AVS Forum banner

3301 - 3320 of 5333 Posts

·
Registered
Joined
·
2,033 Posts
Discussion Starter #3,301 (Edited)
That's good to know. I frequently type from a computer that uses an AMD B450 motherboard. At some point, I planned to purchase an official LG UHD drive if the price was low enough. At least I know now that it should be easy to flash the firmware.

What app did you use to flash it?
 

·
Registered
Joined
·
7,625 Posts
That's good to know. I frequently type from a computer that uses an AMD B450 motherboard. At some point, I planned to purchase an official LG UHD drive if the price was low enough. At least I know now that it should be easy to flash the firmware.



What app did you use to flash it?


Just to be clear I'm talking about friendly drives. The official drives were not recommended. I used the tool on the makemkv forums. I could look into it later to be sure which one.


Sent from my iPad using Tapatalk
 

·
Premium Member
Joined
·
3,998 Posts
Be careful... I don't think modified software utilities can be discussed here.
 

·
Registered
Joined
·
4,586 Posts
480 nits places the knee point (roll-off point) where tone mapping begins at 100 nits for a 10,000 nits scene peak, so it is prevents 0-100 nits from being compressed, which can cause the image to appear flat.

HDR to SDR display mapping and roll-offs are explained somewhat here:

https://forum.kodi.tv/showthread.php?tid=223175&pid=2745194#pid2745194



Torrents are the primary sources that are encoded without black bars. Any 1:1 rip, regardless of aspect ratio, will be clipped correctly by the clipping algorithm. The live algorithm has an issue with blowing out the highlights in some scenes. That is the only difference.



You could, but some scenes will be brighter than they were originally mastered.



Grey blacks are typically a gamma tracking issue. Buying a colorimeter and creating a 2.20 3D LUT with DisplayCal would be your best bet for improving your gamma tracking. HDR to SDR is hard on any display with less than ideal near black gamma tracking because it renders a lot of detail near black. This would be less apparent with SDR content, as SDR doesn't require heavy tone mapping.



I can only recommend matching madVR to the calibrated gamma of your projector to get better blacks, or create a 3D LUT.

I haven't had any issues with color myself. You are trying to do HDR at 50 nits, which is quite difficult if not impossible when the source master can be as high as 10,000 nits. When I compare screenshots I've taken to Blu-ray reviews online, the hue is almost always identical and only the color saturation is slightly different. But I wouldn't say the color is incorrect.
Inputted 50 real peak nits with the assumption that he screen gain is 1.0. A variant of this screen was measured 33% lower than claimed.
Can the X-Rite ColorMunki Display measure screen gain? Is a 1.0 surface required for comparison or can the light coming from the projector be compared to the light from the screen?

Used 200-250 real nits for the LCD monitor instead of 300(specification) because it's 10 years old and I assume the backlight has dimmed. The guide states that Enter the Minimal target nits / real display peak nits (50-150 recommended for most). Is 150 an upper limit or should this be the peak nits of the display? What if a display has 500 peak nits?
Monitor gamma after calibration with the HDR templates is 2.4. 0 compression limit, 100 dynamic tuning, the rest as in your screenshot.

The monitor is an IPS and the ALR coating on the screen do raise the black floor (both have an IPS glow effect). Seen 25fL/85 nits recommended for HDR on projectors, which is not that much. Activated BC (27% more brightness), set the real nits to 67. It did not make that much of a difference. The calibration still set gamma at 2.2.
I could maximize the nits to 116 (with new lamp, closest positioning), and 164 with the other modes (which have BC on), but that would still be less than the monitor.
The issue with the dark scenes being too dark is that it shows up very similar on the monitor, which is supposed to be brighter than the projector.
Desaturation may not be the correct term, but there seems to be a loss of color vibrancy/brightness, most noticeable in dark scenes. Especially when compared to the SDR versions. I understand the algorithm compresses the nits, but there are a lot of dark scenes in movies. Grey blacks due to the display's I can understand, but the colors should be brighter. Perhaps it's a studio's creative decision with the 4K releases, and/or my subjective impression.


Soulnight's tool creates an Original folder when no previous measurements are present. What is stored in this folder, the first measurement?

