HCFR - Open source projector and display calibration software - Page 135 - AVS Forum
Forum Jump: 
 27Likes
Reply
 
Thread Tools
post #4021 of 4035 Old Yesterday, 05:29 AM
Member
 
harlekin's Avatar
 
Join Date: May 2007
Posts: 79
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 12
Quote:
Originally Posted by Stereodude View Post
I'm pretty sure there's a BT1886 topic that you can spam with your tinfoil conspiracy hat theories on the ITU and BT1886 instead of this one.
I'm sorry, either this is science that can be challenged by looking at the impact and the reproducibility of data.

Or this is a believe system where anyone finding flaws that especially impact the range of devices that are available on the market gets shouted out the way you elegantly tried to short out the criticism at hand - trollingly combined with the most severe of accusations > wrong thread. (Essentially doubling down on "dont talk about it here", and "dont talk about it anyways".)

Sorry - I don't pander to

- threads where as a suggestion videos are linked in which a member of the body that formulates the suggestions that then become the standard - clearly doesnt know a thing about the standard he actually was booked to talk an hour about are STILL LINKED as the explanations one should listen to to.

- as alternatives papers and (calibrating software manufacturers) websites are linked in which a surface explanation is given that doesnt add up with the intent (at least for the vast majority of devices out in the market).
--

And especially no explanation is given upon the results that this produces on devices on the market - which display huge variation dependent on their black point only. While brightness doesnt seem to affect the standard at all, and there is no explaination given regarding bt1886 and room light levels.
--

That added to the _incredible_ fact, that this standard is getting shoe horned in 10 years after the first Blurays were released. Which brings up the _really_ sensitive question - what were you guys calibrating your devices to before? (Ups, we totally forgot about gamma compensation..) - which now should be implemented on the end users side rather then at post production?

Last edited by harlekin; Yesterday at 05:38 AM.
harlekin is offline  
Sponsored Links
Advertisement
 
post #4022 of 4035 Old Yesterday, 05:51 AM
Member
 
harlekin's Avatar
 
Join Date: May 2007
Posts: 79
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 12
In a more measured tone:

HCFR is the only open source approach this whole industry is confronted with - as in "look at it and be encouraged to learn what it actually does" - depending on some highly regarded maintainers understanding both theory and implementation.

Now in addition I am pleading for them even to chip in on public explanation. Which normally isn't their job.

Because whats public so far in the bt1886 threads doesnt take into account the actual variations we are seeing at the device level, induced purely by one measurement only - black level.

Why does black level, almost entirely determine the gamma my light tones should be viewed at? Regardless of brightness, contrast or room light.

And if it is truly "compensating" characteristics at mastering - why does it vary that much dependent on solely the black point of the end users device.

Last edited by harlekin; Yesterday at 06:00 AM.
harlekin is offline  
post #4023 of 4035 Old Yesterday, 06:03 AM
Member
 
harlekin's Avatar
 
Join Date: May 2007
Posts: 79
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 12
And what does that mean for the local dimming/pseudo local dimming we see in current TV sets? Does the local dimming setting now impact which gamma points my device should be set to? So is accurate color representation (minimal dE) only reached in certain SCENES?

Questions that ought to be asked when suddenly there is a new default.
harlekin is offline  
post #4024 of 4035 Old Yesterday, 08:05 AM
Member
 
harlekin's Avatar
 
Join Date: May 2007
Posts: 79
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 12
Another fun fact:

Remember Spectracals fact sheet where it states:
"BT.1886 is quickly being adopted in the field, particularly in the broadcast, production, and post-production industries." ?
http://www.spectracal.com/Documents/BT.1886.pdf

Well ask yourself HOW its "quickly being adopted".

If bt1886 is seen as a correction curve, not applied in post, but only on the end users side - that counteracts "most material being produced due to analog standards/mastered on CRTs" the answer quickly becomes...

By them not doing _anything_ apart from them still using their old CRTs/analog camera equipment and forcing gamma correction onto the end user which prior to this had perfectly linear gamma tracking on his devices.

Oh and how quickly they are adopting this, by not doing anything! Its simply astonishing!

And on the end users side? Gamma becomes an approximation (heavily influenced by the least accurate factor measurable - black point, in a comparatively huge scale of variance produced by the black points that devices out there actually have, or in less ideal scenarios - were wrongly measured or approximated) of an approximation (the median CRT that never existed) to then produce "accurate colors" according to something that was first invented as a fudge factor (by definition!), because gamma on LCDs/Plasma was perfectly linear.

Eh - guys?
harlekin is offline  
post #4025 of 4035 Old Yesterday, 08:35 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,299
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 107
Quote:
Originally Posted by harlekin View Post
Found user HDTVChallenged's "BT1886CalcV3.xls" which (if the math is accurate) I find extremely helpful in visualizing changes.

