How accurate are the colors for component cable input to component cable input (SD 480 and HD 1080)?
-
Our native mobile app has a new name: Fora Communities. Learn more.
Color Conversion problems
This weekend Darin and I took a look at the color conversion of the A1. We found a few issues that may interest you.
HD: HDMI to DVI: RGB
This is the scenario where you see plugging the HDMI output of the A1 into the DVI input of something downstream, like a display. In this case the A1 must convert to RGB.
The conversion used by the A1 for HD is 601, which is incorrect.
The conversion used by the A1 for SD at 480p is 601, which is correct. At 720p and 1080i, it is not using 601 so the results are incorrect.
HD: HDMI to HDMI: YCbCr 444
This is the scenario where you are sending HDMI out of A1 to HDMI in of something downstream like a processor or display.
1080i HDMI out to iScan VP30 HDMI is getting proper 709 conversion and it is good to go.
720p HDMI out to iScan VP30 HDMI is getting improper conversion. We verified that the problem is NOT the VP30 by using an Accupel 720p YCbCr as an alternate source and it looks good.
It appears that there is a double color conversion inside the player when sending 720p out as YCbCr 444. It appears to get converted to RGB and then back to YCbCr. That is just a guess at this point.
The above tests were done by using 601 and 709 colors bars in HD and 601 color bars in SD. We viewed the green only channel on the Samsung DLP. You can turn off colors within the service menu. Viewing the green channel makes it easy to see color decoding errors like this.
HD: HDMI to DVI: RGB
This is the scenario where you see plugging the HDMI output of the A1 into the DVI input of something downstream, like a display. In this case the A1 must convert to RGB.
The conversion used by the A1 for HD is 601, which is incorrect.
The conversion used by the A1 for SD at 480p is 601, which is correct. At 720p and 1080i, it is not using 601 so the results are incorrect.
HD: HDMI to HDMI: YCbCr 444
This is the scenario where you are sending HDMI out of A1 to HDMI in of something downstream like a processor or display.
1080i HDMI out to iScan VP30 HDMI is getting proper 709 conversion and it is good to go.
720p HDMI out to iScan VP30 HDMI is getting improper conversion. We verified that the problem is NOT the VP30 by using an Accupel 720p YCbCr as an alternate source and it looks good.
It appears that there is a double color conversion inside the player when sending 720p out as YCbCr 444. It appears to get converted to RGB and then back to YCbCr. That is just a guess at this point.
The above tests were done by using 601 and 709 colors bars in HD and 601 color bars in SD. We viewed the green only channel on the Samsung DLP. You can turn off colors within the service menu. Viewing the green channel makes it easy to see color decoding errors like this.
- Status
- Not open for further replies.
1 - 20 of 87 Posts
"The conversion used by the A1 for HD is 601, which is incorrect.
The conversion used by the A1 for SD at 480p is 601, which is correct. At 720p and 1080i, it is not using 601 so the results are incorrect."
Hi Stacey,
Which one is correct for HD 601 or 709? Its kind of confusing the text here.
The conversion used by the A1 for SD at 480p is 601, which is correct. At 720p and 1080i, it is not using 601 so the results are incorrect."
Hi Stacey,
Which one is correct for HD 601 or 709? Its kind of confusing the text here.
709 is the correct matrix to use for HD. The player is using 601 instead of 709 when connected to a DVI display.
We did not look at the component outputs, sorry.
We did not look at the component outputs, sorry.
Yep - the 601 vs 709 colour space issues are just going to run and run.
Worse still, you can't assume HD=709 and SD=601 - as some HD to SD downconversions don't change the colour space from 709 to 601 - giving you SD video with a 709 colour space, and similarly there are some HD cameras that can use 601 colour space I believe?
Of course when it comes to cameras, they already have RGB to RGB or RGB to YCrCb/YPrPb matrices anyway to compensate for colourimetry issues in their colour splitting and non pan-chromatic sensitivity of their CCD sensors. Most modern cameras allow you to chose from EBU, RAI, ARD, BBC etc. colourimetry (in Europe at least) - as they all have different matrices - and some of them have different black stretch requirements as well.
Worse still, you can't assume HD=709 and SD=601 - as some HD to SD downconversions don't change the colour space from 709 to 601 - giving you SD video with a 709 colour space, and similarly there are some HD cameras that can use 601 colour space I believe?
Of course when it comes to cameras, they already have RGB to RGB or RGB to YCrCb/YPrPb matrices anyway to compensate for colourimetry issues in their colour splitting and non pan-chromatic sensitivity of their CCD sensors. Most modern cameras allow you to chose from EBU, RAI, ARD, BBC etc. colourimetry (in Europe at least) - as they all have different matrices - and some of them have different black stretch requirements as well.
