That sounds like combing and most likely occurs due to inaccurate scaling. I used to get that on my Pioneer plasma until I added the KD HDXplorer card to it.
Seeing Scan Line Artifacts On LCD Flat Panel???
Hey folks,
I figured this forum had the right kind of knowledge for this question.
While watching some displays I've sometimes noticed a certain type of artifact. I'm watching a DVD image and I see what look to be occasional breaking up of the image into interlaced scan lines. This happens even on purportedly progressive displays, such as the LCD I was watching today.
In fact, I was watching a concert DVD on the big Sharp 61" LCD flat panel. I began to notice that almost every time the image switched camera shots, for about the length of a frame or two, the image broke up into what looked like half the scan lines - like watching a CRT tube set in which only 240 lines of the image are being displayed, with the "odd" lines being blacked out.
Once I caught on to it, it was starting to drive me insane. It was like at every shot change from one performer to another, someone had inserted one or two frames of only a 240 un-interlaced frame of the image.
Can someone clear up exactly what I might have been seeing?
Thanks.
Rich
I figured this forum had the right kind of knowledge for this question.
While watching some displays I've sometimes noticed a certain type of artifact. I'm watching a DVD image and I see what look to be occasional breaking up of the image into interlaced scan lines. This happens even on purportedly progressive displays, such as the LCD I was watching today.
In fact, I was watching a concert DVD on the big Sharp 61" LCD flat panel. I began to notice that almost every time the image switched camera shots, for about the length of a frame or two, the image broke up into what looked like half the scan lines - like watching a CRT tube set in which only 240 lines of the image are being displayed, with the "odd" lines being blacked out.
Once I caught on to it, it was starting to drive me insane. It was like at every shot change from one performer to another, someone had inserted one or two frames of only a 240 un-interlaced frame of the image.
Can someone clear up exactly what I might have been seeing?
Thanks.
Rich
- Status
- Not open for further replies.
1 - 20 of 23 Posts
It's called combing, and is caused by the deinterlacer incorrectly assuming two fields are from the same point in time, when in fact they aren't.
The better quality the deinterlacer, the rarer combing is, but even the best deinterlacers can do it occationally with really poorly encoded material. Scene breaks and edits are the best time to see them, as that's where deinterlacers struggle most to work out if something is film-sourced or not.
The better quality the deinterlacer, the rarer combing is, but even the best deinterlacers can do it occationally with really poorly encoded material. Scene breaks and edits are the best time to see them, as that's where deinterlacers struggle most to work out if something is film-sourced or not.
I've seen something similar to what is described on a Sharp 45" 1080p display. It's visible when the image content changes rapidly, which can happen at a scene change but is not limited to that. In this case the artifacts are at the display's native resolution (but then so is the source material). It's subtle and very transient.
In my case, I know that the deinterlacer is not causing combing. The panel is being directly fed a 1080p signal from the deinterlacer which bypasses the outboard Sharp AVC unit.
Has anyone else seen something like this on LCD panels? Does someone here have enough knowledge of how LCD panel updates are handled internally to know if the display's electronics may be causing this?
- Dale Adams
In my case, I know that the deinterlacer is not causing combing. The panel is being directly fed a 1080p signal from the deinterlacer which bypasses the outboard Sharp AVC unit.
Has anyone else seen something like this on LCD panels? Does someone here have enough knowledge of how LCD panel updates are handled internally to know if the display's electronics may be causing this?
- Dale Adams
Quote:
Oh, so we're talking about a different artifact...
Whatever is causing it must be in the processing stages, as LCDs are driven in analogue which should preclude if from doing anything like that. Prehaps Sharp's post-processing treats odd and even scanlines differently with the assuption that most material it's receiving will be 1080i? Just a wild stab in the dark.
| Originally Posted by Dale Adams I've seen something similar to what is described on a Sharp 45" 1080p display. It's visible when the image content changes rapidly, which can happen at a scene change but is not limited to that. In this case the artifacts are at the display's native resolution (but then so is the source material). It's subtle and very transient. In my case, I know that the deinterlacer is not causing combing. The panel is being directly fed a 1080p signal from the deinterlacer which bypasses the outboard Sharp AVC unit. Has anyone else seen something like this on LCD panels? Does someone here have enough knowledge of how LCD panel updates are handled internally to know if the display's electronics may be causing this? - Dale Adams |
Whatever is causing it must be in the processing stages, as LCDs are driven in analogue which should preclude if from doing anything like that. Prehaps Sharp's post-processing treats odd and even scanlines differently with the assuption that most material it's receiving will be 1080i? Just a wild stab in the dark.
I'm sure you are more knowledgable than me but I thought it was caused by bad edits which make the deinterlacer lose the cadence. I notice this artifact often, but only with video or special features.
