In Ubuntu 24.04 I'm getting the following. Cannot install them either since, for ex, libvpx9 is already installed on the system libvpx.so.6 => not found libwebp.so.6 => not found libaom.so.0 => not found libcodec2.so.0.9 => not found libx264.so.155 => not found libx265.so.179 => not found
Aggiornamento, sono riuscito ad ottenere quello che volevo, bordo bianco con sfondo nero delle parole, ma non mi parte il begin time, il progress si ferma al 25%, potrebbe dipendere dal parametro dal parametro Sub Frame Lenght impostato a 2? Perchè per es. , se metto il Sub Frame Lenght a 1 , VideoSubFounder va in crash
Ciao, ho questa situazione, sto usando VSF_intel ultima versione, mi sembra 6.10 o 6.30, ora non ricordo di preciso, ho un video in cui ci sono sottotitoli hardcoded, i sottotitoli sono brevi, ma disposti su due righe, sopra sono bianchi , sotto i sottotitoli risultano gialli, se faccio un run search, vedo solo i sottotitoli bianchi, mentre per quanto riguarda i sottotitoli gialli, si vede solo qualche lettera durante il video, invece effettuando il test, vedo tutto, mi mostra sottotitoli neri su...
Ciao, io ho questa situazione, sto usando VSF_intel ultima versione, mi sembrava 6.10 o 6.30, ora non ricordo di preciso, ho un video, in cui ci sono sottotitoli hardcoded, i sottotitoli sono brevi,ma disposti su due righe, sopra sono bianchi , sotto i sottotitoli sono gialli, se faccio un run search, vedo solo i sottotitoli bianchi, mentre per quanto riguarda i sottotitoli gialli si vede solo qualche lettera durante il video, effettuando il test, vedo tutto, mi mostra sottotitoli neri su sfondo...
if i want insert more color? Tnx
ciao, se volessi inserire più colori?
Hi. Is it possible to change something in settings to improve this? When I want to extract subtitles from video, i usually get more than 3000 pictures, and maybe 500 of them are not junk. Because of that, I have to delete all junk manually before generating .srt subtitle. I noticed that VSF is screenshoting garbage pictures sometimes two in same second of video. All videos are HD in 1080p quality. Here is one screenshot of picture with subtitle and one garbage picture. https://i.postimg.cc/GhyMxLMn/0-13-48-540-0-13-53-579-0025708081415019819201080.jpg...
since it was last updated in 2023, it probably doesn't have support for cuda 12.9 (released in may 2025, or 12.8 from january 2025) and the geforce 5xxx series, i'm guessing that's why it uses cpu only, it needs to be updated to support those too.
since it was last updated in 2023, it probably doesn't have support for cuda 12.9 and the geforce 5xxx series, i'm guessing that's why it uses cpu only.
Why is it not using the GPU even though I’ve already installed all the NVIDIA drivers and CUDA toolkit, and enabled USE CUDA GPU ACCELERATION? If I open the video using FFmpeg, then set FFmpeg HW DEVICES to cuda and USE CUDA GPU ACCELERATION to True, the GPU will be utilized around 20% – 30% and CPU always runs at 100%. If I open the video using OpenCV and set USE CUDA GPU ACCELERATION to True, the GPU is not used at all, and the CPU always runs at 100%. I'm using Windows 10 and CUDA Toolkit 12.9....
Why is it not using the GPU even though I’ve already installed all the NVIDIA drivers and CUDA toolkit, and enabled USE CUDA GPU ACCELERATION? If I open the video using FFmpeg, then set FFmpeg HW DEVICES to cuda and USE CUDA GPU ACCELERATION to True, the GPU will be utilized around 20% – 30%. If I open the video using OpenCV and set USE CUDA GPU ACCELERATION to True, the GPU is not used at all, and the CPU always runs at 100%. I'm using Windows 10 and CUDA Toolkit 12.9. Thank you for your help!
