HCFR - Open source projector and display calibration software - Page 166 - AVS Forum | Home Theater Discussions And Reviews
Baselworld is only a few weeks away. Getting the latest news is easy, Click Here for info on how to join the Watchuseek.com newsletter list. Follow our team for updates featuring event coverage, new product unveilings, watch industry news & more!


Forum Jump: 
 445Likes
Reply
 
Thread Tools
post #4951 of 6505 Old 11-29-2014, 10:07 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,858
Mentioned: 65 Post(s)
Tagged: 0 Thread(s)
Quoted: 982 Post(s)
Liked: 824
Quote:
Originally Posted by pisymbol View Post

So if I turn it off and raise the backlight to hit 35 ftL and then turn if back on and lower the backlight to continually achieve 35 fTL, shouldn't I be able to do a BT.1886 10-pt graysweep?
It would make things a lot easier if you post save files showing what settings you used and demonstrating what you having a problem with. Obviously you are going to get two different gamma curves if you compare local dimming on to local dimming off, independent of what your peak white value is.
zoyd is online now  
Sponsored Links
Advertisement
 
post #4952 of 6505 Old 11-29-2014, 12:15 PM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,925
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 209
Quote:
Originally Posted by pisymbol View Post
Works fine, but that defeats the whole purpose of having an active dimming TV.
In yet another of my "controversial" positions, I've come to the conclusion that schemes such as "local dimming" or "auto iris" do more harm than good. I think it is more important to have a stable transfer function that's been well tuned wrt black-level of the set. IMO, "dimming" is counterproductive, giving the illusion of "better" blacks, when it's actually just increasing the crush-factor of shadow details.
HDTVChallenged is offline  
post #4953 of 6505 Old 11-29-2014, 01:26 PM
AVS Special Member
 
LastButNotLeast's Avatar
 
Join Date: Feb 2007
Location: 08077
Posts: 6,131
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 795 Post(s)
Liked: 565
That was going to be my next question.
Thanks.

Did you really need to quote that entire post in your reply?
Downloadable FREE demo discs: Demonstration Blu-Ray Discs (Independently Authored)
Welcome to AVS - Get out while you still can!
LastButNotLeast is offline  
post #4954 of 6505 Old 11-29-2014, 02:04 PM
Member
 
Kamikaze_Ice's Avatar
 
Join Date: Jan 2007
Posts: 44
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 11
I have a question about the results of comparing my measurements between to the default/built in pattern generator (GDI) and Madashi's MadVR/MadTPG.
I would have tried Overlay, but it just refuses to work (Error # 88760168), but I've not put effort into fixing this issue.

Display: LG 55LM7600, Device: ColorMunki Display, Source: PC (Nvidia GTX 770, Win7 x64), Target range:0-255, Signal Chain: PC>Denon 4520CI>HDTV (note: Denon is set to "Pure Direct" mode so it does not touch HDMI video at all).
I believe I have resolved any PC colorspace quirks (forced 0-255 at all times from Nvidia card). By eye, all grayscale step/ramps look the same when comparing the same file from my phone, consoles (PS3, X360, PCs at work and school set to Adobe's sRGB standard as far as I'm aware), so this is all based on my eyes telling me that I have a correct 0-255 video chain from start to end of my chain.
I use MPC-BE (MPC-HC fork with slightly different GUI), and MadVR is the renderer I use. This does not mean I only care how it looks, I just want everything from the PC to be untouched.


So what is "untouched"? I assumed the GDI would be the answer, and is what I've currently calibrated to. Because I have OCD, I love/hate to check, crosscheck, recheck things and research options and view the opinion of others and add all of that information with my own experiences to make my own conclusion on my results.

In a nutshell, what I'm asking is simply "Why do my GDI results differ from the MadTPG results?" Does MadTPG use a YCbCr (or other?) signal computed at 16-235/0-255 (or whatever you set the display to in MadVR preferences)? Am I right in assuming GDI as being the most faithful to 0-255?

Note: This is in reguards to Gamma. In my screenshot I show the Gamma ramp plots, and put a box around the average/max dE in the header. I'm unsure why the numbers are lower for MadTPG when the plot looks worse and this is the source of my confusion.

