if it had any real significance. Saying that the dE's where different of something was or wasn't in a box, is all relative data. That could be meaningful, or it could've been zoomed in and not meaningful. The difference between a dE of 0.3 and 0.8 isn't impactful in viewing actual content.

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
post #766 of 825 Old 05-17-2014, 09:49 PM
AVS Special Member
 
Doug Blackburn's Avatar
 
Join Date: May 2008
Location: San Francisco - East Bay area
Posts: 3,457
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 25 Post(s)
Liked: 228
Quote:
Originally Posted by tomhahn View Post

Was there ever a reply to this? I have a 6710 , and am contemplating a CM bundle, but this is discouraging, I have to say.

No answer... the problem at lower luminance levels is not bad with Blu-ray, but Is really noticeable on satellite content.

CM isn't setup for auto-cal of the 67 or other 2014 models yet - they are supposed to be working on that now. I didn't have much luck making the projector better than THX mode using controls in the projector. It is REALLY hard to get gamma where you want it without an outboard processor. Based on what I could and couldn't do with the calibration controls, I'm not sure auto-cal is going to be able to do a lot either, though I hope I'm wrong.

The images this projector makes are great... the calibration controls are tricky and time consuming though.

BTW - the 67 and 6710 are identical projectors... the 6710 comes with a spare lamp and 5 year warranty, the 67 has no spare lamp and a 3 year warranty. $500 difference in price. You pays your money...

"Movies is magic..." Van Dyke Parks
THX -- ISF -- HAA
Doug Blackburn is offline  
post #767 of 825 Old 05-17-2014, 10:03 PM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,640
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 175
Quote:
Originally Posted by Doug Blackburn View Post

No answer... the problem at lower luminance levels is not bad with Blu-ray, but Is really noticeable on satellite content.

We have tweaked algorithm for how we handle saturation v. luminance during calibration. This resolves the issue many people were seeing with higher luminances in the blue channel. It creates good dE values, but visible artifacts.

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
post #768 of 825 Old 05-18-2014, 06:14 PM
Member
 
tomhahn's Avatar
 
Join Date: Mar 2002
Location: San Francisco Bay Area, East Bay
Posts: 147
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Doug Blackburn View Post

No answer... the problem at lower luminance levels is not bad with Blu-ray, but Is really noticeable on satellite content.

CM isn't setup for auto-cal of the 67 or other 2014 models yet - they are supposed to be working on that now. I didn't have much luck making the projector better than THX mode using controls in the projector. It is REALLY hard to get gamma where you want it without an outboard processor. Based on what I could and couldn't do with the calibration controls, I'm not sure auto-cal is going to be able to do a lot either, though I hope I'm wrong.

The images this projector makes are great... the calibration controls are tricky and time consuming though.

BTW - the 67 and 6710 are identical projectors... the 6710 comes with a spare lamp and 5 year warranty, the 67 has no spare lamp and a 3 year warranty. $500 difference in price. You pays your money...


Sorry, I should have specified – I would also be getting a radiance, probably the 2021. In that case, your comment about CM not being setup for auto-cal of my 6710 would be not pertinent, correct?
tomhahn is offline  
post #769 of 825 Old 05-18-2014, 06:58 PM
AVS Special Member
 
Doug Blackburn's Avatar
 
Join Date: May 2008
Location: San Francisco - East Bay area
Posts: 3,457
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 25 Post(s)
Liked: 228
Right, if you are going to use a Radiance processor, it can fix most things, even if the display's controls don't help. The one thing a Radiance processor can't fix is undersaturated primary or complimentary colors... but the JVC projectors have modes to pick from where that's not an issue.

"Movies is magic..." Van Dyke Parks
THX -- ISF -- HAA
Doug Blackburn is offline  
post #770 of 825 Old 05-23-2014, 06:12 PM
AVS Special Member
 
p5browne's Avatar
 
Join Date: Jun 2008
Location: Sudbury, Ontario, CANADA
Posts: 4,014
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 183 Post(s)
Liked: 204
New Calman 5.3.5 RC2 Build #1570 now available:

http://color.spectracal.com/downloads

A quick note:

We added a new feature called Pattern Delay Optimization.
There is now a button on the Source tab called Optimize. CalMAN will take some series of readings to help determine the optimal delay for the pattern generator given the generator, display, and meter combination.

We'd like people to play with this new feature as much as possible and give us feedback on how it is working. There are only so many combinations of hardware that we can test internally.

If you do appear to run into issues, turn on the UI/AutoCal logging option, run through it again to get the relevant data in the log, and pass it along to us.
p5browne is offline  
post #771 of 825 Old 06-05-2014, 11:07 PM
AVS Special Member
 
p5browne's Avatar
 
Join Date: Jun 2008
Location: Sudbury, Ontario, CANADA
Posts: 4,014
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 183 Post(s)
Liked: 204
New Calman Version 5.3.5 RC3 Build #1578

http://color.spectracal.com/downloads/

Fixed Issues

June 5, 2014

Further refined the Pattern Delay Optimization. This should help improve the results by removing luminance from the calculations.
Resolved an issue with the i1Pro and i1Pro 2 that would cause CalMAN to be unable to communicate with the meter if CalMAN attempted to connect to it a second time
p5browne is offline  
post #772 of 825 Old 06-09-2014, 06:15 AM
AVS Special Member
 
p5browne's Avatar
 
Join Date: Jun 2008
Location: Sudbury, Ontario, CANADA
Posts: 4,014
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 183 Post(s)
Liked: 204
What is this about 21XX Radiance will put out 4K, but won't accept it on the Inputs - from someone mentioning it on the Calman Forum?
p5browne is offline  
post #773 of 825 Old 06-09-2014, 06:20 AM
AVS Club Gold
 
Mark_H's Avatar
 
Join Date: Aug 2001
Location: England
Posts: 1,933
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 16
It's all in the model numbers...