Stace
Stace
I have a Sony LCD RPTV (1999 vintage) that seems to show this also. It looks like an interlaced picture. The "scan lines" were more noticeable in the dimmer parts of the picture. I constructed a strobe wheel using an old Erector set and that together with a C++ program that put up material timed in millisconds (the set accepts VGA too) seemed to suggest that that LCD panel set has some kind of alternate row refresh, or perhaps to reduce comet trails (slow decay) alternate rows were intentionally refreshed with black on alternate frames. I don't know the exact reason.
The gaps were always dark, not suggestive of combing from a poor de-interlacer.
I read in one of the projector forums here about "peekaboo scan lines" on an LCD projector which could be the same thing.
Video hints:
http://members.aol.com/ajaynejr/video.htm
The gaps were always dark, not suggestive of combing from a poor de-interlacer.
I read in one of the projector forums here about "peekaboo scan lines" on an LCD projector which could be the same thing.
Video hints:
http://members.aol.com/ajaynejr/video.htm
Quote:
Probably. That's not entirely clear from the original problem description, but it most likely is something different. The problem originally reported is most likely a deinterlacing problem, as has already been pointed out. This thread just seemed like a good opportunity to see if anyone has noticed anything similar. 
Quote:
The Sharp panel wants to be fed a 1080p signal, not a 1080i signal. I don't believe it will even display 1080i directly. (Remember, I'm bypassing their outboard video processing box.)
- Dale Adams
| Originally Posted by Carled Oh, so we're talking about a different artifact... |
Quote:
| Whatever is causing it must be in the processing stages, as LCDs are driven in analogue which should preclude if from doing anything like that. Prehaps Sharp's post-processing treats odd and even scanlines differently with the assuption that most material it's receiving will be 1080i? Just a wild stab in the dark. |
- Dale Adams
Quote:
That sounds at least somewhat similar to what I'm seeing. In the case of the 45" Sharp panel, the 'scan line' artifacts are at 1080p resolution, which means you have to be very close to the display to see them at all. I was specifically looking for deinterlacing artifacts so these caught my eye. They definitely appear to be something introduced by the panel's electronics as they're not present in the 1080p image the panel is receiving.
I'll check the thread you mentioned to see if there might be more information there. Thanks.
- Dale Adams
| Originally Posted by Allan Jayne I have a Sony LCD RPTV (1999 vintage) that seems to show this also. It looks like an interlaced picture. The "scan lines" were more noticeable in the dimmer parts of the picture. I constructed a strobe wheel using an old Erector set and that together with a C++ program that put up material timed in millisconds (the set accepts VGA too) seemed to suggest that that LCD panel set has some kind of alternate row refresh, or perhaps to reduce comet trails (slow decay) alternate rows were intentionally refreshed with black on alternate frames. I don't know the exact reason. |
I'll check the thread you mentioned to see if there might be more information there. Thanks.
- Dale Adams
Quote:
Yeah, I know. I was just thinking that Sharp migh have just been expecting most material to be from 1080i sources, and so optimised the internal electronics for deinterlaced 1080i. Or somthing.
| Originally Posted by Dale Adams The Sharp panel wants to be fed a 1080p signal, not a 1080i signal. I don't believe it will even display 1080i directly. (Remember, I'm bypassing their outboard video processing box.) |
There are actually two reasons why one would see combing.
1. Deinterlacer problems (cadence loss, bad edits, etc.).
2. The lesser known problem of the actual LCD being connected through an interlaced interface! My laptop, for example, cannot be blamed for having a bad deinterlacer when running windows... If I move windows around quickly enough, it becomes clear that the LCD in the laptop is actually interlaced (albeit a very fast interlaced connection) and I see very brief combing.
Not sure if it's the first or the second issue, but it's more than likely the deinterlacer at fault here (the big clue is that this happens during scene transitions).
1. Deinterlacer problems (cadence loss, bad edits, etc.).
2. The lesser known problem of the actual LCD being connected through an interlaced interface! My laptop, for example, cannot be blamed for having a bad deinterlacer when running windows... If I move windows around quickly enough, it becomes clear that the LCD in the laptop is actually interlaced (albeit a very fast interlaced connection) and I see very brief combing.
Not sure if it's the first or the second issue, but it's more than likely the deinterlacer at fault here (the big clue is that this happens during scene transitions).
Thanks for the interesting hunches on what was happening on the Sharp. It really was bad enough to drive me nuts once I saw it.
BTW, I notice a few people using the word "combing" to describe the artifact I'm talking about. I wouldn't have thought that was the right word. As I understand "combing," it's a temporal video artifact that happens when both interlaced fields are seen together, but the two fields have been off-set slightly, causing rippling edges etc.
Whereas what I'm describing is seeing one 240-line field at a time (i.e. it appears the second 240 field is being dropped, so I see the black scan lines inbetween the active 240 field lines for a split second).
Yes? No?
Thanks.
BTW, I notice a few people using the word "combing" to describe the artifact I'm talking about. I wouldn't have thought that was the right word. As I understand "combing," it's a temporal video artifact that happens when both interlaced fields are seen together, but the two fields have been off-set slightly, causing rippling edges etc.