Off topic: Interestlingly, my LM7600 has exceptional control over grayscale. 2 & 20 point, and they stack. Enter the service mode and you gain another set of 2 & 20 point controls that all work together while having seperate values from eachother. Is this unusual (service menu + user menu exclusive to one another yet working together)?
Attached Thumbnails
Click image for larger version

Name:	hcfr gdi madtpg comparison.jpg
Views:	121
Size:	368.2 KB
ID:	392218  
Kamikaze_Ice is offline  
post #4955 of 6505 Old 11-29-2014, 02:12 PM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,858
Mentioned: 65 Post(s)
Tagged: 0 Thread(s)
Quoted: 982 Post(s)
Liked: 824
zoyd is online now  
post #4956 of 6505 Old 11-29-2014, 04:48 PM
Member
 
Kamikaze_Ice's Avatar
 
Join Date: Jan 2007
Posts: 44
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 11
Sorry, I can't help it. It's part of my job as a Land Surveyor where one must be a real nitpicker to debate over hundredths of an inch while factoring in measurement error from multiple sources (such a small error can snowball to have a tenth or vastly more for final error, and that can get very bad depending on the job)

While they may be within repeatable error tolerance of eachother, and one couldn't tell the difference between the without measurements, I'd still like to know the "why" it's different. The both generators both return the same general shape on repeated tests (varying by thousandths between the same set), so it's consistently different between the two.

I know you shouldn't pull that string that is sticking up on your clothes, but by gosh it's there I'm gonna examine it. If said string turned out to be from another piece of clothing, why not remove it? (Sorry for the bad analogy. I blame my OCD, I don't mean to nag).


I do have three other questions related to HCFR program funcionality, if you don't mind.
Both questions are from someone using only the internal test pattern generator (GDI) and MadVR/MadTPG. HCFR version is 3.1.6, if it is relevant.

Question 1) View>Test Color. This happens with 10/30/50% columns for gray/red/green/blue.
Steps to reproduce:
starting up HCFR with a fresh file (for the sake of simplicity, I doubt a save from a previous version would matter but I have none to verify that opinion with so I am noting it just incase) then click the 20% gray to set it as your target. Now open the Test Color window, note the RBG value shown (mine states 51/51/51) and click the drop down box--20% gray is already highlighted because it matches the current target value--and select 20% gray and check the value (mine states 51/51/51). No problem here, but now try this with 10% gray (or any other of the 10/30/50 columns), and you will notice the values returned are all different; 10% gray is stated, in the same order as previously performed, as 26/26/26 vs 25/25/25 respectively.
Should they not be the same values?

Question 2) This is a two-part question related to correct use of Continuous Measurements.
a) I have no issues when selecting a value to continuously measure from within the grayscale or primaries & secondary measurement screens, so naturally I assume this interface behavior to behave the same inside the color saturation and color checker pages since they have the same layout.
Should I take this behavior as intentional, ala the arrows "feature" that adjusts the rows in the data table? :P
b) What's up with how both Continuous Measurements, Target/Test Color behave when at the Free Measurement view?
When you select a color, either by clicking it in another measurement screen or using the Test Color window, and perform a Continuous Measure, you would expect that color value to never change outside of user input. This expected behavior is true when you are at any view except Free Measurement view.
Steps to reproduce: Go to any view except Free Measurement and set your color target to 80% Green (0/204/0) and start a Continuous Measurement, after the first measurement is done you do just repeat that same color value test until you decide to stop. Now without changing your color target, switch your view to Free Measurement and start the Continuous Measurements again. After the first 0/204/0 measurement, all measurements after are now 0/255/0.
Is this a possible bug, or am I just using it wrong?


Sorry for being a bother. OCD amplifies my pedantic nature when communicating, so I hope I'm being coherent with my words and I apologize if I've drawn a simple statement or question out longer than it needed to be. Oh damnit, I'm still going on :P

Also, thanks for continuing development.
foodiest likes this.
Kamikaze_Ice is offline  
post #4957 of 6505 Old 11-30-2014, 03:37 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,858
Mentioned: 65 Post(s)
Tagged: 0 Thread(s)
Quoted: 982 Post(s)
Liked: 824
Quote:
Originally Posted by Kamikaze_Ice View Post
While they may be within repeatable error tolerance of eachother, and one couldn't tell the difference between the without measurements, I'd still like to know the "why" it's different.
A combination of instrument noise and display repeatability will limit the precision of the measurement.

Quote:
I do have three other questions related to HCFR program funcionality, if you don't mind.
The internal generator has selectable colors by column when the user is in either the grayscale page or the primaries/secondaries page, any other page is not supported. If the test patch has a different value than the full screen output that's a bug.
zoyd is online now  
post #4958 of 6505 Old 11-30-2014, 05:05 PM
AVS Special Member
 
fafrd's Avatar
 
Join Date: Jun 2002
Posts: 6,011
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 1287
Quote:
Originally Posted by zoyd View Post
A combination of instrument noise and display repeatability will limit the precision of the measurement.



