[DoVi_Scripts] Multi-Function Scripts for Dolby Vision processing and a lot more...

[SamuriHL]

That's a good question. No, not specifically. I assumed, clearly wrongly, that if Resolve was generating the scene cuts they should automagically be aligned. Still learning. :)

[Kuler087]

Yes, they should be aligned. Just trying to see what went wrong.

[SamuriHL]

From what I can see here this really SHOULDN'T be doing the brightness pumping/flashing thing but this is not my area of expertise at all.


Extracting track 0 with the CodecID 'V_MPEGH/ISO/HEVC' to the file 'D:\DoVi_scripts\temp.folder9\BL.hevc'. Container format: HEVC/H.265 elementary stream
Progress: 100%
Reordering metadata... Done.
Parsing RPU file...
{
"dovi_profile": 8,
"header": {
"rpu_nal_prefix": 25,
"rpu_type": 2,
"rpu_format": 18,
"vdr_rpu_profile": 1,
"vdr_rpu_level": 0,
"vdr_seq_info_present_flag": true,
"chroma_resampling_explicit_filter_flag": false,
"coefficient_data_type": 0,
"coefficient_log2_denom": 23,
"coefficient_log2_denom_length": 23,
"vdr_rpu_normalized_idc": 1,
"bl_video_full_range_flag": false,
"bl_bit_depth_minus8": 2,
"el_bit_depth_minus8": 2,
"ext_mapping_idc_0_4": 0,
"ext_mapping_idc_5_7": 0,
"vdr_bit_depth_minus8": 4,
"spatial_resampling_filter_flag": false,
"reserved_zero_3bits": 0,
"el_spatial_resampling_filter_flag": false,
"disable_residual_flag": true,
"vdr_dm_metadata_present_flag": true,
"use_prev_vdr_rpu_flag": false,
"prev_vdr_rpu_id": 0
},
"rpu_data_mapping": {
"vdr_rpu_id": 0,
"mapping_color_space": 0,
"mapping_chroma_format_idc": 0,
"num_x_partitions_minus1": 0,
"num_y_partitions_minus1": 0,
"curves": [
{
"num_pivots_minus2": 0,
"pivots": [
0,
1023
],
"mapping_idc": "Polynomial",
"poly_order_minus1": [
0
],
"linear_interp_flag": [
false
],
"poly_coef_int": [
[
0,
1
]
],
"poly_coef": [
[
0,
0
]
]
},
{
"num_pivots_minus2": 0,
"pivots": [
0,
1023
],
"mapping_idc": "Polynomial",
"poly_order_minus1": [
0
],
"linear_interp_flag": [
false
],
"poly_coef_int": [
[
0,
1
]
],
"poly_coef": [
[
0,
0
]
]
},
{
"num_pivots_minus2": 0,
"pivots": [
0,
1023
],
"mapping_idc": "Polynomial",
"poly_order_minus1": [
0
],
"linear_interp_flag": [
false
],
"poly_coef_int": [
[
0,
1
]
],
"poly_coef": [
[
0,
0
]
]
}
]
},
"vdr_dm_data": {
"compressed": false,
"affected_dm_metadata_id": 0,
"current_dm_metadata_id": 0,
"scene_refresh_flag": 0,
"ycc_to_rgb_coef0": 9574,
"ycc_to_rgb_coef1": 0,
"ycc_to_rgb_coef2": 13802,
"ycc_to_rgb_coef3": 9574,
"ycc_to_rgb_coef4": -1540,
"ycc_to_rgb_coef5": -5348,
"ycc_to_rgb_coef6": 9574,
"ycc_to_rgb_coef7": 17610,
"ycc_to_rgb_coef8": 0,
"ycc_to_rgb_offset0": 16777216,
"ycc_to_rgb_offset1": 134217728,
"ycc_to_rgb_offset2": 134217728,
"rgb_to_lms_coef0": 7222,
"rgb_to_lms_coef1": 8771,
"rgb_to_lms_coef2": 390,
"rgb_to_lms_coef3": 2654,
"rgb_to_lms_coef4": 12430,
"rgb_to_lms_coef5": 1300,
"rgb_to_lms_coef6": 0,
"rgb_to_lms_coef7": 422,
"rgb_to_lms_coef8": 15962,
"signal_eotf": 65535,
"signal_eotf_param0": 0,
"signal_eotf_param1": 0,
"signal_eotf_param2": 0,
"signal_bit_depth": 12,
"signal_color_space": 0,
"signal_chroma_format": 0,
"signal_full_range_flag": 1,
"source_min_pq": 7,
"source_max_pq": 3079,
"source_diagonal": 42,
"cmv29_metadata": {
"num_ext_blocks": 6,
"ext_metadata_blocks": [
{
"Level1": {
"min_pq": 0,
"max_pq": 2893,
"avg_pq": 1377
}
},
{
"Level2": {
"target_max_pq": 2081,
"trim_slope": 1689,
"trim_offset": 2028,
"trim_power": 1509,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 2048,
"ms_weight": 2048
}
},
{
"Level2": {
"target_max_pq": 2851,
"trim_slope": 2003,
"trim_offset": 2048,
"trim_power": 1902,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 2048,
"ms_weight": 2048
}
},
{
"Level2": {
"target_max_pq": 3079,
"trim_slope": 2049,
"trim_offset": 2048,
"trim_power": 2046,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 2048,
"ms_weight": 2048
}
},
{
"Level5": {
"active_area_left_offset": 0,
"active_area_right_offset": 0,
"active_area_top_offset": 264,
"active_area_bottom_offset": 264
}
},
{
"Level6": {
"max_display_mastering_luminance": 1000,
"min_display_mastering_luminance": 1,
"max_content_light_level": 0,
"max_frame_average_light_level": 0
}
}
]
},
"cmv40_metadata": {
"num_ext_blocks": 5,
"ext_metadata_blocks": [
{
"Level3": {
"min_pq_offset": 2048,
"max_pq_offset": 2048,
"avg_pq_offset": 1623
}
},
{
"Level8": {
"length": 10,
"target_display_index": 1,
"trim_slope": 2048,
"trim_offset": 2048,
"trim_power": 2048,
"trim_chroma_weight": 2048,
"trim_saturation_gain": 2048,
"ms_weight": 2048
}
},
{
"Level9": {
"length": 1,
"source_primary_index": 0
}
},
{
"Level11": {
"content_type": 1,
"whitepoint": 0,
"reference_mode_flag": false,
"reserved_byte2": 0,
"reserved_byte3": 0
}
},
{
"Level254": {
"dm_mode": 0,
"dm_version_index": 2
}
}
]
}
},
"rpu_data_crc32": 2450995496
}
Parsing RPU file...

