HELP! Video Play Back Issue *NOW WITH VIDEO CAPS*

[JMWarren]

When I play back FMV on my new 2007FP bright area's seem to have distorted edges. It appears to be an interfearence patter than affects the whole image. Like high speed vertical tearing. It's very tough to explain. Only occurs on FMV and doesn't occur if I drag the video onto my old CRT.

It's almost like it's shaking.

It occurs even when the video is paused. It doesn't seem to happen on the analogue.

I doesn't always seem to happen on DVI either. I can't seem to find the cause, but it happens every now and then.

Looks like its a DVI issues. If I close the file, switch through the inputs and reopen the file it's ok for a min or so then the interfearence returns.

Whats stranger still is if I leave the file open (on pause) with the distortion visible and scroll through the inputs the distortion remains. If i hit stop and play it disapears for a min then is back!

I've added Video. Shot with my webcam, compressed with Dvix

DVI
VGA

ANY IDEAS???

 

[JMWarren]

BUMP!

 

[JMWarren]

Video's Added

 

[rbV5]

I'd say its a bug in the player, codec issue or driver bug. I can't see how the display has anything to do with it.

It sure looks to me like it affects the entire image. It "looks" like the macroblocks are drawing to the wrong part of the overlay. I don't recall seeing anything like that before.

 

[xtknight]

Originally posted by: JMWarren
It occurs even when the video is paused. It doesn't seem to happen on the analogue.

When it's paused, do the artifacts still move or are they just stagnant?

If they move while paused I'm guessing it's a monitor problem then. Or, it could be your video card's DVI transmitter (I doubt that though). You could try changing to 640x480/60Hz and see if it still happens.

 

[JMWarren]

Artifacts still move when paused. It doesn't seem to happen at and below 1280x1024 and not at all on VGA.

 

[xtknight]

Originally posted by: JMWarren
Artifacts still move when paused. It doesn't seem to happen at and below 1280x1024 and not at all on VGA.

Do you notice any other weird things at 1600x1200/DVI? My guess is you just got a faulty batch (the 2007FP is brand new).

 

[JMWarren]

Everything else looks perfect. There are some issues with window's placed over top of video, like tearing. Everything other than video is fine thogh. Dell is going to send out another panel but I'm not so sure it will cure the problem.

I've tried the latest NVidia drives (used Driver Cleaner)

I'm leaning towards a problem with my video card since VGA seems fine. Too bad I don't have another card here (at school) or another DVI flat panel, only my dual input DVI-A/VGA CRT.

 

[rbV5]

How can it be the display when only FMV exhibits the issue? How can the display discern the difference between video pixels and desktop pixels?

Dragging the software accross to the CRT and not exhibiting the issue indicates an issue perhaps with the cards overlay (primary vs secondary display), or maybe with VMR depending on the type of rendering your player is using.

Is it the primary display? What codec is used and what format is the video causing the issue? Different driver make any difference? How does it look in other players? The display itself would be my last suspect to be honest.

 

[JMWarren]

Looks the same in Windows Media Player as it does in Nero Show Time.
If I hook up DVI and VGA to the 2007FP it only shows on the DVI input. Even then it won't always show on the DVI input. If you scroll through the inputs, returning to DVI, it will not show the interference for a min or so.
It's displaying on the primary display. (I shut of the "other monitor/input in the display properties" when testing). It's playing back through DVix and pretty much anything I play back will cause it.

 

[xtknight]

Does any amount of phase/clock adjustment (in OSD) make it visible on VGA?

 

[rbV5]

Looks the same in Windows Media Player as it does in Nero Show Time.

Meaning, it looks OK with "those" players, just not FMV?

When you drug the player across to the other display and it worked...which display was primary and which was secondary?

The DVI on your CRT is still using the analog pins for the signal, correct?

 

[JMWarren]

Yeah, DVI on the CRT is DVI-A

It looks the same in both players as in if it's on DVI its corrupted slightly, if its on VGA (LCD OR CRT) or DVI-A it's fine.

I've tested with every combination of primary and secondary, single.
The results are always the same.

Results are the same regardless of player.

 

[JMWarren]

I think I've hit a wall here. I can't think of anything else to try short of swapping the panel and video card to determine if one is causing the problem. Unfortunatly this isn't an option as I don't have access to replacements.

Any other Ideas?

 

[JMWarren]

I've noticed somthing that may lead me to belive it's the display. The interference pattern is exactly the same if i rotate the panel vertically (physically, then using sofware to match)

It's a horozontal pattern when the monitor is landscape and a vertical patern when the monitor is portrait. The inteference remains the same regardless of input.

