I selected the vulkan port of waifu2x and dandere2x insists on using my integrated gpu instead of my dedicated one, I tried adding "-g": 1 or "-gpu-id": 1 to the json (found here https://github.com/nihui/waifu2x-ncnn-vulkan) but that just causes dandere2x to crash

Do you think dandere2x is fast enough to be able to run in real time to upscale 320x240 to 1920x1080 on a modern gpu? If not with scaling, then only the denoiser?

I've been cleaning up Dandere2x lately, and some code review from others would be helfpul.

Helping with improper syntax, poor documentation, etc etc are all appreciated.

hey there,

the video i want to convert has a framerate of 23.976024 but dandere2x doesnt accept that, since it requires a int. using 24 as a framerate, the video looks stuttery

Is it possible to update the ReadMe about the two video projects that uses Waifu2x?

Not a bug, just saying that I would love to see this on Linux. I'd port it myself if I knew how to, but I don't so...

Sometime when I changed IDE's corrections stopped working.

HI, i want to first say thank you for your continuous effort on this project!
It's really good and useful!

Also, i noticed that there is a new tool called anime4k supported by video2x now.
(https://github.com/bloc97/Anime4K)
It's said to be much faster than waifu2x, would you consider supporting this tool as well?
Anyway, just a small suggestion.

Thank you again!

users have been having issues with incorrect settings. I need to create a script to check to verify if they are correct or not.

Premiere can not import MKV files.

Please add the feature to select the conv_models in the gui

Hello, I think you should really consider taking the time to clean up your mess for building, perhaps use cmake to allow platform independent build file, and perhaps provide a documentation on how to use it. If you do, I may be able to help you do your project, and other people might also help.

Have a nice evening !
Antoine Viallon

Render starts, then after finishing the audio just gets stuck and does not proceed further

INFO:context:Starting new dandere2x session INFO:wrappers.waifu2x_caffe:manually upscaling file INFO:wrappers.waifu2x_caffe:['C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\externals\\waifu2x-caffe\\waifu2x-caffe-cui.exe', '-i', 'C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\demo_folder\\date\\inputs\\frame1.jpg', '-p', 'cudnn', '-n', '3', '-s', '2', '--gpu', '0', '-o', 'C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\demo_folder\\date\\merged\\merged_1.jpg'] INFO:context:Starting Threaded Processes.. INFO:wrappers.waifu2x_caffe:waifu2xcaffe session INFO:wrappers.waifu2x_caffe:['C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\externals\\waifu2x-caffe\\waifu2x-caffe-cui.exe', '-i', 'C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\demo_folder\\date\\differences\\', '-p', 'cudnn', '-n', '3', '-s', '2', '--gpu', '0', '-o', 'C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\demo_folder\\date\\upscaled\\'] INFO:dandere2x_core.merge:Upscaling frame 1 INFO:dandere2x_core.difference:('C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\demo_folder\\date\\', 1, 7320, 30) INFO:wrappers.frame:C:\Users\Anon\Downloads\dandere0.7.3\0.7.3\demo_folder\date\upscaled\output_000001.png dne INFO:wrappers.dandere2x_cpp:['C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\externals\\dandere2x_cpp.exe', 'C:\\Users\\Anon\\Downloads\\dandere0.7.3\\0.7.3\\demo_folder\\date\\', '7320', '30', '4', 'n', '1', '.jpg'] INFO:dandere2x_core.difference:waiting on text INFO:dandere2x_core.difference:<wrappers.frame.Frame object at 0x0000020D22EE9A90> INFO:dandere2x_core.dandere2x_utils:C:\Users\Anon\Downloads\dandere0.7.3\0.7.3\demo_folder\date\inversion_data\inversion_1.txtdne, waiting INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.dandere2x_cpp:finished correctly INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319 INFO:wrappers.waifu2x_caffe:Frames remaining before batch: INFO:wrappers.waifu2x_caffe:7319

The longer I leave it the more of the same log statement it outputs.

sometimes inbetween that log it gives me this, then goes back to looping out that statement earlier

INFO:dandere2x_core.dandere2x_utils:C:\Users\Anon\Downloads\dandere0.7.3\0.7.3\demo_folder\date\inversion_data\inversion_1.txtdne, waiting

However this only seems to be the case for some... video files. Possibly broken?

