Hi!
I installed the trial version Voice AI.
Win10 22H2, Intel i9-9900K, 32GB RAM, UA Apollo Twin QUAD, Cubase 14.0.10
Trial version crashes Cubase every time when generating in Perpetual mode, but works fine when generating for tokens. I am planning to purchase the plugin but I want to use it mainly in the Perpetual mode. But since this mode never works for me (on all models available in the trial version and on different source tracks), I am now wondering whether it is worth buying.
I do the following: I capture a track, select a model from the three available and press the generate button. After that, the beginning of the process 0% is written and Cubase simply disappears as a process in the system, i.e. a complete crash. Rebooting the system, restarting Cubase does not change anything. At the same time, if you generate vocals for tokens, then there are no errors and everything works as it should. I would like to receive feedback from users and/or developers on the problem I have encountered.
Thanks in advance for your help!
4 comments
Similar happening here. It'll process about 60% then just cycle until an error message appears. Upgraded to Cubase 14, still happens.
Hi guys,
Thanks for reporting the issue! We have reported it to our development and it is under investigation. We have reached out to you in individual support request threads for some additional information.
If anyone else sees the issue, please submit a support request too. Many thanks!
Quick followup on this one:
We have a reason to believe that the issue might be related to having SoundID VoiceAI installed on a different/external hard drive. Can you please try to reinstall it on the main C drive instead:
Let us know if that solves the problem, thank you!
Hi everyone, thanks for standing by!
The problem has been resolved, and we have already pushed it out in a new update. Update to the latest 2.0.9, download here.
That concludes the resolution of this specific bug, but we are not done with the development in this area yet: the cache folder will also be addressed in the near future. The cache folder contains the raw audio processing files, so it can get very sizeable. We will implement a custom install path for it too. This will probably be released with an update in a month or two, although I cannot provide any specific ETA for it right now.
Thanks to everyone involved in troubleshooting this - it is highly appreciated!