madVR - high quality video renderer (GPU assisted)

[Clown shoes]

does the shader work or atleast make it better?

I've not had a chance to test that yet as DSPlayer is my weapon of choice.

I am still working with Florian to see if we can get the 3DLUT working so I haven't completely given up hope yet :)

[Clown shoes]

Is anyone here familiar with creating HDR test patterns in Mad TPG?

Under the HDR dropdown I'm not sure what I should be setting the nits to?

Should it be the closest to the maximum my display is capable of (430nits in my case) or is it the peak luminanace (1000 nits in my case) or maybe something else? ;)

[Betroz]

I'm gonna give Windows 8.1 a try on my HTPC (swap out Win10), and may I ask which nVidia driver version works best for MadVr and a GTX 1080 card? I don't play any games on this machine, so MadVr is my only consern.

[Asmodian]

Did you ever find a workaround with your particular display?

I did, I had to put it in any mode beside PC because it's HDR behavior is much better when not in PC mode. However, it is then 4:2:2. :(

I still have to set the peak nits at or above the real peak in the content or it will do that, and some movies go pretty high, I have seen up to 2500 before.

[Clown shoes]

I did, I had to put it in any mode beside PC because it's HDR behavior is much better when not in PC mode. However, it is then 4:2:2. :(

I still have to set the peak nits at or above the real peak in the content or it will do that, and some movies go pretty high, I have seen up to 2500 before.

Ah, that's frustrating. Which TV is it, if you don't mind me asking?

You may be interested to know that I have managed to get rid of nearly all of the clipping by setting a very large number of patches on the HDR profile. The default is 175 but I ran 1100 at Florian's suggestion and nearly all of the clipping has gone. I'm currently running a 1500 patch profile and fingers crossed the clipping will be totally gone this time.

Out of interest have you noticed an overall reduction in luminance since using HDR 3DLUTS? Everything seems just a little bit duller than with passthrough to me.

[Asmodian]

It is LG's 2017 OLED, C7P.

I also used high patch counts, I believe I was using ~1000. Maybe I should try some more but when not using PC mode I got good results (no banding/clipping). It is expected that a 3DLUT (or any software calibration) would lower brightness, the only way to do any type of correction in software is to send <100% because you can never send >100%.

I ended up not using a 3DLUT for HDR, I did a manual calibration of my TV in Calman and simply use passthrough in madVR. Ideally you need a 3DLUT for every video with the peek nits set for the metadata of that video and after the manual calibration, and in home theater mode, my TV does a decent job tone mapping. A 3DLUT crafted for a particular video is a little better but passthough is better than using a "worst case" 3DLUT for everything.

I think there is more difficult HDR content, for example some of the earlier Harry Potter movies have very high peek nits in a few scenes (causing out of gamut issues like this) while most newer movies seem to use a more reasonable peek nits and all look good with the same 3DLUT.

[Clown shoes]

It is LG's 2017 OLED, C7P.

I also used high patch counts, I believe I was using ~1000. Maybe I should try some more but when not using PC mode I got good results (no banding/clipping). It is expected that a 3DLUT (or any software calibration) would lower brightness, the only way to do any type of correction in software is to send <100% because you can never send >100%.

I ended up not using a 3DLUT for HDR, I did a manual calibration of my TV in Calman and simply use passthrough in madVR. Ideally you need a 3DLUT for every video with the peek nits set for the metadata of that video and after the manual calibration, and in home theater mode, my TV does a decent job tone mapping. A 3DLUT crafted for a particular video is a little better but passthough is better than using a "worst case" 3DLUT for everything.

I think there is more difficult HDR content, for example some of the earlier Harry Potter movies have very high peek nits in a few scenes (causing out of gamut issues like this) while most newer movies seem to use a more reasonable peek nits and all look good with the same 3DLUT.

Yeah, well if I'm honest I think I'm going to give up on the 3D LUT route as well. I just tried again with 1500 patches and although it's much better it still doesn't look nearly as good as passthrough or pixel shader math. I may go back and have another crack at calibrating the display much like you have done.

One question I do have for you; is there a reason you've chosen to use Passthrough rather than have MadVR pixel shader math do the work?

