Is the VBI_EN flag working on the ADV7180?
I'm trying to have the same filtering (both luma and chroma) on VBI lines as normal lines but this flag seem to not work, can someone help me?
Yep, there are no known issues with VBI_EN. Setting it to 0 should filter all video lines.
I'm not aware of any dependency there. What other registers are you writing?
I’m using the ADV7180 evaluation with CVBS autodetection mode.
The LUMA (Y) output of lines 17 to 20 has no changes with VBI_EN value.
Are you seeing changes on the output for lines 1 to 16?
What are you expecting to see when you toggle VBI_EN on the output bus?
unfortunately i am not able to test on lines from 1 to 16.
I have a PAL test signal inserter working only from line 17 to 21 and from
line 24 to 315.
What I need is the decoder working in the same mode both in lines 17 to 21
and the others lines and this is what I expect setting VBI_EN to 0.
Watching the Y output I see the ADV7180 try to pass all composite signal on
Y without using the COMB filter to separate chroma component as in lines
Thanks in advance.
Ahh.. so what you mean is that you are expecting to see Cb and Cr data during the VBI but you are only seeing Y? Have you tried the BL_C_VBI setting?
The output is long after the splitting of components... how would you see chroma on the Y in the digital domain? What values of Y are you seeing versus the value you expect is what I meant?
I think the "filter" that you are expecting may not be what that setting actually does.
with BL_C_VBI disabled i see Cb and Cr during VBI but the problem is that
chroma component is present also on the Y component, that’s all.
I need to have Y component filtered as on not VBI lines. It seems, but I can
be wrong, that the COMB filter used to separate chrominance from luminance
is not active during the VBI interval.
Do you agree?
No, the signal is separated into luma and chroma long before any filtering or processing is done. It has nothing to do with the VBI.
I'm trying to understand your assertion that there are chroma components present on the output Y. How are you determining the digital Y values output have chroma in them?
The filtering/scaling that is implied by the VBI has to do with scaling the range of digitized values to account for the range for EAV/SAV and I think the filtering has nothing to do with luma/chroma.
If you had a clean input signal with Y values that were not on the edge of the range, I don't think you would see any change with this setting.
i’m using the EVAL-ADV7180 so I see components not in digital domain but
after the re-encoding in analog domain by the ADV7393.
When I insert a PAL test line on lines 17 to 21 with color (4.43MHz)
component, the color component is present also on the re-coded Y output also
if I disable color decoding with BL_C_VBI.
So you don't know whether the issue is with the encoder or the decoder then. Are you able to put a logic analyzer on the output of the ADV7180 to look at the pixel bus?
That is really the only way to confirm where the problem is. I don't believe your issue has anything to do with the ADV7180.
sorry for my long sleep but i has been involved in other works.
I don't have a logic analyzer, but if i set to 1 the BL_C_VBI, how can the encoder add color components on the VBI lines to the Y output?
The only reason i can guess is the presence of the color components into the luminance decoded signal.
Easily-- noise. There really is no way to sort this out without looking at the pixel bus.
finally i have been able to decode digital lines and now i can say without dubt that on the VBI interval COMB filter is not active so color component is present on the LUMA signal.
In attachment you can see the LUMA signal of the LINE 18 (PAL test signal) with the 4.43MHz signal not filtered, while CHROMA signals are without data, colour not decoded.
At this point i suppose the ADV7180 works only in this mode and to solve the problem i must switch to another decoder chip.
What I recommended was that you look at the digital lines coming out of the ADV7180... what I see in your attachment looks nothing like digital data.
Looking at the pixel bus would be the only way to see what was actually going into the ADV7393 to see what was going on. I really don't see how it's possible for the ADV7180 to blend chroma data with luma data after it's in the baseband as you are asserting.
It's certainly possible you are trying to do something sufficiently out of scope of the applications for ADV7180 that it's not an appropriate part.
the screenshot is simply the analog display of the digital line data.
From your last sentence i understand you don't have a sufficient knowledge of the component.
As this is the only technical support yours italian field application engineer has suggested me, i think AD support is not useful if i can't talk with someone who can say if the component does or doesn't a specific task.
The question is very simple, what i want is absolutely normal, i want all lines of PAL/NTSC video processed in the same mode.
By digital data-- I mean 1's and 0's aligned with a VCLK. The ADV7180 has a pixel port output with bits for luma and chroma. What you have shown is not that and it is not useful to check your assertion that chroma has been magically blended on top of luma.
The ADV7180 simply cannot do what you claim it is doing. Chroma data is either killed-- in which case nothing comes out on the Cb/Cr pixel bus or it is decoded in which case it comes out on the Cb/Cr bus. In either case, it does not come out on the Y bus. It's really that simple, and those are the only options. This is why I keep asking for the pixel bus output. Without that, there really is nothing else we can do to help you.
The ADV7180 is designed to process normal video which video is in the active region and not in the blanking region. Blanking and Active region are different, and it is not normal to want to process them in the same way. Blanking typically contains nothing or digital data-- not video.
after reading carefully the ADV7180 datasheet i found my problem.
The register 0xEB say "Controls position of first active (comb filtered) line after VBI" and i can select lines 22 to 25, but i need to select line 7.
I think this isn't possible with this component.
Is there another decoder where i can enable comb filter from line 7 ?
Unfortunately there is no way to comb filter the blanking region. The control is there for differences between standards of where the active region starts, as the description states. Treating blanking the same as active is not a normal application for any of our video decoders.
I have the same problem with the ADV7188, the lines from 1 to 22 are not fully transparent. These lines are useful for PAL broadcasters to estimate the link quality without disturbing the video. I try to set VBI_EN = 0 and BL_C_VBI = 0 but the result is not good... Do you know if there is a mean to treat these test lines as the active lines (by using the hidden registers for example) ?
There is no way to treat the VBI region as active. It's either filtered as digital data or passed through as-is without the filter.
Retrieving data ...