or Connect
AVS › AVS Forum › Video Components › Home Theater Computers › Avisynth Script for doubling video frame rates
New Posts  All Forums:Forum Nav:

Avisynth Script for doubling video frame rates - Page 61

post #1801 of 1968
If you are using the correct files I posted, have updated to the latest Visual C++ and are using the latest nvidia drivers, then you are set. That script is optimised for hyper-threading i7's which you've got so everything should be fine.
post #1802 of 1968
@widezu, what do I need to change if anything to get this working on a 4 core/4 thread phenom II chip?

I'm guessing SetMTMode(1,15) maybe change the 15 to 6?
post #1803 of 1968
No change it to 7. What GPU do you have?
post #1804 of 1968
Thanks, and I use an AMD HD 5830.
post #1805 of 1968
How long does it take to mux a video with using http://www.spirton.com/convert-videos-to-60fps/ method of actually converting each video to always play 60fps?
post #1806 of 1968
It depends on the video resolution, the codec settings, etc. The only way to really know is to do it yourself.
For me if I use the fastest codec settings on HD video I convert at around realtime (60fps), so a 44 minute video takes 44 minutes to convert. However my CPU is faster than most so just try it to see.
post #1807 of 1968
So not worth it then. That would take me forever! I have over 500 movies and over 500gbs of TV shows too!

I just want my system to play movies and TV with the smoothness these scripts and SVP provide but with NO artifacts.

Is that possible with an I7?
post #1808 of 1968
Nothing is perfect. If there was something that created perfect interpolated frames, there would not be a thread like this.

We have the most optimised scripts at the moment which are improving over time. I'm satisfied with just playing 720p with the regular script and 1080p using the low end script.
post #1809 of 1968
Quote:
Originally Posted by thejulianath20 View Post

So not worth it then. That would take me forever! I have over 500 movies and over 500gbs of TV shows too!

I just queue them up and run the conversions overnight. My computer is always on for other things anyway so it doesn't require any extra time for me
post #1810 of 1968
You can queue them in an order? Like once one is done the other will start right away?

I would have loved that feature for MakeMKV! I just got done turning all of my iso's into mkv's now! It took weeks when I had time.
post #1811 of 1968
I hope it's not a cross-posting

SVP 3.0.4
post #1812 of 1968
Where do I find the scripts that SVP uses?

I want to see my SVP setup in script form so to compare it to the scripts you guys are releasing here.

Also that way I can see if getting an I7 cpu is worth it or not!
post #1813 of 1968
Right-click on the systray icon and select Interface->Advanced, then right-click the icon again and click Show->AVS-script from last play
post #1814 of 1968
I don't know if this is too much and what exactly I should put here for the script... But here is what I pull up for each of my main 3 profiles:

1080p:

svp_scheduler=true
SetMemoryMax(1024)
svp_cache_reduce=true
svp_flow_lib="C:\\Program Files (x86)\\SVP\\Plugins\\MVTools2\\libflowsse.dll"
multinum=2
multiden=1
algo=21
block=false
ml=80
search=4
sadml=0
sadgamma=1
blend=true
dist=2
pel=1
sharp=2
Recalc=0
blkh=32
blkv=32
overlap=0
overlapv=0
wid_den=1
hei_den=1