Summary:
Frames: 186363
Profile: 8
DM version: 2 (CM v4.0)
Scene/shot count: 3225
RPU mastering display: 0.0001/1000 nits
RPU content light level (L1): MaxCLL: 1058.23 nits, MaxFALL: 144.96 nits
L6 metadata: Mastering display: 0.0001/1000 nits. MaxCLL: 0 nits, MaxFALL: 0 nits
L5 offsets: top=264, bottom=264, left=0, right=0
L2 trims: 100 nits, 600 nits, 1000 nits
L8 trims: 100 nits
L9 MDP: DCI-P3 D65

Parsing RPU file...
Exporting scenes list...
Done.
Indexing, please wait... 100%
Writing index... done.
Deleting TEMP folder...


Start time: 18:34:13.12
End time: 18:53:55.33
Time taken: 0:19:42


Scenes:
https://pastebin.com/embed_js/D6ApMBWd

Does any of this look suspect to you? Thanks for your help!

[Kuler087]

You have to check those scene cuts with the video, see: https://www.youtube.com/watch?v=lIO_ZfT-LnY

[SamuriHL]

Got it. I'll work on that tomorrow and see what's going on with it.

[en6ads]

Apologies if this has been answered earlier I have been out for several days and am catching up.

If I understand this correctly: Converting (appending) cm2.9 RPUs to cm4.0 for cm4.0 mastered movies will always be better than using CM_Analyze or Resolve to generate new DV from the BL if no web DV exists?

Edit: Here's an example:
Sisu.US-BD.P7.FEL.Generated.T3.OSC.bin (Oct 3rd 2025)
Sisu.US-BD.P7.FEL_Original_L1_CMV4.0_Flag_RPU_Edited.bin (Mar 28th 2026)
Would you recommend using the newer 'Flag' RPU for this movie?



