Community Q&A

Welcome to Audiokinetic’s community-driven Q&A forum. This is the place where Wwise and Strata users help each other out. For direct help from our team, please use the Support Tickets page. To report a bug, use the Bug Report option in the Audiokinetic Launcher. (Note that Bug Reports submitted to the Q&A forum will be rejected. Using our dedicated Bug Report system ensures your report is seen by the right people and has the best chance of being fixed.)

To get the best answers quickly, follow these tips when posting a question:

  • Be Specific: What are you trying to achieve, or what specific issue are you running into?
  • Include Key Details: Include details like your Wwise and game engine versions, operating system, etc.
  • Explain What You've Tried: Let others know what troubleshooting steps you've already taken.
  • Focus on the Facts: Describe the technical facts of your issue. Focusing on the problem helps others find a solution quickly.

High computation cost for vorbis decoding. Is it possible to keep decompressed SFX in memory?

0 votes

Hello, we are having performance issues on Switch with vorbis decoding when many SFXs must be played at once.
Attached screenshots with the most expensive functions.
We would like to know if it's a good idea and if it's possible to decode some SFXs once (the most used and shortest ones, like gunshots) and keep them in memory uncompressed.
We are using our version of UE4.27 and WWise version 2022.1.9.8365.
Thanks!Image1image2

Edit: I've also seen this page https://www.audiokinetic.com/en/courses/wwise301/?source=wwise301&id=Saving_a_Decoded_SoundBank for unity that allows to save decoded soundbanks from the unity editor that will be used instead of the encoded once. Is there a way to do this in UE4?

in General Discussion by Louis P. (100 points)

1 Answer

0 votes
Hello,

The callstacks you quoted exhibit signs of being compiled in Debug (i.e. no optimization). I'd recommend evaluating performance with the Profile or Release configuration. Well, Profile is best during development given you can connect Wwise with it.

Also, these performance numbers need more context: how many sounds are active? Check the Wwise Profiler for the "Physical Voices" counter. A reasonable number on Switch would be around 150. Use Virtual Voices and Limits to manage that. If you are running 1000 voices, sure, decoding cost will be major.

You can obviously use another Codec for very short files (ADPCM or PCM) if you'd like.
by Mathieu J. (Audiokinetic) (7.1k points)
Powered by Question2Answer
...