You are not logged in. Please login or register.
SmoothVideo Project → Posts by dawkinscm
> But also mpc-hc is very old
wat?
One of the reasons mpc-be came out was because HC was a dead project. But I forgot that it has been restarted. I've updated my original post.
I have an I7-8086K and a RTX 4090. I watch movies on MPC-HC with Vapoursynth, MPC internal renderer, and Internal LAV filter. My performance is very odd, with huge spikes then dropping to zero. It is making movies unwatchable. It is not an issue when SVP is not running. Any help would be greatly appreciated!
You may have to use mpv instead. If you still have issues, even with mpv then there's something wrong and as @flowreen91 says you can try re-installing SVP.
Im already use mpv player since SVP4 Pro buy, I custom mpv for my preferences, so I cant get extra performance at all? I test d3d11va and RTX 4070 dies at 720@24>144 realtime playback, but d3d11va-copy didnt
The reason you get better performance with d3d11va over d3d11va-copy is because with d3d11va SVP can't properly process the frames. d3d11va-copy copies frams to system ram where SVP has full access to them. So unless SVP has changed and somehow has direct access to GPU memory, you need to use copy. But you probably already know this and you were just joking.
I deleted mpv after one day. Horrible GUI and doesn't work with the MadVR renderer.
What made you think that it should work with MadVR? Running mpv is equivalent to running mpc-be+MadVR and with that flexibility there is a learning curve. Yes it is text based but you can also download gui's for it. There are also gui based players that use mpv as a display engine.
But SVP comes with baseline mpv.conf file so unlike mpc-be, it will just work with zero extra configuration.
VLC is the "for dummies" media player and doesn't have much customization. mpv goes too far extreme in the other direction and is for those who think FFmpeg command line is the greatest encoder. MPC-HC/BE is in the middle and that's good enough for me.
mpc-be is a great media player and a nice middle ground, but mpv just works more efficiently with SVP. mpv is open source and uses code from other oepn source projects including ffmpeg and even VLC "for dummies".
Is MPC-HC better for GPUs like RTX 4080? I'm on MPV but screen is 1080p so don't think it really matters right now. I'm thinking of getting a 4K screen soon, though, so maybe MPC-HC is better for 4K?
No. One of the reasons I switched from mpc to mpv is that it works better with SVP at any resolution.
MAG79 wrote:Gippy thanks for information.
In the past D3D11 Native value in LAV decoder was incompatible to SVP.
Need to check if it works correctly and SVP does its work fine (24 to 60 fps conversion).
---
I've checked. D3D11 works good with SVP.
In LAV decoder filter properties it indicates as Active Decoder: d3d11 cb directDoes d3d11va-copy same thing D3D11 in mpv player? How to force D3D11 thru svp and mpv player, a bit confused, I want more performance!
For SVP, whichever decoder you use in mpv it needs to be a copy-back decoder where the frames are copied back to RAM, wtherwise SVP can't process them. d3d11va-copy is one of the better Windows copy-back decoders. If you want to start using mpv then use the SVP installer to install it then use the default mpv.conf that comes with SVP. It's not perfect but it will 100% work.
While employing MPV with Rife 2, I've noticed seamless performance, yet I'm unaware of regularly dropping "output" frames. Despite my GPU hovering around 80%, the root cause eludes me. Moreover, I stumbled upon the "ffdshow" parameter under Advanced Settings, which is currently set to "true". Is this configuration accurate?
I've asked the ffdshow question myself a long time ago. I don't remember getting an answer so I just set it to false. With mpv there could be a number of reasons for frame drops including a bug in the stats page, the hwdec you are using or something else. Also make sure that the SVP index is not lower than 1.0 because if it is then you may have to use a less compute heavy version of Rife. You are just going to have to troubleshoot it. If it's not affecting what you are seeing on screen then maybe don't worry about it too much until you see a problem
This is good to know, but what are the differences between the models? When should one pick a particular one? What are the different models best at? What are the quality differences?