There was a thread which explained the MaxCLL, MaxFALL, and all the other values, but can't find it.

(FALL algo) in this post Link claims And if real nits is lower than the no compression limit, it does not play any role except as lower limiter. So at 50 peak nits the 50 is a limiter for no compression?

With the exception of dynamic tuning/100 rather than 50, is this accurate?
Just a quick reminder on the FALL algo formula:
Below the "no compression limit":
Target nits = Peak nits
Above the "no compression limit ":
Target nits = no compression limit + %dynamicTuning/50 × 2×Fall
So the non compression limit is used as an offset/start value in the formula.
Also important, the true "no compression limit " which is being used is equal to:
max (entered real nits, entered no compression limit.)
So when neo-xp says zero no compression limit, he actually uses 150 (his entered real nits) instead.
If someone has only 50nits like myself, I would still advise to use a no compression limit of 150 since it is also used as an offset. Minimum should be 100 nits ( and I have probably already hardcoded that minimum).

Used 0 so far, and in dark scenes there is no difference. So is the compression limit 50 nits (for 50 real peak nits)?

In the formula for the target peak nits, does this part Max(no compression limit, min target / real display peak nits) refer to which ever is greater, the no compression limit or real nits?

What about in the Live algo, what if compression limit is greater than peak nits?

Not sure what overblown looks like ( with dynamic clipping in the Live algo), but in medium to bright scenes activating it (with 100 recovery strength) gives a brightness boost, similar to going from 100 to 50 dynamic tuning. Is this effect bad, because I like the increased brightness.
The files put through Soulnight's utility are darker (with/without dynamic clipping) (either FALL/BT2390) vs the Live algo without dynamic clipping. Algo's settings: 50 peak nits, 100 dynamic tuning, 100 clipping strength, dynamic target nits selection, Preset 3.
What does dynamic clipping do? In this post Link it's suggested that it cuts some of the highlights. The statistics in the tool also indicate this.
I can't see any difference with the algo (FALL or BT2390) between dynamic clipping or without. In the two attachments, under master luminance the values are different - with the algo on the left, without algo on the right. Picture 1 is FALL algo, picture 2 is BT2390. Also can't see any difference, at least the files I've tried, between the the FALL and BT2390 algos.
What are % decompressed pixels? (in the expanded statistics).
 

Attachments

·
Registered
Joined
·
2,033 Posts
Discussion Starter #3,305
Just to be clear I'm talking about friendly drives. The official drives were not recommended. I used the tool on the makemkv forums. I could look into it later to be sure which one.


Sent from my iPad using Tapatalk
MakeMKV now supports official LG UHD drives and UHD friendly drives:

https://www.makemkv.com/forum/viewtopic.php?f=16&t=18346

As far as I know, you use the same software to roll back the firmware. I've heard and am hoping they are more reliable at reading every UHD disc on the first try without issue because they are designed for that purpose.

I'll pass along the fact that you can use AMD motherboards to flash friendly drives without even having to use SATA 1 or IDE.
 

·
Registered
Joined
·
2,033 Posts
Discussion Starter #3,307
Inputted 50 real peak nits with the assumption that he screen gain is 1.0. A variant of this screen was measured 33% lower than claimed.
Can the X-Rite ColorMunki Display measure screen gain? Is a 1.0 surface required for comparison or can the light coming from the projector be compared to the light from the screen?

Used 200-250 real nits for the LCD monitor instead of 300(specification) because it's 10 years old and I assume the backlight has dimmed. The guide states that Enter the Minimal target nits / real display peak nits (50-150 recommended for most). Is 150 an upper limit or should this be the peak nits of the display? What if a display has 500 peak nits?
Monitor gamma after calibration with the HDR templates is 2.4. 0 compression limit, 100 dynamic tuning, the rest as in your screenshot.