Scratch all my questions except for one.

Details on how a Black (0,04 cd/m2) White (120/140 cd/m2) device reacts under bt1886 compared to a lower (0,004 cd/m2) black level device - regarding "room light conditions".

Can you factor out room light at all?
I don't understand the question. From my perspective the process of "calibration" assumes that you are using the display in an appropriately light-controlled environment. Therefore, if one were to place the display in a less than appropriately controlled environment, the value of a "calibration" tends toward zero.

In other words, if you insist upon installing your "reference level" plasma in a solarium there's not much the "community" can do for you.

The point of BT1886 is simply to mimic the EOTF characteristics of CRTs on non-CRT devices, because, as it turns out, the simpler power-law method is not exactly the "correct" way of doing that emulation.
HDTVChallenged is offline  
post #4026 of 4035 Old Yesterday, 03:19 PM
AVS Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 5,416
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Liked: 110
Wow, 11 out of the last 13 posts are from harlekin. No offense, but that's not a healthy ratio. Also, this thread is about HCFR specific things, not about general calibration topics. Discussion about whether BT.1886 is a good thing or not should go into a separate thread, please. I'm not interested in such a BT.1886 discussion, just about HCFR. Thank you.
ConnecTEDDD and spacediver like this.
madshi is online now  
post #4027 of 4035 Old Yesterday, 05:04 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 555
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 19 Post(s)
Liked: 70
Quote:
Originally Posted by harlekin View Post
As a visual reference:

How your black level decides upon what your gamma should look like according to bt1886:
Yep. So what ?
gwgill is offline  
post #4028 of 4035 Old Yesterday, 06:28 PM
Member
 
harlekin's Avatar
 
Join Date: May 2007
Posts: 79
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 12
Just to make it publicly known - I find it almost unnerving to see the kind of attitude all those questions remain glanced over. The reaction that was closest to being "on point" was a mere "so what?".

So let me just sum up what no one here finds problematic or unsettling at all:

- Changing the content reproduction chain mid standard, and doing so without much public recognition. If gamma correction was overlooked for this extensive periods (ten years), it is nothing short of a scandal.
- Reintroducing gamma compensation as part of the "end user realm", instead of accounting for it in post.
(Therefore ruining devices with a perfectly linear gamma (what we are aiming for), that could be set according to room light for example.
- Doing it in a way which produces an even higher variance between displays - dependent on solely their black level.
(The logic behind it, that compensates for black crush on displays where neither 2.2, nor 2.4 would introduces any - but now we are calibrating them at 2.0 and 2.1 at the low end.)
- Not having any opinion on the huge nature of variance in curves that bt 1886 produces (*quote*So what?*endquote*)
- Remaining silent about bt1886 not even adhering to other standards regarding roomlight - in fact brightness remains largely ignored, also gamma always tops out at about 2.35
- Introducing more uncertainty into the calibration process, by approximating an entirely made up approximation.
- Not making clear that bt1886 is used as a compensation curve on end user devices - which deviates in its nature according to the black point (= introduces new errors).
- Using in large parts marketing language to sell others on the concept of it.

I might have not been surprised that some people get a blank stare when they hear the words "algorithmically drawn curve", but just because math is used to present a "solution" it doesnt mean that suddenly all becomes better. It also doesnt mean that you shouldn't look at what is promised and whats actually happening.

Displays with great black points are almost configured at a linear gamma of 2.35, perfectly fit for dark room viewing, where gamma 2.35 in lighter scenes makes perfect sense. I call that a confirmation bias on part of what the standard was aimed for.

Displays with "commercial black points" get a gamma of 2.0 to 2.1 applied to darker areas, but brights still are reaching 2.35 - which makes them problematic for dim room viewing and produces entirely different picture characteristics, even in brighter environments.

When we are introducing gamma variance willfully and thereby changing color targets arent we saying that now all displays with blackpoints that arent excellent also get a penalty on color reproduction just because their tracking should be less linear to go along with a less linear gamma curve? Should we now ask from display producers to compensate for that - therefore also introducing variance, where those displays would have a perfectly similar color representation at a linear gamma function? > Are we making most TV less accurate (color wise). And less suitable for both bright room as well as dark room viewing at once? Well I guess thats also a way to sell new TVs...

The problem regarding local dimming (how its implemented) and plasmas becoming extinct is also known - and with that much resting solely on the black point, we are establishing this standard now - to introduce more variance in calibrations yet again?

And for what? Just so the gamma compensation can be said to be in the end users responsibility - and all previous bluray mastering wasn't inaccurate at all, it was just that we have forgotten to make our perfectly linear gamma devices "more like those old CRTs"...
-

Also I might emphasize that "postings in a row" without being able to actually establish ANY kind of discourse arent "unhealthy" a priori, as upvoted by some users in here - if they are actually filled with content, arguments, comparisons.