Why is it not using the GPU even though I’ve already installed all the NVIDIA drivers and CUDA toolkit, and enabled USE CUDA GPU ACCELERATION? If I open the video using FFmpeg, then set FFmpeg HW DEVICES to cuda and USE CUDA GPU ACCELERATION to True, the GPU will be utilized around 20% – 30%. If I open the video using OpenCV and set USE CUDA GPU ACCELERATION to True, the GPU is not used at all, and the CPU always runs at 100%. I'm using Windows 10 and CUDA Toolkit 12.9. Thank you for your help!
Why is it not using the GPU even though I’ve already installed all the NVIDIA drivers and CUDA toolkit, and enabled USE CUDA GPU ACCELERATION?
sao mình cài đủ driver nvidia và cuda toolkit rồi mà ko thấy nó dùng gpu vậy bạn, đã bật USE CUDA GPU ACCELERATION
Mình cũng thử cái argument đó rồi mà nó không chạy, bạn có chạy được ko
-uc trong chế độ CLI hình như chỉ chạy với tuỳ chọn -ccti thì phải..
I have a similar problem. As I am using the VSF graphical user interface, I can see my GPU workload is 100%, but for CLI with "-uc" I can not see the same GPU workload when looking at the Task Manager. Is there any possible reasons for this?
I have this parameters to run VSF with cmd: start /wait "" %VIDEOSUBFINDER_PATH% -c -r -ccti -i !VIDEO_FILE! -o !OUTPUT_FILE! -te %CROP_TOP% -be %CROP_BOTTOM% -le %CROP_LEFT% -re %CROP_RIGHT% --use_cuda -uc I have -uc and --use_cuda. But I don't understand why my CUDA isn't activated. In Task Manager, with the GUI I can see the GPU is being utilized, but not for CLI. Does anybody have any idea what the possible cause might be?
"I have seen the software you developed, and it's very good. I also have an idea of what you're trying to do, but..."It seems that the author has stopped developing this software.
It seems that the author has stopped developing this software.
First of all, thank you for adding a program VideoSubFinder to work on command-line Command-line works fine the problem is we don't know when to finish exporting images VideoSubFinderCli.exe -c -r -i "video.mp4" -o "ResultsDir" -te 0.169261 -be 0.00589391 -le 0 -re 1 example: in command-line C:\Release_x64\VideoSubFinderWXW.exe -c -r -i "The Royal Tailor.mp4" -o "ResultsDir" -te 0.169261 -be 0.00589391 -le 0 -re 1 video progress %50 - elapsed time 00h:0m:55s
Hi, I use VSF to extract text from subtitles hardcoded video. Most of the extract images are fine, but maybe 1 out of 10 lines, the letter "E" is missing on the text. When I click on "Test", the text is fine. But when I run search, then Create Cleared Text Images, the output isn't exactly the same. How can I fix it?
Hi, I use VSF to extract text from subtitles hardcoded video. Most of the extract images are fine, but maybe 1 out of 10 lines, the letter "E" is missing on the text. When I click on "Test", the text is fine. [url=https://imgbb.com/][img]https://i.ibb.co/rx2BtPH/Capture-d-cran-2025-01-24-103555.png[/img][/url] But when I run search, then Create Cleared Text Images, the output isn't exactly the same. [url=https://imgbb.com/][img]https://i.ibb.co/wBB650f/Capture-d-cran-2025-01-24-103518.png[/img][/url]...
This video is no longer available. Is there a copy?
https://www.4shared.com/s/fToCaC7tQge I am running on Linux Mint latest version
I also have the same problem with the Linux distro that I use Zorin OS 17.1 , can you help me?
I also use RTX 3050 and also get all CPU resources
Menu File --> Set Priority --> Below nomal or IDLE Tab Settings --> OCR Settings -->Number Of Parallel Tasks;(For Run Search), Number Of Parallel Tasks;(For Create Cleared TXTImages)
Hello! At present, the utilization rate of VSF is 100%. How can I reduce the CPU utilization through software Settings? thank you!