The monitor is an IPS and the ALR coating on the screen do raise the black floor (both have an IPS glow effect). Seen 25fL/85 nits recommended for HDR on projectors, which is not that much. Activated BC (27% more brightness), set the real nits to 67. It did not make that much of a difference. The calibration still set gamma at 2.2.
I could maximize the nits to 116 (with new lamp, closest positioning), and 164 with the other modes (which have BC on), but that would still be less than the monitor.
The issue with the dark scenes being too dark is that it shows up very similar on the monitor, which is supposed to be brighter than the projector.
Desaturation may not be the correct term, but there seems to be a loss of color vibrancy/brightness, most noticeable in dark scenes. Especially when compared to the SDR versions. I understand the algorithm compresses the nits, but there are a lot of dark scenes in movies. Grey blacks due to the display's I can understand, but the colors should be brighter. Perhaps it's a studio's creative decision with the 4K releases, and/or my subjective impression.


Soulnight's tool creates an Original folder when no previous measurements are present. What is stored in this folder, the first measurement?

There was a thread which explained the MaxCLL, MaxFALL, and all the other values, but can't find it.

(FALL algo) in this post Link claims And if real nits is lower than the no compression limit, it does not play any role except as lower limiter. So at 50 peak nits the 50 is a limiter for no compression?

With the exception of dynamic tuning/100 rather than 50, is this accurate?
Just a quick reminder on the FALL algo formula:
Below the "no compression limit":
Target nits = Peak nits
Above the "no compression limit ":
Target nits = no compression limit + %dynamicTuning/50 × 2×Fall
So the non compression limit is used as an offset/start value in the formula.
Also important, the true "no compression limit " which is being used is equal to:
max (entered real nits, entered no compression limit.)
So when neo-xp says zero no compression limit, he actually uses 150 (his entered real nits) instead.
If someone has only 50nits like myself, I would still advise to use a no compression limit of 150 since it is also used as an offset. Minimum should be 100 nits ( and I have probably already hardcoded that minimum).

Used 0 so far, and in dark scenes there is no difference. So is the compression limit 50 nits (for 50 real peak nits)?

In the formula for the target peak nits, does this part Max(no compression limit, min target / real display peak nits) refer to which ever is greater, the no compression limit or real nits?

What about in the Live algo, what if compression limit is greater than peak nits?

Not sure what overblown looks like ( with dynamic clipping in the Live algo), but in medium to bright scenes activating it (with 100 recovery strength) gives a brightness boost, similar to going from 100 to 50 dynamic tuning. Is this effect bad, because I like the increased brightness.
The files put through Soulnight's utility are darker (with/without dynamic clipping) (either FALL/BT2390) vs the Live algo without dynamic clipping. Algo's settings: 50 peak nits, 100 dynamic tuning, 100 clipping strength, dynamic target nits selection, Preset 3.
What does dynamic clipping do? In this post Link it's suggested that it cuts some of the highlights. The statistics in the tool also indicate this.
I can't see any difference with the algo (FALL or BT2390) between dynamic clipping or without. In the two attachments, under master luminance the values are different - with the algo on the left, without algo on the right. Picture 1 is FALL algo, picture 2 is BT2390. Also can't see any difference, at least the files I've tried, between the the FALL and BT2390 algos.
What are % decompressed pixels? (in the expanded statistics).
That is a Goddamn mess. I'll have to read through it and reply some other time.
 

·
Registered
Joined
·
7,625 Posts
That is a Goddamn mess. I'll have to read through it and reply some other time.
Well onkyoman was good for a laugh today lol, don't know why this was to funny to me but it was.
 

·
Registered
Joined
·
227 Posts
Guys, considering latest fiasco with previous Win Update and necessity to roll back, with lots of hustle attached, I was wondering if it safe to update to this one: https://www.techradar.com/news/wind...ch-and-adds-gpu-temps-reading-to-task-manager and it will not screw HDR as before:

I need some help here, latest Windows 10 May 2019 Update broke HDR settings or something, any HDR move is now displayed in magenta colours, and even if I tried to switch on HDR in Windows (HDR for games) it does the same, I searched quickly in the forum but could not find any solution.

Anyone came across this?
 

·
Registered
Joined
·
7,625 Posts
Anybody tone mapping with a ryzen 2400g? Was wondering if it could be done with basically everything else turned off. Thinking about the smallest Kodi box I could build and still have the ability to use madvr with tone mapping. Preferably without a dedicated GPU.

My 1060 handles it fine. And still renders at like 24. Seems like I could have less power and be ok. Not sure what the minimum is though.

