101 Topic by Fanty1972

(30 replies, posted in Using SVP)

Hi there! I need some help/advice from people who know something about these things.
Something happend (either to me or the computer). I seem unable to watch video without SVP because its totaly unwatchable and I dont think that its because I got used to SVP, it must be something else.

Its like this:
Tearing lines shutter and jump like craisy on normal 24FPS video (without SVP and I made sure that Avisynth is off). Specially camera movement causes image shuttering that is so extreme that I cannot imagine that this was before. I think its because all the imagery does exactly what the tearing like does: dancing like mad.

(its the question if my computer does this with SVP too or not. I think it does, just 5 times as fast (120FPS), so that it looks like 1 line not jumping. But its slightly wider than the paused lines. So it does!

I oberve this behavior in all videos, in VLC, MPC or Potplayer and even Flashplayer in a browser.

Unfortunantely I just sold my old computer and have no "neutral" mashine avaible to check if this is indeed "normal" behavior and I just got used to smoother things (I doubt it) Also, isnt the MPC Tearing test suposed to be smooth and unjumpy in 24 or 25 FPS?

Windows recovery says there are no recovery points.... uh... (strange, I am sure I did some. Last was before K-Lite Codec Pack install, totaly sure).... strange   sad

Well I uninstalled the last things I did to check if it helps: SVP (and everything that was installed with it), K-Lite Codec Pack and the GPU driver. No change so far.

Any ideas?  sad

Go to forum: Using SVP

102 Topic by Fanty1972

(4 replies, posted in Using SVP)

Hi there, its me again.  big_smile
I tried googling it, if it was already asked (Forum seems not to have a search on itself) but did not find it at once, so I open this topic here.
So many questions.... sorry. wink

OK, I finaly found the description of the shaders. And it says Shader 13 is like shader 11+"dynamc median".
And Shader 21 is shader 11 with a masking system to reduce "halo" effects, while shader 23 is shader 21 with a more complicated masking system that masks "halos" more succesfull, but loses smoothness. Right? Well whatever.

I checked a scene of a Music video of Lindsey Stirling where she attacks Peter Hollens with a fiddlestick that is used like a sword.

The fiddlestick has quiet anoying artefacts in that scene (it took a 32 grid to get rid of it completely) I checked the same scene with shader 11, 13, 21 and 23 and it looked absolutely identical in 11, 21 and 23 while it looked much better (not perfect, but .... almost like a 24 Grid with the other shaders even through a 12 was used.

So yeah ok I understand it looks identical because 21 and 23 are 11 with a halo masking. What interests me is, why isnt there a 13 with a halo masking? the "dynamic median" (whatever that is) seems to really improve the shader, so why arent 21 and 23 not based on 13 but on the older (lower number?) 11?

Go to forum: Using SVP

103 Reply by Fanty1972

(3 replies, posted in Using SVP)

Rimsky wrote:

Fanty1972
Can it be, that the second picture just happens to be a "original" frame instead of an interpolated one?

i think first frame is the origiinal

oh of course I also meant the first, not the second. smile

Go to forum: Using SVP

104 Topic by Fanty1972

(3 replies, posted in Using SVP)

Until now I thought, how the frames look like is only dependant on the settings and not on the hardware. (at how much FPS one can be with these settings, beeing dependant on hardware but not the look of one frame)

Minutes ago I stumbled over this:
http://www.avsforum.com/t/1326189/smoot … t_22945621

Is it really like this or are the thoughts of that poster missleaded and the difference in the screenshot is just random? (it does not seem to be the same frame, the position of the car is different. Can it be, that the second picture just happens to be a "original" frame instead of an interpolated one?  hmm

Go to forum: Using SVP

105 Reply by Fanty1972

(3 replies, posted in Using SVP)

Thanks. That worked.
I had a error message that SVP wasnt allowed to replace the file but after starting it with administrator privilegs it worked.

But I wonder how it can become broke when it already worked?