I suppose you can choose different models for different content if you wish. But no one can answer this question for you. You will have to test each model with different content. Later models have less artefacts and better motion handling. So maybe choose the one that works best for you and stick with it?
BTW my latest tests suggest that as good as Rife 4.14 is for fast movement, Rife 4.13 is actually a little better in reducing the double images.
Guys,
I still haven't seen a full explanation about the different between the 'normal' and 'lite' models of the same versions? What situations should I pick the 'lite' version over the 'normal' version of a model?
The purpose of the "lite" versions were never fully clear to me except for Rife 4.14 "lite" which is different to every model that came before it. But there is no answer to your question, you just have to try them and see which one works best for you. As of right now, if you don't have a high end card then don't bother with anything above Rife 4.9. Rife 4.13 is computationally heavy even for a 4080. Rife 4.14 "lite" is not "lite" at all unless you have a card with high memory bandwidth like a 'maybe' a 3070ti and above.
RTX 3060 Ti, i5 12400f, 32 GB of RAM.
"Lite" versions works better everytime. From newest versions with "lite" (4.12 and so on) usual versions dropped frames and it's like don't have enough power or overloading, because it's not working on 60 FPS everytime (48 FPS works OK), but i didn't do any test, just saw the FPS and my internal feelings. So, now i'm on 4.12 "lite" version and it's great.
Newest 4.13 and 4.14 and lite too don't work properly in any scenarios with my GPU because of overloading or something. So, now i'm on 4.12 "lite" with hope for better versions in the future.
Rife 4.14 "lite" is different to the other "lite" models. It uses a technology that requires more GPU bus bandwidth but not as much compute power. Rife 4.13 "lite" is similar to regular Rife 4.9 so if you struggle with 4.9 you will struggle with 4.13. I think maybe Rife 4.14 needs more compute power and memory bandwidth than the 3060ti can provide.
roshin1401 wrote:Im noob to this software. While using TensorRT, a cmd popups doing compilation. Does this occur for every video file opened? or only once? And also please suggest best rife model for my gpu. Its a 3070 laptop.
I'm having the same problem and unfotunately this post wasn't responded to. I've put the .OONX files in the "models" and/or 'Rife" folders as instructed and yes, they do show up in the AI models pulldown in the SVP manager, but when I select it and try to run a video, a CMD box pops up and after a few seconds goes away. The video then plays but is not converted. Nor does the little green text in the lower left corner come up telling you what setting you're using, etc. Tres bizarre. Plus, if I try to transcode the video it doesn't work either. Says conversion failed.
It seems pretty straitforward, I don't what I'm doing wrong here. Please help!
I keep saying this but it's worth repeating. If you are new to this then don't do anything you see here because we are experimenting. SVP works out of the box. Just uninstall everything, go to the SVP Rife AI page https://www.svp-team.com/wiki/RIFE_AI_interpolation and follow the instructions. All the information you need about the Rife, the DOS box and everything else is on that page. Once you have done that then have a look at the SVP Transcoding page https://www.svp-team.com/wiki/Manual:SVPcode. I don't transcode, but once you have followed those instructions there are others on here who can help you with any specific questions you have.
537.58 eliminated the pan stutters for me.
It made no difference for me when I tried it but I do like the options that come with the installer. So I installed the latest version because of Cuda 12.3 but with the options. It works at least as well as before with or without 537.58 but with a couple of non SVP related improvements.
Other modes such as sync 24fps video to screen refresh, or fps x 2.5, or whole multiples, is really taxing on the GPU. An overclocked 4090 can't do it, even on the highest performing RIFE (4.4/4.6).
I tried 2.5 and it was pretty bad but then if an overclocked 4090 can't do it then I've got not chance 
Thanks for the input, so looks like I'll just stick with the 3090 until 5090 is released.