The internal generator has selectable colors by column when the user is in either the grayscale page or the primaries/secondaries page, any other page is not supported. If the test patch has a different value than the full screen output that's a bug.

I have been learning to calibrate using HCFR and the GCD patterns. HFCR is a fantastic package for beginning DIY calibrators so to whoever contributed to putting it together, thanks.


I have wanted to try to calibrate the 'near black' but GCD only has near-white. I have never used the pattern generation. re there pros and cons of using the internal pattern generation versus using GCD? Will the internal pattern generation allow me to measure near-black? Are there ofther sources of the near-black patterns?


I believe I may also have the same question as that posted above - is there a reason that 'continuous measure' cannot be displayed when using color checker to look at real-time impact of primary/secondary changes on specific color accuracy?

Last edited by fafrd; 11-30-2014 at 08:47 PM.
fafrd is online now  
post #4959 of 6505 Old 11-30-2014, 06:05 PM
Senior Member
 
22point8's Avatar
 
Join Date: Mar 2012
Location: London, UK
Posts: 252
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 139 Post(s)
Liked: 64
Quote:
Originally Posted by fafrd View Post
I have been learning to calibrate using HCFR and the GCD patterns. HFCR is a fantastic package for beginning DIY calibrators so to whoever contributed to putting it together, thanks.


I have wanted to try to calibrate the 'near black' but CDS only has near-white. I have never used the pattern generation. re there pros and cons of using the internal pattern generation versus using GCD? Will the internal pattern generation allow me to measure near-black? Are there ofther sources of the near-black patterns?


I believe I may also have the same question as that posted above - is there a reason that 'continuous measure' cannot be displayed when using color checker to look at real-time impact of primary/secondary changes on specific color accuracy?
You can, if you have already measured the colour and its displayed on the CIE triangle, if you display that pattern i.e light skin and look at the CIE diagram and activate free measures it will be a yellow dot on the diagram, it will move around as you adjust the cms, colour, tint or flesh tone setting (samsung).

@zoyd , I discovered that the flesh tone setting influences the luma of the colorchecker SKIN patterns.
22point8 is offline  
post #4960 of 6505 Old 11-30-2014, 08:49 PM
AVS Special Member
 
fafrd's Avatar
 
Join Date: Jun 2002
Posts: 6,011
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 1287
Quote:
Originally Posted by 22point8 View Post
You can, if you have already measured the colour and its displayed on the CIE triangle, if you display that pattern i.e light skin and look at the CIE diagram and activate free measures it will be a yellow dot on the diagram, it will move around as you adjust the cms, colour, tint or flesh tone setting (samsung).

@zoyd, I discovered that the flesh tone setting influences the luma of the colorchecker SKIN patterns.

I knew that, but it is not as good as the primary/secondary matrix display where you can see x-y error, luminance, and dE. I believe the color chart is only providing a very subjective measure of x-y error...
fafrd is online now  
post #4961 of 6505 Old 11-30-2014, 10:19 PM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,925
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 209
Quote:
Originally Posted by fafrd View Post
I have wanted to try to calibrate the 'near black' but GCD only has near-white. I have never used the pattern generation. re there pros and cons of using the internal pattern generation versus using GCD? Will the internal pattern generation allow me to measure near-black? Are there ofther sources of the near-black patterns?
The AVS-HD disk has the near-black patterns. Just be sure to check the "use round down" box with the AVS disk.
fafrd likes this.
HDTVChallenged is offline  
post #4962 of 6505 Old 12-01-2014, 07:58 AM
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,830
Mentioned: 1 Post(s)
Tagged: 1 Thread(s)
Quoted: 296 Post(s)
Liked: 727
Quote:
Originally Posted by fafrd View Post
I have wanted to try to calibrate the 'near black' but GCD only has near-white.

Hi, Ted's LightSpace CMS Calibration Disk @ the CalMAN's Session there is available a Chapter for 6-Point Near Black.

It contains the following patches:

0% Black (RGB Triplet: 16.16.16)
0.5% Gray (RGB Triplet: 17.17.17)
1% Gray (RGB Triplet: 18.18.18)
2% Gray (RGB Triplet: 20.20.20)
3% Gray (RGB Triplet: 23.23.23)
4% Gray (RGB Triplet: 25.25.25)
5% Gray (RGB Triplet: 27.27.27)

These targets are matching the HCFR Engine's RGB Triplets also.

HCFR has Near Black measurements. (for 0/1/2/3/4/5 exept the 0.5%)

So HCFR users will have to skip the 0.5% and measure all the others.

This has been confirmed by Zoyd at this post.

The precision of the disk has been confirmed by professional calibrator who examined using his DVDO AVLab TPG the disk for digital errors and he measured the Near Black Patterns also, among with others.