Go to forum: Using SVP

106 Topic by Fanty1972

(3 replies, posted in Using SVP)

I mentioned this in another topic allready. Now I get it the third time (in like 3 weeks).  hmm

I dont know what was up the first time. But the second time it was after a crash of MPC.
Now in the third time it suddenly happend after I used FRAPS and then SUPER. No crash, everything fine... but SVP wont work anymore and tell me "there is no function named set mt mode".

The only working method to fix this, that I know of, is re-installing SVP... but why is this happening again and again to me?  neutral

Go to forum: Using SVP

107 Reply by Fanty1972

(10 replies, posted in Using SVP)

MAG79 wrote:

Fanty1972
Did you test your monitor with dynamic resolution tests? Did you see many contours at your display without SVP?

I dont even know what dymanic resolution is.  wink

The display has something "strange" called "Lightboost technology", wich is meant to lose less brishtness with 3D. Dont know if this has any backside.

Tests in magazines claimed the display aswell as the GPU the best avaible for mobile mashines (middle of 2012). They complained about the sound (really crappy compared to my former laptops, well with external speakers its kind of ok), the weight (8kg) and the low battery capacity (60 minutes idle, 30 Minutes if doing demanding things)

But I anyways have this problem with laptopdisplay aswell as beamer and it does not show up with madVR (wich I wanted to avoid because even in lower than default settings I get framedrops at 120Hz with it, what I dont get with EVR)
Kind of funny. One has framedrops but a perfect testline and the other had 0 Framedrops but this doubling testline... sigh. And and lets not forget that Shader 13 doesnt have the problem aswell. (11,21 and 23 show it)

Whats your opinion about shader 23 compared to shader 13 anyways?
I like SVP shader 13 for its artifacts free picture.

Hmm ok. At the moment I am unsure about the two. I was satisfied with 13 until I tried to rewatch Battlestar Galactica with it. BSG is full of anoying artefacts (space scenes are ok, but anything else... haha), really so far I dont know anything else that causes SPV to freak out so much as BSG. I blame the concept of "let us make the show look like a random guy filmed all this with his smartphone camera" (maybe plus Parkinson illness). The forrest scenes on Kobol apeared almost unwatchable (forrest background, scene chance every 1 or 2 seconds and shaking the camera like mad  mad  ). Shader 23 eased the pain with those. It cleary did better.

But I frequently watched other things (not BSG so far) with 13 again and sometimes had the feeling that it looks better than what I am used to (23).  hmm

Go to forum: Using SVP

108 Reply by Fanty1972

(10 replies, posted in Using SVP)

@MAG
Hmm. Its the same for me with the excepton Haali for splitter (its the MPC setup from the SVP installation)

I think one cant see anything special in the video anyways. Its 24FPS (filmed with a photocamera actually) and had to long shuttertime (one can see 6 lines in each frame)

Strange thing is anyways, I fail to see this effect in the movie itself somehow. Well ok, most things move not as fast as the MPC tearing testline.
The SVP Test line moves slower, wich results in the second (showing and vannishing) overlapping with the main one about 50%

Whats your opinion about shader 23 compared to shader 13 anyways?

Go to forum: Using SVP

109 Reply by Fanty1972

(10 replies, posted in Using SVP)

Rimsky wrote:

Fanty1972
1. can you make a short video for a 5-10 seconds, please?

I made one (only 4 sec) but then I realised that I cant show you. I deleted my Youtube account after Google made that Google+ crap and all. wink

2. is this "bug" appears only on the 23shaider?

Simple Lite, Simple and Complicated. Not in the others.

3. did you try to change videorender: madvr, evrcp?

ok, the bug is visible with EVR-CP. It DID NOT seem to to show up in madVR, though I dont know if the FPS was right (MPC was at 116 and stil raising it slowly towards 120, something I only have with madVR, not with EVR)
The line is also not as calm as with EVR-CP and "standart" shader.