EDIT## Seems like it is unnecessary, HDR isn't great on my Sammy when set to passthrough, but on my TCL's it looks as expected. Still might bite the bullet and build a mini pc with something like a mini 1070.
 

·
Registered
Joined
·
7,625 Posts
if you want it small look at zotac zbox.


Yeah those are sweet. But the diy guy in me has a hard time buying a prebuilt. Mainly because of cost. I think I worked out my HDR issue though. Probably going to buy a little Odroid to run Coreelec to play around with.
 

·
Registered
Joined
·
7,625 Posts
Can someone give me a pointer here. I am trying to having a group for 720p60 and 720p24.

I thought I had done it right, but when I play a video that is 720p 23.976, the profile for standard 720 is being used because it is using NGU low instead of NGU high which I have set in the 24p profile.

I have confirmed this by adjsting the NGU setting in the standard 720p profile.
 

Attachments

·
Registered
Joined
·
2,033 Posts
Discussion Starter #3,315 (Edited)
Probably going to buy a little Odroid to run Coreelec to play around with.
I just found out about the ODROID-N2 recently. It looks like a good replacement for the Rasberry Pi with actual support for HDR passthrough on a Linux operating system. I'm curious what you can do with the Google Play Store mentioned in the Odroid documentation.

The specs look promising and the price is very good.
 

·
Registered
Joined
·
2,033 Posts
Discussion Starter #3,316
Can someone give me a pointer here. I am trying to having a group for 720p60 and 720p24.

I thought I had done it right, but when I play a video that is 720p 23.976, the profile for standard 720 is being used because it is using NGU low instead of NGU high which I have set in the 24p profile.

I have confirmed this by adjsting the NGU setting in the standard 720p profile.
You are missing (deintFps 25) and (srcHeight > 1080) "2160p60"
else if (deintFps > 25) and (srcWidth > 1920) "2160p60"

else if (deintFps 720) and (srcHeight 720) and (srcHeight 25) and ((srcWidth > 1280) and (srcWidth 960) and (srcWidth
 

·
Registered
Joined
·
7,625 Posts
I just found out about the ODROID-N2 recently. It looks like a good replacement for the Rasberry Pi with actual support for HDR passthrough on a Linux operating system. I'm curious what you can do with the Google Play Store mentioned in the Odroid documentation.

The specs look promising and the price is very good.
Yeah I was debating whether it would be better to use Android or CoreElec. With android you could get stuff like Netflix, but with it not being certified the program is gimped so not really worth it imo. Have you seen the kit that just came out preloaded with coreElec? Pretty sweet deal ready to go out of the box. I think I am going to get it. If anything just to play around, maybe do some retro gaming on it.

https://ameridroid.com/collections/new/products/odroid-n2-coreelec-edition
You are missing (deintFps 25) and (srcHeight > 1080) "2160p60"
else if (deintFps > 25) and (srcWidth > 1920) "2160p60"

else if (deintFps 720) and (srcHeight 720) and (srcHeight 25) and ((srcWidth > 1280) and (srcWidth 960) and (srcWidth
 

·
Registered
Joined
·
2,975 Posts
madVR by default is supersampling 720p make sure it doesn't do that and you should be good using jinc for the last scaling step is not really worth it too.
 

·
Registered
Joined
·
227 Posts
Guys, since Geforce 1080ti come affordable, I have had and idea of upgrading my MSI Geforce 1060 6G with MSI 1080ti 11Gb, but is it really worth it?, I have most of madVr 1080->4K up-conversion settings at its mid range at the moment. Is it worth $500 upgrade? It is 5.5/6 meters viewing distance with 180" screen.
 

·
Registered
Joined
·
2,033 Posts
Discussion Starter #3,320
When I Google the Sony VPL-HW65ES in your signature, it says it is a 1080p projector. In that case, you don't need the extra horsepower.

If you had a true 4K projector, it could be worth the investment. But I would wait until the next build is released to determine what you will be upgrading for. Tone mapping HDR10 content at maximum quality, adding a 3D LUT and increasing your upscaling settings can all be worth it when you are using a projector and screen of that size.
 
3301 - 3320 of 5333 Posts
Top