i have python 3.6 installed as well as python 3.7, problem is that 3.6 is the default, and you batch files just use python ...

would there be any way to get compatibility with multiple python installations?

scipy is on legacy code - imread doesnt work

Yeah It deleted the suspended video instead of continuing

hey there,

any chance to get a version that works on linux? i normally run all cudnn stuff on my ubuntu, so it would be nice not having to use windows for this.

i know you said its easier to release for windows, just putting this here to show theres interest

hey there,
i noticed that especialyl for larger videos the harddrive usage is really high. so high in fact that i had to abor because my harddrive was full 👎

there appear to be too many versions of a image saved. for example you could delete the input image and the original image once the upscaled one is done

also, the debug images could be deleted once they are used (maybe have a config setting for this)

i hope theres something that can be done about this

Issue posted here as requested from: https://www.reddit.com/r/Dandere2x/comments/ea2w4m/could_not_upscale_first_file_check_logs_file_to/ffk628f?utm_source=share&utm_medium=web2x

Dandere2x Version 1.7.1

Process:
1 - Load dandere2x.exe as Administrator.
2 - Select video + settings
3 - Start Upscale

Console outputs the message "Could not upscale first file.. check logs file to see what's wrong"

Log Files - https://drive.google.com/open?id=14MP8whh5jwjniNirxEggb2-b45-R5-Mj

As a side note, if I try and run waifu2x-ncnn-vulkan from the externals dir on its own that doesnt work either, but if I download and run https://github.com/f11894/waifu2x-ncnn-vulkan-GUI/ that works just fine.

I tried up-scaling a video and Python threw an exception in thread 'dandere2x_thread' and a type error . I'm not sure what's causing the problem and I changed the settings a few times, but I keep getting the same results. I have Python 3.7.3 installed on my PC.

Hey, few questions / suggestions here

1. Is it still possible to use that same build pattern for the lattest versions ? ( as described in https://github.com/aka-katto/dandere2x/releases/tag/0.1.4.5 )
2. Would you be interested in automating the build process ? (I'd)
3. Do you want me to make it into a small .sh file or anything like that, so that it download deps, and make it all ? (or a makefile even if you want)

.split requires spaces to work .-.

after encoding, the video time increased was 2-13 it became 2-39

[matroska,webm @ 000001b15a50d280] Element at 0x179ffdb ending at 0x17a023e exceeds containing master element ending at 0x179ffcf

Hello, i'm using a 2k display and using scale 150% on windows

However, under this setting, the gui displays abnormally like this

If i change the scaling to 100%, it went back to normal, but the font would be way too small without scaling.

• tell people to delete workspace and stuff if it freezes up

Exception: STATUS_ACCESS_VIOLATION at rip=0010043F0C1
rax=0000000000000000 rbx=0000000000000000 rcx=0000000000000001
rdx=000000A30000008B rsi=00000006000847D0 rdi=0000000000000000
r8 =0000000000000001 r9 =000000000000003C r10=0000000600095D60
r11=00000000000002D0 r12=00000000000004EC r13=0000000600084380
r14=00000006000963A0 r15=0000000000000001
rbp=000000000000C4E0 rsp=00000000FFFFC670
program=C:\Users\bad\Desktop\0.9\externals\dandere2x_cpp.exe, pid 11428, thread main
cs=0033 ds=002B es=002B fs=0053 gs=002B ss=002B
stackdump

rn releases are not explained and ppl are getting errors because they're not installing correctly - provide install instructions!

I set denoise level 0,but the program still using denoise level 3.
I try to delete noise0 model , but when I set denoise level 0 ,the program is still running.
when I delete noise3 model , i set denoise level 0/3,the program is crash.
I'm using waifu2x-ncnn-vulkan.
And I hope you can add some options to choice photo mode or anime mode.waifu2x-caffe has some model which is optimized for photo.

it is currently broken!

users have issues with running program, a debug tool would help

I wanted to test this out as I got video2x working but it just takes too long. I don't need perfect results but want decent with some speed. Tried this out and when I used to gui to set everything up it all seems good but then after no more than 25 frames in it gives me this error. The frame file isn't there in the inputs directory so it's not a matter of not seeing it its a matter of failure to place or extract the frame.
Windows 10 using the waifu2x-caffe type.
Let me know if there is anything else you want to know.