I think I need to scan my library for:

-Mastered in CM2.9 with CM2.9 RPU with no web DV to inject/upgrade [is CM_Analyze / Resolve the best option here for 4000nit MDL, what about 1000nit MDL?]
-Mastered in CM2.9 with CM2.9 RPU with web DV [just inject/upgrade RPU with web]
-Mastered in CM4.0 with CM2.9 RPU with no web DV to inject/upgrade [Append to CM4.0, do not generate new with CM_Analyze / Resolve?]
-Mastered in CM4.0 with CM2.9 RPU with web DV [just inject/upgrade RPU with web]

Would be amazing if your script could parse libraries into these types of lists.

I have a bunch of generated RPUs using CM_Analyze over time, but I did not check master CM version first. Luckily I archive all the original RPUs from disk so it will be easy replace.

Lastly, what logic is used to determine the cm master version?

Many thanks for everything.



These new RPUs are effectively identical to using the Ugoos “append CM v4.0” option. For RPUs that were originally mastered in 4.0 but delivered in 2.9, and where no web source exists to restore the 4.0 metadata, it is preferable to retain the original L1 values rather than generating new metadata.

because of:

https://i.ibb.co/2738txqX/firefox-MTh9-Jw-N24-I.png

[Kuler087]

you can batch plot(6-4) all your RPU, and it will be written if the rpu was mastered in cmv4.0.


https://i.ibb.co/KxttRktg/xnviewmp-82-Ao-Phl-Lmx.png

[olinko]

Hi all,

I've been using dovi_scripts for a while now and I have a question.
I use it mainly because I have an LG OLED that only supports Dolby Vision via MP4 files.

To fix this, I use dovi_script gui.
I put one of my Dolby Vision MKVs in it, select MP4 output, and simply click on dovi_muxer, and that's it.
I end up with a working Dolby Vision MP4 that actually launches Dolby Vision mode once I play it on my TV.

But it seems surprisingly too easy to me, so I'd like to know if this is really the right method to do it?
Is the MP4 file I get really a Dolby Vision file, or does it trigger the mode on my TV but it's not "real" Dolby Vision?

Thanks to anyone who takes time to explain this to me. :thanks:

I came here with the exact situation and exact same question EXCEPT I don't know how to convert from P8 to P5 properly, the GUI starts generating absurdly large temp files I don't have the space for

[Kuler087]

P8 to P5 requires re-encoding with DEE, and DEE is designed for large mezzanine files, not already compressed HEVC video. Unless your TV is super old, it should support P8 anyway.

[Kontrarian]

Wow, I just watched the first Pirates of the Caribbean with my boys tonight... what a horrible UHD!
I didn't do anything to it, just watched the straight up HDR10 UHD.
I plotted it and the movie is essentially SDR wrapped in an HDR container, probably an upscale too.
Disney... go figure.

[DMD]

Wow, I just watched the first Pirates of the Caribbean with my boys tonight... what a horrible UHD!
I didn't do anything to it, just watched the straight up HDR10 UHD.
I plotted it and the movie is essentially SDR wrapped in an HDR container, probably an upscale too.
Disney... go figure.
I confirm the upscaled version
https://www.blu-ray.com/movies/Pirates-of-the-Caribbean-The-Curse-of-the-Black-Pearl-4K-Blu-ray/303535/

https://www.hdnumerique.com/dossiers/1007-test-4k-ultra-hd-blu-ray-black-pearl.html

[Kontrarian]

I just watched the second movie with them tonight.
We literally watch movies in complete darkness and it was still way to dim in FMM.
I switched it to Cinema Home and disabled all the processing I could in that mode and it was watchable.
As a bit of a purist I hate to do that but I think they just authored these disc incorrectly.
The only ever time I've had to switch out of FMM before was Heat 4K.
My brain won't accept sunny outdoor scenes with clear skies looking like they're overcast.

@Kuler087
Did a firmware update ever give us something so we can remove the dithering and/or noise on 5 series OLEDs?
It's almost impossible to see during the vast majority of movie scenes but it's pretty noticeable in very low nit scenes.
Reminds me of my old 50" Panasonic plasma.

[Kuler087]

All the POC are 100–200 nits; only the last one is 1000 nits on BD and 4000 nits on streaming.

