76 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dawkinscm wrote:
RickyAstle98 wrote:
dawkinscm wrote:

I'm not talking about Smooth Motion which is why I specifically mentioned DLSS4. I'm not sure how useful Smooth Motion would be if it is as bad as described above.

Im already playing games with DLSS4 SR without new driver, Transformer model looks awesome!

I watched a deep dive into the 5090 architecture and boy is it AI focussed and not games focussed at all. They could have made a cheaper card which actually performed better for games. But instead they made an AI card that can be used for games. This could be great for Rife. Unfortunately this probably only applies to the 5090 because the 5080 is so far behind it in performance, that some argue it's really a 5070.

Unlike previous generation, 4070 was 50% of 4090 performance on 4K average output, according to 5080/5090 review tests, I think 5070 would be 45% of 5090 now!

Go to forum: Using SVP

77 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dawkinscm wrote:
Xenocyde wrote:
dawkinscm wrote:

The new Nvidia app enables some of the better DLSS4 algorithms for all Nvidia cards. Not sure if they meant to do that but apparently it works pretty well.

I got the latest version and don't see any Smooth Motion option in it.

I'm not talking about Smooth Motion which is why I specifically mentioned DLSS4. I'm not sure how useful Smooth Motion would be if it is as bad as described above.

Im already playing games with DLSS4 SR without new driver, Transformer model looks awesome!

Go to forum: Using SVP

78 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dawkinscm wrote:
RickyAstle98 wrote:
dlr5668 wrote:

Does anyone have 50xx? It comes with new interpolation https://imgur.com/a/c7irsaU

According to NVIDIA descriptions, this is driver-based AI model delivers smoother gameplay by inferring an additional frame between two rendered frames on GeForce RTX 50 Series GPUs!

If thats not DLSS Multi Frame Generation at all, we have brand new driver-based interpolation now? What a shame, no FG for all RTX users!

For games without DLSS Frame Generation support, NVIDIA Smooth Motion is a new option for enhancing your experience! *NVIDIA

The new Nvidia app enables some of the better DLSS4 algorithms for all Nvidia cards. Not sure if they meant to do that but apparently it works pretty well.

Actually not, I dont like NVApp at all, performance metrics works glitchy! Also one more thing: NVIDIA Smooth Motion comes to 40 series soon, future updates!

Go to forum: Using SVP

79 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dlr5668 wrote:

Does anyone have 50xx? It comes with new interpolation https://imgur.com/a/c7irsaU

According to NVIDIA descriptions, this is driver-based AI model delivers smoother gameplay by inferring an additional frame between two rendered frames on GeForce RTX 50 Series GPUs!

If thats not DLSS Multi Frame Generation at all, we have brand new driver-based interpolation now? What a shame, no FG for all RTX users!

For games without DLSS Frame Generation support, NVIDIA Smooth Motion is a new option for enhancing your experience! *NVIDIA

Go to forum: Using SVP

80 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dawkinscm wrote:
RickyAstle98 wrote:
dawkinscm wrote:

Which GPU reviewers are reviewing SVP?

Thats was my personal reques for Linus benchmark, If you dont believe me, not my problem!

When did belief become an issue?  Linus is not the most trustworthy person but I don't have an opinion on this issue one way or the other. I simply asked who was the reviewer.

Nevermind!

Go to forum: Using SVP

81 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

oriento wrote:

where did linus posted his reply to your PERSONAL request ?
in public or in PM ?

PM

Go to forum: Using SVP

82 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

oriento wrote:

where is posted that review ?

PM

Go to forum: Using SVP

83 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dawkinscm wrote:
RickyAstle98 wrote:
Hottea wrote:

Do you guys think something like 5090 finally get us above 4k 60hz with newest models or there are another bottlenecks like DDR bandwidth like some ppl have been saying?

According to 5090 reviewers: the 5090 runs 4.25 heavy 42% faster than RTX 4090 card!

Which GPU reviewers are reviewing SVP?

Thats was my personal reques for Linus benchmark, If you dont believe me, not my problem!

Go to forum: Using SVP

84 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

Hottea wrote:

Do you guys think something like 5090 finally get us above 4k 60hz with newest models or there are another bottlenecks like DDR bandwidth like some ppl have been saying?

According to 5090 reviewers: the 5090 runs 4.25 heavy 42% faster than RTX 4090 card!

Go to forum: Using SVP

85 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dawkinscm wrote:
RickyAstle98 wrote:
scb wrote:

What is the lossless scaling algorithm vs SVP vs SVP RIFE, all like?

RIFE better overall but 8 times heavier than SVP4 basic algo!

LSFG better than SVP4 basic algo!

But you were able to combine Rife and LSFG?

I do RIFE x3 + LSFG x2 instead of RIFE x6 now, 36% GPU usage vs 84% with same smoothness!

1280x720 output 24>144 (RTX 4070)

Go to forum: Using SVP

86 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

scb wrote:

What is the lossless scaling algorithm vs SVP vs SVP RIFE, all like?