Below are the results of that testing: (his original post is here)


Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #4963 of 6505 Old 12-01-2014, 08:42 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,858
Mentioned: 65 Post(s)
Tagged: 0 Thread(s)
Quoted: 982 Post(s)
Liked: 824
Quote:
Originally Posted by fafrd View Post
I knew that, but it is not as good as the primary/secondary matrix display where you can see x-y error, luminance, and dE. I believe the color chart is only providing a very subjective measure of x-y error...
Hi - I answered this in my response to Kamikaze_Ice, only grayscale and primaries/secondaries are currently supported for automatic targeting. I plan to add saturation patterns (including the color checker) at some point but it may have to wait in order to support a new feature in ArgyllCMS.
zoyd is online now  
post #4964 of 6505 Old 12-01-2014, 06:31 PM
Advanced Member
 
HX_Guy's Avatar
 
Join Date: Sep 2003
Location: Phoenix AZ
Posts: 692
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 175 Post(s)
Liked: 40
Hey guys we've been having an ongoing discussion in the F8500 thread on this weird issue but I figured I'd ask in here too and get it in front of some eyes that don't read that thread.

I recently calibrated my 64F8500 using HCFR, AVS HD 709 disk, i1D3 meter and my Panasonic blu-ray player (which only plays the AVS disk at 60p, not 24p - it's a known issue). I was able to get my greyscale and gamma curve where I wanted it and thought the picture looked great.

Then yesterday I received my new Denon 1713UD blu-ray player and to my surprise all my previous settings were now way way off. I thought maybe the Denon was malfunctioning, so I went out and bought a Sony blu-ray to also test out but that gave me the same results as the Denon, so the Panasonic is the odd one out it seems.

The tests give completely different results based on if you are playing at 1080p 60 frames vs 1080p 24 frames which is very weird, no? My Panasonic doesn't play the AVS HD disk at 24 frames but I can force it into 24 frames through my receiver's conversion vs pass-through.

Which is the correct one to use then if you want to calibrate for regular television as well which comes in at 720p 60 frames and 1080i 60 frames?

I did a bunch of back to back testing between the Denon and the Panasonic (didn't include the Sony once I confirmed it gave very similar results to the Denon) and here is what I got. What do you guys make of this?


Denon 60p (top) vs Panasonic 60p (bottom):











Denon 24p (top) vs Panasonic 24p (bottom):








HX_Guy is offline  
post #4965 of 6505 Old 12-01-2014, 06:33 PM
Member
 
Kamikaze_Ice's Avatar
 
Join Date: Jan 2007
Posts: 44
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 11
Quote:
Originally Posted by zoyd View Post
The internal generator has selectable colors by column when the user is in either the grayscale page or the primaries/secondaries page, any other page is not supported.
I propose a consideration for changing this behavior. I understand HCFR is, for lack of a better word, for "advanced" users, but some aspects could be improved upon.
Some examples of candidates are:
1) How Continuous Measure & Test Color interact with eachother and other measurement pages. Since grayscale or primaries/secondaries would most likely be one of the first pages a user would interact with, and are supported, one would expect things to behave just as they have up to this point, as there is nothing informing the user that functionality has changed.
2) The arrow buttons that add/remove rows to the measurements window. I like pressing all the buttons and seeing what does what, in general, but until clicking them I would never know, let alone expect, that those arrows were for another window table. I first assumed it was a generic next/previous/up/down button linked to the drop down menu because of their relative location and being in the same window table.
3) Zooming when viewing charts/graphs. Mouse wheel support and working keyboard shortcuts (Num pad +/-, while identified as a shortcut for zooming, does not work for me.) The CIE chart does have mouse wheel support, the scroll/zoom action is reversed (Typically it's user relative instead of display relative). A higher maximum zoom would be nice to have as well.

Please note that I am not demanding these, only proposing these and other quirks be given consideration for some point in the future.

Quote:
Originally Posted by zoyd View Post
If the test patch has a different value than the full screen output that's a bug.
I think I'm misunderstanding your response, so please bear with me as I try to clarify my observation. And to help, see the attached picture.
I've added a magenta box to highlight the color values shown in the Test Color window.
Ignoring the difference in naming (50% white vs 50% grey), you can see that there is a discrepancy in the values listed between the 50% white grayscale table listing and the 50% grey palette preset swatch. As I mentioned previously, this difference applies to the 10/30/50% columns of the Test Color palette.


Another quirk I've discovered is the Advanced>Test Patterns>Display Patterns do not comply with the primary monitor selected by either the system or the active HCFR file (changing the internal generators Target Screen does nothing).
The Animated white/black tests behave as expected (changing the target screen works).


