Selur's Little Message Board

Full Version: Resize Keeps Changing on its own when Starting a Batch Queue, Please fix this.
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3 4 5 6 7 8
Checked, i9 9900k is too old to have efficiency cores. Smile
Are you using Vapoursynth or Avisynth?
" FilterInterlanceHandeling on Slower." <- what?
I suspect QTGMC or the interlace handling is at fault.
I have another issue dunno if its the new version, but some of my videos are getting 10minutes cut if i have Vapoursynth Enabled Filter. If i dont have any enabled they encode just fine dunno what the issue is if its the new version of hybrid or not i might have to go back to old version and check when i have the time. Because i did the same video before.

This is the option that was tanking the fps, but the FPS are still lower than usual:

[Image: VJCdzsk.png]
I was talking about this:

[Image: CnQBRQX.png]


EDIT:

If i noticed this earlier maybe would have made quicker problem solving instead of rencoding every time. It cuts like 10k Frames. 49k on input 39k on output 31k in preview.

[Image: eNRID2q.png]


Edit2:

Reverted back to old version, everything works fine same settings video is full and the slow preset doesnt kill the FPS to 10 instead its 40FPS at 100% cpu. Somethng is screwed with last version.



Edit3:

Another thing to note CAS GLSL Shader makes the whole Video GREEN on last version but not in the previous version. Latest updated defiantly broke something.
About the speed: try a different source filter and compare the Vapoursynth Script Views to see if anything other than the source filter changed.
About CAS: probably depends on the combination and order of filters, so also check the Vapoursynth Script View.
From your description, I can't reproduce the problem.


Cu Selur
(12.10.2024, 08:10)Selur Wrote: [ -> ]About the speed: try a different source filter and compare the Vapoursynth Script Views to see if anything other than the source filter changed.
About CAS: probably depends on the combination and order of filters, so also check the Vapoursynth Script View.
From your description, I can't reproduce the problem.


Cu Selur

I just updated again and used the profile from the previous version no issues 40FPS Full Video 100% CPU. I Restarted Hybrid and followed a video, setting up the Profile from the video instead of loading it. Same settings all of it 10FPS Green Video dropping to 30% CPU. So something is bad in the default configuration that i am not seeing that has changed ? What changed from the previous version ?
Here is my config you can try it if you want not sure if you will be able to replicate it. Bascilly if u load the config the encoding is fine, but try loading it then screenshot/record the settings, reset hybrid and input the same stuff u will get the bad results. Maybe the best way is to do that then save the config and compare the text file for differences to find what is causing this. I am actually gonna provide both:

Config 2024.06.16 Working:

https://www.mediafire.com/file/ya333n5xy...9.xml/file

Config 2024.09.29 Bad one:

https://www.mediafire.com/file/0scidh542...e.xml/file

I see the filter order changed but i think that is just because u added bunch of new stuff. I never touch that i leave it how it is default.

Edit Differences from what i could gather:

Code:
<HybridData name="vsProToonMotionMaskTh2" value="10"/>
<HybridData name="vsProToonOnRegion" value="false"/>
<HybridData name="vsProToonOnRegionHeight" value="0"/>
<HybridData name="vsProToonOnRegionWidth" value="0"/>
<HybridData name="vsProToonOnRegionX" value="0"/>
<HybridData name="vsProToonOnRegionY" value="0"/>
<HybridData name="vsProToonOnlyRange" value="false"/>
<HybridData name="vsProToonPrewittEdgeMaskBinarize" value="false"/>
<HybridData name="vsProToonPrewittEdgeMaskBinarizeThreshold" value="0.5"/>
<HybridData name="vsProToonPrewittEdgeMaskInvert" value="false"/>
<HybridData name="vsProToonPrewittEdgeMaskShow" value="false"/>
<HybridData name="vsProToonRanges"/>
<HybridData name="vsProToonRectangleMaskCropBottom" value="0"/>
<HybridData name="vsProToonRectangleMaskCropLeft" value="0"/>
<HybridData name="vsProToonRectangleMaskCropRight" value="0"/>
<HybridData name="vsProToonRectangleMaskCropTop" value="0"/>
<HybridData name="vsProToonRectangleMaskInvert" value="false"/>
<HybridData name="vsProToonRectangleMaskShow" value="false"/>
<HybridData name="vsProToonRetinexEdgeMaskBinarize" value="false"/>
<HybridData name="vsProToonRetinexEdgeMaskBinarizeThreshold" value="0.5"/>
<HybridData name="vsProToonRetinexEdgeMaskInvert" value="false"/>
<HybridData name="vsProToonRetinexEdgeMaskShow" value="false"/>
<HybridData name="vsProToonRetinexEdgeMaskTCSigma" value="1"/>
<HybridData name="vsProToonRobertsEdgesBinarize" value="false"/>
<HybridData name="vsProToonRobertsEdgesBinarizeThreshold" value="0.5"/>
<HybridData name="vsProToonRobertsEdgesHigh" value="255"/>
<HybridData name="vsProToonRobertsEdgesInvert" value="false"/>
<HybridData name="vsProToonRobertsEdgesLow" value="0"/>
<HybridData name="vsProToonRobertsEdgesShow" value="false"/>
<HybridData name="vsVinverseRectangleMaskCropBottom" value="0"/>
<HybridData name="vsVinverseRectangleMaskCropLeft" value="0"/>
<HybridData name="vsVinverseRectangleMaskCropRight" value="0"/>
<HybridData name="vsVinverseRectangleMaskCropTop" value="0"/>
<HybridData name="vsVinverseRectangleMaskInvert" value="false"/>
<HybridData name="vsVinverseRectangleMaskShow" value="false"/>
<HybridData name="vsVinverseRetinexEdgeMaskBinarize" value="false"/>
<HybridData name="vsVinverseRetinexEdgeMaskBinarizeThreshold" value="0.5"/>
<HybridData name="vsVinverseRetinexEdgeMaskInvert" value="false"/>
<HybridData name="vsVinverseRetinexEdgeMaskShow" value="false"/>
<HybridData name="vsVinverseRetinexEdgeMaskTCSigma" value="1"/>
<HybridData name="vsVinverseRobertsEdgesBinarize" value="false"/>
<HybridData name="vsVinverseRobertsEdgesBinarizeThreshold" value="0.5"/>
<HybridData name="vsVinverseRobertsEdgesHigh" value="255"/>
<HybridData name="vsVinverseRobertsEdgesInvert" value="false"/>
<HybridData name="vsVinverseRobertsEdgesLow" value="0"/>
<HybridData name="vsVinverseRobertsEdgesShow" value="false"/>
<HybridData name="vsVinverseSobelEdgeMaskBinarize" value="false"/>
<HybridData name="vsVinverseSobelEdgeMaskBinarizeThreshold" value="0.5"/>
<HybridData name="vsVinverseSobelEdgeMaskInvert" value="false"/>
<HybridData name="vsVinverseSobelEdgeMaskShow" value="false"/>
<HybridData name="vsVinverseSobelFastEdgeMaskBinarize" value="false"/>
<HybridData name="vsVinverseSobelFastEdgeMaskBinarizeThreshold" value="0.5"/>
<HybridData name="vsVinverseSobelFastEdgeMaskInvert" value="false"/>
<HybridData name="vsVinverseSobelFastEdgeMaskShow" value="false"/>
<HybridData name="vsVinverseUseMasked" value="false"/>
<HybridData name="vsVinverseUseMerge" value="false"/>
<HybridData name="yadifmodEdeint" value="NNEDI2"/>
Did you read the post you quoted? Huh

Quote:Maybe the best way is to do that then save the config and compare the text file for differences to find what is causing this.
it is not
Compare the used scripts if you want to see what changed.
Like I wrote, it's probably the source filter.