2XXX = 2K input
2X2X = 2K in/2K out
2X4X = 2K in/4K out

Presumably there will be a:

4XXX = 4K input

At some stage...

My cinema: The Cave!

My kit: 15' 2.35:1 Screen Research CP2 4-way mask, Sony vw1000es, Lumagen 2144, Meridian 861/621/7x5500/2xSW5500

Mark_H is offline  
post #774 of 825 Old 06-09-2014, 07:01 AM - Thread Starter
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,430
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 508
Quote:
Originally Posted by Mark_H View Post

It's all in the model numbers...

2XXX = 2K input
2X2X = 2K in/2K out
2X4X = 2K in/4K out

Presumably there will be a:

4XXX = 4K input

At some stage...

It will be released a Radiance PRO Model based on HDMI 2.0 Input/Output Transmission Chipsets with HDCP 2.2 support. We will have more news about it at the Q4 2014.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is offline  
post #775 of 825 Old 06-13-2014, 02:01 PM - Thread Starter
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,430
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 508
CalMAN 5.3.0 RC4 Build: 1581 Released @ 13 June 2014

New Features

Added support for the NEW MediaForge™ On-Demand Steaming Pattern Generator

Fixed Issues

Reverted to previous SDK for the i1Pro 2 to resolved issues experienced with RC2 and RC3.

Resolved an issue that would cause CalMAN to crash when loading certain workflows.

Download CalMAN 5 for Business 5.3.5 RC4 (Build 1581)

Download CalMAN 5 for Home Video Calibration 5.3.5 RC4 (Build 1581)

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is offline  
post #776 of 825 Old 06-19-2014, 03:30 PM - Thread Starter
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,430
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 508
Lumagen Radiance (XE/XE+/XE3D/XE/XE+/XE3D/Mini-3D/XD/XD3D) New 060214 Firmware Update


Release Notes


Precision improvements were made for color accuracy including test patterns. (Update time ~9 minutes @57k)

Download Link

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is offline  
post #777 of 825 Old 06-19-2014, 11:32 PM
AVS Special Member
 
p5browne's Avatar
 
Join Date: Jun 2008
Location: Sudbury, Ontario, CANADA
Posts: 4,014
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 183 Post(s)
Liked: 204
Colours Made Better

All Model Colour Pattern improvements with resulting Deeper, and Bolder Colours

PQ was seen by Eyeballing after a DDC Reset - will check my results tonight from last night's calibration. Darn issues with 100% Blue rising after the 125 Point 3D LUT calibration. Being it was 3AM when finished, will leave and check tonight as well.

Looking good!

Last edited by p5browne; 06-20-2014 at 10:03 PM.
p5browne is offline  
post #778 of 825 Old 06-24-2014, 03:07 AM - Thread Starter
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,430
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 508
Lumagen Radiance (XE/XE+/XE3D/XE/XE+/XE3D/Mini-3D/XD/XD3D) New 060914 Firmware Update

Release Notes

Fixes up a few bugs in some of the test patterns in the previous release. (Update time ~3 minutes @57k)

Download Link

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is offline  
post #779 of 825 Old 06-26-2014, 07:26 AM
AVS Special Member
 
p5browne's Avatar
 
Join Date: Jun 2008
Location: Sudbury, Ontario, CANADA
Posts: 4,014
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 183 Post(s)
Liked: 204
Quote:
Originally Posted by ConnecTEDDD View Post
Lumagen Radiance (XE/XE+/XE3D/XE/XE+/XE3D/Mini-3D/XD/XD3D) New 060914 Firmware Update

Release Notes

Fixes up a few bugs in some of the test patterns in the previous release. (Update time ~3 minutes @57k)

Download Link
Note: FW # 061014 for the 20XX and 21XX Models
p5browne is offline  
post #780 of 825 Old 07-03-2014, 12:56 PM - Thread Starter
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,430
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 508
CalMAN 5.3.5 (Final Release) Build: 1597 Released @ 03 July 2014

New Features

Added support for Video Levels on the EIZO for 1D LUTs.

Added new Monitor-Direct workflow for HP DreamColor Z27x and EIZO ColorEdge monitors. This supports the internal LUTS of these monitors as well as the creation of ICC Profiles.

Changes

CalMAN now supports the JVC x5 and x6 series CMS DDC controls in native ranges.

Performance improvements were made on the Radiance for 3D LUTs

Resolved Issues

Resolved issue which caused a crash when profiling meters in Single Pass.

Download Link:

Download CalMAN 5 for Business / ColorChecker 5.3.5 (Build 1597)

Download CalMAN 5 for Home Video Calibration 5.3.5 (Build 1597)
Manni01 likes this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is offline  
Reply Display Calibration

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