LoadPlugin ("C:\\Program Files (x86)\\SVP\\Plugins\\MVTools2\\mvtools2.dll")
SetMTMode(3,5)
ffdShow_source()
Assert(width==1920 && height==800,"Frame size "+string(width)+"x"+string(height)+" instead of 1920x800")
delim=":"
WriteFileStart("C:\\Users\\Buckners\\AppData\\Roaming\\SVP\\Logs\\avsFPS.txt", "FramerateNumerator()", "delim", "FramerateDenominator()")
SetMTMode(2)
src=last
wid_den==1 ? last : \\
hei_den==1 ? LanczosResize(int(width/wid_den/2)*2,height) : LanczosResize(int(width/wid_den/2)*2,int(height/hei_den/2)*2)
super=MSuper(pel=pel, hpad=blkh, vpad=blkv, sharp=sharp)
finest=pel==1 ? super : MFinest(super)
backward_vec=MAnalyse(super, isb=true, blksize=blkh, blksizev=blkv, overlap=overlap, overlapv=overlapv, search=search, searchparam=dist, pelsearch=dist, plevel=0)
forward_vec=MAnalyse(super, isb=false, blksize=blkh, blksizev=blkv, overlap=overlap, overlapv=overlapv, search=search, searchparam=dist, pelsearch=dist, plevel=0)
backward_vec = Recalc==0 ? backward_vec : MRecalculate(super, backward_vec, blksize=blkh2, blksizev=blkv2, overlap=int(overlap/4)*2, overlapv=int(overlapv/4)*2, search=search, searchparam=dist)
forward_vec = Recalc==0 ? forward_vec : MRecalculate(super, forward_vec, blksize=blkh2, blksizev=blkv2, overlap=int(overlap/4)*2, overlapv=int(overlapv/4)*2, search=search, searchparam=dist)
smooth_video=MSmoothFps(super, backward_vec, forward_vec, finest=finest, num=FramerateNumerator(last)*multinum,den=FramerateDenominat or(last)*multiden, algo=algo, ml=ml, sadml=sadml, sadgamma=sadgamma, block=block, blend=blend)
## == MSmoothFps_2.avs ==
mi = MSmoothFps(super, backward_vec, forward_vec, finest=finest, num=FramerateNumerator(src)*2,den=FramerateDenominator(src), algo=algo, ml=ml, sadml=sadml, sadgamma=sadgamma, block=block, blend=blend).SelectOdd
mi = (wid_den==1)&&(hei_den==1) ? mi : mi.BicubicResize(src.width, src.height, b=0, c=0.75)
smooth_video = interleave(src,mi)
smooth_video


distributor()



720p:

svp_scheduler=true
SetMemoryMax(1024)
svp_cache_reduce=true
svp_flow_lib="C:\\Program Files (x86)\\SVP\\Plugins\\MVTools2\\libflowsse.dll"
multinum=2
multiden=1
algo=21
block=false
ml=80
search=4
sadml=0
sadgamma=1
blend=true
dist=4
pel=2
sharp=2
Recalc=0
blkh=16
blkv=16
overlap=4
overlapv=4
wid_den=1
hei_den=1


LoadPlugin ("C:\\Program Files (x86)\\SVP\\Plugins\\MVTools2\\mvtools2.dll")
SetMTMode(3,5)
ffdShow_source()
Assert(width==1280 && height==696,"Frame size "+string(width)+"x"+string(height)+" instead of 1280x696")
delim=":"
WriteFileStart("C:\\Users\\Buckners\\AppData\\Roaming\\SVP\\Logs\\avsFPS.txt", "FramerateNumerator()", "delim", "FramerateDenominator()")
SetMTMode(2)
src=last
wid_den==1 ? last : \\
hei_den==1 ? LanczosResize(int(width/wid_den/2)*2,height) : LanczosResize(int(width/wid_den/2)*2,int(height/hei_den/2)*2)
super=MSuper(pel=pel, hpad=blkh, vpad=blkv, sharp=sharp)
finest=pel==1 ? super : MFinest(super)
backward_vec=MAnalyse(super, isb=true, blksize=blkh, blksizev=blkv, overlap=overlap, overlapv=overlapv, search=search, searchparam=dist, pelsearch=dist, plevel=0)
forward_vec=MAnalyse(super, isb=false, blksize=blkh, blksizev=blkv, overlap=overlap, overlapv=overlapv, search=search, searchparam=dist, pelsearch=dist, plevel=0)
backward_vec = Recalc==0 ? backward_vec : MRecalculate(super, backward_vec, blksize=blkh2, blksizev=blkv2, overlap=int(overlap/4)*2, overlapv=int(overlapv/4)*2, search=search, searchparam=dist)
forward_vec = Recalc==0 ? forward_vec : MRecalculate(super, forward_vec, blksize=blkh2, blksizev=blkv2, overlap=int(overlap/4)*2, overlapv=int(overlapv/4)*2, search=search, searchparam=dist)
smooth_video=MSmoothFps(super, backward_vec, forward_vec, finest=finest, num=FramerateNumerator(last)*multinum,den=FramerateDenominat or(last)*multiden, algo=algo, ml=ml, sadml=sadml, sadgamma=sadgamma, block=block, blend=blend)
## == MSmoothFps_2.avs ==
mi = MSmoothFps(super, backward_vec, forward_vec, finest=finest, num=FramerateNumerator(src)*2,den=FramerateDenominator(src), algo=algo, ml=ml, sadml=sadml, sadgamma=sadgamma, block=block, blend=blend).SelectOdd
mi = (wid_den==1)&&(hei_den==1) ? mi : mi.BicubicResize(src.width, src.height, b=0, c=0.75)
smooth_video = interleave(src,mi)
smooth_video


distributor()



