# GTA 6's HDR Calibration is 𝗕𝗥𝗢𝗞𝗘𝗡… Here's What to Do Instead

**Channel:** HDTVTest
**Source:** https://www.youtube.com/watch?v=d3UaVoP6dPg
**Transcript page:** https://www.withtranscript.ai/video/d3UaVoP6dPg

## Chapters

- 0:00 — Introduction: The 2,000 Nits Paradox
- 1:03 — Max Brightness Setting Explained
- 2:17 — Paper White Interactions and Black Levels
- 3:48 — Dark Room Calibration and Sponsor Segment
- 5:48 — Paper White Tuning and Final Recommended Settings

## Transcript

**[0:00]:** To get 1,000 nits in GTA 6 actual HDR gameplay, you have to set max brightness to 2,000. No, that's not a mistake. And no, my reference monitor isn't broken. Let me demonstrate. Here, I've set max brightness to 1,000 nits. And Rockstar's test pattern measures exactly as advertised. But inside the actual game, the sun measures only 600 nits. Nothing has changed on the max brightness slider. The only other thing I've adjusted is one other setting on the very same HDR calibration screen. In other words, GTA 6 HDR calibration test pattern is telling you one thing while the game is doing something else entirely. I've measured all of this using my Canon reference monitor with inbuilt HDR analysis tool and validated the results on a Sony BVM HX3110 mastering monitor costing £30,000 sterling. So, by the end of this video, you will have some idea of how to obtain the correct settings for your display, even though they may not make sense on paper.

**[1:03]:** First things first, GTA 6 doesn't use the console level HDR calibration settings on the PS5 or the Xbox Series X, so the in-game settings directly and exclusively dictate the game's HDR10 presentation. There are three HDR calibration settings in GTA 6: Max brightness, black levels, and paper white. And they are far more interconnected than what's suggested by Rockstar's instructions. Max brightness lets you set the peak brightness inside the game. Basically, you're telling the game's HDR tool mapper how bright your display can go so that highlights are fitted within its capabilities instead of being clipped. The test pattern, a Rockstar logo inside a white square is actually cleverly designed. The white background always measures slightly brighter than the selected value and the logo marginally darker. So in theory on a TV or monitor with HGIG or source the base tone mapping which hard clips at the display's peak brightness instead of rolling off you would engage HGIG increase max brightness until the logo disappears then

**[2:07]:** go back one step at which point the logo should look as faint as the second line on your mistress's pregnancy test. But there's one major problem.

**[2:17]:** The max brightness value only carries across accurately into the game if paper white stays at its default setting of 50 or higher. Once you reduce the paper white setting, which is what we should do to get closer to a less washed out paper white of between 200 and 300 nits, then the actual in-game peak brightness would measure lower even though the test pattern still indicates the same as the max brightness value. In addition, the black levels setting is also influenced by the paper white setting. At the default paper white setting of 50 and above, changing black levels generally does not elevate the black floor. It's kept at zero nits. However, once paper white goes below the default of 50, then increasing black levels from the minimum value of zero would gradually raise the black floor. It seems that the more you touch up top, the more sensitive it becomes down below.

**[3:13]:** Now given that an elevated black floor can compromise the true blacks of OLED, wash out the image or exacerbate all sorts of near black artifacts on both OLED and full array local dimming LED LCD displays. My initial instinct was to set black levels to zero for a black floor of precisely zero candelas per square meter. However, in the actual game itself, particularly during the prologue cutscene where Jason first meets Lucia, the shadows looked a little too contrasty with some black crush, which was also confirmed on our Canon reference monitor.

**[3:48]:** So, my recommendation is to turn off all the lights in your room. Lower black levels all the way down to zero. Then press the relevant button on your controller to hide the UI, allowing your eyes to dark adapt and become more sensitive to shadow detail. From there, gradually increase black levels until the sheerbot background first becomes discernible. On my Sony HX3110 mastering monitor, the black levels value I arrived at was 20. If you are playing in a brighter room, man, you should really switch off your lights and play in the dark since the HDR10 PQ standard is absolute. But if you can't control your lighting, then just follow my instructions until you can first see the checkerboard background. I'm saying this with as much enthusiasm as me on a Friday afternoon team's call. While we cannot take Rockstar's test pattern at face value, here's something you can verify for yourself.

**[4:44]:** Zeskit, a leading HDMI cable manufacturer who has kindly sponsored this video, has launched its Maya 96 series of HDMI 2.2 cables, offering a staggering 96 Gbits per second of bandwidth, doubling the 48 GBs per second provided by HDMI 2.1 and therefore paving the way for support of resolutions up to 12K or even 16K. Since my last video, the Maya 96 has been officially certified under the HDMI forums Altra 96 program. So, you can scan the QR code on the certification label to verify that you are getting the genuine article even though there's currently no HDMI 2.2 source device and display on the market for me to test this ZESTI cables to their full rated bandwidth. That said, HDMI 2.2 2 is backward compatible and I've been using Zestate certified HDMI cables extensively in my in-depth TV reviews and side-by-side comparisons and they have never let me down.

**[5:43]:** I will leave a link in the YouTube description below. Thanks again for your support.

**[5:48]:** Okay, the paper white setting lets you set the diffused white level. In other words, the brightness of a white object such as a sheet of paper as opposed to light sources and specular highlights. And Rockstar has provided one bright and one dark test image for you to adjust paper white. The default paper white setting of 50 caused the midtones to be over brightened not only on the brighter test image but also in certain sequences in game. Ironically leading to a flatter image due to a more compressed dynamic range. For playing in a very dark room, ideally paper white should be close to the reference white level of 203 nits recommended by the ITUR BT2408 standard for 108 HDR10 content. But we have no indication from Rockstar or from my testing as to which object or element inside either test picture should be the target to be measured.

**[6:40]:** So I decided to measure the white in the UI and also any fade to white in the cinematic cutscenes. At the default paper white setting of 50, these elements measured 350 nits, which also happens to be the cap in the opening splash screen. So I'm confident I'm on the right track. The subtitle brightness measured just a tad dimmer, which is acceptable. There's another issue with lowering the paper white setting though because this also reduces the maximum peak brightness inside actual gameplay. Put another way, the max brightness setting and the paper white setting interact with each other and any particular paper white value under the default setting of 50 would mean that the in-game peak brightness would be noticeably lower than the max brightness setting even though the test pattern still measures correctly. With all these permutations at play, I had to go through the process of first dialing in my target paper white luminance by measuring the UI and the fade to white transition before increasing max brightness to achieve a specific

**[7:42]:** pick luminance based on in-game measurements and then readjusting the paper white because it would have been increased. One useful shortcut I discovered was that this spotlight in the darker nighttime calibration image measured exactly the same brightness as the sun in the actual game. This made the whole testing process considerably more efficient since I no longer had to keep jumping back and forth between actual gameplay and the HDR calibration menu. It will take me some time to cross-ch checkck these results using the relevant test patterns, actual gameplay and cinematic cutscenes both on my Sony HX3110 reference monitor and on various consumer displays. But in the meantime, here is a good set of starting values that should work reasonably well for HDR displays capable of 1,000 nits or brighter. Set max brightness to 2,000 nits, black levels to 20, and paper white to 35. Please let me know in the comments how these settings look on your particular TV or monitor and I will take your feedback into account when putting together a more comprehensive brand specific HDR settings guide which you can watch

**[8:48]:** by clicking here once it's available.