Quote:
Originally Posted by HDTVChallenged View Post
In yet another of my "controversial" positions, I've come to the conclusion that schemes such as "local dimming" or "auto iris" do more harm than good. I think it is more important to have a stable transfer function that's been well tuned wrt black-level of the set. IMO, "dimming" is counterproductive, giving the illusion of "better" blacks, when it's actually just increasing the crush-factor of shadow details.
I may be misinterpreting you, so let me apologize in advance.
Note that this is my opinion, and solely based on how MY display, LG 55LM7600 (WLED IPS panel with edge-lit with, I think, 12 or so local zones), behaves in regard to local dimming.
Using my ColorMunki Display, and backlight increased to produce white at ~100 nits/cd m^2 for my viewing space, black measures 0.14 nits/cd m^2. This is extremely noticable, and being an edge-lit display you want to puke anytime the overall screen is dark from the uneven backlighting. You want that local dimming. Three settings (low/med/high), and off, are your options. High and Medium are very aggressive and signifigantly change the screen, and while I've not fully explored their effects outside of a few that produced banding on gradients immediately, I think one should stay away from such aggressive settings as they appeared to affect ALL stimulus levels with a variable range depending on apl and how much black per zone. I thought about trying to log this behavior, but I lack the experience and knowledge to even know how to observe useful information.
Low, however, is much nicer to the displayed scene, only effecting ~≤30% white levels. Using the same settings as above, black now measures 0.045 nits/cd m^2@field, 0.071 nits/cd m^2@10%window 20% APL. High with fields and lower backlight for a darker room can get low enough to not measure with my ColorMunki.
With my display, Local Dimming is not producing the "illusion" of better blacks, it actually produces better blacks by lowering the backlight in zones but not turning it off (the latter of which I would consider an illusion. Since Local Dimming at Low only affects the the bottom end, I think one should take that into consideration when adjusting grayscale/gamma if one plans to use said feature AND it their display behaves similar to mine.
If you use Local Dimming, with similar behavior, you might as well adjust that affected area for dark screentime to preserve as much white balance/gamma as you can, I mean if that zone is dark enough to trigger dimming then adjusting for SOME dimming has to be better than not doing so (read: higher chance of crushing) and all the screen zones that don't dim aren't effected.

I'm still exploring options to account for this dimming myself, unknown if Local Dimming is triggered by zone APL, luminocity, average zone black level + contrast, a mix of the mentioned or something else.
Attached Thumbnails
Click image for larger version

Name:	test color.jpg
Views:	83
Size:	190.1 KB
ID:	396370  

Last edited by Kamikaze_Ice; 12-01-2014 at 07:45 PM. Reason: Forgot response to HDTVChallenged
Kamikaze_Ice is offline  
post #4966 of 6505 Old 12-02-2014, 12:21 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,925
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 209
Quote:
Originally Posted by Kamikaze_Ice View Post
I may be misinterpreting you, so let me apologize in advance.
No ... you're probably not misinterpreting me.

Quote:
Note that this is my opinion, and solely based on how MY display, LG 55LM7600 (WLED IPS panel with edge-lit with, I think, 12 or so local zones), behaves in regard to local dimming. Using my ColorMunki Display, and backlight increased to produce white at ~100 nits/cd m^2 for my viewing space, black measures 0.14 nits/cd m^2. This is extremely noticable, and being an edge-lit display you want to puke anytime the overall screen is dark from the uneven backlighting. You want that local dimming.
I understand your concern. The problem here is that it's nearly impossible to predict the behavior of the dimming effect, which makes it difficult/impossible to adjust the EOTF/Gamma response. The good news for you is that (I'm guessing) probably your LG has at least a 10pt greyscale adjustment, so you may have a chance of incorporating the dimming into the overall EOTF/gamma response. To do this, you are going to have to calibrate with the dimming turned on *and* you are going to have to experiment with different grey targets types and sizes (APL targets vs 4% area targets vs. etc, etc.) Basically, it's the same thing that plasma owners go through with ABL issues, only at the dark end instead of bright end of the luminance curve. The problem is finding a target type that will represent actual program material. Then again, if the display is treating "local" areas differently, then there is simply no consistent EOTF/gamma response to be found. So ...

