Author Topic: BUG v1.02.11 Color decoding is wrong most of the time  (Read 14752 times)

July 16, 2009, 12:34:20 PM
Reply #15

GuyWD

  • Jr. Member

  • Offline
  • **

  • 88

July 21, 2009, 11:37:01 AM
Reply #16

khkg

  • Newbie

  • Offline
  • *

  • 5
Hokeysmoke:  What programs are you using to see if files are properly flagged or not?

July 21, 2009, 02:57:25 PM
Reply #17

GuyWD

  • Jr. Member

  • Offline
  • **

  • 88
Hokeysmoke, can you please post a BT.601 flagged video as well?

July 21, 2009, 08:17:13 PM
Reply #18

Hokeysmoke

  • Newbie

  • Offline
  • *

  • 29
Here is a 7z archive of 3 1080p23.976 MPEG files.  They are all identical except for the flags.  They are color bar patterns against a reference gray background.  The top bars are encoded with BT.601 values and the bottom bars are encoded using BT.709 values.  When decoded properly, the brightness of the colors will match the brightness of the gray background.  Typically this is easiest to see using color filters, although you can also just visually compare them against each other to figure out how they're being decoded.

cscmatrix1080p23.m2v is unflagged
cscmatrix709flag1080p23.m2v is flagged BT.709
cscmatrix601flag1080p23.m2v is flagged BT.601

I needed to use 7z archiving to fit the attachment into the limitations of this forum.  7z is open source and I use 7-zip from 7-zip.org to compress and decompress these files.
« Last Edit: July 21, 2009, 09:27:56 PM by Hokeysmoke »

July 21, 2009, 09:40:07 PM
Reply #19

Hokeysmoke

  • Newbie

  • Offline
  • *

  • 29
Hokeysmoke:  What programs are you using to see if files are properly flagged or not?

To detect the flags I use DGIndex for MPEG files, h264_parse for h.264 files, and vc1_info for VC-1 files.  For VC-1 and H.264 files I generally create a file with TSMuxer using only the first 10 seconds of the video.  This makes the process much faster.

October 01, 2009, 06:53:28 PM
Reply #20

Hokeysmoke

  • Newbie

  • Offline
  • *

  • 29
This is still not fixed in v1.02.11.

October 13, 2009, 04:28:51 PM
Reply #21

robin of the greenwood

  • Newbie

  • Offline
  • *

  • 3
Does anyone know if the new WDTV Live suffers from the color decoding error that effects the original WDTV?

October 17, 2009, 05:19:08 AM
Reply #22

wiiBox

  • Guest
Does anyone know if the new WDTV Live suffers from the color decoding error that effects the original WDTV?

If it does, that's the least of it's problems in terms of video quality.  On the Live, video playback has lost all it's life compared to the original wdtv.  Everything looks quite drab, especially in shadowed faces.  Some searching revealed it seems I'm not the only one.  I've only had it for a day or so but my experience shows a few glaring problems with PQ.  The one I mentioned earlier and one that seems to show up only on certain videos.  Sort of a grid artifact that shows up during certain colour transitions in the recorded video.  But definitely not there on wdtv1 when I had tested it.  And SD content specially suffers from soft edges.  It's not a huge difference and will probably be fixed once they get that grid bug and fix the decoding engine where most of these bugs probably originate.

I wonder if wdtv1 was anything like this at launch, and I don't know if I want to find out how long it took to get it to where it is now.  I mean, it's only in the last month or so that wdtv1 gained idx/sub support, and not very good support, at that.  Granted, it's not officially listed.

Anyways, I guess I'm saying, if you're at all picky about your AV equipment, you're in a world of hurt when it comes to media players.  Personally, I don't know how I'm supposed to keep the Live unit when video looks so dreadful on it.   ???

October 18, 2009, 02:43:00 PM
Reply #23

wiiBox

  • Guest
After some more time it seems I may have over-stated my case.   ::)

The grid patern turned out to be Tversity was transcoding mkv files on me.  For some reason, I thought upnp/dnla apps and devices were supposed to comunicate and exchange info, and in this case, transcode only when needed.

The other observations were from memory since I returned my wdtv1 right after I got the Live.  In retrospect, and for the sake of curiosity, maybe I should have compared them side by side before returning it.  However, I was able to compare the Live output to my Xbox 360 (something I did with the wdtv1, too) and I have to say the differences are about the same.  I still think the Live output is a bit drab and may end up getting a tweaking before it's too long.

(edit: glaring spelling mistake!)
« Last Edit: August 18, 2011, 05:00:43 PM by wiiBox »

October 18, 2009, 07:43:13 PM
Reply #24

khkg

  • Newbie

  • Offline
  • *

  • 5
Does anyone know if the new WDTV Live suffers from the color decoding error that effects the original WDTV?

Does anyone know of *any* media players that don't suffer from this issue.  I've sadly taken to converting my .mkvs to Xbox360 compatible format  :(

November 06, 2009, 02:48:41 PM
Reply #25

wiiBox

  • Guest
Does anyone know if the new WDTV Live suffers from the color decoding error that effects the original WDTV?

Does anyone know of *any* media players that don't suffer from this issue.  I've sadly taken to converting my .mkvs to Xbox360 compatible format  :(

Actually, on the 360 you can manually choose the component colour space.  I guess you can use that as a work-around.  It's under display settings.

OT:  I finally got around to testing some of those test files and since no one has mentioned it, I can confirm that the WDTV Live does suffer from this problem.  The alternating test file lights up like a xmass tree!

Odd how on the computer the colour ramps don't even blink.  Ah, the weird and wacky world of video.  You learn something new every day!  It took me a while to get my head around the whole black level thing!  It seemed perfectly straight forward at first glance, but it was not.

PS: I know it's a relatively old thread but any update would be greatly appretiated.  Especially from GuyWD!   ::)

November 15, 2009, 09:12:58 PM
Reply #26

khkg

  • Newbie

  • Offline
  • *

  • 5

January 23, 2010, 09:03:27 AM
Reply #27

Hokeysmoke

  • Newbie

  • Offline
  • *

  • 29
Odd how on the computer the colour ramps don't even blink.  Ah, the weird and wacky world of video. 

I don't know about the software player you're using, but PowerDVD 7.3 always uses BT.601 no matter what the flag reads.  This makes the test file stable, but wrong all of the time.  So the test file is not the only thing you should check to make sure your player is working right.

January 23, 2010, 08:39:44 PM
Reply #28

wiiBox

  • Guest
Odd how on the computer the colour ramps don't even blink.  Ah, the weird and wacky world of video.

I don't know about the software player you're using, but PowerDVD 7.3 always uses BT.601 no matter what the flag reads.  This makes the test file stable, but wrong all of the time.  So the test file is not the only thing you should check to make sure your player is working right.
Hey Hokeysmoke!

I hear what you're saying, my player seems to detect by resolution and I can manually switch back and forth between sd and hd colour space.

BTW, have you tried the patch for your NMT?  Is it out to the public or still in beta?

Also, do you know if Realtek players also suffer from this problem?

January 24, 2010, 09:00:57 AM
Reply #29

Hokeysmoke

  • Newbie

  • Offline
  • *

  • 29
have you tried the patch for your NMT?  Is it out to the public or still in beta?

Also, do you know if Realtek players also suffer from this problem?

The patch is not even in beta yet, although a few people are trying it out.  Syabas suggested that it could be ready in about a week.

The XTreamer, which is Realtek player, apparently does not suffer from this issue at all, although it uses BT.709 for all HD files.  This is certainly preferred to using BT.601 most of the time.