Cu Selur
(12.10.2024, 18:03)Selur Wrote: [ -> ]Did you read the post you quoted?  Huh

Quote:Maybe the best way is to do that then save the config and compare the text file for differences to find what is causing this.
it is not
Compare the used scripts if you want to see what changed.
Like I wrote, it's probably the source filter.

Cu Selur

Yeah i read it sorry, i thought it was something else i tried, u meant as in the source tab filter. Yes you are correct speed is back to 40FPS 100% The option "Prefer BestSource" Was enabled by default which i didnt notice, it was killing the fps even tho it was set on CPU. After disabling it issue has been solved 40 FPS Encoding 100% CPU and the Time of the video jumped from 21min back to 27min.

Reencoding video looks like it will be still green its encoding at 22kBit/s. Its green in the Preview. Disabling "CAS GLSL" fixes it. But i wanna use it, i used it in the previous version with the same setup with no issues. I dunno what order i should be changing, I might mess it up even more. Dunno what should i be looking for in the VapourSynth script. I will have to go back again to see what changed in the script i guess that would be a start.

As for the green picture, it is green even with the previous config (On Latest Version) i just noticed. I was wrong earlier when i said it was good. Only the FPS Was good and the full video.

Edit:

I went back and now the screen is Green there too... This make no sense, I literally encoded 3 videos this morning with it. Why is it green now ? Sorry for bothering with this dumb stuff but this is so random .....

I also see Engine Directory in the last version, Can Hybrid use .engine TENSSOR to upscale ?

Edit2:


No fukin shot dud. Having Media Player Classic in the Background makes the encode green if you have "CAS GLSL" selected. I have never seen this shit before wtf, i was like hmm lemme close all the programs and try again.. This can be one of two things. RTX Super Resolution being active on MPClassic or Sound Exclusivity BS. Thanks for you time sorry, if you can just tell me if i can use Tensorrt with this so i dont have to deal with docker anymore .
Quote: Its green in the Preview. Disabling "CAS GLSL" fixes it.
You are probably using GLSL somewhere,... (yes, apparently in MPC-HC)

Quote:Can Hybrid use .engine TENSSOR to upscale ?
Through the torch- and vs-mlrt-addon, Hybrid does support a bunch of filters which can use TensorRT. (The addons are on my GoogleDrive, which is linked on the download page.)

Quote: RTX Super Resolution being active on MPClassic or Sound Exclusivity BS.
Basically, anything that does exclusively grab the shaders will kill glsl filters.
This will probably also happen if you use the SDR => HDR option in the GPU and your media player settings. Wink

Quote:Yes you are correct speed is back to 40FPS 100%
Happy that is solved then.

Cu Selur
(12.10.2024, 19:06)Selur Wrote: [ -> ]
Quote: Its green in the Preview. Disabling "CAS GLSL" fixes it.
You are probably using GLSL somewhere,... (yes, apparently in MPC-HC)

Quote:Can Hybrid use .engine TENSSOR to upscale ?
Through the torch- and vs-mlrt-addon, Hybrid does support a bunch of filters which can use TensorRT. (The addons are on my GoogleDrive, which is linked on the download page.)

Quote: RTX Super Resolution being active on MPClassic or Sound Exclusivity BS.
Basically, anything that does exclusively grab the shaders will kill glsl filters.
This will probably also happen if you use the SDR => HDR option in the GPU and your media player settings. Wink

Quote:Yes you are correct speed is back to 40FPS 100%
Happy that is solved then.

Cu Selur

I setup the AI Stuff, First i wanna say this program is crazy good, good job. It is really underRated. I have a question on some of the stuff since they don't really have a description. Sorry for the bombardment of questions. If you have the time i would really appricate the help. Thank you. Just to let you know ive used VSGAN styler00dollar docker and want to make the switch.