I don't have it in my hand how fast or how on topic others respond to the contemplations I bring forward.

Theres a difference between accusing me of liking my own post showing up and in not responding in any constructive (or even counter argumentative) manner in the first place so my lists of accusations gets longer and longer, and longer - with every bit of additional information I process on my part.

Its a friendly way of saying - "catch up".

But dont worry. As most of the people here gladly are celebrating whenever a default change in the standards they calibrate to occurs and have no ambition whatsoever to understand or discuss the effects of a new "algorithmically drawn curve" to be implemented, I won't bother you much more.

If I don't get any takers - I have to accept it.

With one merit. I can call color calibration (at least for the current home cinema standard) a pseudo science that should rather be denounced than encouraged.

Someone care to sell me on a 3D LUT box?

Last edited by harlekin; Yesterday at 06:41 PM.
harlekin is offline  
post #4029 of 4035 Old Yesterday, 06:45 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 30 Post(s)
Liked: 65
Harlekin, this thread is not your personal soapbox. If you have particular issues you'd like to discuss, then please post them in the appropriate thread and/or create new threads. This thread is dedicated to a discussion of HCFR software.
spacediver is offline  
post #4030 of 4035 Old Yesterday, 07:33 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 555
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 19 Post(s)
Liked: 70
Quote:
Originally Posted by harlekin View Post
But dont worry. As most of the people here gladly are celebrating whenever a default change in the standards they calibrate to occurs and have no ambition whatsoever to understand or discuss the effects of a new "algorithmically drawn curve" to be implemented, I won't bother you much more.
There was no official standard Video EOTF before BT.1886, just a an encoding standard and CRT display behaviour "expecation". The advent of alternate display technologies such as LCD heightened the problems the lack of a such standard can cause, so the introduction of something to fill that glaring hole can only be regarded as a good thing. It is not a magic wand to cure all that can go wrong with presenting color images to end users via consumer equipment in the widely varying situations they find themselves in though.

I'll leave it to yourself to comment on your own technical understanding of EOTF's - many people responding here already understand them and their interaction with the image encoding, display devices and viewing conditions quite well enough.
gwgill is offline  
post #4031 of 4035 Old Today, 02:24 AM
Member
 
harlekin's Avatar
 
Join Date: May 2007
Posts: 79
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 12
As imagined I found zoyds perspective on this very enlightening.

As luck would have it - it is also well documented here on the forums. All problematic aspects already have been touched upon - especially the really astonishing oversight of gamma impacting color reproduction, when establishing the standard. So essentially someone slipped up and no one noticed until we were 10 years into Bluray production on linear gamma devices as the viewing standard.

*break for applause*

How power law gamma calibration can lead to crushed blacks

Now whats still puzzling me is the amount of near black level correction that bt1886 produces on different end market devices that I will call "consumer standard".

By what is presented the part of bt1886 that is "a CRT correction curve" is mostly the 2.4 exponent, while the black level correction part is in there to combat perceptual black crush - that I didn't see nor I can measure on my black level 0.04 cd/m2 device at power 2.2 gamma at home. But bt1886 sure does compensate against it!

As a fact, adhering to bt1886 standards (in the dark parts of the image) - worsens my contrast, worsens my greyscale, brightens my blackpoint - all while dE at near black arent substantialy different, or outrageous in both cases (before gamma is applied?).



So why again are we configuring for this "black compensation"? And why are we allowing for it to introduce measurable variance where prior to this there was none? Also at very dark blacks bt1886 gamma basically becomes 2.35, so why is there no educated opinion on bt1886's performance on various, slightly brighter black points.

Wait - that question was asked two years ago, also.
How power law gamma calibration can lead to crushed blacks

The only new glimmer of light in this whole matter is that zoyd mentioned that in one test run color performance on one system color errors werent extensibly worse than when meassured against powerline 2.4, so cheers to that guys. You just messed up gamma for (8 to) 10 years, but it didnt impact colors "that much" and only now we are fixing it against a "best guess recommendation" - with a black compensation bias that might be a tad overreaching on certain black points, or not or "so what" - who cares no one looked at it in practice.

And the variance it introduces is just rebranded "perceptually desired".

Last edited by harlekin; Today at 02:49 AM.
harlekin is offline  
post #4032 of 4035 Old Today, 02:39 AM
Member
 
harlekin's Avatar
 
Join Date: May 2007
Posts: 79
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 12
Quote:
Originally Posted by gwgill View Post
The advent of alternate display technologies such as LCD heightened the problems the lack of a such standard can cause, so the introduction of something to fill that glaring hole can only be regarded as a good thing.
No kidding.

