Previous Isaac versions had support for audio, including a neat text to speech synthesizer. Just noticed that this is missing from the latest Isaac.
It doesn’t seem like anyone outside of Nvidia could migrate the feature from the older Isaac to the newer Isaac, because some libraries for which the source code isn’t available need to be recompiled with Jetpack 4.5 support.
Not asking for new features, but really hoping Nvidia could issue an update for Isaac with audio and Jetpack 4.6 support.
If that’s a longer term project, how about making the source for the audio libraries available so that we can declare it as an external feature that is pulled in and compiled by bazel automatically, like the RealSense camera library.
Thanks so much
Most of the audio packages were removed in Isaac SDK 2020.1NX because they were not being maintained in JetPack 4.4 at the time. Some of these audio packages have been incorporated into Riva. We’ll consider resurrecting these audio gems for future releases and our other SDKs.
OK even if Jetpack 4.51 support could be added it would be nice and we would appreciate that. Overall audio is important for a robot, and we have lots of robots running on Isaac SDK 2021.1 with Jetpack 4.51
EDIT: You mentioned Riva but not quite sure how that would be added to a robot (like Kaya or Carter) running Isaac SDK. Would Riva require a separate Jetson processor ? On Carter that would use more power and reduce run time and on a small robot like Kaya there’s no room for a second Jetson.
Or would it require an always-on Wifi or cellular internet connection to a cloud server ? That would not work for our applications, which are designed to be autonomous with only occasional connection to the internet.