Noise: the only fix is to watch everything in DV. HDR10 is shit on the G5.
DV has chrominance overshoot and scanlines sometimes, though.

[daffie]

If I understand this correctly: Converting (appending) cm2.9 RPUs to cm4.0 for cm4.0 mastered movies will always be better than using CM_Analyze or Resolve to generate new DV from the BL if no web DV exists?

Edit: Here's an example:
Sisu.US-BD.P7.FEL.Generated.T3.OSC.bin (Oct 3rd 2025)
Sisu.US-BD.P7.FEL_Original_L1_CMV4.0_Flag_RPU_Edited.bin (Mar 28th 2026)
Would you recommend using the newer 'Flag' RPU for this movie?


From what I learned and know the newer Flag RPU is always better than a Generated RPU.

From best to worse quality :
1. cmv4.0 added via websource
2. Flag RPU
3. Generated RPU

If anything I said is incorrect I'll be happy to be corrected by @Kuler087

[ZiROX]

I'm quite confused with the new CMv4.0-Flag feature. Maybe someone can help me understand the benefits of it or explain what I'm missing in my test scenario.

Hardware used: Sony XR-83A90J OLED and Ugoos AM6B+

So far I had generated CMv4.0 via cm_analyze for all my CMv2.9 UHD-BD movies. I have compared some scenes from Hannibal (2001) using the Universal UK UHD-BD as the differences there are quite strong:

Watching the scene beginning at timestamp 14:03 I get a really dark picture watching the original CMv2.9 version.
Playing the version with generated CMv4.0 metadata (Analysis Tuning 2) as I had used to before the picture is A LOT brighter on my TV, escpecially in that scene.
As it is a "Mastered in CMv4.0" disc I used DoVi_Scripts to append the CMv4.0 flag to the RPU und reinjected it. Playing that version results in the darkened picture again, I don't see a difference in comparison to the original CMv2.9 version.

I have attached the DoVi-Plots for the original version and my generated CMv4.0 metadata.

So what am I missing? How is the movie supposed to look? Is the dark picture the correct/intended look? So far I thought my TV was just not capable of playing the correct brightness with the original disc and generating CMv4.0 would be helping in that case.
So should I stick to generating new metadata or use the new method of appending the flag? Would the result on a newer TV differ from my experience comparing the original and the "flagged" version?

[Kuler087]

Wow, the original metadata for this movie looks very wrong for MEL(and 1000nits MDL). So in this case, the generated metadata will definitely look brighter and better than the original (CMv4.0 flag or not).
My Blu-ray of this movie is FEL, which expands the brightness up to 10,000 nits. But your disc is MEL, so the 1,000-nit base layer cannot be expanded to 10,000 nits, and applying that tone-mapping metadata to a 1,000-nit layer will make it overly dark.

[daffie]

Wow, the original metadata for this movie looks very wrong for MEL(and 1000nits MDL). So in this case, the generated metadata will definitely look brighter and better than the original (CMv4.0 flag or not).
My Blu-ray of this movie is FEL, which expands the brightness up to 10,000 nits. But your disc is MEL, so the 1,000-nit base layer cannot be expanded to 10,000 nits, and applying that tone-mapping metadata to a 1,000-nit layer will make it overly dark.

I guess that really is an exception?
The generated RPU is in most cases inferior to the RPU flag method. Correct?

[Kuler087]

I guess that really is an exception?

Yes, but it's not the first time I've seen original metadata that makes no sense.

The generated RPU is in most cases inferior to the RPU flag method. Correct?

If the original was mastered with the old 2.9/4.0 algorithm, the generated RPU using the newer (2023) analysis tuning may be better. The new algorithm is far more tolerant to compression noise and detail-less specular highlights, whereas the old algorithm could unnecessarily make the image dimmer (more conservative).


https://i.ibb.co/XrsvCLMH/firefox-ODVTOQE7-RV.png

[daffie]

How can you see If the original RPU was mastered with the old 4.0 algorithm?

If the original RPU has 100, 600 and 1000 nits L2 values and it mentions "mastered in cmv4.0" that means it's the NEW 4.0 algorithm?

[Kuler087]

When the RPU is clearly mastered in 4.0 (10 nits avg_pq floor), aside from the release date, there’s no reliable way to tell whether the analysis was performed using the new tunings or Legacy.