Hello. On v6,the -ccti clear text images option is no longer there. What happened to it? Thanks.
Thank you for the reply. I got the -ccti option from the provided working example of 5.x versions. It works fine on 5.x. Though, that is not the issue. The issue is that the entire option was removed n 6.x. -ccti, /ccti,, and --create_cleared_text_images no longer work. They are not mentioned in the help of 6.x versions as opposed to 5x versions where they are mentioned.
I remember there is only this parameter: /ccti, --create_cleared_text_images Create Cleared Text Images
Hello. On v6,the -cctp clear text images option is no longer there. What happened to it? Thanks.
1) Run Search 2) Create Cleared Text Images 3) Abbyy Fine Reader OCR many thanks
Hi After creating the srt, the srt is showing lots of errors and mistakes, to a point I would have to go through the whole srt and the video and check all the lines thoroughly again. The font is a standard font used on the subtitles, am i doing something wrong is there a setting i need ammeding? is it abby fine ocr that is causing the issue? many thanks Adam
It seems to have something to do with my GPU. After I switch gpu from AMD in-built to gtx4060, it worked, but it doesn't explain it won't affect x86 version. The software should work on CPU by default if I remrember correctly.
x86 is working on the same environment, strangely, x64 version can't. My windows 11 is 64bit. What reason could it be? Evrytime I try to open it, it generates a report.txt. VideoSubFinderWXW.cpp:194 Starting program... VideoSubFinderWXW.cpp:195 CVideoSubFinderApp::Initialize... VideoSubFinderWXW.cpp:197 CVideoSubFinderApp::Initial LoadSettings... MainFrm.cpp:1252 CMainFrame::LoadSettings(): starting... MainFrm.cpp:1254 CMainFrame::LoadSettings(): ReadSettings(g_GeneralSettingsFileName, g_general_settings)......
I found this video.. In the 5:13 time segment there is a setting for the program, try applying it https://youtu.be/_VUBY61sccg?t=313
What settings are best to get images with the least amount of repetition, no blank images, and no missing images with subtitles? My subtitle in white...
Please elaborate how should I do it exactly. Thank you
Try reducing the number of running threads in the settings
Is it just me or these task is CPU intensive, even I have CUDA enabled? My CPU always goes around 90C degree. Is there a way to use more GPU (I have RTX 3050 Laptop) and less with CPU.
What are the steps you take?
Hi First of all, thank you for coming up with a great piece of software. Compared to a friend of mine who is doing the same process, same video for some reason when when i get to the ocr of the video, it comes up with an extra 1000 files. Any ideas why or if im doing something wrong? Thanks in advance. Adam
is there good settings for 6.10 release to get better subtitle detection? for now like 50% of RGBImages are images with out subtitles. And one thing looking one by one 1000+ images and delete all those that are empty but if its more its pain in the ass.
Is it possible to add a progress bar in CLI mode?
There will be only some sentences, more than 90% working perfectly.
OCR in Thai Lang Sometime vowels not show
Hi I've read the Thread "With Background" but maybe the posts are old and something has changed on the latest versions, and I haven't found any answers. Here's my problem: on two videos the background color was very light (Icelandic film with snow and fog), this led VideoFinder to extract Subs with many missing letters or with signs instead of words. In these cases is there any configuration change that can be done to highlight the text better? Thank you...
VSF 6.10 ignores subs near the end of the video
In command line, you can only select the range of the image region from 0 to 1.0 : -te 0.5 -be 0.1 , not like in the application, it would be nice to support more accurate selected region in the cli.
Is there a method to enable CUDA support for running the search subtitle, or is there a specific configuration or command line option? If not, are there any plans to incorporate this feature in the future? Thank you.