And how much of the Bluray Home Cinema standard was expected to be viewed on devices with a non linear gamma response close to 2.4 (avg) - because the last time I checked - CRTs in the consumer space weren't capable of displaying FullHD content in native resolution - therefore the number of customers buying Blurays just to watch them on non linear gamma devices may have been somewhat limited, all the way - even ten years back.

To again phrase it in a way business would understand. "A somewhat limited target demographic."

And it only took 10 years until the "not even a standard by now" "fix" (which consisted of a best guess with a black compensation bias, all being applied on the end user level (to "destroy" his perfectly 2.2 tracking reproduction unit) ) gets first rolled out.

So how far are we actually educating calibrators?
http://referencehometheater.com/2014...gamma-correct/

Well just about there I'd say?

Also there is no educated opinion on bt1886 regarding room light conditions (should it be able to shift, should it not), and no detailed view on what it actually does in real world scenarios when "compensating for black crush" - but thats okay, because the curve just is at a dandy 2.35 on all those Kuros no one can buy anymore.
-

edit:

@mods and fellow readers - I'm toning down my voice and posting frequency from here on - I understand that it got a bit overbearing over the last pages - also, I've about as clear of a picture on the the matter as I could get using preexisting resources on the forum (linked in my posts.). Thanks for bearing my outburst. As always, getting to know a new facet of a truth can be a disturbing process for the individual.

Last edited by harlekin; Today at 03:18 AM.
harlekin is offline  
post #4033 of 4035 Old Today, 03:56 AM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 555
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 19 Post(s)
Liked: 70
Quote:
Originally Posted by harlekin View Post
And how much of the Bluray Home Cinema standard was expected to be viewed on devices with a non linear gamma response close to 2.4 (avg) - because the last time I checked - CRTs in the consumer space weren't capable of displaying FullHD content in native resolution - therefore the number of customers buying Blurays just to watch them on non linear gamma devices may have been somewhat limited, all the way - even ten years back.
The answer, is all of them. I hardly think selling Blueray's marked "Only to be viewed on brand XXX LCD displays set to some new EOTF, and by the way all your existing displays won't work" is going to fly.

But don't expect any further responses from me.
gwgill is offline  
post #4034 of 4035 Old Today, 04:46 AM
Member
 
harlekin's Avatar
 
Join Date: May 2007
Posts: 79
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 12
Ten years after the Bluray was first introduced. Ten years.

Also you are intentionally obfuscating the relational statement made here. ALL LCD and Plasma manufacturers were _targeting_ mostly linear power line 2.2 even in movie modes (until very recently (Sony f.e. now goes for 2.4)) - so there was a defacto standard that got mostly ignored over most of the lifetime of Bluray production so far - while now we are "fixing it" by tuning them to a gamma curve of an invented "proto CRT" - with questionable black level correction (and no clue how room light should relate if gamma is not at near perfect 2.35 linear progression).

And I shouldn't expect any responses on the matter any more? How very unexpected judging from the previous course of the exchange.

The calibration community proves to be transparent, open and proactive - and not at all making up clutches as they go a long - once more.

One more promise on my side - that is as far as I will go confronting anyone here on a personal level in regard to this matter, and only because the honor was first presented to me by the dear user gwgill.
harlekin is offline  
post #4035 of 4035 Old Today, 05:00 AM
Member
 
omenII's Avatar
 
Join Date: Jun 2005
Location: UK
Posts: 148
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 10
I ran a quick a few quick measurements on my Sony 55W905(900) over the weekend and the results were mostly perfect. Delta E's across the greyscale were all under 1 without any white balance adjustments at all and gamma came in at a near ruler flat 2.1 when set to zero in the TV menu's. I only used AVS709 full saturation patterns for primary colours but Red and Blue were spot on with Green a bit undersaturated. The TV only has a +/- colour control and no CMS so there wsn't anything I could do with the Green unfortunately.

But I'm going to put a bit more time into it this week and use the GCD patterns instead and measure the 75% colours as well as do other input sources. A couple of questions if anyone can answer them for us please :-

My set only has 2-point controls so I can't calibrate to BT.1886 gamma. I prefer 2.2 so will stick with that but do I use gamma with black compensation or power law? I used black compensation on my last full calibration a few years ago but that was an LG plasma with 0.20 cd/m2 black levels where as my W905 is coming in a under 0.05 cd/m2.

What setting does HCFR have to be in for best relationship between my meter (i1D3) and telly eg. LED White IPS or LED RGB? Just wondering if the fact that my set uses Quantum Dot RGB backlighting for the triluminous colours will have a bearing on this?

Finally, how do I calibrate 3D? I think I need to put one eye of the glasses in front of the meter but is there special patterns I need to use as well?

Cheers.
omenII is online now  
Reply Display Calibration

User Tag List

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