Spatializer vs 3D Obj Spatializer

Hello!

I have various events in fmod projects that can swap between using regular spatializer and 3D obj spatializer based on user options.

That generated a lot of mixing problems, as there’s a huge difference between how both sound - there is a big volume and eq difference. (I know that both spatializers operate differently and one is specifically designed to leverage object metadata for object based formats, yet that doesn’t help much in the end)

Also noticed a bugy behaviour that adding just a gain effect with 0dB on it before the chain with 3d object spatializer, somehow makes it suddenly louder (no actual gain added).

Anyone has any ideas how to approach those mix differences and where’s that wierd behaviour with gain jumps coming from?

Currently, I’m monitoring on regular 5.1 surround system withouth any spatial formats;

Also, since 3D object spat. is format based, does that mean it will work very different (and sound) between different formats, say dolby atmos vs windows sonic etc? I’m trying to get unified results between tools and formats, I understand there nuances between different formats, yet I need to be sure that wherever it will go it will at least give me sth that works properly. Would appreciate any help.

Another thing that bothers me is - how the system handles situation when there are to many objects? Platforms have different limits, currently it’s pretty much 128 objects across most of it, but what happens when there will be more? Does it use stealing or is it seperate or format based?

And last question - how does that actually incorporate bus structure and routing in case od 3D obj spatializer? Going by documentation it should completely obey it and go directly to master, yet from my experience it doesn’t really work that way - I can solo group and be able to hear only those 3d objects that go there, also they seem to react properly to snapshots changes controlling different groups;

Would apreciate any tips on that!

Cheers,
Marcin

I have not been able to reproduce this issue. What version of FMOD Studio are you using?

It cerainly sounds different- I am finding that the Object Spatializer plays back about 5dB louder with Winsonic than the FMOD Spatializer does. We don’t really have any control over what the external spatializer hardware/software does with the signal after we hand it over, but there might be more we can do to try to get the level more consistent. I will discuss with the Dev team and see what they think. Otherwise, you could apply some gain reduction yourself and use Platform Exclusion to control the amount. Though I can see how this would be tedious if you are using a lot of object spatializers, and targeting multiple platforms.

The same voice stealing behavior also applies to object spatialized sounds, and in addition, if we run out of object spatializers, we swap in an FMOD Spatializer instead.

The Object Spatializer gets its FMOD_DSP_OBJECTPAN_OUTPUTGAIN value updated in response to mixer changes, so parent Bus and VCA volume adjustments will still be applied. You will notice however, that no DSP effects on the parent bus are applied to Object Spatialized sounds: Object Panner.