1. What does Stepped Resize do ? Does it upsclale using the above model again ? 

2. Why does it say 48x48 to 852x480 in the corner when the input is 720x480 ?

3. Is it possible to get the video to be rescaled first to 852x480 (16x9) then upscaled, because now its upscaling DVD (720x480) which would ruin the Aspect Ratio, since DVD starts at more of a square, what would be the best course of action here if i want a DVD times x4 upscaled, rescale to 852x480 then run Hybrid again ?

4. Why cant i use Tensorrt Cores for RealCugan but i can for ESERGAN whats up with that. I would be really nice if it was present on CUGAN. I understand VSGAN is present but it throws an error when selecting it, Picture below.

5. Is VGAN not using Tensorrt only vsmlrt ?, No duplicate aware in VSGAN, assume since its not vsmlrt VSGAN like i used before ?

5. I would want to apply RealCugan x2 Model twice, which i know is available in VSGAN (Maybe i want VSMLRT instead?) not possible in RealCugan Section?

6. What is Sync Gap in REALCugan and would that reduce quality if switched to VSGAN/vsmlrt ? (I just tried VSMRT with cugan x2 onnix slim fp16 clamp and results seem better)

7. What is the order of the Multi Upscale since i want denoise x3 first then no denoise mode to go through, i assume it goes in order of selection when browsing for it ?

8. You cant load Custom Engine Models ? Because i am not sure what the log is cmd is blank, but from what i learned if u use the same min opt max Shapes as the video upon engine creation it would speed up the upscale process. How would this program know the shapes it needs to make when i give it two x2 models to upscale, unless it makes high res shape engine which would worsen the FPS Encoding, there is a case also with RealCugan x3 where it needs to be the exact shape of the video (same min opt max shapes) otherwise the engine wont create.

9. I guess above problem is solved by resizing the video back to orignal state, but than wouldnt i be losing tons of detail by cramping 2 UPSCLAES in a 480p video lmao.

10. Why are there 2 VSGAN in Frame and Others ? and which model is Cugan Pro Denoise LIght ? x1 Denoise x2 Denoise or x3 Denoise ?

[Image: 7c5kgkA.png]
[Image: FWbyRSf.png]
[Image: 7JFDSDR.png]
Quote:1. What does Stepped Resize do ? Does it upsclale using the above model again ?
Stepped resize is a sub-option for the selected resizer.
What is does is instead of directly resizing to the target resolution XXXX x YYYYY it applied the resizer multiple times / steps.
In the intermediate steps not the target but an intermedia resolution is used as target, for example:
Code:
# Step 1: current: 640x352, target: 1280x704
# Resizing using fmtconv - spline64
clip = core.fmtc.resample(clip=clip, kernel="spline64", w=1280, h=704, interlaced=False, interlacedd=False) # resolution 1280x704 before YUV420P8 after YUV420P16
# Step 2: current: 1280x704, target: 1920x1056
# Resizing using fmtconv - spline64
clip = core.fmtc.resample(clip=clip, kernel="spline64", w=1920, h=1056, interlaced=False, interlacedd=False) # resolution 1920x1056 before YUV420P16 after YUV420P16
Usually it does not make sense to use this on a machine learning based resizer, but for conventional resiziers.

Quote:2. Why does it say 48x48 to 852x480 in the corner when the input is 720x480 ?
48x48 is used while the input resolution is not known (no source loaded), should not happen when a source is loaded
The label indicates the resolution before and after the resizing.
If you for example additionally used "Filtering->Vapoursynth->Misc->Script->Lower res. before resize" it might indicate: "640x352 -> 320x176 -> 1920x1056" First is the base resolution for the resizing (after crop). Second is the resolution adjusted by 'Lower res. before resize' and the last is the target resolution for the resizing.