Overview

In anime, sometimes movement is very dramatic, being as low as 2-3 frames per second.

Right now step size is hard coded and preforms decent, but I would imagine better dandere2x performance if stepsize could be determined.

hello again

i briefly looked through the code and noticed you were using multithreading
this is not the best you can do while multithreading certainly is a improvement over normal python for IO tasks like website requests, its better to use multiprocessing for processor hungry tasks! why? cause multithreading isnt paralell!

article about it: https://medium.com/contentsquare-engineering-blog/multithreading-vs-multiprocessing-in-python-ece023ad55a

In Multithreading, threads run in parallel.

❌ FALSE !

-> Actually in a ThreadPool, only one thread is being executed at any given time t.

i hope im not bothering too much

I don't have the log open and it shouldn't be used by another process.

recent call last):
File "gui_driver.py", line 24, in run
File "wrappers\dandere2x_gui_wrapper.py", line 40, in start
File "dandere2x.py", line 156, in run_concurrent
SystemExit: 1

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "gui_driver.py", line 28, in run
File "wrappers\dandere2x_gui_wrapper.py", line 27, in start
File "shutil.py", line 513, in rmtree
File "shutil.py", line 397, in _rmtree_unsafe
File "shutil.py", line 395, in _rmtree_unsafe
PermissionError: [WinError 32] The process cannot access the file because it is being used by another process: 'C:\Users\Home\Downloads\1.5.1_release\workspace\default\dandere2x.log'

Not a real issue with this project but could maybe be interesting.

Vapoursynth (http://www.vapoursynth.com) is a video manipulation software that uses python scripts to create video output (raw frames) that you can then pipe into ffmpeg or whatever encoder you like. After that you would need to mux the audio from the original video since vapoursynth doesn't support audio, but this is pretty easy (ffmpeg -i videowithouadio.mp4 -i videowithaudio.mp4 -map 0:v -map 1:a -c copy finalvideo.mp4 or something)

Using it would (theoretically) have some benefits:

• No need to save intermediate images to the disk and then use those for waifu2x, you could do that directly. The interpolation stuff should also be possible to do with vapoursynth functions (and/or plugins).
• Batching/multi threading.
• General speed, I think the vapoursynth plugins for waifu2x are more optimized for video than vanilla waifu2x at least. So would in combination with dandere2x be even faster.
• Ability to optionally do fancy vapoursynth video manipulation stuff like sharpening, edge cleaning, denoising and more.

There are two vapoursynth plugins that I know of:
https://github.com/HomeOfVapourSynthEvolution/VapourSynth-Waifu2x-caffe and
https://github.com/kice/vs_mxnet .
Not sure if it would be easy (or even possible) to hook dandere2x into these but might be interesting to consider at least.

Some log statements are non-intuitive or untrue

Would it be possible to enable the software to make use of systems that have multiple GPU's / SLI setups?
In the past i've split my video's frames into two bins, then set 2 copies of Waifu2x-caffe going, with each set to render on each seperate GPU. (I have 2 GPUs)

I found there's a fairly straightforward way to do it. waifu2x-caffe-cui.exe just needs the --gpu argument followed by which GPU to use, starting with 0. So a system with 2 graphics cards would have one copy going with the "--gpu 0" argument, and another instance with "--gpu 1"

If you could possibly add this it'd be much appreciated :D. keep up the awesome work!

Hi CardinalPanda, thanks for making this tool!
It's really much faster compared to the lossless one, wonderful work!

I'm currently trying to upscale a 720x480 video with block size 30
The result, however, is in 1440x960, which does not preserve the original scale?
Am i doing something wrong or is there a manual setting for scale that need to be set?

Please help me out, and thanks again for making this awesome tool!

I don't have enough storage on my C: Drive to use save temp files. Can I move the temp files to a different Drive/Location?

not sure whats going on here. Will need to look into

I try to upscale a 720(w)*1280(h) video , the lower side of the video will turn black with some flickering white spots.

ffmpeg -i input.mp4 -metadata:s:v rotate="90" -codec copy output.mp4
When I rotate the video using ffmpeg, the output returns to normal.

could you maybe make a overall progress bar based on the amount of finished images?

it would be really nice to have something that at least shows how many out of the total images are finished