 

[JMWarren]

I can eliminate the problem if I reduce the hardware acceleration to level 3 (3 from the left, 4 from the right)!

This also disable DirectDraw and Direct3D accelerations as well as cursor and advance drawing accelerations.

Does this give any more clues?

 

[xtknight]

What led me to believe it was the monitor was because DVI showed it and VGA didn't show it. But if the hardware acceleration reduction fixes it then it could be a video card problem. I don't think it would be software, because as you said, it occured on DVI and not VGA. Unless you just can't see it in VGA mode? It's possible VGA is obscuring that effect.

What if you put the hardware acceleration back to maximum and go in WMP and disable High Quality mode? That's under Tools->Options->Performance->Advanced->Use video mixing renderer. Just leave Use VMR and Use Overlays checked but not Use High Quality mode. That's under the Video Acceleration group. Also just leave YUV, RGB, primary surfaces checked. Then click OK, restart WMP, and open the file and see if it still has the problem. And use the 2007FP at native resolution, 60 Hz, and primary display (only display preferable). Check both DVI and VGA. You've tried your system running at stock non-overclocked, right?

 

[JMWarren]

High quality mode was disabled already...

I tried enabling it and it fixed the problem

When I enable High Quality it Disables Overlays...is that normal?

I think we're getting there!

 

[xtknight]

Yeah that's normal. I guess the monitor being a problem was a misdiagnosis and you just don't notice it in VGA analog mode. Unless there's some weird video card problem. So your settings now are High Quality on yet hardware acceleration in the middle (whatever you set it as before?) Or you're at full hardware acceleration now?

 

[JMWarren]

Setting are High Quality ON, Full Hardware acceleration. This eliminates the problem.
Settings of High Quality OFF, Level 3 Hardware Acceleration also eliminate the problem.

I'm 99.9% sure it isn't there on VGA. Its not there on DVI for the first minute or so after an input switch, then shows up. I'm convinced its a video card issues...oh well this 6600GT won't be with me too much longer.

Couple of other questions. Does anyone know what Burnin Mode would do in the Factory Service menu? How about DDC-CI (ISP)? Both have only on or off settings.

Thanks!

 

[xtknight]

Burn-in mode likely sets all crystals to active (full white in the case of your S-IPS screen). Different than image persistence as in pixel burn-in though. I believe crystals need to be set to inactive (full black in your case) to solve image persistence or stuck pixel problems. Ironically they both have the same name but are completely opposite. I wouldn't change the settings, but they probably wouldn't damage anything. OK, I probably would change them because I'm the curious type of person.

DDC-CI? I think that will enable you to control your monitor from software (like NVIDIA Direct Adjustment under Coolbits) or with softMCCS. But of course I'm not sure. It could also be for updating firmware, the ISP (In System Programming).

 

[JMWarren]

Those were my guess as well. Setting either of them to ON seemed to have no effect.

There's also a debug setting (on/off) but neither setting seems to change anything.

Thanks for all your help, I guess I'll just leave the HQ box ticked and Full acceleration on.

Now to see if the replacment panel from Dell fixes my Gradient and Dead Pixel problem. I've got the feeling this one is a dud.

 

[rbV5]

Is it with all video files, or just a specific format (divx, mpeg....) or decoder (cyberlink, nero, ffdshow....)

Does full screen or windowed make a difference? How about stock clocks on the vcard or cpu?

Does it pass the dxdiag tests?

 

[JMWarren]

It's with all videos, even some game intro videos.
Full Screen and Windowed are the same. Overclocking makes no difference.
Tests pass. Enabling High Quality in WMP seems to have done the trick though.

Thanks Everyone. It's not too often that these kinds of things stump me, but when they do I know there's always someone on here who's got more experience.

 

[rbV5]

NEW INFO!!! SOLUTION BUT WHY?

HQ mode in WMP uses D3D and VMR9 to render the video rather than the Overlay. This indicates an issue with the Overlay either through a driver issue, issue with DirectX, or even how WMP is handling the overlay (software issue), or even a hardware issue with the overlay(vcard)

The fact that simply enabling VMR9 rendering in WMP fixes the issue means that WMP is responsible for the video rendering in every case (including game intro videos that exhibit the issue).

Nero Showtime should still exhibit the issue unless its now using VMR9, if another application uses overlay, it likely will exhibit the issue, so your workaround may not work for every case. More than likely it is software, probably graphic card drivers, less likely DX or WMP, and even less likely graphics hardware.