Ok ... just spit-balling ... I might try nearly filling up a single zone ... or half of a single zone with the grey target, etc and leave the rest of the screen area black, then calibrate with local dimming turned on. Or ... ... ... I'm giving myself a headache, better quit typing before I blow a gasket.
HDTVChallenged is offline  
post #4967 of 6505 Old 12-02-2014, 03:35 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,858
Mentioned: 65 Post(s)
Tagged: 0 Thread(s)
Quoted: 982 Post(s)
Liked: 824
I think calibrating with local dimming on would be a waste of time. If the effect can be simulated by varying the backlight (and that this actually changes the white point) I would do a few calibrations with local dimming off and at various backlight levels. I would then average the resultant gain/bias settings together.
zoyd is online now  
post #4968 of 6505 Old 12-02-2014, 09:33 AM
AVS Special Member
 
xvfx's Avatar
 
Join Date: Feb 2013
Location: UK
Posts: 1,201
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 349 Post(s)
Liked: 195
Quote:
Originally Posted by xvfx View Post
I meant to ask before and had forgotten. What are the files colors_500.csv and 250.csv for? Was this the reason why I got an error trying to run the skin setting? Have they to be placed in the HCFR main folder?
Quote:
Originally Posted by zoyd View Post
They are example random color sets. Rename one of them to colors.csv and place in the main folder. Then choose "user" from the color checker drop down to measure them.
Seems I still can't get rid of this problem. Soon as I run the SKIN drop down and try to do a sweep. Up pops the Error: during generator initialisation. For DVD images.

I'm using Power Law, CIE76uv Absolute Y w/gamma and CIE94.
xvfx is online now  
post #4969 of 6505 Old 12-02-2014, 09:40 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,858
Mentioned: 65 Post(s)
Tagged: 0 Thread(s)
Quoted: 982 Post(s)
Liked: 824
zoyd is online now  
post #4970 of 6505 Old 12-02-2014, 09:46 AM
AVS Special Member
 
xvfx's Avatar
 
Join Date: Feb 2013
Location: UK
Posts: 1,201
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 349 Post(s)
Liked: 195
So it's only user? Oh and colors_250 file is missing… Can't access it.
xvfx is online now  
post #4971 of 6505 Old 12-02-2014, 10:05 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,858
Mentioned: 65 Post(s)
Tagged: 0 Thread(s)
Quoted: 982 Post(s)
Liked: 824
These saturation measurement selections are only available when using the view images generator:

AXIS (8 luminance levels per primary color)
SKIN (24 PANTONE skin colors)
USER (colors defined in colors.csv)

I'll reupload the 250 color example after sourceforge is back online.
zoyd is online now  
post #4972 of 6505 Old 12-02-2014, 03:51 PM
AVS Special Member
 
fafrd's Avatar
 
Join Date: Jun 2002
Posts: 6,011
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 1287
Quote:
Originally Posted by zoyd View Post
Hi - I answered this in my response to Kamikaze_Ice, only grayscale and primaries/secondaries are currently supported for automatic targeting. I plan to add saturation patterns (including the color checker) at some point but it may have to wait in order to support a new feature in ArgyllCMS.

Thanks.

HCFR is a very advanced calibration program, but here are two areas I am struggling with:

1/ Primary/Secondary calibration - my TV offer Hue, Saturation, and Brightness controls but while Calman shows distinct errors for Hue, Saturation and Brightness which make it clear which aspect is most responsible for the total dE errors of each primary or secondary, in HCFR, I only see delta xy and delta E, which make it more difficult to calibrate (especially brightness) - is there any way to see/understand this additional level of detail within HCFR?

2/ I've seen that most are recommending calibrating my TV (P70) at 75% and I assume that this means using the HDTV Rec 709 (75%/75%) Color space standard rather than the straight HDVT Rec 709. On the GCD disk I see the patterns for Primaries/Secondaries 75% Saturation and 75% Luminance which I believe is what I should be using to calibrate CMS RGBCMY. After this, when I want to run Saturation Sweeps, I am unsure what to do - first rerun Primary/Secondaries using HDTV Rec 709 and then do the sweeps at 100% Luminance, or is there a better way? Advice on the flow to use to best calibrate a TV with a primary that cannot be accurately calibrated at 100% Saturation would be greatly appreciated.
fafrd is online now  
post #4973 of 6505 Old 12-02-2014, 04:19 PM
Member
 
Kamikaze_Ice's Avatar
 
Join Date: Jan 2007
Posts: 44
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 11
Quote:
Originally Posted by HDTVChallenged View Post
Ok ... just spit-balling ... I might try nearly filling up a single zone ... or half of a single zone with the grey target, etc and leave the rest of the screen area black, then calibrate with local dimming turned on. Or ... ... ... I'm giving myself a headache, better quit typing before I blow a gasket.
This is basically where I'm at., the headache part :P
All the unscientific testing I have done up to this point, as in measurements randomized measuring at various zones, APL, average zone luminance, and just visual classification/guesswork as to how everything related, I did start to notice a pattern which leads me to believe for specific behavior types of local dimming one could log meaningful data to determine the most optimal testing methods to have the best setting overall.