720x576:


svp_scheduler=true
SetMemoryMax(1024)
svp_cache_reduce=true
svp_flow_lib="C:\\Program Files (x86)\\SVP\\Plugins\\MVTools2\\libflowsse.dll"
multinum=2
multiden=1
algo=21
block=false
ml=80
search=4
sadml=100
sadgamma=1
blend=true
dist=4
pel=2
sharp=2
Recalc=0
blkh=16
blkv=16
overlap=8
overlapv=8
wid_den=1
hei_den=1


LoadPlugin ("C:\\Program Files (x86)\\SVP\\Plugins\\MVTools2\\mvtools2.dll")
SetMTMode(3,5)
ffdShow_source()
Assert(width==624 && height==352,"Frame size "+string(width)+"x"+string(height)+" instead of 624x352")
delim=":"
WriteFileStart("C:\\Users\\Buckners\\AppData\\Roaming\\SVP\\Logs\\avsFPS.txt", "FramerateNumerator()", "delim", "FramerateDenominator()")
SetMTMode(2)
src=last
wid_den==1 ? last : \\
hei_den==1 ? LanczosResize(int(width/wid_den/2)*2,height) : LanczosResize(int(width/wid_den/2)*2,int(height/hei_den/2)*2)
super=MSuper(pel=pel, hpad=blkh, vpad=blkv, sharp=sharp)
finest=pel==1 ? super : MFinest(super)
backward_vec=MAnalyse(super, isb=true, blksize=blkh, blksizev=blkv, overlap=overlap, overlapv=overlapv, search=search, searchparam=dist, pelsearch=dist, plevel=0)
forward_vec=MAnalyse(super, isb=false, blksize=blkh, blksizev=blkv, overlap=overlap, overlapv=overlapv, search=search, searchparam=dist, pelsearch=dist, plevel=0)
backward_vec = Recalc==0 ? backward_vec : MRecalculate(super, backward_vec, blksize=blkh2, blksizev=blkv2, overlap=int(overlap/4)*2, overlapv=int(overlapv/4)*2, search=search, searchparam=dist)
forward_vec = Recalc==0 ? forward_vec : MRecalculate(super, forward_vec, blksize=blkh2, blksizev=blkv2, overlap=int(overlap/4)*2, overlapv=int(overlapv/4)*2, search=search, searchparam=dist)
smooth_video=MSmoothFps(super, backward_vec, forward_vec, finest=finest, num=FramerateNumerator(last)*multinum,den=FramerateDenominat or(last)*multiden, algo=algo, ml=ml, sadml=sadml, sadgamma=sadgamma, block=block, blend=blend)
## == MSmoothFps_2.avs ==
mi = MSmoothFps(super, backward_vec, forward_vec, finest=finest, num=FramerateNumerator(src)*2,den=FramerateDenominator(src), algo=algo, ml=ml, sadml=sadml, sadgamma=sadgamma, block=block, blend=blend).SelectOdd
mi = (wid_den==1)&&(hei_den==1) ? mi : mi.BicubicResize(src.width, src.height, b=0, c=0.75)
smooth_video = interleave(src,mi)
smooth_video


distributor()



Remember I have an older quadcore q6600 with 4gb ram. I would love to play all 3 profiles with no artifacts.

So can I upgrade these scripts/svp with an I7 and get NO artifacts or very very little?
post #1815 of 1968
Will an I7 make that much of a difference that will allow for no artifacts?
post #1816 of 1968
The CPU does not determine the amount of artifacts. It's the settings used in the scripts. Just because a script loads more on the CPU, does not mean that it is a better script.

Yes you can get an i7, it will help with CPU load but if you use the same script, the picture will be the same.

There is no such thing as NO artifacts. You have to understand that MVTools is not magic, it does not create perfect frames. It was made primarily for denoising and frame recovery not frame creation. We have people over at SVP who seem to know what they are doing and are working on better things which is great but we are all noobs here.

Simply put: asking for NO artifacts will gain you nothing.

I have attached what should be decent settings for 1080p. For 720p and the rest, what you should do is activate the 'Recalculation of vectors'. Those are the best settings that I recommend.

