YomiNinja v0.9 RC2 - Download
Added 2024-10-22 08:32:40 +0000 UTCHello everyone! This is the release candidate (preview) for v0.9. I’m currently finalizing the guides for this version, as well as the next poll. This post will be updated soon with more details.
Enhancements:
Added image preprocessing options to the OCR Templates menu.
Added a new screen capture method for up to 10x faster screenshots.
Improved screenshot quality, potentially providing a slight boost to OCR accuracy in some cases.
Browser extensions can now access screenshots, fixing the Yomitan mining issue (you need to reinstall Yomitan).
Initial support for the Wayland window system (tested on Nobara KDE and Ubuntu GNOME).
Added a button next to each hotkey setting to copy the equivalent CLI command.
Added a button to refresh the capture sources list.
Experimental API for controlling the app from third-party applications.
The hotkey for moving and resizing the overlay is now customizable.
Updated the built-in Yomitan extension to version 24.10.7.1.
Updated the built-in 10ten Reader extension to version 1.22.0.
Added a MangaOCR GPU acceleration patch (CUDA).
Bug fixes:
Fixed an issue where the overlay was displayed on the wrong screen in Entire Screen mode on Linux and macOS.
Fixed an issue where fullscreen mode wasn’t detected on Linux.
Fixed incorrect overlay coordinates on Linux Mint.
Access More TTS Voices on Windows
1. Create a system restore point for the C: drive (optional but recommended).
2. Install VOICEVOX (video 0:34 - 1:44).
3. Install VOICEVOX SAPI Adapter (video 1:44 - 2:35).
4. Right-click the apply_tts_fix.bat file and select "Run as administrator".
5. Run VOICEVOX.
6. Run YomiNinja and verify if the new voices are available.
Microsoft Natural Voices:
1. Create a system restore point for the C: drive (optional but recommended).
2. Add natural voices in Narrator.
3. Download NaturalVoiceSAPIAdapter.
4. Run the installer.
5. Check "Include Narrator natural voices".
6. Check "Include Microsoft Edge online voices".
7. Ensure "All supported languages" is selected.
8. Click "Install" (64-bit).
9. Right-click the apply_tts_fix.bat file and select "Run as administrator".
10. Run YomiNinja and verify if the voices are available.
The apply_tts_fix.bat file is a script I created to fix a Chromium (and consequently Electron) bug. The script allows Chromium-based applications to access third-party TTS voices.
VOICEVOX support for macOS and Linux will be available in YomiNinja starting with v0.9.1.
Getting Started With YomiNinja
⚠ Back up your Yomitan settings and dictionary collection before updating!
Password: NzQ5OWZiMmQtMDYzOC00ZTA1LThiZTMtMTcyZjkzNjJjZWNi
Windows:
macOS:
Linux:
Please note that for Linux users, there's a dependency on xdotool.
Comments
Hi, are you still developing this? There haven't been any replies to messages in the discord and no updates for a few months. Hope all is well and you can continue developing this as it is hands down the best OCR.
Casey
2025-01-13 08:08:28 +0000 UTCIs there a way to speed up MangaOCR? It takes 17 seconds for it to scan text. Google lens was much faster, but that no longer works in Yomininja.
Pocket Monster GB
2025-01-04 02:59:11 +0000 UTCThis soft kicks ass.
2DTengu
2024-12-21 10:41:12 +0000 UTCIs Paddleocr no longer an option for "primary ocr engine"? Just wondering, i have been using an older public version (the one that first introduced yomitan support) for the last few months - updated today to 0.9 and OCR seems to be acting erratically, mainly that i can't seem to get it to recognize text and i wonder if this is the reason why, my only options are mangaocr and the two from google.
january
2024-11-11 22:55:00 +0000 UTCHi, my Discord User is #Sorsos I haven't been able to use the new version because the GPU acceleration is not working, may you please help me?
Sonto
2024-10-24 22:05:42 +0000 UTCThanks for the new update, it seems to be broken on Wayland though. Whenever I hovered over text it would get all jittery and wouldn't let me look anything up. It still works fine on X11.
Koyuki
2024-10-23 06:42:33 +0000 UTCHello, thanks for the update. I found out some bugs in this preview, for Now I've rolled back to 0.8.3. It crashes when refreshing extensions and while using paddleOCR for Chinese it almost didn't recognize any character anymore, and I also couldn't install the Zhongwen extension, didn't try for another one. I don't know if I'm the only one though.
kaine nier
2024-10-22 22:07:56 +0000 UTCSame thing is happening to me as well.
Seff Mcc
2024-10-22 19:56:36 +0000 UTCAs a guy who is a listening-main, I feel that I will use the TTS quite extensively. Sometimes It's hard to understand what's written, but if someone reads it out loud, it becomes clear.
Yury Katkov
2024-10-22 11:35:15 +0000 UTCYomitan seems to be bugged in this release. Trying to configure it just gives me a black screen.
VirtualBoost
2024-10-22 11:12:56 +0000 UTC