3840x2160 was watchable on the 4080, but with slow pan stutters. I updated the Nvidia Driver to latest version released yesterday and used the v14Test3 TensorRT library (Cuda 12.3) and the latest vslmrt.py with Rife 4.14v2 lite. It's still dropping frames but the slow pans are now mostly smooth. The 4090 should have no problem so you don't need to wait for a 5090.
I currently don't have any intention of using SVP with 4K but this should help to clarify your options.
Blackfyre wrote:Question, regarding RIFE I assume the 4070, 4070 Super, 4070 Ti Super, etc are all better than the RTX 3090 with regards to RIFE yes?
I'm asking because I can sell my RTX 3090 used for around $1100 in Australia, and just buy an RTX 4070 Super or 4070 Ti Super for example on special for around the same price.
3090 can no longer handle 4K at 2x with the latest RIFE models sometimes, if the 4070 and higher can handle it, I am willing to do the switch, despite having half the vRAM.
Tested! My 4070 provide 2x 4.14v2 (and lite) under 3200x1824opt (3520x1984opt unstable) At exact 4K nah, so I think 4070Ti close enough for latest models!
Windows 10 fresh (3 month) and NVIDIA driver 546.17 without mods...
4080 with Rife 4.14v2 lite plays 3840x2160 at 2x SVP Index 1.0 GPU 99-100% with frame drops. Mostly smooth but some slow pans a little jerky. 4080 can barely do it so 4070Ti will not do it.
537.58 is super stable indeed. You can also use https://www.techpowerup.com/download/te … leanstall/ to install it without bloat (GFE and other crap)
SVP, games with FSR3+DLSS2 mod - all works great
I installed 537.58 and tbh it made little obvious difference. But the install tool was useful to completely fix some Nvidia driver issues I knew about but had only partially fixed. I just made another back up of everything. Thank you
Question, regarding RIFE I assume the 4070, 4070 Super, 4070 Ti Super, etc are all better than the RTX 3090 with regards to RIFE yes?
Not necessarily. I would test 1.14 lite first because the 3080/3090 has a wider memory bus than even the 4070ti/4080 so the 3090ti might work better with Rife 1.14 lite than anything less than a 4090. Hopefully someone can help with other card comparisons.
Gentlemen,
Consistent reports are emerging regarding recent nvidia drivers causing stuttering in apps and games. It is related to the DPC latency.
Current advice is to roll back to 537.58 drivers (dated October 2023).
More info in the COMMENTS in this thread:
https://www.reddit.com/r/nvidia/comment … iscussion/I believe I have seen these stutters in SVP too. Rolling back to the recommended driver fixed the issue. Best staying on the older driver till a confirmed fix has been integrated...
Looking through that thread, there are conflicting reports as to which Nvidia drivers are causing problems. There are also reports of problems being caused by Windows 11. I downgraded to Windows 10 months ago after having lots of issues. I switched to Nvidia Studio drivers as they are more stable with less changes and less likely to cause an issue. But I'm not a gamer. I also turned off Windows GPU scheduling on the recommendation of a VR dev.
dawkinscm wrote:cemaydnlar wrote:v1.14 v2 is the smoothest version of them all for me somehow...
v1.14 v2 is so smooth that I made sure to back everything up
I meant v1.14 lite v2 sorry
i am using 3070. That might be the case i appreciate your answer bud thx.
It's OK. We were both talking about 1.14 lite v2 and it is very smooth
v1.14 v2 is the smoothest version of them all for me somehow...
v1.14 v2 is so smooth that I made sure to back everything up
It gave a weird error while caching but is somehow working without doing that caching cmd again.2nd i don't see any improvement for gpu usage. İt's the same as v.14. Is it normal?