RIFE better overall but 8 times heavier than SVP4 basic algo!

LSFG better than SVP4 basic algo!

Go to forum: Using SVP

87 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

narkohol wrote:

vs-mlrt has been updated to 15.8 at https://github.com/AmusementClub/vs-mlrt/releases/

Besides support for RTX 50-series, are there any improvements we can expect on current GPUs by using this new file?

No improvement in realtime performance across all models, almost the same transcoding speed (0.8% slower), but new heavy models runs 5% faster now!

RTX 4070

Go to forum: Using SVP

88 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dawkinscm wrote:
RickyAstle98 wrote:

Part of my message got lost. I was asking how are you actually running LSFG? I have not seen anything except for a steam package.

I will upload quick video YT guide, I cant record generated frames from LSFG overlay!

https://youtu.be/8cIXy2IuMAs

Oh I see. It's part of the new Nvidia app.

Third-party software, not part of new NVIDIA App actually!

Go to forum: Using SVP

89 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

Part of my message got lost. I was asking how are you actually running LSFG? I have not seen anything except for a steam package.

I will upload quick video YT guide, I cant record generated frames from LSFG overlay!

https://youtu.be/8cIXy2IuMAs

Go to forum: Using SVP

90 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

dawkinscm wrote:
RickyAstle98 wrote:
raider10 wrote:

Do you do it with lossless scaling?

As you can see, I did RIFE x3 (1st action) on movie, then (LSFG3 x2) ontop!
I'm probably missing something obvious but how are you applying it? I only see a Steam implementation, nothing else.

The LSFG3 x3 alone: looks choppier and less accurate than RIFE x3 mode, takes 8 times less resources!

Thats better solution for 4K videos for example! For HFR monitor owners SPECIFICALLY!

Usage: when your GPU able to do 3x RIFE with 4K input, theres enough GPU headroom to RIFE x2 + LSFG3 x3 ontop without 100% usage!

For 144Hz users (1080p output / 4070 power levels): RIFE x3 + LSFG x2 = looks very smooth (24>144 / GPU usage 70% ~120W)

For 240Hz users (1080p output / 4070 power levels): RIFE x2.5 + LSFG3 x4 = looks very smooth and accurate (24>240 / GPU usage 80% ~130W)

For 360Hz users (1080p output / 4070 power levels): RIFE x3 + LSFG3 x5 = looks very smooth and accurate (24>360)

All tests done with 4.4v2 model (old) for GPU usage local measurement, newest models almost 2 times heavier, the multiplier decision, gentlemens, is up to you!

Go to forum: Using SVP

91 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

raider10 wrote:
RickyAstle98 wrote:
flowreen91 wrote:

Guys, can anyone check if they made any significant improvements since last time by testing the 20x frame generation on 4k videos?
https://www.reddit.com/r/nvidia/comment … n_30_with/

LSFG3 vs LSFG2 on any multipliers - smoother and better! Quality increased, same fast transitions now look more accurate, not just a fluid!

Im using RIFE x3 + LSFG x2 for 24>144 movies conversion, usually GPU usage looks like RIFE x4 with 6x real FLUIDITY rate!

On videos with LSFG3 only, looks pretty normal, conversion from 24FPS is less accurate, but smooth enough!


Do you do it with lossless scaling?

As you can see, I did RIFE x3 (1st action) on movie, then (LSFG3 x2) ontop!

The LSFG3 x3 alone: looks choppier and less accurate than RIFE x3 mode, takes 8 times less resources!

Thats better solution for 4K videos for example! For HFR monitor owners SPECIFICALLY!

Usage: when your GPU able to do 3x RIFE with 4K input, theres enough GPU headroom to RIFE x2 + LSFG3 x3 ontop without 100% usage!

Go to forum: Using SVP

92 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

Vequa wrote:

Can someone please post updated instructions on installing and using V2 rife models?
I tried to get it done based on this thread but all the significant information is either from 2023 with dead links, doesn't work or is impossible to understand for us common folk.
I attempted to add the "4.25 heavy" v2 model with and without replacing "vsmlrt.py". It appears in the "AI model" dropdown menu but no matter what I do I can't get it to work.

Rename the 4.25 heavy to 4.30 for example, worked for me without replacing vsmlrt.py a long time ago!

Go to forum: Using SVP

93 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

flowreen91 wrote:

Guys, can anyone check if they made any significant improvements since last time by testing the 20x frame generation on 4k videos?
https://www.reddit.com/r/nvidia/comment … n_30_with/

LSFG3 vs LSFG2 on any multipliers - smoother and better! Quality increased, same fast transitions now look more accurate, not just a fluid!

Im using RIFE x3 + LSFG x2 for 24>144 movies conversion, usually GPU usage looks like RIFE x4 with 6x real FLUIDITY rate!

On videos with LSFG3 only, looks pretty normal, conversion from 24FPS is less accurate, but smooth enough!