I appreciate all the information I've found in this forum, being a lurker here for years, I love the informative posts by people who go out of their way to explain things in detail when they could have made a simple and to the point reply.


And while I appreciate the information and opinions of others, I still wish to respectfully disagree as far as Local Dimming and calibrating my display. My observations of my specific set seem to be less random so I believe calibrating should consider Local Dimming to a degree. I figure doing like Zoyd mentioned a poll and average of results would be better than nothing, and I doubt spending tremendous time to determine my displays "optimal" solution would offer extremely minimal results for the time invested.

As every display is different, I'm only trying to leave my opinion encouraging others to experiment with their own display if they are interested in using Local Dimming. Maybe their display will be have similar, or extremely different than mine, but without first hand observations one shouldn't exclude the possibility of another outcome simply because Mr. Reputable Source and his Degree in Science has observed X, thus hold X as true. I mean I'm not trying to say anyone is lying, but if everyone believed everyone science would never evolve.

In short, Thanks for the help, but I want to be stupid sometimes for fun, but atleast I'm a smarter kind of stupid

*I hope no one takes this the wrong way, my OCD causes me to write at length for a simple sentance saying thanks, seriously. This is why I lurk more than participate in group discussion. Damnit fingers stop typing!*
Kamikaze_Ice is offline  
post #4974 of 6505 Old 12-02-2014, 04:34 PM
Member
 
Kamikaze_Ice's Avatar
 
Join Date: Jan 2007
Posts: 44
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 11
Quote:
Originally Posted by fafrd View Post
Thanks.

HCFR is a very advanced calibration program, but here are two areas I am struggling with:

1/ Primary/Secondary calibration - my TV offer Hue, Saturation, and Brightness controls but while Calman shows distinct errors for Hue, Saturation and Brightness which make it clear which aspect is most responsible for the total dE errors of each primary or secondary, in HCFR, I only see delta xy and delta E, which make it more difficult to calibrate (especially brightness) - is there any way to see/understand this additional level of detail within HCFR?
I think you might want to try the "Use HSV level bars" option (Advanced>Preferences...>General tab>"User Measures" option group has the check box.
This will change the RGB bars to Hue Saturation and Value/Brightness/Luminosity, and is what I use for my display's CMS (each primary and secondary has a +/-50 slider to adjust each for Saturation, Tint and Luminance values). Just adjust according to the bar values similar to doing white balance.

Quote:
Originally Posted by fafrd View Post
2/ I've seen that most are recommending calibrating my TV (P70) at 75% and I assume that this means using the HDTV Rec 709 (75%/75%) Color space standard rather than the straight HDVT Rec 709. On the GCD disk I see the patterns for Primaries/Secondaries 75% Saturation and 75% Luminance which I believe is what I should be using to calibrate CMS RGBCMY. After this, when I want to run Saturation Sweeps, I am unsure what to do - first rerun Primary/Secondaries using HDTV Rec 709 and then do the sweeps at 100% Luminance, or is there a better way? Advice on the flow to use to best calibrate a TV with a primary that cannot be accurately calibrated at 100% Saturation would be greatly appreciated.
I've been searching lightly for a same general answer for the whole 75/75 process. I just assumed one would use the standard 709 profile for the initial setting of primaries (and secondaries if CMS control allows) and then switch to 75/75 for error adjustment as the error focus would shift to the most common range of color.

As I don't use any calibration disc, should one change the inbuilt pattern generator's "Pattern Intensity %" when doing this? My assumption was no, but would not mind having a definite answer quicker than searching.
Kamikaze_Ice is offline  
post #4975 of 6505 Old 12-02-2014, 04:42 PM
AVS Special Member
 
fafrd's Avatar
 
Join Date: Jun 2002
Posts: 6,011
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 1287
Quote:
Originally Posted by ConnecTEDDD View Post
Hi, Ted's LightSpace CMS Calibration Disk @ the CalMAN's Session there is available a Chapter for 6-Point Near Black.

It contains the following patches:

0% Black (RGB Triplet: 16.16.16)
0.5% Gray (RGB Triplet: 17.17.17)
1% Gray (RGB Triplet: 18.18.18)
2% Gray (RGB Triplet: 20.20.20)
3% Gray (RGB Triplet: 23.23.23)
4% Gray (RGB Triplet: 25.25.25)
5% Gray (RGB Triplet: 27.27.27)

These targets are matching the HCFR Engine's RGB Triplets also.