[yogibeezwax]

I'm sorry if this has been addressed... If tried to skim this long thread but I would really appreciate some answers to Fullscreen exclusive vs fullscreen windowed;
Is there any discerning difference in performance?
I have a 1070 TI strix with a i7-4790k and 16 GB ram and I've had several issues with "failed exclusive mode" when running files in a playlist..and at worst, when moving the mouse sometimes the screen goes blank and I'm forced to hard reset the PC to get out of it..However, running in fullscreen windowed doesn't give any issues whatsoever. I've read that using exclusive mode isn't a must anymore..long story short; is fullscreen windowed with d3d11 as good as exclusive?

[Asmodian]

One question I do have for you; is there a reason you've chosen to use Passthrough rather than have MadVR pixel shader math do the work?

What work? Pixel shader math simply keeps the TV in one tone mapping mode while passthrough tells the TV different metadata but the TV still does its tone mapping either way. I don't notice an improvement having madVR's pixel shaders pre-processing for the TV and I always default to less processing if I don't see an improvement.

I use very different display settings for HDR, via its HDR mode, so I don't use pixel shaders (or 3DLUT) to convert to SDR. ~680 nits is way too high for SDR. :p

is fullscreen windowed with d3d11 as good as exclusive?

Yes, I like it better myself. :)

Edit: There is currently limited 10-bit support in fullscreen windowed d3d11 so use exclusive if you need 10 bit. I recommend sticking to 8 bit output in most cases anyway but it is good to know if you are testing 10 bit output.

[yogibeezwax]

What work? Pixel shader math simply keeps the TV in one tone mapping mode while passthrough tells the TV different metadata but the TV still does its tone mapping either way. I don't notice an improvement having madVR's pixel shaders pre-processing for the TV and I always default to less processing if I don't see an improvement.

I use very different display settings for HDR, via its HDR mode, so I don't use pixel shaders (or 3DLUT) to convert to SDR. ~680 nits is way too high for SDR. :p



Yes, I like it better myself. :)

Edit: There is currently limited 10-bit support in fullscreen windowed d3d11 so use exclusive if you need 10 bit. I recommend sticking to 8 bit output in most cases anyway but it is good to know if you are testing 10 bit.

Don't think I can do 10 bit with my panasonic vt65..never seen 10-bit when playing any source so I'll stick with 8-bit and windowed mode..hate the exclusive "switching" so I guess I'll go with that until I buy a new TV (have my eye on the new Q9n 65"..might wait for microled..but I'm an impatient bastard;))

[el Filou]

Don't think I can do 10 bit with my panasonic vt65..the Panasonic plasmas accept 10- and 12-bit but it's not worth the hassle and the potential for quality loss if Nvidia messes something up in their drivers.

[ryrynz]

the Panasonic plasmas accept 10- and 12-bit but it's not worth the hassle and the potential for quality loss if Nvidia messes something up in their drivers.
Waste of time on this set, greyscale is no better than 8bit with dithering, I looked myself. As mentioned many times already most displays have zero benefit using 10bit.

[Manni]

Waste of time on this set, greyscale is no better than 8bit with dithering, I looked myself. As mentioned many times already most displays have zero benefit using 10bit.

When you compare native 10bits vs 8bits with a display able to properly handle 10-12bits (i.e. natively) and the dithering is done by MadVR without the GPU doing its thing behind its back, you can’t look for banding on a pattern. There most likely won’t be any. You have to look for noise in real content. There should be a bit more noise in 8bits dithered than in 10bits, but there shouldn’t be more banding (provided the source is 10bits and properly mastered) as MadVR’s dithering is excellent. Also I suppose that by greyscale you mean a greyscale ramp, as 10bits shouldn’t haven any effect on calibration.

[oldpainlesskodi]

Just wanted to update and share.

I have finally managed on Nvidia to get 23.976 4k RGB 10bit..yep, not 12, but 10 (both SDR and HDR) (confirmed also by my AV receiver on-screen info).

I discovered that using the display port (which is 1.4 on the 1080), and using the Club 3D CAC 1080 Display Port 1.4 HDMI 2.0 A (B) Hdr Active displayport to hdmi adapter, i can now choose between 8, 10 and 12bit in NVC @ 4k (10 and 12bit upto 30hz).

Using the adapter also puts the driver into a slightly different mode, in as far as Set Content Type is no longer available etc. Also, the white point scale and crispness is different than straight hdmi (which I prefer over straight hdmi). Some of the Nvidia drivers will only report upto 30hz being available using the adapter, but 397.76 displays all modes upto 60hz. The trade off is, as with some previous drivers, you cannot set up a custom res at 10bit (defaults to 8), so thats a down side - I guess it's about which trade offs you can live with.