Hey there, i was trying to play around with a script that interact with the CLI of the program, but the instruction are kinda confusing, first, does this program uses cuda when it searches for the subtitle? i tried -uc or /uc but still uses my cpu, or only when it creates text from it? Secondly, how can i run search and create folder for RGBImages only in command line, seems like it creates a lot of folders that i don't need are they necessary ? Lastly what's the value of <str>Thank you.
Hey there, i was trying to play around with a script that interact with the CLI of the program, but the instruction are kinda confusing, first, does this program uses cuda when it searches for the subtitle? i tried -uc or /uc but still uses my cpu, or only when it creates text from it? Secondly, how can i run search and create folder for RGBImages only in command line, seems like it creates a lot of folders that i don't need are they necessary ? Lastly what's the value of <str> Thank you.</str>
Hey there, i was trying to play around with a script that interact with the CLI of the program, but the instruction are kinda confusing, first, does this program uses cuda when it searches for the subtitle? i tried -uc or /uc but still uses my cpu, or only when it creates text from it? Secondly, how can i run search and create folder for RGBImages only in command line, seems like it creates a lot of folders that i don't need are they necessary ? Lastly what's the value of <str></str> Thank you.
Hey there, i was trying to play around with a script that interact with the CLI of the program, but the instruction are kinda confusing, first, does this program uses cuda when it searches for the subtitle? i tried -uc or /uc but still uses my cpu, or only when it creates text from it? Secondly, how can i run search and create folder for RGBImages only in command line, seems like it creates a lot of folders that i don't need are they necessary ? Thank you.
I use Real-ESRGAN 2x on the RGBImages folder. (Only at resolutions lower than 720p) After that I generate TXTImages and send them to google drive.
I use Real-ESRGAN 2x on the RGBImages folder. After that I generate TXTImages and send them to google drive.
Hello, regarding "I usually do this in anime, so when the resolution is too low I use Real-ESRGAN to enhance them before uploading them to google with the python script made by Abu3safeer" Do you use Real-ESRGAN on RGBImages or on TXTImages? (it looks that you upload to google "Restored RGBImages" or TXTImages generated on "Restored RGBImages")
It is possible, because the program accepts the command line, I made a simple .bat where it loops in the folder with the videos and generates the images individually and then uses the python script that does the google OCR, when it finishes, it cleans everything and goes to the next video. I usually do this in anime, so when the resolution is too low I use Real-ESRGAN to enhance them before uploading them to google with the python script made by Abu3safeer. I've done dozens of batch OCRs, never had...
And a tip for combining all ‘part’ Google Docs. I tried the Google application ‘PDF Mergy’. It can easily combine all the Google Docs into one PDF in which you just press CTRL+A to copy the entire subtitle and insert it into ‘join_txt_results.txt’.
Thanks, will try your alternative solution. Did a small test with the following picture. Most text are aligned to the left, but a few subs are centered. It seems that GoogleDocs did a fine job on those. But wouldn’t the best solution be to crop all subtitle pictures and align them to either left or right depending on language?
Hello, agree, will try to add this feature, but currently there is alternative solution, just checked on left aligned video, you can set through GUI "Join Images Split Line Text" to something like (by addding multiple space to the end): "[sub_id] " As result you will get:
I have a set of subtitles where most of text is left aligned. Creating a ImagesJoined picture and upload it to GoogleDrive and do OCR with GoogleDoc does not give a good result. I then tried to move the [%%] numbers to the left as well. And the GoogleDocs does a much better job. See the 2 attached files and try to upload them to Google Drive. Maybe adding an option to align the numbers to left, center or right would help.
Also try to check by defining number of threads for run search == 1 (or for clear images == 1). The issue can be related to not enough memory.