[daffie]

When the RPU is clearly mastered in 4.0 (10 nits avg_pq floor), aside from the release date, there’s no reliable way to tell whether the analysis was performed using the new tunings or Legacy.


Only thing we can say for sure is that if the 4k is released before 2023 it's almost certain Legacy.

If the 4k is released in 2023 or later it will probably be the new tunings.

Correct assumption?

[Kuler087]

before September 2023, yes...
Anything after, maybe, maybe not. I'll use the cmv4.0 append for any movie mastered in 4.0, regardless of the use of tuning.

[ZiROX]

Okay, so I just picked the wrong movie/disc for my tests... I just used this one as it was always a good example for what difference the generated CMv4.0 metadata could make on my setup. Just good to know that it just was for all the wrong reasons :D

Thanks for your answer :)

[daffie]

I'll use the cmv4.0 append for any movie mastered in 4.0, regardless of the use of tuning.

Why?

Because you also said :


If the original was mastered with the old 2.9/4.0 algorithm, the generated RPU using the newer (2023) analysis tuning may be better. The new algorithm is far more tolerant to compression noise and detail-less specular highlights, whereas the old algorithm could unnecessarily make the image dimmer (more conservative).

The generated RPU may be better - so why do you still use the cmv4.0 append on all movies mastered in 4.0 ? That seems strange to me to do...
Because if the original RPU was mastered with the old algo, the cmv4.0 append could be worse (dimmer) than the Generated one.

[Kuler087]

Because that's too much work and G5 doesnt need tone mapping for most content.

[daffie]

Because that's too much work and G5 doesnt need tone mapping for most content.

Ok, thought so.

For me with a G1 it's a different story.

[SamuriHL]

Yea and if HDR is truly fixed on the G6, I may start using that more rather than generating DV for them. Definitely for anything under 2000 nits. Anything over 2000 nits and you get more bang for buck with DV CMv4.0. For CMv2.9, I'll be doing the same as Kuler087 and just setting it to always.

[Kuler087]

I'm hearing the G6 is still too bright near black, so DV cmv4.0 might still be the way to go.

Your TV might not have enough usage yet, but can you try these files? On the G5, DV has the scanlines dithering effect and HDR10 flickers because it's too bright near black.

DoVi (https://drive.google.com/file/d/1Vc-DqY8Ug5s7BZK_PCDuxsLuIJAeJQGp/view?usp=drive_link)
HDR10 (https://drive.google.com/file/d/1pi8e5dAwNF2i_8NO6q7DvKkg8k9wBxIi/view?usp=drive_link)

[SamuriHL]

Yea I'll take a look tonight. Right now my display is overbright on both DV and HDR and until it's broken in and I can calibrate it I won't know for sure if it's fixable or not. But I'll see what I can in those files once it's dark enough in here.

[Kuler087]

thanks I'd be curious about the low luminance dithering noise too. The G5 HDR10 has much more noise than DV under 10nits. Also way worse in Green cyan and Yellow


Dovi low lum test file (https://drive.google.com/file/d/17GgXn7PTB3pFEvbpqXiVUowPa8kdKN5Q/view?usp=drive_link)
HDR10 low lum test file (https://drive.google.com/file/d/14VnigMRBxGqnqiqmtxj4-36WyjfAKfq_/view?usp=drive_link)

For reference, the C2 has no dithering noise at all, see: https://www.avsforum.com/threads/lg-2025-g5-series-oled-owners-friends-thread-general-help-settings.3323209/page-864?post_id=64475783#post-64475783

[SamuriHL]

I don't recall my G2 having any either. I have the next few days off after I get off work today so I should be able to do a whole bunch of stuff. I badly want to see how good this new HDR pipeline is now that it's 12 bit processing on the G6. They claim it fixes a bunch of stuff but we'll see.

[dkangel]

hello
weird thing happened to me.
i download a movie with wrong L5 (0.0.0.1320) so i correct it to 0.0.0.0
but doing that the framerate of the resulting hevc file change from 23.97 to 25 so i had to use dovi_script again to correct it
it's the first time i see that, what could be wrong with this file ?
thank you

[Kuler087]

Itunes webdl?
If not already corrected, the framerate bitstream information is missing, and it will default to 25fps.