Whereas what I'm describing is seeing one 240-line field at a time (i.e. it appears the second 240 field is being dropped, so I see the black scan lines inbetween the active 240 field lines for a split second).
Yes? No?
Thanks.
Quote:
Correct - that's the standard definition.
Quote:
That definitely sounds like something different. While I suppose it's possible for the deinterlacer to do something like that, it's not the typical type of deinterlacing artifact you see when the deinterlacer makes a wrong decision about which 2 fields should be weaved together, or whether any 2 fields should be weaved at all.
Now, if it's actually at the resolution of the 480i source rather than the display (which I assume is much greater than 480p), then it would have to be introduced at a processing stage which deals with the 480i signal. This is normally the deinterlacer. If the artifacts are at the display's resolution - i.e., more finely grained than 480 lines - then it would be introduced at a processing stage after the deinterlacer.
Any chance you could take a photo of the effect (which could be difficult, I know) and post it?
- Dale Adams
| Originally Posted by R Harkness BTW, I notice a few people using the word "combing" to describe the artifact I'm talking about. I wouldn't have thought that was the right word. As I understand "combing," it's a temporal video artifact that happens when both interlaced fields are seen together, but the two fields have been off-set slightly, causing rippling edges etc. |
Quote:
| Whereas what I'm describing is seeing one 240-line field at a time (i.e. it appears the second 240 field is being dropped, so I see the black scan lines inbetween the active 240 field lines for a split second). |
Now, if it's actually at the resolution of the 480i source rather than the display (which I assume is much greater than 480p), then it would have to be introduced at a processing stage which deals with the 480i signal. This is normally the deinterlacer. If the artifacts are at the display's resolution - i.e., more finely grained than 480 lines - then it would be introduced at a processing stage after the deinterlacer.
Any chance you could take a photo of the effect (which could be difficult, I know) and post it?
- Dale Adams
Actually taking a photo could give convincing proof and would be easy -- if the camera can be manually set to use a shutter speed less than 1/60'th second! It would take more than one shot since synchronizing with the LCD refresh is a matter of luck.
I called it combing, I could be wrong on this, but the effect affects more than the edges of an image but does not go all the way across the screen. Something like this:
http://neuron2.net/LVG/inthead.jpg
http://neuron2.net/LVG/inthead.jpg
So we can can cross combing off our list.
Your problem would almost certainly be somewhere in the processing stage of your screen.
If one supposes Sharp staggers the refresh of pixels to try and mask the poor response speed of LCD (or some other reason), then it could be conceivable for something like that to happen. Kinda like tearing except...not, I guess.
But yeah, if you could pull off a photo of it then we should be able to at least have a good idea what's happening.
Your problem would almost certainly be somewhere in the processing stage of your screen.
If one supposes Sharp staggers the refresh of pixels to try and mask the poor response speed of LCD (or some other reason), then it could be conceivable for something like that to happen. Kinda like tearing except...not, I guess.
But yeah, if you could pull off a photo of it then we should be able to at least have a good idea what's happening.
As I said, I get "combing" (or interlacing) on my LCD display on my laptop, although it has no deinterlacing stage at all. I believe it's part of the addressing issue of the display, put enough FPS through and you start to see it.
I've never seen anything like that on consumer LCD displays, though.
I've never seen anything like that on consumer LCD displays, though.
Kei,
Your picture is of combing from straight-weave de-interlacing.
Your picture is of combing from straight-weave de-interlacing.
Allan,
I know, but I guess I was trying to explain that it was not just ont he edges but covered a wider area. Oddly enough, I could catch it sometimes even in fully white background. This was on a plasma, and may not be the same issue as it went away after I changed scalers.
I know, but I guess I was trying to explain that it was not just ont he edges but covered a wider area. Oddly enough, I could catch it sometimes even in fully white background. This was on a plasma, and may not be the same issue as it went away after I changed scalers.
Quote:
Actually Kei that looks very much like what I was seeing. As I said, it was flashing so fast it was difficult to know exactly what it was. But as I said it was a concert DVD (I was watching this in a store) and I'd stare at a performer. When the shot changed then I'd see those scan lines. I think in fact that I likely only saw the scan lines over the image of the performer, vs the whole frame, but since the performers were typically isolated against a black background I assumed it was happening over the full frame (so I can't be sure).
But it looked very much like the picture you've posted - a flash of scan lines over a person as the shots switched. (And I think I've seen it in brief flashes with certain content on my older Panasonic plasma too).
| Originally Posted by Kei Clark I called it combing, I could be wrong on this, but the effect affects more than the edges of an image but does not go all the way across the screen. Something like this: http://neuron2.net/LVG/inthead.jpg |
But it looked very much like the picture you've posted - a flash of scan lines over a person as the shots switched. (And I think I've seen it in brief flashes with certain content on my older Panasonic plasma too).
1 - 20 of 23 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
961 Replies
BASSAHOLIC
529 Replies
bass addict
401 Replies
Recommended Communities
