- This topic has 11 replies, 5 voices, and was last updated 9 months, 3 weeks ago by
.
-
Posts
-
https://www.testufo.com/blacklevels . Only in the dark with a raised black level can i see the moveing 1s. I can adjust 5% and see the black background change but not the see the moveing ufos. I think I see the 2 black square on the bottom even in gamma 2.2 black compensated even in the light. Perhaps my display is not 8bit. It does show a lot more in 7 bit. I prefer black to be black so I think 5% gain at red +4 and green +1 is best.
Browser maybe maping HTML colors to something. As a general rule not reliabe test.
Test in MS paint with a set of preconfigured PNG images for testing raw display colorspace behavior.Thanks. I was looking for setup the page wanted. Finaly say it was Adobe SRGB. I ran it in hdr mode, and it had experimental web platform features needed to be turned. It was for hdr mode but it made sdr mode show that 2 better.
Indeed browsers are color managed bunches. I run my own experiments and thanks for the help. I’m just figuring what looks best. A lot of the screen is not less than 4 RGB in some shows. I watch a lot of movies with night and caves in them and black uniforms. Black makes a difference and even in games. I noticed a game ignored and set it own screen scaling. That was a tale it resets the gamma trc. I checked it running the game and it was linear. Now I figure it is better to run without a profile but calibrated to rec709 gamma 2.2 or 2.4 .
Sorry if I have no direct answer to your question.
I want to thank you however for the link, though. My IPS panels perceived black levels are somewhat sensitive to viewing angle and viewing height, and this is a great tool to optimize these factors.I thought it was a good for perceiving how things look. It shifts on my screen to. You can really see what being not at 90 degree straight line to screen does. My right side of screen is tilted toward desk more and shows more blacks on the bottom. The left side is visible when you turn the head right.
IPS is cheaper and older. Hopefully it will last longer. It also does not have image burn in. I remember even CRT back in school in 1989 got burned in. Burning in a screen never happened to me.
I’m able to see the animation on an OLED, in the dark, after my eyes adjust for a few seconds.
Thanks. I was reading how you have to dark calibration on OLEDs. It is way darker on a OLED so it really needs to be in the dark. I did a Display Cal calibration in gamma 2.4 0% input offset and can see the animation on some parts of the screen. lol No screen is perfect. At 2.2 gamma i can see the animation in light just barely. Eyes have to adjust. I manualy set calibration 2.4 50% input offset for the hardware calibration so Displaycal did not have much to adjust and it added to RGB 1 to 50 from linear VCGT. Nice to have a bigger than 256 RGB 1. Normally it takes the 256 down to 190s and makes crush worse but calibrated 1st it did not have to. TV is SRGB gamut so no profile sometimes not needed. The profile does change the color though. On Windows 11 24h2 Vizio V5-555 J01.
Noticed HCFR software has a 0 rgb to 1 rgb comparision on its 1% black advanced patterns. The bottom and top of the screen 1,1,1 and the middle is 0,0,0 . Windows Screen colorpicker in Windows toys is nice to see the color numbers on the screen. I never knew HCFR had a pattern 1,1,1 on it. The bottom and top I always thought was screen going brighter on the edges.
I have spent only a few minutes trying the black and white tests in a couple of environments this test is tricky WRT to variables of color management policy in display, OS, and browsers. I suggest being cautious with results unless you are prepared to get into many details of color policy and rendering for any given setup.
Some simple questions:
- What results should be expected for a given system setup?
- Why are your specific results occurring?
- Are the results good or bad?
There are no obvious answers to these questions at that site.
Here’s an objective warning sign that thinking behind the black test is poor: the use of the language “PLUGE” to describe the parallel bars part of the test window is a hopeless misappropriation of that term and demonstrates a degree of ignorance of the history of display systems and test practices. This single word calls into question the entire presentation.
Similarly, the lagom.nl black/white tests are patch difference, but more comprehensible because the presentation is so simpler: It’s just numerical R=G=B levels presented in a browser window, so it’s easier to focus on the all the system policy that leads to a result without all the stuff flying around on the screen.
To give an example of how a seemingly useful patch test can go totally wrong, consider the “gamma calibration” test on lagom.nl based on visually matching brightness strips to determine average gamma. The test is modeled after the Adobe Gamma control panel shipped with Photoshop in the 1990s. It’s from the CRT era and can never work on an LCD if fractional resolution scaling is used and may not work in some conditions such YCbCr4:2:x connected displays even if 1:1 display resolution is used.
Regardless of the distinction of any particular patch test, it’s unwise to assume anything is going wrong because you saw something on the internet! And even at its best, this test won’t help you with explain why particular numerical levels lead to any particular visibility on the face of you display. Without knowledge of the bigger picture of signaling and coior policy, the results don’t mean anything.
This stuff suffers from Brandolini’s Law: It’s more work to debunk than to seek deeper understanding and methods, which leads to forum with post like “my OLED shows patch 2 but my LCD only shows patch 3,” absent of needed additional context.
Left unchecked, we end up with weird lore like hanging your HDMI cables from rare Brazilian hardwood trellises to get that last level to appear. I just made a 90s era hi-fi quip, but you can still find people selling gold-plated ethernet cables for richer more detailed audio:
Like many of you, I woke up thinking “You know what, I feel like listening to some audiophile Ethernet cables today!” So I brew up my morning coffee, line up the cables, and rub my hands vigorously.
Many audiophiles who stream their music via Spotify, Qobuz, or Tidal don’t realize the sonic impact of Ethernet cables. And that’s understandable. I mean, “it’s just digital” – right?
It is ok. Its just a test not a presentation. It shows how environment effects quality. A browser test is good browser test. It helped me know if it was really clipping blacks or not. It just blacks so low you can not see them in light or reflective light on screen from light in front of screen. Good to adjust your tv angle too.
A pro would use a test for their intend audience. The internet use web browsers. I had no questions. Just comparing what it should look like. It tests if the hardware is good or not. It appears I have an 8-bit screen and not 7-bit. Red and Green not need to go up from default 5% white in tv controls. If there was clipping it might. Wrong test to see what amounts to adjust. I use numbers for that exactly. HCFR gamma graph did not match numbers so I had to visual check. One test means nothing but everything off means something. Things like not seeing the dark in games or movies. Losing details is bad to me. Light is the problem sometimes. Many things are. That is why the test is not a presentation.
I thin they used the word Pluge because it was a good keyword. Gamers look for keywords and not very smart.
I have spent only a few minutes trying the black and white tests in a couple of environments this test is tricky WRT to variables of color management policy in display, OS, and browsers. I suggest being cautious with results unless you are prepared to get into many details of color policy and rendering for any given setup.
+1000
I have configured firefox (about:config) to render HTML colors and untagged images to sRGB, and my display profile says g2.2 (BPC) although it is not all the way to RGB 0 (1000:1) and I can see full screen the moving near black patterns and barely see “1”. Not because of display, good behaved IPS but just 1000:1, but rather by the lift done by color management from 2.2 TRC to sRGB TRC.
Without color management on a “keep 2.2 all the time you can in black to white ramp” this near black images are going to be “packed” in a tiny Y nit range dute to 1000:1, So I doubt I could see RGB 2-3 non color managed.Calibration visual tests, best done with non color managed apps (unless you want to test if CM is working properly)