[dkangel]

yes itune webdl, so it's normal....
but
Video
ID : 1
Format : HEVC
Format/Info : High Efficiency Video Coding
Format profile : Main 10@L5@High
HDR format : Dolby Vision, Version 1.0, Profile 8.1, dvhe.08.06, BL+RPU, no metadata compression, HDR10 compatible / SMPTE ST 2094 App 4, Version HDR10+ Profile B, HDR10+ Profile B compatible
Codec ID : V_MPEGH/ISO/HEVC
Duration : 1 h 31 min
Bit rate : 23.9 Mb/s
Width : 3 828 pixels
Height : 1 596 pixels
Display aspect ratio : 2.40:1
Frame rate mode : Constant
Frame rate : 23.976 FPS

and after correction of L5

Video
ID : 1
Format : HEVC
Format/Info : High Efficiency Video Coding
Format profile : Main 10@L5@High
HDR format : Dolby Vision, Version 1.0, Profile 8.1, dvhe.08.06, BL+RPU, no metadata compression, HDR10 compatible / SMPTE ST 2094 App 4, Version HDR10+ Profile B, HDR10+ Profile B compatible
Codec ID : V_MPEGH/ISO/HEVC
Duration : 1 h 27 min
Bit rate : 25.0 Mb/s
Width : 3 828 pixels
Height : 1 596 pixels
Display aspect ratio : 2.40:1
Frame rate mode : Constant
Frame rate : 25.000 FPS

it seems that framerate info is there in the original file

also that with original file

Summary:
Frames: 603
Profile: 8
DM version: 2 (CM v4.0)
Scene/shot count: 3
RPU mastering display: 0.0001/1000 nits
RPU content light level (L1): MaxCLL: 258.54 nits, MaxFALL: 10.05 nits
L6 metadata: Mastering display: 0.0001/1000 nits. MaxCLL: 421 nits, MaxFALL: 283 nits
L5 offsets: top=0, bottom=0, left=0, right=1320
L2 trims: 100 nits, 600 nits, 1000 nits
L8 trims: 100 nits
L9 MDP: DCI-P3 D65
Base Layer (HDR10+P8): min: 0.0050 cd/m2, max: 1000 cd/m2(BT.2020), MaxCLL: 817nits, MaxFALL: 288nits
Resolution\FPS: 3828 x 1596 @ 23.976

after L5 correction

Summary:
Frames: 603
Profile: 8
DM version: 2 (CM v4.0)
Scene/shot count: 3
RPU mastering display: 0.0001/1000 nits
RPU content light level (L1): MaxCLL: 258.54 nits, MaxFALL: 10.05 nits
L6 metadata: Mastering display: 0.0001/1000 nits. MaxCLL: 421 nits, MaxFALL: 283 nits
L5 offsets: top=0, bottom=0, left=0, right=0
L2 trims: 100 nits, 600 nits, 1000 nits
L8 trims: 100 nits
L9 MDP: DCI-P3 D65
Base Layer (HDR10+P8): min: 0.0050 cd/m2, max: 1000 cd/m2(BT.2020), MaxCLL: 817nits, MaxFALL: 288nits
Resolution\FPS: 3828 x 1596 @ 25.000

[SamuriHL]

It may be Saturday before I get to looking at those videos. Doing a movie night tomorrow night and had other priorities that took too much time tonight. I'm chilling with John Wick 2. Dolby rocks!

[Kuler087]

it seems that framerate info is there in the original file


Just because mediainfo reports a framerate doesnt mean the bitstream is correct. Write a new framerate in 9-2

[Kuler087]

It may be Saturday before I get to looking at those videos. Doing a movie night tomorrow night and had other priorities that took too much time tonight. I'm chilling with John Wick 2. Dolby rocks!

Looking forward to your impressions, even though I’m not upgrading my G5 anytime soon.

[SamuriHL]

Looking forward to your impressions, even though I’m not upgrading my G5 anytime soon.

There's no reason to upgrade a G5. Hell I didn't even really want to upgrade the G2 yet but I will say that in 4 years, LG has certainly been busy. One of the most impressive displays I've ever seen.

[SamuriHL]

I finished John Wick 2 (wow btw) and had a minute to check those 4 videos. Oy what a mess.

