21 replies to this topic

[Pr2]

Hello,

 

Tested on the VU+ Solo4K.

 

In the infobar there is an icon that display the image format (by default it is set back to 4:3 on every zap).

 

If you zap to an HD channel the icon turns into HD but switching to another HD channel the icon remains on 4:3.

In fact once a image format is detected,  switching to another channel with the same image resolution (HD to HD, 4K to 4K) always report 4:3 in the infobar.

 

If you want the proper image format you need to switch between different image resolution, so switching form HD to 4K will properly display 4K, from 4K to HD will display HD but from 4K to 4K will display 4:3 from HD to HD will display 4:3.

 

I was first thinking that this problem occurs between 4K channels that where I saw it first, until I test between HD channels and saw that we have the same bug there.

 

Remark: the channel  "Test technique UHD NRJ12"  (DVB-S2 10713 H 23500 3/4 8PSK 19.2E) that use an image format of 3830x2152p50 is reported as an HD channel instead of 4K.

 

Pr2

[Dimitrij]

My mistake https://github.com/O...08ff6fdbd5c00f4

 

for info...

enigma2-plugin-extensions-vcs solved this problem until canceled patch

Edited by Dimitrij, 25 November 2016 - 21:12.

[Pr2]

No problem the forum is there to report problem and OpenPLi team is always fast to fix bugs that's also why I'd love to use OpenPLi images. 

[littlesat]

I'm afraid also a combi with a driver madness

[Pr2]

A restart of just OpenPLi (not a full restart of the box) fix the problem on the channel you watch (of course bug is back when you zap).

But indeed i just check and the ET-9000 doesn't seems to have this bug.

[DeBeek]

Same problem on DM8000

 1_0_19_1B1C_802_2_11A0000_0_0_0.jpg   132.73KB   6 downloads  1_0_1_44D_8A6_2210_EEEE0000_0_0_0.jpg   173.09KB   6 downloads

[littlesat]

So should we revert that patch?

[Dimitrij]

So should we revert that patch?

seem yes

[DeBeek]

indeed   yes.

[littlesat]

And all about driver maddness. so the drivers cue the video resolution and only forward changes... So we in python need to cue them again to get this actually right...
Bad choise somehow from the driver builders to cue this stuff....

[Pr2]

Cannot you change the way this variable is set to avoid resetting it to 4:3 on every zap? This will fix the problem and should not arm other drivers that always report the right resolution on every zap.

[Pr2]

Remark: the channel  "Test technique UHD NRJ12"  (DVB-S2 10713 H 23500 3/4 8PSK 19.2E) that use an image format of 3830x2152p50 is reported as an HD channel instead of 4K.

 

Hi,

 

Thanks the problem is solved with the roll-back of the patch.

 

But the wrong detection for "Test technique UHD NRJ 12" remains, this channel is reported as HD while it should be 4K, due to this, it report an image format of 3830x2152p50 instead of  3840x2160p50

 

Thanks,

 

Pr2

Edited by Pr2, 27 November 2016 - 12:36.

[Erik Slagter]

Ah, so the test should be > HD instead of == UHD.

[littlesat]

Wierd.... Because actually that reaolution does not exist... I'm afraid a driver issue....

[Pr2]

Ah, so the test should be > HD instead of == UHD.

 

Indeed I think that ranges should be used, this would be more specific and taking into account that Japan start testing 8K too :-)

 

Something like this:

< 650 is SD 

< 1620 =  (1080+2160)/2  is HD  (this will cover 720 and 1080)

<3240 = (2160+4320)/2    is 4K UHD (this cover values around 2160)

>3240  is   8K UHD (this cover values around 4320)

 

Then you don't need to get precise value to select the rigth indication.

[Erik Slagter]

Wierd.... Because actually that reaolution does not exist... I'm afraid a driver issue....

There is no such thing as a resolution "that does not exist". You can encode to any resolution you like. Whether it's DVB compliant, that's another story. Besides that, with anamorphic resolutions, the encoded resolution can actually be different from the displayed resolution, through the sample aspect ratio.

[Erik Slagter]

 

Ah, so the test should be > HD instead of == UHD.

 

Indeed I think that ranges should be used, this would be more specific and taking into account that Japan start testing 8K too :-)

 

Something like this:

< 650 is SD 

< 1620 =  (1080+2160)/2  is HD  (this will cover 720 and 1080)

<3240 = (2160+4320)/2    is 4K UHD (this cover values around 2160)

>3240  is   8K UHD (this cover values around 4320)

 

Then you don't need to get precise value to select the rigth indication.

Yes, something like that. Be careful with SD resolutions, because they're less tightly regulated, everything under 720x576 is allowed (in steps of 16).

 

Also I actually very much dislike the "4k" denomination. It simply isn't correct, as 3840 != 4k, 4k resolution = 4096, with a slightly different aspect ratio. Also DVB has no standard for "4k", it does for "UHD".

[littlesat]

t simply isn't correct, as 3840 != 4k, 4k resolution = 4096, \

-> Sorry that is not the reason why it is called 4K..... however sometimes it is indeed mentioned that way....

[littlesat]

https://github.com/O...9b6bd39436dd78b

[Erik Slagter]

t simply isn't correct, as 3840 != 4k, 4k resolution = 4096, \

-> Sorry that is not the reason why it is called 4K..... however sometimes it is indeed mentioned that way....

UHD has multiple resolutions: 1920x1080p (!), 3840x2160p and 7680x4320. None of these are officially called "4k" (or "8k" or "2k" for that matter). 4k = 4096 x 2160. I'd prefer to use the proper name "UHD", as neither 1920x1080p nor 7680x4320p are actually used in DVB, it's perfectly clear what's meant.