At least on my set up, the colours seem more faithful, and no more over blown colours in HDR mode.

I know there is a lot of debate about 8bit with dithering etc, but, just wanted to share.

Update:

Ok, maybe i'm not doing the test correctly somehow, but using the 4.4.4 Chroma test image (link https://linustechtips.com/main/uploads/monthly_2016_11/TextTest_4k.png.8fc940e703b15b869281af8c8d22f67e.png ) it also appears to pass the test at 23hz....?

[sim11]

Hi, I've following this thread for years now, but can't seem to find the answer to this, I know for sure someone on this thread will know the answer to this for sure.

When trying to calibrate using madTPG as a pattern generator, i have noticed no matter what changes i make to gamma etc on windows 7, madTPG pattern gamma etc remains untouched and does not change in line with the windows 7 gamma etc. Internal pattern gamma etc change in both calman and hcfr but not in madTPG. Thanks sim11

[ryrynz]

Also I suppose that by greyscale you mean a greyscale ramp, as 10bits shouldn’t haven any effect on calibration.

Yeah I used the 2160p Gradient_Split test file from avsforum. Plasmas are noisy and the higher the bitdepth the noisier it becomes, so it makes it very hard to see any difference.

Don't think I can do 10 bit with my panasonic vt65..never seen 10-bit when playing any source so I'll stick with 8-bit and windowed mode..hate the exclusive "switching" so I guess I'll go with that until I buy a new TV

I'd recommend setting 'error diffusion option 2' and 'change dither for every frame' This resulted in the least amount of noise on my VT60, surprisingly changing the dither patterns should increase the noise but on plasma it appears to have the opposite effect, go figure.

[Siso]

Is this a bug or using NGU standart-medium or any NGU, always delays 1 frame. Video card is GTX 1050 ti, 390.65 win7 64 ultimate. No problems with super xbr-100 AR. Rendering times are low around 14 ms.

[Warner306]

I own an OPTOMA UHZ 65 laser video projector that handles very well the HDR, in fact with madvr I'm going to passthrough .... yesterday out of curiosity I wanted to use the HDR to SDR conversion, the image is more brighter, but the colors appear washed out, even if you set it in "this display is already calibrated" BT.2020, what can it depend on?....I use MPC-BE.... I also wanted to know from the owners of AMD video cards which is the most stable driver for HDR?

Increase the target nits until the image gets darker. You will get better results with a projector by leaving more room for tone mapping by providing a higher target nits than your actual peak nits.

[Warner306]

Is this a bug or using NGU standart-medium or any NGU, always delays 1 frame. Video card is GTX 1050 ti, 390.65 win7 64 ultimate. No problems with super xbr-100 AR. Rendering times are low around 14 ms.

What do you mean by delays one frame? NGU will shift a pixel or two but frames should be rendered the same. Maybe it is taking a second longer at playback start to fill the queues?

[Warner306]

I'm sorry if this has been addressed... If tried to skim this long thread but I would really appreciate some answers to Fullscreen exclusive vs fullscreen windowed;
Is there any discerning difference in performance?
I have a 1070 TI strix with a i7-4790k and 16 GB ram and I've had several issues with "failed exclusive mode" when running files in a playlist..and at worst, when moving the mouse sometimes the screen goes blank and I'm forced to hard reset the PC to get out of it..However, running in fullscreen windowed doesn't give any issues whatsoever. I've read that using exclusive mode isn't a must anymore..long story short; is fullscreen windowed with d3d11 as good as exclusive?

It won't change rendering times. Currently, there is a bug where 10-bit HDR passthrough isn't possible with Nvidia cards without fullscreen exclusive mode. This is considered a driver problem. Otherwise, it is a matter of preference.

[brazen1]

Any chance of adding toggle for RRN (reduce random noise) hotkey?

[huhn]

denoise toggle.

[Siso]

What do you mean by delays one frame? NGU will shift a pixel or two but frames should be rendered the same. Maybe it is taking a second longer at playback start to fill the queues?

I have enabled "delay playback start until render queue is full", and it just delays 1 frame, it says it in the OSD stats. This happens when using NGU only, I have no idea why is that, no other problems with other chroma scallers...The delayed frame is somewhere around 10-15 min of the movie.

