Log in Register FAQ Memberlist Search pcHDTV Forum Index
pcHDTV Forum

pcHDTV Forum Index -> General pcHDTV topics -> Real 59.94 fields per second 1080i Goto page Previous  1, 2, 3, 4, 5  Next
Post new topic  This topic is locked: you cannot edit posts or make replies. View previous topic :: View next topic 
PostPosted: Mon Aug 21, 2006 10:17 am Reply with quote
ubikdood
 
Joined: 24 Jan 2006
Posts: 18




Scott Larson wrote:
I'm talking about the 4496 driver I was using before. These problems sounded like hardware to me too so I bought an FX5500. I had the exact same problems with it too so I put the quieter FX5200 back in. It still could be a motherboard problem or who knows what else.


So you get geen and pink squares with driver 4496 (right ?). What other driver version(s) did you test and also got green and pink squares + infinite loop ?
View user's profile Send private message
PostPosted: Mon Aug 21, 2006 10:56 am Reply with quote
Scott Larson
 
Joined: 15 Oct 2003
Posts: 713
Location: Portland, OR




I haven't tried any others since I couldn't get anything but 4496 to work with my display until this week. If the latest driver gives me the same problem as one from two years ago, I tend to believe they never fixed it. Does Nvidia tend to fix things and break them again?
View user's profile Send private message
PostPosted: Tue Aug 22, 2006 4:02 am Reply with quote
ubikdood
 
Joined: 24 Jan 2006
Posts: 18




Scott Larson wrote:
I haven't tried any others since I couldn't get anything but 4496 to work with my display until this week. If the latest driver gives me the same problem as one from two years ago, I tend to believe they never fixed it. Does Nvidia tend to fix things and break them again?


As you can see from my test results, yes. Smile
(i.e.: 8xxx has worse XvMC performance than previous versions)
In your case, I would at least test one driver from each "generation" : 6xxx, 7xxx, etc.

If your issues (XSyncSurface infinite loop + green/pink squares) are easy to reproduce but you can't make a certain driver to instantly work with your 1366x768 plasma/lcd, you could just grab a TFT or CRT and test the thing.
View user's profile Send private message
PostPosted: Tue Aug 22, 2006 9:02 am Reply with quote
Scott Larson
 
Joined: 15 Oct 2003
Posts: 713
Location: Portland, OR




Unfortunately they aren't easy to reproduce at all. It took me a few days of heavy system usage to make them happen. That's a lot of trouble to go through when it's unlikely that the problems were fixed in one version and then rebroken another, especially since I haven't heard of anyone else experiencing these problems. I'm assuming it's my motherboard or something unique to my system that's causing them.

Yes, the video board has its own interrupt. Smile
View user's profile Send private message
PostPosted: Tue Aug 22, 2006 11:01 am Reply with quote
ubikdood
 
Joined: 24 Jan 2006
Posts: 18




Scott Larson wrote:
That's a lot of trouble to go through when it's unlikely that the problems were fixed in one version and then rebroken another, especially since I haven't heard of anyone else experiencing these problems. I'm assuming it's my motherboard or something unique to my system that's causing them.


I vote for a motherboard exchange. Good luck.

Scott Larson wrote:

Yes, the video board has its own interrupt. Smile


By all means, Sir. I would not dare to think otherwise Smile


Last edited by ubikdood on Mon Sep 11, 2006 3:13 am; edited 1 time in total
View user's profile Send private message
PostPosted: Tue Aug 22, 2006 4:02 pm Reply with quote
Scott Larson
 
Joined: 15 Oct 2003
Posts: 713
Location: Portland, OR




Or an additional motherboard. Now that I have an Nvidia driver that works with my HDTV and a 2.6 kernel, it's probably time to get a 64 bit CPU and motherboard which I've been putting off ever since the new Nvidia drivers stopped working. I have plenty of extra memory and drives laying around.
View user's profile Send private message
PostPosted: Sun Aug 27, 2006 1:20 pm Reply with quote
ubikdood
 
Joined: 24 Jan 2006
Posts: 18




Scott Larson wrote:
Did you confirm that the X server CPU usage was due to the mplayer text output?


Just for the record, I re-tested the clips using driver 8762 and the new 8774 and had the same results (growing cpu, maxed at 100%). Redirecting mplayer's output to /dev/null didn't help.
View user's profile Send private message
PostPosted: Sun Aug 27, 2006 1:26 pm Reply with quote
ubikdood
 
Joined: 24 Jan 2006
Posts: 18




Added test results for nvidia driver versions 7184 and 8774 (both for kernel 2.4.33 and 2.6.17).
View user's profile Send private message
PostPosted: Sun Aug 27, 2006 1:45 pm Reply with quote
ubikdood
 
Joined: 24 Jan 2006
Posts: 18




Also, while playing 59.94 bob-deint streams, I was getting a slight bump every 3 seconds. Mplayer was not dropping any frame or field, so I found this very strange.

