I love talking about games in general, so I thought why am I not sharing my vision with others? Obviously, I don't expect to get millions of views, but as I've said talking about games, design, art etc. I love this stuff, so as a hobby I've decided to use my Youtube channel a bit more actively. If you love these sort of stuff, check out this video below and give me your thoughts!

Also a little update, there will be a "Beta" testing soon for my game, so stay tuned for that. I am almost there!

See you soon cultist friend!

Seven new fully supported languages have been added to Orbyss, in time for the release in September 2025!

Total of 12 languages:

• English

• French

• German

• Spanish

• Italian

• \[NEW] Portuguese (Brazil)

• \[NEW] Portuguese (Portugal)

• \[NEW] Polish

• \[NEW] Russian

• \[NEW] Chinese (Simplified)

• \[NEW] Japanese

• \[NEW] Korean

Already available in the demo version.

Have fun!

I love talking about games in general, so I thought why am I not sharing my vision with others? Obviously, I don't expect to get millions of views, but as I've said talking about games, design, art etc. I love this stuff, so as a hobby I've decided to use my Youtube channel a bit more actively. If you love these sort of stuff, check out this video below and give me your thoughts!

Also a little update, there will be a "Beta" testing soon for my game, so stay tuned for that. I am almost there!

See you soon cultist friend!

• New Content: "Something stirs within the Abyss" - Plabs content creator quote

• Large animated Titan K. Mark award effect added

• Hidden Steam Achievement added

• Slight effect adjustments for game balance

• Performance optimizations

I've done it. It's out there. For you, to play and enjoy. My horse racing roguelike card drafting game.

And it works on SteamDeck, as the above video shows. Wow! Just wow!

What does the demo offers?

Racing with 3 or 4 horses, about 50 charms that modify the game mechanics, and unlimited number of races. If you opt-in for telemetry, I will be able to analyze the data and further improve the game, but that's optional - and all data is always anonymous! You can in fact see the code for it on GitHub if you are a programmer or fellow indie game developer.

Enjoy the game, and I'm always open to feedback and suggestions for it. And if you find any bugs, please also don't hesitate to reach out through the Discussions hub.

Cheers,

Mateusz

• Fixed a small problem with rolling enhanced materials from lootboxes.
  
  
• Fixed a problem with finding artifacts sometimes not unlocking anything.
  
  
• Soul locking artifacts now save, so you don't have to make sure they are selected before each reincarnation.
  
  
• Should no longer be locked and enabled after reincarnation
  
  
• Tribulation window should now update after reincarnation
  
  
• Soul forging should now give correct xp during offline forging
  
  
• Fixed Changing physique not updating physique icon in adventure panel
  
  
• XP pills now have double base Xp value and higher scaling for tiers, will listen to more feedback regarding it.
  
  
• Fixed a problem where afk cultivating wasn't working.
  
  
• Fixed some bugs with unarmed combat attack speed bonus.
  
  
• There is now a buff icon for stacking attackspeed.
  
  
• Adjusted modifiers for defeating tribulations for soul power, is now 20% 60% 100% and 150% base rate.
  
  
• Buffed droprate for spirit saplings
  
  
• Fixed mining soul activity. Now you need to select a node and press start mining, to not interrupt any other main activity. Will likely add some more clarity to what is selected in the near future.
  
  
• Added information for spirit tree effect in stat page

Hello everyone!

This update is one of the biggest evolutions in LiveCap history. We've massively improved both accuracy and speed at the same time!

Remember those moments when keyboard typing sounds were mistaken for voice? Or when it missed the beginning of your speech? And that frustration when subtitles took forever to appear even after you spoke...remember?

Well, we fixed it all. 😎

We completely rebuilt LiveCap's voice detection method. Instead of simply looking for silence, it now uses actual Voice Activity Detection - advanced AI technology that can distinguish your voice from everything else.

The result? False detections dropped from 67% to less than 5%. That's not a typo. It really got that much better.

We switched the default Japanese engine to NVIDIA's latest model. Faster, more accurate, just better all around. It's that simple.

We added a real-time noise gate that filters out background noise before it reaches speech recognition. Mechanical keyboards and air conditioner sounds won't bother you anymore.

• New: MistralAI Voxtral - Supports 8 languages!

• New: NVIDIA Parakeet Japanese Edition - The best experience for our Japanese users

While speaking, transcription updates in real-time every 2 seconds. No more waiting until you finish talking!

Implemented a new silence detection system that completes transcription the moment you stop speaking. No more waiting after "umm..." 40% faster processing achieved.

Automatically skips long silent sections, intelligently removes short noises. Efficiently processes only what's needed, zero wasted time.

With our newly developed FileVADProcessor, even long videos are processed at super speed. 1-hour stream archive? Done before you finish making coffee. Really 190x faster.

(1 minute might be theoretical though...)

Now you can process multiple video files in parallel. Got 10 stream archives? Process them all at once. Uses your CPU to the fullest.

LiveCap now displays a beautiful splash screen during startup. No more wondering if it's loading - you can see exactly what's happening!

Whisper stopped complaining about CUDA. Finally grew up.

Now fully browser-based and actually readable. The previous one was...well, let's forget about it.

Fixed the mystery phenomenon where the same microphone appeared 5 times. Really, Windows...what were you doing?

You can now display the original text along with translations!

Japanese audio → English subtitles + Japanese original text displayed simultaneously in VRChat. Perfect for multilingual streaming and language learning streams.

Completely fixed the LLVM error that occurred in some environments. LiveCap won't suddenly disappear anymore. It wasn't a poltergeist after all.

ReazonSpeech no longer breaks with long audio. Stream archive transcription is now worry-free.

Audio device selection is now actually remembered. No more reselecting every time.

Switching from VRChat to OBS subtitles now works without restart. Safe to switch during streaming.

Translation labels now display correctly. No more garbled text.

• \[Common] Optimized memory usage - No more memory hogging

• \[Common] Improved CPU efficiency - Now kinder to your CPU

• \[File] Large files run smoothly - Heavy files feel light

• \[Stream] Translation flickering eliminated - Easy on the eyes

• \[Common] Instant engine switching - Zero wait time

• \[Common] More intuitive settings screen - You won't get lost

Remember when I said LiveCap was good? With this update, both accuracy and speed are really good.

Thanks to the VAD State Machine (new voice detection system) and FileVADProcessor (ultra-fast file processing engine), LiveCap is like magic. Try talking, typing on your keyboard, then talking again. See how it tells the difference. And try processing a 1-hour video. It'll be done in 19 seconds.

This is our masterpiece. 🎉

---

Ready to experience the new LiveCap? Update now and let us know what you think!! We're waiting for your reviews and messages on Discord!

P.S. - The in-app documentation accessible from the help menu is still being updated and may contain some outdated information. We're working hard on it! If something seems off, trust the app's actual behavior over the documentation. 📚

P.P.S. - File mode parallel processing performance varies depending on your PC's CPU cores, but it'll definitely be faster on any PC! 💨