Quote:
The source colorimetry info is embedded within the bitstream and the HD-DVD specs indicate precisely how to handle the colorimetry processing. Since what is going on doesn't seem to meet the HD-DVD specs, I would hope the firmware upgrade will fix it.
| Originally Posted by sneals2000 Yep - the 601 vs 709 colour space issues are just going to run and run. |
Interesting. I was wondering if these types of tests are typically performed by manufacturers prior to release? I am surprised that something as basic as this (or is it?) could be overlooked.
Quote:
Yep - this sounds to be the case with the player in question.
However with 601 and 709 both in use - and some legacy HD gear running 601 (especially in Europe) and SD derived from 709 without 601 conversion - there are going to be errors made in the colour space flagging as well I would expect.
| Originally Posted by kjack The source colorimetry info is embedded within the bitstream and the HD-DVD specs indicate precisely how to handle the colorimetry processing. Since what is going on doesn't seem to meet the HD-DVD specs, I would hope the firmware upgrade will fix it. |
However with 601 and 709 both in use - and some legacy HD gear running 601 (especially in Europe) and SD derived from 709 without 601 conversion - there are going to be errors made in the colour space flagging as well I would expect.
Quote:
Given all the new things these players do, I can easily see something like that being missed. At least the player is software upgradable.
| Originally Posted by Ilka Interesting. I was wondering if these types of tests are typically performed by manufacturers prior to release? I am surprised that something as basic as this (or is it?) could be overlooked. |
Quote:
I couldn't have put it better.
| Originally Posted by umr That is tragic. |
Quote:
This is the same crap that happened with the JVC 5U D-Theater/D-VHS VCR.
I can't believe this is happening again.
| Originally Posted by sspears This weekend Darin and I took a look at the color conversion of the A1. We found a few issues that may interest you. HD: HDMI to DVI: RGB This is the scenario where you see plugging the HDMI output of the A1 into the DVI input of something downstream, like a display. In this case the A1 must convert to RGB. The conversion used by the A1 for HD is 601, which is incorrect. |
I can't believe this is happening again.
would like to know how bad to the PQ?
Quote:
That's a shame. Is there any future planning on looking into the Component outputs ?
Bruce
| 709 is the correct matrix to use for HD. The player is using 601 instead of 709 when connected to a DVI display. We did not look at the component outputs, sorry. |
Bruce
Quote:
http://www.avsforum.com/avs-vb/attac...chmentid=32137
| Originally Posted by imwho88 would like to know how bad to the PQ? |
Quote:
The color primaires in the elemenatry stream is listed as 709 on all titles, which is incorrect as all films currently released are color corrected with SMPTE C primaries. Unless the metadata is correct, it is useless.
I will create some 709 bars with 601 and 709 matrix info and see if this data is actually passed on to the device doing the color conversion.
| Since what is going on doesn't seem to meet the HD-DVD specs |
I will create some 709 bars with 601 and 709 matrix info and see if this data is actually passed on to the device doing the color conversion.
Quote:
Can you please check this with component, as well? Thanks!
| Originally Posted by sspears The color primaires in the elemenatry stream is listed as 709 on all titles, which is incorrect as all films currently released are color corrected with SMPTE C primaries. Unless the metadata is correct, it is useless. I will create some 709 bars with 601 and 709 matrix info and see if this data is actually passed on to the device doing the color conversion. |
Quote:
Ah, the joys of bringing a new format to market, eh? 
Why do they use SMPTE C for HD content?
| Originally Posted by sspears The color primaires in the elemenatry stream is listed as 709 on all titles, which is incorrect as all films currently released are color corrected with SMPTE C primaries. Unless the metadata is correct, it is useless. |
Why do they use SMPTE C for HD content?
Quote:
Society of Motion Picture and Television Engineers. These are, after all, motion pictures, correct? Seems reasonable that they would use this standard.
Vern
| SMPTE |
Vern
Quote:
Because home video is color corrected on Sony BVM monitors. The HD Sony BVM monitors are using SMPTE C phospors. One day we will move beyond, one day.
| Why do they use SMPTE C for HD content? |
1 - 20 of 87 Posts
- Status
- Not open for further replies.
AVS Forum
A forum community dedicated to home theater owners and enthusiasts. Come join the discussion about home audio/video, TVs, projectors, screens, receivers, speakers, projects, DIY’s, product reviews, accessories, classifieds, and more!
Full Forum Listing
Top Contributors this Month
View All
WilliamR
909 Replies
BASSAHOLIC
512 Replies
bass addict
399 Replies
Recommended Communities