[huhn]

delaying a frame when starting playback is not a problem.

[Siso]

delaying a frame when starting playback is not a problem.

I always reset the frames counter when I go into fullscreen, this happens 10-15 min after playback.

[brazen1]

Thank you huhn. I see you don't need to toggle on/off. 'Increase' essentially turns on and 'Decrease' turns it off.... and all 10 steps in between. NICE!

[Seedmanc]

Is there a way for madvr to automatically disable GSYNC during playback like it does with GPU gamma ramps? Disabling it manually every time is annoying and with it enabled at the same time with smooth motion it goes bonkers.

[Georgel]

madVR is out of updating?

I am getting freakouts and crashes with madVR has crashed every ten minutes watching anime upscaled from FHD to 4K. Tried multiple file types, multiple encodes, and I tried keeping the algorithms very friendly. Not sure why it happens.

This is what my crash dump thingy looks like

lavvideo!CLAVPixFmtConverter::convert_yuv420_px1x_le+0xb7
lavvideo!CLAVPixFmtConverter::Convert+0x19e
lavvideo!CLAVVideo::DeliverToRenderer+0x7a5
lavvideo!CLAVVideo::Deliver+0x1e7
lavvideo!CDecAvcodec::DecodePacket+0x6d0
lavvideo!CDecAvcodec::Decode+0x192
lavvideo!CDecBase::Decode+0x113
lavvideo!CDecodeManager::Decode+0x53
lavvideo!CLAVVideo::Receive+0x115
lavvideo!CTransformInputPin::Receive+0x59
lavvideo!CDeCSSTransformInputPin::Receive+0x1b
lavsplitter!CBaseOutputPin::Deliver+0x1c
lavsplitter!CLAVOutputPin::DeliverPacket+0x4d3
lavsplitter!CLAVOutputPin::ThreadProc+0x181
lavsplitter!CAMThread::InitialThreadProc+0x36
lavsplitter!strcspn+0x9d
kernel32!BaseThreadInitThunk+0x24
ntdll!__RtlUserThreadStart+0x2f
ntdll!_RtlUserThreadStart+0x1b

[Asmodian]

The last update was less than a month ago! How often do you expect updates from a free project with one developer working in his spare time? :p

Sorry, no idea why you are getting crashes but what player are you using? Do you have any non-default values set in LAV Video?

[ryrynz]

I am getting freakouts and crashes with madVR has crashed every ten minutes watching anime upscaled from FHD to 4K.

Assuming it's related to MadVR? Try another renderer.

[truexfan81]

anyone else get a huge madvr performance drop after upgrading to the windows 10 1803 build? i'm running a ryzen7 1700 with 16GB ram, and an evga 1050ti ftw, and I cannot get the present queue to fill, I've turned off everything except smooth motion, and set the scalers to spline36AR. its showing 0.21ms present time for a 392p30 video. decoder queue, and upload queue are both full, render queue is at 1-2/7 present queue is at 0-3/3. monitor is running at native 1080p60

[cyberscott]

anyone else get a huge madvr performance drop after upgrading to the windows 10 1803 build? i'm running a ryzen7 1700 with 16GB ram, and an evga 1050ti ftw, and I cannot get the present queue to fill, I've turned off everything except smooth motion, and set the scalers to spline36AR. its showing 0.21ms present time for a 392p30 video. decoder queue, and upload queue are both full, render queue is at 1-2/7 present queue is at 0-3/3. monitor is running at native 1080p60

Multiple individuals, including me have this issue, in varying degrees.
If you use the "search this thread" feature at the top of this page for "windows 10 1803" You'll have at least two pages of information to go through.
No full fix for those of us affected beyond rolling back to the previous Windows version. I've reduced a few setting and also set my render queue (GPU) to 7 and my present queue to 6 as they will not fill up beyond that. I have everything working with the slight settings reduction in quality and with the queues set to what I mentioned.
You will just have to adjust settings while watching a video until you can find something that will work with your hardware to fill up the reduced queues. For me, dropping settngs in "artifact removal" helped me.
I'm sure madshi will be able to come up with solutions once he can allocate some time for madVR. He has a real job beyond madVR and committments and bills to pay. :)

[Betroz]

anyone else get a huge madvr performance drop after upgrading to the windows 10 1803 build? i'm running a ryzen7 1700 with 16GB ram, and an evga 1050ti ftw, and I cannot get the present queue to fill, I've turned off everything except smooth motion, and set the scalers to spline36AR. its showing 0.21ms present time for a 392p30 video. decoder queue, and upload queue are both full, render queue is at 1-2/7 present queue is at 0-3/3. monitor is running at native 1080p60

I too have this issue with the queues not filling up and dropping frames. My fix was to remove Windows 10 and install Windows 8.1. Now MadVr works as intended. If Madshi will be able to fix the problem, then maybe I'll switch back to Win10 later. But the thing is that the next big update to Win10 things can brake again...! Win8.1 will stay the same.