HCFR has Near Black measurements. (for 0/1/2/3/4/5 exept the 0.5%)

So HCFR users will have to skip the 0.5% and measure all the others.

This has been confirmed by Zoyd at this post.

The precision of the disk has been confirmed by professional calibrator who examined using his DVDO AVLab TPG the disk for digital errors and he measured the Near Black Patterns also, among with others.

Below are the results of that testing: (his original post is here)

Ted,


that seems like a good disk - thanks for the pointer. I will be willing to spend $30 for a good calibration disk but want to continue to progress on my learning curve a bit more before making that commitment. Will analyze your website more closely and get back to you with any questions.
ConnecTEDDD likes this.
fafrd is online now  
post #4976 of 6505 Old 12-02-2014, 04:43 PM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,858
Mentioned: 65 Post(s)
Tagged: 0 Thread(s)
Quoted: 982 Post(s)
Liked: 824
Quote:
Originally Posted by Kamikaze_Ice View Post
As I don't use any calibration disc, should one change the inbuilt pattern generator's "Pattern Intensity %" when doing this? My assumption was no, but would not mind having a definite answer quicker than searching.
That's correct, leave at the internal pattern settings at 100% when using the special color space.
zoyd is online now  
post #4977 of 6505 Old 12-03-2014, 12:42 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,925
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 209
Quote:
Originally Posted by zoyd View Post
I think calibrating with local dimming on would be a waste of time. If the effect can be simulated by varying the backlight (and that this actually changes the white point) I would do a few calibrations with local dimming off and at various backlight levels. I would then average the resultant gain/bias settings together.
That might be a saner way to go. I was thinking more about the "fitting" the gamma/EOTF curve more correctly.

Considering a simple power-law curve, one really wouldn't want the gamma at 2.2 from 40% to 100% with 2.4 @30%, 2.6 @20% and 2.8@10% which is pretty much the type of curve that "dimming" technologies produce. And we know from BT1886, that this is actually the *opposite* of the curve one really needs visually. Hence, my position that dimming is somewhat counterproductive ... furthermore, "local" dimming seems to somewhat pointless if there are "high" luminance objects/areas on the screen. So ... YMMV ... To be honest, this may be a no-win situation.

PS: I wouldn't have an issue with the 0.14Nit black-level, it sounds like the "unevenness" is the bigger problem and that may not be fixable.
HDTVChallenged is offline  
post #4978 of 6505 Old 12-03-2014, 01:47 AM
Senior Member
 
Andrei_VVB's Avatar
 
Join Date: Mar 2010
Posts: 247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 28
OT: a better way to increase the perceived black level will be to have some sort of ambient lighting, preferably behind the set and calibrate and watch without dimming activated.
Andrei_VVB is offline  
post #4979 of 6505 Old 12-04-2014, 07:46 AM
Advanced Member
 
vega509's Avatar
 
Join Date: Apr 2003
Location: The Northeast
Posts: 955
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 123 Post(s)
Liked: 49
Quote:
Originally Posted by HDTVChallenged View Post
That might be a saner way to go. I was thinking more about the "fitting" the gamma/EOTF curve more correctly.

Considering a simple power-law curve, one really wouldn't want the gamma at 2.2 from 40% to 100% with 2.4 @30%, 2.6 @20% and 2.8@10% which is pretty much the type of curve that "dimming" technologies produce. And we know from BT1886, that this is actually the *opposite* of the curve one really needs visually. Hence, my position that dimming is somewhat counterproductive ... furthermore, "local" dimming seems to somewhat pointless if there are "high" luminance objects/areas on the screen. So ... YMMV ... To be honest, this may be a no-win situation.

PS: I wouldn't have an issue with the 0.14Nit black-level, it sounds like the "unevenness" is the bigger problem and that may not be fixable.
I've been playing with exactly this on my Samsung 8550. I have had some success engaging the SmartLED option and correcting the gamma curve with the 10pt adjustments. The key on this set was to adjust 100% until most of the other points are as close as possible, then work on the rest. It didn't take long once I discovered this. The drawback is the gamma level will be limited to a narrow range, 2.29 to 2.38 no matter where it is set once the gamma curve is corrected.
Attached Thumbnails
Click image for larger version

Name:	gamma.jpg
Views:	94
Size:	453.9 KB
ID:	401618  
vega509 is offline  
post #4980 of 6505 Old 12-05-2014, 11:12 AM
Advanced Member
 
vega509's Avatar
 
Join Date: Apr 2003
Location: The Northeast
Posts: 955
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 123 Post(s)
Liked: 49
when calibrating 3D mode, should the glasses be turned on? or left off and just placed in front of the meter?
vega509 is offline  
Sponsored Links
Advertisement
 
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