4. what is your hardware: cpu/gpu/mb/ram/ssd/hdd/etc

A Asus G75 "Republic of Gamers" laptop in the 3D edition, buyed about 1 year ago.
i7@3,2Ghz
NVIdia GTX 670M (3GB)
8GB RAM
System (Win 7 64Bit) on SSD HD
120Hz 3D Display (1080p) ----alternative: Acer HD5360 Beamer (120Hz, 3D capable, 720p), NVidia 3D Glasses


I just realised, its NOT happening at 60Hz, only above (3x, 4x or to 120Hz)

EDIT:
Oh and.... something else seems wrong:
While testing around, MPC crashed and after it, SVP wont work anymore (There is no function called ....) had to reinstall SVP to make it work again. Thats the second time that this happend to me, since I use it (several weeks). Hmm.

Go to forum: Using SVP

110 Reply by Fanty1972

(10 replies, posted in Using SVP)

Ok, here are 2 shoots (on one image) from a beamer picture on the wall, which show the effect:

http://img15.imageshack.us/img15/7879/y96v.jpg

One pair of lines is stable, the other pair of lines comes and goes (left picture shows it half disapeared)

Go to forum: Using SVP

111 Reply by Fanty1972

(10 replies, posted in Using SVP)

MAG79 wrote:

Can you give a screenshot?

Hmm... I cant capture the effect on a screenshot.  hmm

Maybe I try to take a picture with a digicam later, but not right now, I need to drive somewhere now.

Go to forum: Using SVP

112 Topic by Fanty1972

(10 replies, posted in Using SVP)

I just realised that with Shader 23, the lines of a "tearing test" are somewhat doubled for me.

With SVP build in "tearing test" there is one line, moving softly and another one right beside the other one wich frequently shows and disapears.

when I use the MPC-HC "tearing test" both lines have blinking copies in more distance to the main ones than in SVP (maybe because the speed is higher).

Is thi because of the masking that this shader uses? Could it be that it also affects the images of the video? I ask this because I still wonder if shader 23 has anyhting "bad" (besides needing more render power) compared to the standart shader.

Hmm

Go to forum: Using SVP

113 Reply by Fanty1972

(109 replies, posted in Using SVP)

Rimsky wrote:

The best settings for curren system SVP apllyed during install

I have a question about "Complicated". Is it more bad than "standard"? And what exactly is the advantage of "Standard" over "complicated", except for consuming less hardware power?

Atm I happen to prefer "G4" Settings + Complicated over pure G4 Settings, because it masks the bow wash effect of objects plowing though the sea of complicated backgrounds. Also, it does sound interesting that it uses the motion data from 4 frames. Wouldnt something like that not improve the reproduction of a courve like motion?

Hmm. Maybe that masking that is included into that shader has similiar problems like the "mask artefacts" function? Like replacing large parts of the image with the non-smoothed original frame and by this killing smoothness?

Hm. Also, if its a movie that includes people swinging aorund long thin objects like... uh... swords, (light)sabers, or muskets it seems that a motion grid of 24 is superior to 12 (small grids seem to destroy those long and thin objects).

Go to forum: Using SVP

114 Reply by Fanty1972

(39 replies, posted in Using SVP)

Uh... it took me a while to realise what was wrong.
It all started with me recognising a tearing line though the picture in a single video file. It also was running with lower frame rate and I thought something else was strange behaving.

The file had the name "Happy Hour" and because of that, was recognized as "Half over under 3D"  lol

Go to forum: Using SVP

115 Reply by Fanty1972

(10 replies, posted in Using SVP)

Thanks for your help Chainik!
I think the thread could be deleted.

Go to forum: Using SVP

116 Reply by Fanty1972

(10 replies, posted in Using SVP)

Huh.... one second.
Now after many times starting this and that (and with different players)... both are recognized as 1280x720 now.  hmm
wtf.....

need to check after a restart  neutral

Go to forum: Using SVP

117 Reply by Fanty1972

(10 replies, posted in Using SVP)

Chainik wrote:

ffdshow properties -> info
it's not "mediainfo" but what ffdshow thinks about the video


I find something in "Info and CPU" ("Info und Prozessor" for me, I have German version). It says for both files the same.
Hm.