[Warner306]

So many lines of code have been written into madVR over the years. It could be a real mess if Microsoft has made significant changes to way software interacts with the OS. The fact it is not happening to everyone is quite strange.

Have you tried wiping the drivers and reinstalling? The driver upgrades sometimes don't always go as planned. I'm not sure if Windows forced a driver upgrade or not.

[Betroz]

Have you tried wiping the drivers and reinstalling? The driver upgrades sometimes don't always go as planned. I'm not sure if Windows forced a driver upgrade or not.

I did a clean install of Win10 1803, and tried 3 different nVidia drivers (after a DDU cleanup too). I had no problems with the 1709 build of Win10, so yes something changed with 1803. I even tried an older build of MadVr with NNEDI3 128 settings, and that worked fine under Win10 1803.

Now I'm using the newest version av MadVr under Win 8.1 and NGU is working as it should. Besides Madshi have recommended Win 8.1 several times in the past, so now I see why. An ever changing OS like Win10 is not allways the best thing it seams.

[zaemon]

Same here. Went back to Win 8.1 since custom modes were not working and I had pretty bad repeated frames. Even though custom modes are not working either, I get excellent results now (1 frame repeated every hour). I’m running the latest drivers for my GTX.

[foozoor]

Time to switch to D3D11/12 and MediaFoundation!

[huhn]

feel free to rewrite everything.
maybe start with decoders and don't forget the media player.

[el Filou]

I tested with Windows 10 v1803 and I encountered the queue length issue too.
Here's a folder with OSD and GPU-Z screenshots:

https://drive.google.com/open?id=13Gtasnn4dRPBFcmUofKrJ3qhpZIbywhI

Test system: Core i5-2520M (HD Graphics 3000 12 EUs @ 1300 MHz, 35W TDP); latest versions of Windows, Intel drivers, MPC-HC, LAV, and madVR; all software 64-bit; graphics power plan set to Maximum Performance and verified using GPU-Z render test (1300 MHz, 9,6W @ 100% GPU).
EDIT: screenshots taken with DXVA2 Native decoding, as that's what gives the best performance with Intel iGPUs.
EDIT2: I never noticed any issue like this with Windows 1709 obviously, but I can't go back to that version to test more thoroughly as it's an Insider install, and I have no desire to test 1803 on my 'production' PCs. :rolleyes:

Observation: the problem isn't restricted to NGU, actually all scaling algorithms can fail to fill up the render queue but they do it at different queue lengths depending on their complexity.

- NGU @ render queue length 8: render time is 75% of the frame time which should be OK, GPU is at 71% load but only at half clock although not limited by TDP. Render queue 1-3/8, and madVR drops 13 frames on ~45 seconds of playback.

- NGU @ Q4: now fills up a bit more at 2-3/4, and no frames are dropped anymore. Frame render time and GPU clock/load are identical.

- Bilateral Sharp: this the lower complexity scaler closest to NGU Low, and @ Q9 it can fill up the queue but it can't @ Q16 even though once again the GPU isn't even at 1/3 its tested maximum capability.

- Jinc: uses 30% less power, can fill the queue @ Q16, cannot @ Q24.

- Bicubic: can fill the queue even @ Q24.

Things I tried:
- Disabled Windows' Game Mode
- Disabled Windows' Full Screen Optimisatons for MPC-HC
- Unchecked "delay playback start until render queue is full"
- Used D3D9 instead of D3D11
- Tried software and QuickSync decoding
but none of that did have any impact.

GPU power definitely isn't the issue, because switching to higher complexity scalers makes it reach higher clocks successfully.

Only idea: could it be a VRAM issue? I noticed all the scenarios that don't drop frames used ~165-180 MB of VRAM in this test, i.e. NGU at QL 2-3 is using the same amount of VRAM as Bicubic at QL 23-24 but nothing that would need more than that amount - like NGU at higher queue lengths - will fill up the queue.
Could memory allocation be the source of the problem?

