1080i material not deinterlacing under 1.0.1.0 (1 Viewer)
- Thread starter Andrew H
- Start date
Add me to this list.
GPU: ATI 4550
GPU: ATI 4550
Despite of the errors, that .ts file plays smooth in MP1.0 Final but stuttery to hell in MP 1.0.2
NB: those using onboard HD3200 and not using a phenom can increase performance by increasing their HT in bios (overclocking it). The reason is that the "road" between the shared memory and the hd3200 is dependent on the HT speed and it's a lot lower if you don't have a phenom. If you overclock the FSB by 10-20% you can increase the HT from 1000hz (default) to 1100-1200Hz makes a huge difference accessing shared memory, as this chip can struggle a bit with 1080i and the little bit extra helps a lot.
I use the following settings for my X2 4200 with DDR800
VGA Core = 600 (default = 500)
FSB = 220 (default = 200) - on my Mobo the HT = FSBx5
CPU Multiplier = x9 (default = x11, FSB x multiplier = CPU Freq = 1980Hz)
Everything else is auto
The multiplier will be different depending on your cpu frequency, I decreased the multiplier as i dont want to overclock my cpu as its plenty fast enough already. use whole numbers only for the multiplier. i have a fan blowing on the gpu. you might want to just try a 10% increase in the FSB first. The GPU is pretty stable you can overclock it past 700 some people have had it at 900 or more, but i dont think overclocking the gpu core helps i think increasing the HT helps a lot more.
Different video codecs handle the corrupted stream differently. Especially when testing some bug in MP the stream shouldn't cotain any errors. So, please upload a new sample, althou I'm still pretty sure that this is only a performace issue with lower powered GPU's and 1080i.
- Thread starter
- #65
UPGRADING my Video card cleared this up!
I have been unsuccessful at getting any video to play from the .ts clip you posted earlier. I attempt to play it in WMP, MPC-HC, and drag it directly into GraphEdit and never get a video stream :-(
I think the tip you gave about overclocking may be a help for some. I can't say I'm disappointed in the performance of the N9500GT MD512 D3 aside from I may could have saved $15 by getting the DDR2 version but splurged for the DDR3 after I'd read "how much better the DDR3 version worked given its higher RAM speed." Assuming I actually receive the $30 MIR I can't say $40 is too much to pay... although $25 would have been a better value if the DDR2 version performs adequately. (I'd bet it would given I'm only interested in video and not gaming.) I did run a Video Card Stability Test which yeilded a 215 fps benchmark -- ten-times faster than the 20~25 they report for my NVidia 7050PV / nForce 630a on-board I was using! Oddly enough, the ATi AIW 9600 (xt) I (still) use on my OLDEST system (still at MP 0.2.3.0) they rate and I tested at 61 fps... who'd o' thunk that my OLDEST system would outperform my 7050PV. Incidentally, they report the ATi HD3200 Graphics at 32~36 fps.
I have been unsuccessful at getting any video to play from the .ts clip you posted earlier. I attempt to play it in WMP, MPC-HC, and drag it directly into GraphEdit and never get a video stream :-(
I think the tip you gave about overclocking may be a help for some. I can't say I'm disappointed in the performance of the N9500GT MD512 D3 aside from I may could have saved $15 by getting the DDR2 version but splurged for the DDR3 after I'd read "how much better the DDR3 version worked given its higher RAM speed." Assuming I actually receive the $30 MIR I can't say $40 is too much to pay... although $25 would have been a better value if the DDR2 version performs adequately. (I'd bet it would given I'm only interested in video and not gaming.) I did run a Video Card Stability Test which yeilded a 215 fps benchmark -- ten-times faster than the 20~25 they report for my NVidia 7050PV / nForce 630a on-board I was using! Oddly enough, the ATi AIW 9600 (xt) I (still) use on my OLDEST system (still at MP 0.2.3.0) they rate and I tested at 61 fps... who'd o' thunk that my OLDEST system would outperform my 7050PV. Incidentally, they report the ATi HD3200 Graphics at 32~36 fps.
1080i_H264_DVBT_MP1.0_Final_Recorded_using_tvserver_config_manual_control.ts (23mb)
can't post 1.0.2 recording of 1080i tv, as i've gone back to 1.0 final
i noticed 1.0.2 stuttered when playing recordings made by 1.0 final prior to upgrade
this recording was made directly using tvserver configuration utility
size = 23mb unfortunately not a lot of panning in this shot bc its a dum talk show.
can't post 1.0.2 recording of 1080i tv, as i've gone back to 1.0 final
i noticed 1.0.2 stuttered when playing recordings made by 1.0 final prior to upgrade
this recording was made directly using tvserver configuration utility
size = 23mb unfortunately not a lot of panning in this shot bc its a dum talk show.
If someone can show me how to properly edit one of these .ts files to get a sample, I'd be happy to suppy one. I recorded The Tonight Show last night which plays 100% perfectly smooth in MPC-HC but is jittery as all get out in MP.
Have that problem too (https://forum.team-mediaportal.com/...r-90/upgrade-1-0-0-1-0-2-stuttering-tv-60800/), but I'm only using SDTV.
I read through this thread but I did not see any solution. Did I skip something?
weissollo
I read through this thread but I did not see any solution. Did I skip something?
weissollo
- Thread starter
- #69
Scroll up to #65 where I said a new Video Card has fixed me... and #54 where a previously unannounced texture writing methodology has been disclosed as the reason. I looked at the test I'd mentioned and see your 6200 A-LE reports 23fps... my 7050PV rates in at 20~25fps. I'd suggest a new video card would do you right... save the $15 over the DDR3 and let us know how the N95GT-MD512-OC does.
Oh... in reading my log I see where my new N9500GT is using pixel adaptive deinterlace -- apparently a real power hungry GPU feature over BOB.
Bottom line: I spent $40 ($70 - $30 MIR) to restore the function I had before the developers changed their methodology. On another note, I can play the HD-DVD King Kong in PowerDVD with 70% CPU with solid video -- an improvement from 98% CPU and occasional dropped frames/jittery playback.
Oh... in reading my log I see where my new N9500GT is using pixel adaptive deinterlace -- apparently a real power hungry GPU feature over BOB.
Bottom line: I spent $40 ($70 - $30 MIR) to restore the function I had before the developers changed their methodology. On another note, I can play the HD-DVD King Kong in PowerDVD with 70% CPU with solid video -- an improvement from 98% CPU and occasional dropped frames/jittery playback.
- July 1, 2007
- 274
- 77
- Home Country
-
New Zealand
So that's the end of the issue, is it: expecting people to throw more money into their HTPCs? I'm not interested in having to fork out around $100, my currency, to fix something that only appears to be a problem in Mediaportal (maybe I'm wrong on this: haven't researched yet, but I doubt it as ALL my external players have NO issue with 1080i playback. It is around 20-28% max. CPU load with CUDA support inside MP RC4).
I'll stick with RC4 as it works perfectly stable. & when it appears to become too outdated for my needs (ie. the plugins aren't supported or I require more extras) I guess I'll have to revisit the front-end I used before I discovered MP, as I understand that has also progressed considerably since those days. shame, as I have come to love this front-end. Hell, I may even need a new card by then & given the devs progression, it may have to be as high as a 295GTX.
I'll stick with RC4 as it works perfectly stable. & when it appears to become too outdated for my needs (ie. the plugins aren't supported or I require more extras) I guess I'll have to revisit the front-end I used before I discovered MP, as I understand that has also progressed considerably since those days. shame, as I have come to love this front-end. Hell, I may even need a new card by then & given the devs progression, it may have to be as high as a 295GTX.