Decided to take a look around the forums and here's what happened:

- Found this very cool utility in avsforum named JudderTest (windows, requires PowerStrip running) ;

- Booted into windows and installed powerstrip (required by JudderTest) ;

- Ran judder test and... found some judder. In my case, the VGA's vertical frequency is one thing, the plasma's vertical frequency is another. Gaaah !

- Fiddled a bit with powerstrip advanced timing options, stretched the screen here and there until I found a rate at which the PDP stops producing judder : ~ 59.510 Hz (your mileage may vary)

- Extracted the XFree86 modeline from Powerstrip (very nice!), and went back to linux.

- In the same forum I found an even cooler utility named juddertest for Linux ! Tested it and no judder was observed. Good.

- Tried mplayer with a 59.94 bob-deint stream and got judder, but this time was different : mplayer was dropping fields ;

- Changed my mplayer xvmc scripts to include switch "-speed 59510/59940". No more judder ! Smooth as silk...
View user's profile Send private message
PostPosted: Sat Sep 09, 2006 9:16 pm Reply with quote
nr
 
Joined: 16 Aug 2005
Posts: 18
Location: Malden, MA




ubikdood wrote:
Just finished testing several nvidia drivers. Interesting results.

Code:
...
7676    | 2.6.17  | 35      | 0      | constant cpu usage *
...
8774    | 2.6.17  | 35      | 0      | growing cpu. maxed at 100% (due to X)


This experiment was an incredible service. I had upgraded my Nvidia driver when I installed a new kernel, and I was getting bad results with 720p and was completely hosed with 1080i. Reverting to 7676 fixed everything.

Quote:
Also, making driver 7676 compile with kernel 2.6.17 was not very hard...

Additional note: compiling against the stock Debian kernel, I had to get in and explicitly turn off APM support in the header file, otherwise my nvidia.ko would not insmod because it referred to a symbol in an apm module.

Does anybody know how to communicate what's going on to the people at nVidia? The latest driver understands my display much better, but the CPU problem is a deal-breaker. I would love to see a new nVidia driver that goes back to 35% CPU use.
View user's profile Send private message Visit poster's website
PostPosted: Mon Sep 11, 2006 3:18 am Reply with quote
ubikdood
 
Joined: 24 Jan 2006
Posts: 18




Quote:
Does anybody know how to communicate what's going on to the people at nVidia? The latest driver understands my display much better, but the CPU problem is a deal-breaker. I would love to see a new nVidia driver that goes back to 35% CPU use.


The only forum I know where some communication takes place between nvidia card owners and nvidia's support is "NVIDIA Linux Forum" at http://www.nvnews.net/vbulletin/ . Perhaps one could report this issue there, although 99% of the forum's threads relate to OpenGL, 3D performance, stability and driver installation problems...
View user's profile Send private message
PostPosted: Sat Oct 21, 2006 7:40 pm Reply with quote
aspchdtv
 
Joined: 21 Oct 2006
Posts: 3




I just tried this patch.

It worked very nicely - except when there was something displayed from the OSD.

Then the framerate seemed to slow down and motion was very jerkey, and the sound got out of sync.

As soon as the OSD was turned off the video played very fast till it was all caught up and the audio synced up as well.
View user's profile Send private message
PostPosted: Sun Oct 22, 2006 8:17 pm Reply with quote
nr
 
Joined: 16 Aug 2005
Posts: 18
Location: Malden, MA




aspchdtv wrote:
I just tried this patch.

It worked very nicely...


What version of the Nvidia driver are you using?
View user's profile Send private message Visit poster's website
PostPosted: Sun Oct 22, 2006 8:43 pm Reply with quote
aspchdtv
 
Joined: 21 Oct 2006
Posts: 3




nr wrote:
What version of the Nvidia driver are you using?


8774-3

From debian testing.

On kernel 2.6.15.
View user's profile Send private message
PostPosted: Mon Oct 23, 2006 9:11 am Reply with quote
Scott Larson
 
Joined: 15 Oct 2003
Posts: 713
Location: Portland, OR




Are you using a single display or a dual display? Do you see other X applications slowing down when they try to display something? With the 8xxx versions of the Nvidia driver, it can take up to five seconds for an xterm to display one new character when mplayer is running on my system. It's like using a 300 baud modem. Shocked
View user's profile Send private message
Real 59.94 fields per second 1080i
  pcHDTV Forum Index -> General pcHDTV topics
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
All times are GMT - 7 Hours  
Page 4 of 5  
Goto page Previous  1, 2, 3, 4, 5  Next
  
  
 Post new topic  This topic is locked: you cannot edit posts or make replies.  


Powered by phpBB © 2001-2003 phpBB Group
Theme created by Vjacheslav Trushkin