Low Luminance DV: it is way WAY WAY overbrightened below 0 nits. That's depressing. Probably able to fix it in calibration.
Low Luminance HDR: This is actually better. It's still too bright below 0 nits but better than DV by far. Should be fairly easy to clean this up in calibration.
Scan lines DV: Oh yea definitely can see those up close. Not as noticeable at distance, of course
Scan lines HDR: Not noticeable even up close and personal. NICE

It would appear HDR is quite impressive out of the box. These are factory default FMM with 13 hours on the panel. I like!

P.S. I've turned OFF VS10 HDR->DV now as I do NOT think that's worth it on this panel given what they've done.

[Kuler087]

Just like the C2/G2 at launch, DV was way too bright… Typical LG, repeating the same mistake almost every year. The next firmware will likely make FMM accurate across all three modes, as it should be.
The overtracking is probably masking the dithering noise, so please check again once it’s fixed.

As for the scanlines sample, do you see flickering in HDR10? On both the C2 and G5, it looks terrible because it comes out of black too quickly.
DV scanlines rarely show up, but when it does, it's annoying. Usually, with poorly graded movies with lifted blacks.

[SamuriHL]

The G6 is coming out of black way slower. No flickering at all. It's actually quite impressive for an uncalibrated display.

Will definitely check again once I calibrate it. We'll see how many hours I get in this weekend as I'm off for 4 days. LOL

[Kontrarian]

Not that my C5 77" matters much in a G5/G6 conversation but the low nit dithering/noise is really something.
Up close it looks like snow coming down with diagonal lines.
I can't see it from my couch though in detail but I can tell that it's not 100% black and that's even with the L5 active area.
Made me think that L5 was not working properly, but it turns out it's just low nit dithering and noise.
I have a Kodi skin (running at 1080p120 SDR) with a black background and white text and the background is 100% black, but playing a movie I get that dithering / noise in the L5 active area.
And I know L5 is working properly because it passes all the L5 test videos.
I wasn't expecting perfection from my C5 and I knew that low nit noise and dithering were going to be there to offset overshoot but yeah, once you see it then you know.
Again, I can't see it from my couch, I have to be pixel peeping to actually see it, but from my couch I can tell that the black is not black.

[Gatorman3385]

I've added a feature request at Blackmagic to add the ability to modify per-clip blanking through the Scripting API to Davinci Resolve. I'm not overly optimistic, but maybe with replies that show support for the idea they'll add it.

https://forum.blackmagicdesign.com/viewtopic.php?f=33&t=234984

[Kuler087]

The G5 noise in DV isn’t really an issue at a normal viewing distance.
The scanlines are far more distracting and easily noticeable from 8–10 ft in a dark room, but thankfully, they don’t occur very often in real content.

On the other hand, HDR10 with the noise and near-black flickering is terrible(and way worse than the C2).

[Kuler087]

I've added a feature request at Blackmagic to add the ability to modify per-clip blanking through the Scripting API to Davinci Resolve. I'm not overly optimistic, but maybe with replies that show support for the idea they'll add it.

https://forum.blackmagicdesign.com/viewtopic.php?f=33&t=234984

Just being able to import an XML that automatically splits the timeline into shots based on scene cuts, with all the metadata intact, would be ideal.
That’s what I would have asked for.

Right now, when you import the XML, it pulls in the tone-mapping metadata per shot, but not the blanking, and the timeline itself isn’t actually split into individual shots.

I wouldn’t automate that in a script though, I prefer verifying blanking manually. In my experience, Disney’s variable L5 metadata often has mistakes, even on recent titles like Tron: Ares.

[Gatorman3385]

That's fair. I can add that to the thread, or reply to the one you created. Do you have a link?

[Kuler087]

That would make the imported metadata fully editable (trims, new shot analysis, and it would even show which tuning analysis was used)
I did not make any request on their forum... I said: That’s what I would have asked for.

[Gatorman3385]

Hah! Reading was never my strong suit. :)

[Gatorman3385]

I took your advice and added a separate feature request.

Import XML that includes all metadata (https://forum.blackmagicdesign.com/viewtopic.php?f=33&t=235076&sid=0bfcc31d5c57351b0ee93ae12bfa2aaa)

[Kuler087]

Thanks, that would be amazing.