HDR-issue with 2017 LG OLEDs or with the Vero?

Can you clue me in why? All 3 files have different max display luminance values.

Was your Vero 4K+ plugged into the same HDMI port as your FireTV stick on your TV? If not, check if you have the same picture preset for both inputs when you tested. The preset on your TV for the Vero should not have Dynamic Tone Mapping enabled (or some other Dynamic-whatever feature). DTM ignores HDR metadata and does a frame-by-frame analysis to apply its own interpretation of the best tonemapping.

I’ve never fully trusted the Shield, it had color issues since the beginning and for quite some time.

Because Vero is defaulting to 1000nit in the metadata.

I’m confused. I thought the Vero is supposed to correctly transmit 1000 or 4000 or whatever it says in the mastering display luminance metadata. Isn’t that sort of the point of this thread? Sam said in this thread the Vero only falls back to 1000 when there’s no metadata (or the SEI parser cannot extract this data).
What am I missing?

You’re missing that in most cases the SEI parser cannot extract the data so the default is transmitted. It’s an issue we’ve been struggling with and have been trying to get some answers from AMLogic for a while.

So it is unsurprising that 1000nits is transmitted. It is surprising that displays don’t apparently always respond in the way expected.

What?! This is a brand new piece of information to me. This makes all of this testing useless. Even more so as it seems the SEI parser is randomly working or not working as clearly the same files work for some people on here but not for others.
It sounds like things won’t change with the upcoming kernel(?)

Wouldn’t a better solution be to extract the HDR data yourself (MediaInfo lib or something) in that case and transmit that data instead of the 1000 fallback? Otherwise there’s really no point in keeping the Vero 4K :neutral_face:

Actually I have no data to back up the ‘most’. What Vero can and can’t do was discussed at length in HDR too dark on TV. Somewhere in there it says some seis are being read and some aren’t and it’s clear what was done with the default mastering data at that time.

We looked at trying to extract the sei data by other means as a workaround but it wasn’t practical. It would be even less practical for HDR10+.

Here we go again…

Go where? I just want my shit to work. After 2 years.

1 Like

Yes the recent news are slightly discouraging to say the least. I was also under the impression that the only thing between now and perfect HDR playback ( non Dolby vision) on the vero is the new kernel people were talking about.

Even more concerning for me is that I can’t just say oh shit happens and buy Some other device to use until maybe vero catches up,or maybe can I ?

If I’m properly informed Apple TV 4K plays only Netflix and co content perfectly but for proper 4k HDR remux with Dolby vision or HDR+ playback from file id need to buy one of the oppo clones or try to find some guy that doesn’t know what he has and is selling his oppo for a reasonable price because new in box oppos now sell for 5x or 6x of the original price

I still didn’t pull the tiger and buy a flagship 4k tv but once I’ll do it I’ll want to make damn sure that the content I’m watching is presented the way it was intended to because if not why spent a few thousand on a tv and a thousand more on professional calibration if then the media player isn’t capable to transfer what’s on the disc to the tv correctly. Might as well hold off a nother few years and spend the money on something else instead

1 Like

I already did, gonna pick it up later today :slight_smile:

Both are connected to my Denon amp which goes to my TV.
Both used the same Technicolor (expert) presets and had DTM turned off.

We have two issues:

  • Vero is typically sending 1000nit maximum display lumi and no MaxCLL/MaxFALL. The new kernel does pass through all static metadata in tests done so far so will solve that.
  • despite Vero sending consistent static metadata (what it sends depends only on the source), the results reported above, all using the same sources, show there is a variation in response of ppl’s displays, depending on model, year and settings

It’s the second one I’m scratching my head over. User-defined tonecurves should help, and we know how to do that.

1 Like

Let us know if it plays everything exactly as you expect and I’ll ask Sam to put it on the test bench.

1 Like

hm, isn’t this:

and this:

kind of… contradictory? Or did you mean to write “we were struggling with”? To me it sounded like even with the new kernel the metadata won’t be correctly transmitted if the SEI parser randomly and inexplicably decides to not parse that data.

No - the metadata and parsing is solved in the new kernel.

What’s puzzling us is whether this will keep everyone happy, or we will still see varied responses.

3 Likes

That’s good to hear. Hopefully we get to test it soon, the return period on my newly bought device ends in 14 days and I’d rather keep the Vero :grimacing:

What device do you consider to be a good alternative to Vero4K+? I have heard about Odroid N2 but have not read a comprehensive comparison between the two.

I understand that none of these boxes are a “perfect” solution for everyone but with tech evolving so fast there are bound to be some devices which check more boxes compared to others.

It wouldn’t make sense (at least in my mind) for @HDRpotato to recommend the N2 as a new solution if he has issues with the SEI parser. Both devices implement the same SEI parser and CE on the N2 also implements OSMC’S CS/SD switching approach in their kernel, albeit with some caveats.

It may be hasty to look towards new hardware considering the improvements we have around the corner.

Sam

I don’t think this is the right place to talk about this.

Oh I am definitely not looking to replace the Vero4k+ I have and use as a daily driver. It was just a question to keep myself informed. As I said there’s no one-size-fits-all box as far as I know but Vero, for me, checks most of the boxes that I care about.

I read some info a few days ago about N2 but never researched the product in any great detail. This thread got me curious if N2 did “HDR” any better than Vero4k+ as I have a LG B7 OLED as well. I haven’t tested the files @HDRpotato posted - next week probably.

Thanks for the explanation about OSMC/N2 HDR handling. Glad you guys are working on further developments.