Go to forum: Using SVP

118 Reply by Fanty1972

(10 replies, posted in Using SVP)

Chainik wrote:

well... may be the text from "ffdshow info" page would be more helpful
since SVP gets video info from ffdshow

Uh... cant find where media info could be found in it?

Go to forum: Using SVP

119 Reply by Fanty1972

(10 replies, posted in Using SVP)

hm

Go to forum: Using SVP

120 Reply by Fanty1972

(10 replies, posted in Using SVP)

hmmm

Go to forum: Using SVP

121 Topic by Fanty1972

(10 replies, posted in Using SVP)

well... bla

Go to forum: Using SVP

122 Reply by Fanty1972

(109 replies, posted in Using SVP)

Hm. I just realised how a driver does change the number of dropped frames.

I still used 306.97 because it was the driver that was optimized for World of Warcraft Pandaria and went back to it after trying 4-5 newer ones after I realized it came with more input lag or less FPS in WOW.
Now, a year later I gave the 331.65 a chance, because I hoped to get rid of dropped frames in video playback.

Well I tried it one day.... it was worse. It increased the number of dropped frames to 27, where it was 0 before. I am back to my beloved 306.97, all fine again.  big_smile

Go to forum: Using SVP

123 Reply by Fanty1972

(109 replies, posted in Using SVP)

Hm.

I think I have problems to keep 0 dropped frames at all.
I looked at the results of complete show episodes I watched.

1. TV Show 22 Minutes, 720p source, 720p target, EVR-CP@120Hz: 1 dropped frame
2. another episode of the same show, same settings: 0 dropped frames
3. one more episde but with madVR at default settings and 60Hz: 1 dropped frame
4. one more episode, same settings as in 3: 0 dropped frames

I am kind of becoming paranoid with this now  tongue

Does 1 dropped frame in a 22 minute show mean, the complete show is at a continious slight stutter or something or is it only a single jerk in the whole show?

Go to forum: Using SVP

124 Reply by Fanty1972

(109 replies, posted in Using SVP)

Chainik wrote:

so you were using too heavy settings in madVR for 120 Hz...

I already used lower settings than default.

All scaling at "bicubic" and sharpness 50 (there was no scaling anyways with the values I postet, it was 720p material, projected with a 720p beamer)
Overlay windowed mode (much less dropped frames for me)

I also use Reclock.

Hm I just tried several settings with a (saved to mpg4) Youtube Video:
Its 720p and I try it at my laptop screen in 1080p, so its scaled.

http://youtu.be/nwkYaxkDO3c

All with G4 Settings (recommended)

madVR with default settings @120Hz = 909 Frames dropped
madVR with default settings @ source x 4 = 82 Frames dropped
madVR with default settings @60Hz = 0 Frames dropped

madVR with "Overlay Windowed mode" @120Hz = 22 Frames dropped
madVR with "Overlay Windowed mode" @ source x 4 = 0 Frames dropped
madVR with "Overlay Windowed mode" @60Hz = 0 Frames dropped

madVR with all my settings @120Hz: 10 Frames dropped
madVR with all my settings @ source x4 : 0 Frames dropped
madVR with all my settings @60Hz: 0 Frames dropped

EVR-cp @120Hz = 0 Frame drops

Go to forum: Using SVP

125 Reply by Fanty1972

(109 replies, posted in Using SVP)

Chainik wrote:

Fanty1972
Ah ok, the default render (EVR-Custom present) is at CPU 16% and GPU 23% when doing 120 FPS.

well, and what about frame drops?

OK, I did a 20 minute test with EVR-Custrom present @120 FPS = 0 frame drops

One difference is also, that MP-HC actually reaches 120FPS instead of like 119,8 FPS with madVR

Go to forum: Using SVP