If the messages come after the "profiling" statement then my understanding is that the warnings are part of the profiling process. As for the performance, that might depend on you GPU memory bandwidth. The Nvidia 3070Ti and above start at 600GB/s memory bandwidth. The 4070Ti has the same memory bandwidth as a 4080 at 700GB/s. But he older 3080 and 3090 series cards might perform better with 4.14 "lite" than the 4080 because of higher memory bandwidth. Below this class of card YMMV.
The new technique used for the "lite" versions is potentially a little more accurate than the normal versions but I suspect that GPUs with low internal memory bandwidth will have issues with v14 "lite". I'm keeping it for now and will see what happens further down the line.
Edit:
The figures from @aloola and @RickyAstle98 show that this is a heavy model. But now everything has kind of bedded in, Rife v1.14 "lite" (v2) is using a little less GPU resources than before, down from 94-100% with Rife v1.14 to between 86-92% with "lite" while "so far" retaining the same quality as before. It's also very smooth. YMMV on GPUs with limited memory bandwidth.
So far v14.4 lite uses more GPU than the non lite version. I thought it was may be because of the changes I made but I reverted back to SVP standard configs and it's the same. Maybe there is an issue they need to fix.
Update:
It uses about the same amount of GPU resources but it works in a different way which makes it more dependant on memory bandwidth than GPU resources. I've read up a little on the new methodology and it is interesting the result is that the "lite" version may not be as "lite" before.
There may be fixes and enhancements in the latest v14 releases which is not in v13. If you're willing to help test, I would try downloading https://github.com/AmusementClub/vs-mlr … 4.test2.7z and unzipping
vstrt.dlland the folder
vsmlrt-cudainto
C:\Program Files (x86)\SVP 4\rife
I've updated and I don't see any obvious differences. But I have thought about updating for the latest fixes to TensorRT and Cuda so thanks
Current default SVP installation has 4.6/4.9 v1 models which means average user never saw this jiggling issue.
In the future when SVP devs will want to add the v2 models they will surely update the default old TensorRT libraries too so your average user will never see this issue.Until then we should follow @cws guide on how to update our own TensorRT library from https://github.com/AmusementClub/vs-mlrt/releases if we want to play around with latest RIFE models:
cws wrote:Right, TensorRT 9.x is in "beta" and not officially released. Still, v13 was released when TensorRT 8.5 was the latest version, when there is a new release of TensorRT 8.x (which is 8.6.)
There may be fixes and enhancements in the latest v14 releases which is not in v13. If you're willing to help test, I would try downloading https://github.com/AmusementClub/vs-mlr … 4.test2.7z and unzipping
vstrt.dlland the folder
vsmlrt-cudainto
C:\Program Files (x86)\SVP 4\rife
If true then this is not so much a vsmlrt fix as it is a Cuda/TensorRT fix. I think the last time I tried this my Tensor libraries weren't updated but I might use the info above and try again.
Both SSimDownscaler and KrigBilateral I get from here and are frequently updated.
Yep. That's where I got the last copy from too. It's configured for Catmull Rom. The ones I saw configured for Mitchell must have been local variations by other users.
ontop glsl-shader="C:\Users\Username\AppData\Roaming\mpv\Shaders\SSimDownscaler.glsl"
FYI While I think SsimDownscaler by default was configured for Mitchell, some configs out there have it configured for Catmull Rom.
It's true there's no significant performance improvement, but there may be fixes and other minor improvements.
Specifically, when using v13 (with TensorRT 8.5 / CUDA 11.8) with the v2 models, I can see some blurriness and pixels shifting in the bottom right corner if there is some obvious static content there like a logo. If this is something you can reproduce, try using the latest v14 (with TensorRT 9.x / CUDA 12.x)
Thanks. If you see a particular issue that you want fixing then this does make sense. Especially if you are saying that v14 fixes this issue for you. I still do testing of new models, changing code etc but I think SVP is stable and works well now so I would not recommend most people do this. Why fix something when it ain't broken ?
SmoothVideo Project → Posts by dawkinscm
Powered by PunBB, supported by Informer Technologies, Inc.