Quote:3. Is it possible to get the video to be rescaled first to 852x480 (16x9) then upscaled, because now its upscaling DVD (720x480) which would ruin the Aspect Ratio, since DVD starts at more of a square, what would be the best course of action here if i want a DVD times x4 upscaled, rescale to 852x480 then run Hybrid again ?
That makes no sense.
If your source PAR isn't set correctly, let Hybrid know the correct source PAR. (check the Vapoursynth Preview, to see whether you selected the right PAR)
It will then make sure that source par is kept during resizing.
If you want to change the output PAR i.e. to 1:1 let Hybrid know and it will make sure that you can only resize so that no distortion of the image takes place. (or is kept as low as possible, keeping the mod in mind)

Quote:4. Why cant i use Tensorrt Cores for RealCugan but i can for ESERGAN whats up with that. I would be really nice if it was present on CUGAN.
Hybrid will only offer options the used filter offers.
Quote:I understand VSGAN is present but it throws an error when selecting it, Picture below.
That is a bug, I didn't take stepped resizing into account when adding colorfix. => will look at it after work

Quote:5. Is VGAN not using Tensorrt only vsmlrt ?
Hybrid will only and can only offer options the used filter offers.
Quote:, No duplicate aware in VSGAN, assume since its not vsmlrt VSGAN like i used before ?
The duplicate aware option is something I implemented, due to a user requires for VSGAN, which uses https://github.com/Selur/VapoursynthScri...esizing.py. It probably could be extended for VSLMRT, but I haven't written code for it or tested it so far.


Quote:5. I would want to apply RealCugan x2 Model twice, which i know is available in VSGAN (Maybe i want VSMLRT instead?) not possible in RealCugan Section?
You could use stepped resizing, or use the model (assuming it is in a compatible format) in VSGAN or VSMLRT multiple times.
(Usually applying a machine learning multiple times makes no sense.)

Quote:6. What is Sync Gap in REALCugan and would that reduce quality if switched to VSGAN/vsmlrt ? (I just tried VSMRT with cugan x2 onnix slim fp16 clamp and results seem better)
see: https://github.com/Kiyamou/VapourSynth-R...cnn-Vulkan
Yes, that option is not available when using CUDAN through VSMLRT or VSGAN and thus can influence the output quality.

Quote:7. What is the order of the Multi Upscale since i want denoise x3 first then no denoise mode to go through, i assume it goes in order of selection when browsing for it ?
Yes, it should use the order the models are selected.


Quote:8. You cant load Custom Engine Models ? Because i am not sure what the log is cmd is blank, but from what i learned if u use the same min opt max Shapes as the video upon engine creation it would speed up the upscale process. How would this program know the shapes it needs to make when i give it two x2 models to upscale, unless it makes high res shape engine which would worsen the FPS Encoding, there is a case also with RealCugan x3 where it needs to be the exact shape of the video (same min opt max shapes) otherwise the engine wont create.
You can see the code Hybrid creates in the "Vapoursynth Script View". Hybrid does not change shapes parameters or other stuff.

Quote:9. I guess above problem is solved by resizing the video back to orignal state, but than wouldnt i be losing tons of detail by cramping 2 UPSCLAES in a 480p video lmao.
If you apply multiple resizers, Hybrid will simply apply one after the other and add the end make sure that the 'resize output resolution' meats the target resolution.
So if you apply three x2 resizers on a 640x320 source while aiming for 1920x1080, the first would resize to 1280x640, the second to 2560x1280 and the third would resize to 5120x2560 and then at the end Hybrid would resize that to 1920x1080.


Quote:10. Why are there 2 VSGAN in Frame and Others ?
In the Others-tab each time a model is applied, the output is resized to the input resolution before applying the model.
It is mainly meant for 1x_... models.
Quote:and which model is Cugan Pro Denoise LIght ? x1 Denoise x2 Denoise or x3 Denoise ?
No clue. Don't think Hybrid comes with those models.

Cu Selur

Ps.: Like I wrote, I'll look into color fix and stepped resizing after work.
I had a look at the 'found no field' popups, they don't appear here.
So I guess I fixed them in the current dev version.
=> Please try the latest dev (dev versions are in the 'experimental'-folder of my GoogleDrive which is linked on the download page) version and let me know if the problem is still present or already fixed like I suspect.

Cu Selur
Pages: 1 2 3 4 5 6 7 8