Search options (Page 5 of 18)

dawkinscm wrote:

RickyAstle98 wrote:

My actual scenarios
v14test3 320fps / v15.2 310fps (4.4v2 with 1.78 dynamic shape and 320 fps with 2.35 dynamic shape)
Starting models 4.16 everything gets worse, just a performance drops variables from 6 to 12% respectively, the movie dynamic shapes always faster (even when the targeted resolution is slightly higher)

So it works better with Performance Boost disabled? Am I understanding this correctly?

Yes and no! Easier explanation - depending on the aspect ratio of the sources, depending on the resolution and the model used, performance can be either increased to 2% or degraded to 12% with or without TRT boost option! Seek performance is night and day with boost enabled! I did not regret that I switched to a new library!
So - with 15.2 you gets maximum 12% performance degradation, but practically is just 6% of that!

..on average he got around an 8 fps improvement. But this stat should be taken in the context of dev notes that suggest there might be be specific situations where the performance might be slightly worse.

aloola wrote:

v15.2 gives me a tiny FPS boost
v14test3 ~150fps vs ~ v15.2 ~158fps, also fixed dynamic shape (Performance boost: off) bug

15.2 changelog:

    Upgraded to TensorRT 10.3.0.

    Fixed performance regression of RIFE and SAFA models starting with vs-mlrt v14.test4. This version may still be slightly slower than vs-mlrt v14.test3 under some conditions, however.

Looking forward to @aloola benchmarks for the new TensorRT like he did before:
https://www.svp-team.com/forum/viewtopi … 324#p84324

erriep2703 wrote:

# Deband filter. Always turn on for anime.
deband=yes # Default values are 1:64:16:48

I see that these settings are present in an outdated guide:
https://kokomins.wordpress.com/2019/10/ … fig-guide/
<meta property="article:modified_time" content="2024-02-08T06:53:17+00:00"/>

Probably we should use more recent configs from 2024:
https://kohana.fi/article/mpv-for-anime
https://artoriuz.github.io/blog/mpv_upscaling.html
https://iamscum.wordpress.com/guides/vi … /mpv-conf/
https://thewiki.moe/tutorials/mpv/

dawkinscm wrote:

Xenocyde wrote:

I haven't noticed any smoothness problems since I don't watch anime. I only tested with my usual football scenes
and 4.21 still has the same edge artifacts as 4.20 in those scenes.

BTW, 4.22 is out. Maybe they fixed it with this one. Testing now.

4.22 is the fix he's talking about. It's a little better than 4.21 but overall it has the same artefacts.

Ah thought it was a 4.21 update, they don't actually mention 4.22 in the comments.

Xenocyde wrote:

Can confirm the edge artifacts are still there in 4.21.

The devs fixed the smoothness issue:
https://github.com/hzwer/Practical-RIFE … 2275333756

v15.2 gives me a tiny FPS boost
v14test3 ~150fps vs ~ v15.2 ~158fps, also fixed dynamic shape (Performance boost: off) bug

15.2 changelog:

    Upgraded to TensorRT 10.3.0.

    Fixed performance regression of RIFE and SAFA models starting with vs-mlrt v14.test4. This version may still be slightly slower than vs-mlrt v14.test3 under some conditions, however.

Hello,

I'm having a bit of trouble getting SVP to run with MPV with other scripts enabled. If I run SVP4 with just the "Open file with MPV", it works fine. But I wanted to add Anime4K script to MPV so I can upscale old anime to higher quality as I had done before. But whenever I install said script, SVP stops functioning and won't detect MPV, even if launched from SVP4. If I uninstall it, it works. Is there some kind of conflict between the 2?

Thanks in advance!

dawkinscm wrote:

As I said before, machine generated scores doesn't guarantee the best looking picture because human perception and preference isn't objective. The other thing I said is that much of this is designed specifically for Anime and that definitely applies to ArtCNN. So hopefully you are an Anime watcher

Also don't forget there are different levels to it, for example I would never use strongest FSRCNNX because it would introduce too much sharpness for non-Anime. However, the one I do use, is always very helpful in improving the sharpness just a little a bit because I have the TV sharpness set to 0 on the LG C2.

TV sharpness is much worse than these AI shaders btw, so actually using the shaders as an alternative and toning down the TV ones definitely helps.

In most cases though, I imagine people here use monitors, which automatically have sharpness set to 0.

It also wouldn't make sense to use these upscalers on a 27" or 32" 4K monitor, as again like @dawkinscm says, it would look too sharp.

However, watching on a 42" screen, and I have been using 42" and 43" screens for almost a decade now, I started using FSRCNNX in its very early days and have not looked back, it is great.

So, a lot of factors play a role, no harm in testing and seeing what you prefer.

For reference, I use FSRCNNX_x2_8-0-4-1.glsl, which is the 8- and it's very low sharpness compared to the 16- and 32-.

I just had SVP trigger in "ignore HFR sources," but I'm not really sure why this happened. I'm watching S10 of the new Doctor Who series and all episodes appear to be encoded at 25 fps according to MPV. However, starting with Episode 6, SVP is triggering the ignore HFR sources profile after saying it is playing at 50 fps. How does SVP identify some videos as HFR and some as non-HFR if all are 25 fps? Also, if I raise the video frame rate threshold from 47 to 60 on the ignore HFR sources profile, the RIFE profile gets triggered with fixed 60 fps, but some videos exhibit strange microstuttering as if they are not truly running at 60 fps, probably because of the threshold conversion. Any way to fix this?

dawkinscm wrote:

Nowadays I mostly use in-built scalers.

I found this link with upscaler examples really useful:
https://github.com/LitCastVlog/Plex-GLSL-Shaders

the ultimate goal is to improve visual quality, so we will always try to pick the combination that makes it look better overall
can u please explain which in-built scalers combination u use so we can try that too?
i preffer to set max visual quality + resize down until GPU stops dieing, than watch pure image native 24 fps. xD

Xenocyde wrote:

What are these for exactly?

Scaling methodology.

Dithering methodology.

The shaders at the bottom (which have to be downloaded and added into a shaders folder separately) improve quality. FSRCNNX is an upscaler, while KrigBilateral is a chroma upscaler, SSimDownscaler is for downsampling.

You can read more about them on their pages. But also, if you want an article explaining them, as well as better shaders that are more demanding which I cannot run with an RTX 3090 while using RIFE, have a read through this: https://artoriuz.github.io/blog/mpv_upscaling.html