Go to forum: Using SVP

94 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

RAGEdemon wrote:

This is probably a stupid question however, with AI software, I have noticed that on a 4090, RT is faster than CUDA but not by much.

A 4090 has 128 RT cores and >16000 coda cores.

Is it true that with SVP, when RT cores are being used, CUDA cores are not really being used, and vice versa?

In that case, half the power of the GPU is wasted.

Would it be possible to implement both TensorRT AND CUDA simultaneously to double the performance?

Thats waste of time, recently tested RIFE low/mid workload while playing game, the game doesnt use RT cores for example, lost about 65% peak performance ingame, thats was Counter-Strike 2 game!

In both cases lost averages about 60% performance drop, which means over half the power of the GPU is wasted!

Go to forum: Using SVP

95 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

oriento wrote:

i'm with 4.25v2, you're using 4.4v2 ?

The difference between 4.4 and 4.25 twise as much by performance, but without artefacts! Thats was example!

Go to forum: Using SVP

96 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

oriento wrote:

thanks, for those with rtx 4090, what fps do you get when transcoding 4k video ?

The RIFE model dependant!
Older the model - better performance, newer backwards!
My RTX 4070 can handle 4K source, transcoding 24>48 sequence at 48FPS speed - 4.4v2 performance!

Go to forum: Using SVP

97 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

narkohol wrote:
RickyAstle98 wrote:

Something weird happens, did 1080p test now, 6 delayed frames, then tried 1440p test, got 16 delayed frames!
Man, I used 4.25 heavy, thats model twice heavier than 4.6 (by performance)!

What's your GPU model? Nvidia drivers version and Windows version?

The 11 months old (bought 11 months ago) RTX 4070 GPU model, 565.90 graphics driver, Windows 10 19045.5011 OS build!

Go to forum: Using SVP

98 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

narkohol wrote:
RickyAstle98 wrote:

Only 2 delayed frames for 24 seconds with VSync jitter always sit under 0.200 (59.940Hz display mode / GSync off)
Only 5 delayed frames for 24 seconds with VSync jitter always sit under 0.320 (60Hz display mode / GSync off)

Ahhh...   now I see what it's happening here.

https://i.postimg.cc/MpRK7xy9/image.png

I am talking about 1440p output. 720p or 1080p are also fine here. This was with Model 4.22_lite v2, I think:

https://i.postimg.cc/T3r5VcpR/image.png

but 1440p:
https://i.postimg.cc/CK7ZtLGX/image.png

So, nowhere near maxing the GPU % usage, but very noticeable stuttering/hitching/dropped/delayed frames.


This is with Model 4.6, much lower GPU % usage, not even 50%, but still a lot of stuttering/hitching/dropped/delayed frames:
https://i.postimg.cc/T3QwR216/image.png

I don't get why it doesn't work smoothly for 1440p, when it's not using even 50% of the GPU.
And going back to 1080p resolution in 2024 is just too big of a sacrifice. The trade-off of resolution/smoothness at 1080p doesn't pay off...

Something weird happens, did 1080p test now, 6 delayed frames, then tried 1440p test, got 16 delayed frames!
Man, I used 4.25 heavy, thats model twice heavier than 4.6 (by performance)!

Go to forum: Using SVP

99 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

narkohol wrote:
RickyAstle98 wrote:

I have smooth 60FPS playback!

Default SVP settings and RIFE engine, mpv player...

You mean 'default' with no scaling at all and no special settings in MPV??

MPV doesn't count Dropped frames properly. Even when you have a stutter fest playback it always says 'Dropped Frames: 0 (decoder) 0 (output)'

I had to use   --hr-seek-framedrop=no --framedrop=no --video-sync=display-tempo    to display 'Delayed' frames, and then:

https://i.postimg.cc/T3QwR216/image.png

https://i.ibb.co.com/jfNYT8S/2024-11-04-212432580.png

Only 2 delayed frames for 24 seconds with VSync jitter always sit under 0.200 (59.94Hz display mode / GSync off)
Only 5 delayed frames for 24 seconds with VSync jitter always sit under 0.320 (60Hz display mode / GSync off)

Go to forum: Using SVP

100 Reply by RickyAstle98

(2,447 replies, posted in Using SVP)

narkohol wrote:
RickyAstle98 wrote:

I have smooth 60FPS playback!

Default SVP settings and RIFE engine, mpv player...

You mean 'default' with no scaling at all and no special settings in MPV??

MPV doesn't count Dropped frames properly. Even when you have a stutter fest playback it always says "0"

I had to use --hr-seek-framedrop=no --framedrop=no --video-sync=display-tempo  to display 'Delayed frames', and then:

https://i.postimg.cc/T3QwR216/image.png

I dont need display-tempo/dislay-resample at all, GSync monitor, just did same test and see no delayed/mismatched or even dropped frames, perhaps thats hardware issue, not MPV related, I can clearly state that because my friends has 4070/4080s and have different behavior with RIFE engine with exactly the same settings, so thats impractical, OSD counts bad, see drops by my own, there was none!

Go to forum: Using SVP