Hello, not sure but possibly it is blocked by Antivirus which sometimes happen at first run (Avast). Try to use this version (with extended logging) with opening video by both OpenCV and FFMPEG (ffmpeg has more logging in case of video open) https://drive.google.com/file/d/1ytJN986zIM8KrouVzsnoR2MMZ2lV3_5C/view?usp=share_link and provide report.log file and screen short for stuck. Does it stuck at the beginning of "Run Search" or it passed part of video and stuck at some moment or something else...
Hello, not sure but possibly it is blocked by Antivirus which sometimes happen at first run (Avast). Try to use this version (with extended logging) https://drive.google.com/file/d/1ytJN986zIM8KrouVzsnoR2MMZ2lV3_5C/view?usp=share_link and provide report.log file and screen short for stuck. Does it stuck at the beginning of "Run Search" or it passed part of video and stuck at some moment or something else?
Hello, not sure but possibly it is blocked by Antivirus which sometimes happen at first run (Avast). Btw how stuck is looked? Can you provide report.log + error.log (if it was generated) + screen short?
My computer processor is AMD Ryzen 5 PRO 2500U w, it can't be processed in videosubfinder version 6.10 using opencv, after opening the video via opencv and setting the region, it gets stuck when I click start I don't know what is the reason? The system version of windows 11 already has the runtime library collection package installed. The problem mainly occurs with version 6.0, but not with version 5.
Ability to choose subtitle color like SubRip
Already supported.
Thank you, I have opened VSF ^^
Okay, try to use this package: https://drive.google.com/file/d/1-VgdOQM-CoGaBwqVaa_jkWSQflztDmI4/view?usp=share_link
libvidstab.so.1.1 => not found This is what displays in the log.txt file
Hi Kobayashi, Can you please run LD_LIBRARY_PATH=. ldd ./VideoSubFinderWXW | grep "not found" >log.txt And provide result: log.txt According your screen some dependencies now are missed after update.
Hi Simeon Kosnitsky, I am having an error at Endeavouros, after I update the ingredients in Endeavouros to the latest version, they include Linux header, after updating I cannot open VSF, although I am still using it before.
New 6.10 version was released. Some Notes: VideoSubFinderWXW_intel.exe mostly faster then VideoSubFinderWXW.exe on i9900KF In case of 6.10: run_search: intel: real 1m9.142s msvs: real 1m22.828s clear_txt_images: intel: real 3m29.468s msvs: real 3m44.479s Also 6.10 working slower then 6.00 due to fix for small subtitles support [5f1730] : run_search: 6.10 Intel: real 1m9.142s 6.00 Intel: real 1m8.261s clear_txt_images: 6.10 Intel: real 3m29.468s 6.00 Intel: real 3m9.085s Main changes: *) Added join...
VSF 6.00 won't load positions setting from cfg file
Fixed in 6.10 version
Made fixes for Linux support
Updated -h support on Windows through CLI
Made fixes for Linux support
Works, thx.
Thank you! I will try it!
Fixed. Try to use this version: https://drive.google.com/file/d/1fTyzIbgbirV2u8Pe0-cPytd8Abs01J1n/view?usp=share_link
Reproduced, made minor fix but after it can be a little more false detections, also file names naming format was changed (not compatible with previous version generated), try to use this version. https://drive.google.com/file/d/1fTyzIbgbirV2u8Pe0-cPytd8Abs01J1n/view?usp=share_link
Made fix for CLI support
Made fix for Load Settings support
Made minor fix for support small subtitles like 'I ...' (but after can be more false detections)
Updated chn localization provided by Shijie Zhang
Fixed. Try to use this version: https://drive.google.com/file/d/1nbif21-PhsfpPKTtMj9a81mRiyQqr_rr/view?usp=share_link
Reproduced, made minor fix but after it can be a little more false detections, also file names naming format was changed (not compatible with previous version generated), try to use this version. https://drive.google.com/file/d/1nbif21-PhsfpPKTtMj9a81mRiyQqr_rr/view?usp=share_link
Thanks. Reproduced, will fix today.