Technical problems and automatic creation of subtitles with speech recognition

[cutegitcat]

Dear URUWorks/TeroSubtitler development team,

I would like to provide feedback on software version 1.01.7. Unfortunately, your software does not work with the operating system of either my Apple laptop or the Ubuntu-Linux laptop. The portable version of your software version 1.01.7 works best under Windows10 with some limitations.

Here I summarise where and what the problem is:

Apple Laptop – macOS Ventura, Version 13.6.5
The installation wizard cannot download the functions of the additional tool: Whisper.cpp. After waiting a few minutes, no download process is displayed. With Faster-Whisper, on the other hand, the tools and the language pack for speech recognition can be downloaded without any problems.
I added an mp4 video and ran it with Faster-Whisper. Unfortunately, the software was automatically interrupted after a few seconds. The automatic transcription did not work. However, it was tried several times, but always with the same result. The software cannot be terminated correctly, but must be cancelled via 'Abort'. There is an "Access violation". This software version 1.01.7 is therefore unusable on Apple laptops.

Ubuntu-Linux Laptop – Jammy Jellyfish, Version 22.04.4 LTS
After unpacking the ZIP file and clicking on 'Tero', an installation wizard appears. However, there are no download buttons to download any of the elements mentioned. Instead, you have to go to the settings. Additional tools can be downloaded manually in the left-hand navigation bar under 'Tools'. Unfortunately, there are no instructions on which correct file to download (due to compatibility must fit) and where to save it. The software version 1.01.7 is not usable on an Ubuntu-Linux laptop.

Windows10
The portable version of the software version 1.01.7 works best on Windows10. One of the biggest major shortcomings after automatic transcription with Whisperer can be noted: Subtitles do not fade in and out correctly according to the spoken text. In many video media, subtitles are displayed by default when the spoken voice begins (time code: in/on) and fade out at the end of the spoken voice (time code: out/off).

If all the shortcomings and gaps mentioned above could be eliminated, I would recommend it to everyone, both in their free time and during their studies or further education and at work.

I would like to thank all those involved in the development of the software and wish them every success.