I hope this helps. If you thought I was too hard on you I'm sorry. I get lots of people just plainly asking me for a zero artifact script and I get really fed up, if I had one, don't you think that I would put it up already?
LL
post #1817 of 1968
I understand cpu doesn't determine the output. Obviously its the script! But my old q6600 can't handle the intense scripts for 1080p. That is why I ask if an I7 could handle the faster more intense, more accurate scripts. But yes thank you for help.




I wonder what the comparison between the best script with an I7 and a brand new TV with 240hz motion flow built in? Which looks better? Or does it look even better with both running (tv 240hz and SVP together)?
post #1818 of 1968
Quote:
Originally Posted by thejulianath20 View Post

I understand cpu doesn't determine the output. Obviously its the script! But my old q6600 can't handle the intense scripts for 1080p. That is why I ask if an I7 could handle the faster more intense, more accurate scripts. But yes thank you for help.

Yes, it could handle them better. i7s are good.

Quote:
Originally Posted by thejulianath20 View Post

I wonder what the comparison between the best script with an I7 and a brand new TV with 240hz motion flow built in? Which looks better? Or does it look even better with both running (tv 240hz and SVP together)?

The AviSynth scripts are better quality than the TV interpolation. The best thing to do is to run them both, so the last option. Interpolate with AviSynth to 60fps then the TV goes from 60 to 240 (or 600 or however much your TV does)
post #1819 of 1968
Thank you Widezu69 and Subjunk.

What setting in SVP is the best to fiddle with to remove the most artifacts in fast motion scenes? Because in camera panning and normal motion, everything looks fluid but fast motion looks very noticeable with artifacts.
post #1820 of 1968
Quote:
Originally Posted by thejulianath20 View Post

fast motion looks very noticeable with artifacts.

Do not use algo=21, 23. The masking is sometimes, let's say, "strange". It looks like this: .

We're working on this , use algo=13 instead.
LL
post #1821 of 1968
^^I second that, algo=13 introduced by MSmoothFPS is very good. I have to thank the SVP team for this.

@thejulianath20 Can your CPU not handle the settings I posted for you there? Those are basically the same settings that I use in my scripts and I have a lowly laptop i7. 2.7Ghz on 4 cores/8threads. If you have some money, try getting an i7, you won't need anything huge. What is your budget? Or what is it that you got your eye one?
post #1822 of 1968
I don't have a specific one in mind yet... I will be building another htpc here before the end of the year. I want it to be small, probably micro or mini atx with no video card. So I want a new intel cpu that can handle all of the video formats I through at it!

I have not tried your setting yet, widezu69. I will tonight, but if you have an I7 with 8 threads, then I can safely bet that my setup can't handle those settings yet.



How can I change the Algo setting in SVP?
post #1823 of 1968
The settings I posted use algo=13
post #1824 of 1968
@widezu Those settings work great for my 720p profile! But my 1080p can't even get anywhere near that....


What can I change around on this to look "better" without maxing my cpu? I wish gpu accel would work for me, but I get bad screen tears with it on.
LL
post #1825 of 1968
Hmm, I hate to say it buddy but you probably need a new CPU. My advice when building your next HTPC, stick in a cheap graphics card. You'll save so much more CPU power as a lot of the work is offloaded to a GPU.

You could probably get away with a using your quad at the moment but just sticking a cheap GPU in and turning on OpenCL.
post #1826 of 1968
I have a great gpu in already! I have a ATI Redeon HD 5570! Please tell me how OpenCL works and I can finally offload some of the work to my GPU!

Like I said before I tried gpu acceleration with SVP and it caused screen tears so I turned it off. It also didn't save much cpu processing.
post #1827 of 1968
Hmm, do you have Aero enabled? Enable GPU acceleration from the SVP menu. Perhaps install reclock, open up Media Player Classic, press O, down the bottom is the Audio Renderer option, select reclock. Try playing something again with OpenCL.
post #1828 of 1968
Still the same result...

I don't know how reclock for audio will help lower the cpu usage... But I did it and turned on GPU acceleration in SVP and I had random tears again and my CPU usage was higher by about 10%-15%.
post #1829 of 1968
Reclock is not there to help reduce CPU usage. It is used to sync the video to your refresh rate removing tearing. Have you perhaps tried a different renderer? Haali or MadVR?
post #1830 of 1968
Should I download OpenCL for my AMD graphics card?
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Home Theater Computers
AVS › AVS Forum › Video Components › Home Theater Computers › Avisynth Script for doubling video frame rates