Cheers.

[Betroz]

Observation: the problem isn't restricted to NGU, actually all scaling algorithms can fail to fill up the render queue but they do it at different queue lengths depending on their complexity.


Well if you test an older version of MadVr, you'll see that NNEDI3 works so that the queues does fill up with that.

[ryrynz]

Madshi, could you allow for chapter markings to be seen on the exclusive seek bar please?

[yogibeezwax]

Thanks for all the replies on my issue with exclusive mode! I'd quote ppl, but I'm a bit distracted with family stuff right now (you all know who you are) and I truly appreciate it!

[el Filou]

Well if you test an older version of MadVr, you'll see that NNEDI3 works so that the queues does fill up with that.IIRC, NNEDI3 used an OpenCL interop. Maybe that's where the difference in behaviour is from.

[Betroz]

IIRC, NNEDI3 used an OpenCL interop. Maybe that's where the difference in behaviour is from.

Maybe yes, but only Madshi can say for sure.

[Seedmanc]

Actually it goes bonkers with GSYNC even without Smooth Motion enabled sometimes.

[sat4all]

hi guys,

I’m trying to calibrate my lg oled hdr mode and i need a bit of advice.
I’ve already created a 3d-lut while setting both displaycal and madtpg to 400nits. But i have a problem with some movies, take The Shallows UHD as example, with the 3d-lut on it’s blindly bright while with either passthrough or hdr process using pixel shaders math it looks normal.
Using hdr 3d-lut is no choice for me, because my tv color gamut is set to wide (native) for sdr 3d-lut calibration and my tv share the same preset for both sdr and hdr so without hdr calibration while tv gamut set to wide, colors will be off.
My Display hdr peak luminance is arround 400nits so i have set displaycal to 400cd/m and hdr process, do i have to set madtpg hdr tap to 400nits too or something else?

[Clown shoes]

hi guys,

IÂ’m trying to calibrate my lg oled hdr mode and i need a bit of advice.
IÂ’ve already created a 3d-lut while setting both displaycal and madtpg to 400nits. But i have a problem with some movies, take The Shallows UHD as example, with the 3d-lut on itÂ’s blindly bright while with either passthrough or hdr process using pixel shaders math it looks normal.
Using hdr 3d-lut is no choice for me, because my tv color gamut is set to wide (native) for sdr 3d-lut calibration and my tv share the same preset for both sdr and hdr so without hdr calibration while tv gamut set to wide, colors will be off.
My Display hdr peak luminance is arround 400nits so i have set displaycal to 400cd/m and hdr process, do i have to set madtpg hdr tap to 400nits too or something else?

OK, so firstly if you're creating an HDR 3D LUT in Displaycal (using the Madvr HDR profile presumably) with a 400 nit max display you should set the peak luminance on the 3D LUT tab to around 1000 nits.

In MadTPG you are emulating an HDR movies metadata, so set the minimum to 0 and the max to 4000.

Personally I increase the patches from the default 175 to just over 1000 as I had some issues with using less, however you may be fine with the default.

Remember to set the peak luminance in Madvr to 1000 nits also.

Let me know how you get on with that :)

[sat4all]

@Clown shoes
thanks much, i will give it a shot and let you know.

[stefanelli73]

OK, so firstly if you're creating an HDR 3D LUT in Displaycal (using the Madvr HDR profile presumably) with a 400 nit max display you should set the peak luminance on the 3D LUT tab to around 1000 nits.

In MadTPG you are emulating an HDR movies metadata, so set the minimum to 0 and the max to 4000.

Personally I increase the patches from the default 175 to just over 1000 as I had some issues with using less, however you may be fine with the default.

Remember to set the peak luminance in Madvr to 1000 nits also.

Let me know how you get on with that :)

then I have a vpr where the peak of luminance is around 120 nits if I want to make a 3dlut "process hdr ...." how many nits should I set in DISPCAL and in madtpg?

[Asmodian]

Using hdr 3d-lut is no choice for me, because my tv color gamut is set to wide (native) for sdr 3d-lut calibration and my tv share the same preset for both sdr and hdr so without hdr calibration while tv gamut set to wide, colors will be off.

HDR uses wide gamut natively, why would your TV in wide gamut mode cause colors to be off for HDR content?