I’m trying to hook up some new move SFX in a game I’m working on and I’ve been running into a weird behavior in FMOD Studio 2.03.09 with AHDSR.
Whenever I play multiple instances of the same event in fast succession, the release of the first instance affects the new instances and I can’t hear them until the first AHDSR finished. There is no limits or stealing happening and I could confirm it is the release of AHDSR.
I don’t know if this is a bug in this version or if I’m doing something wrong but this bug is being really noticeable whenever I use longer releases for SFXs that repeat often so I’d appreciate any help!
Unfortunately, I haven’t been able to reproduce the issue you’re describing. Can I get you to provide a simple series of steps that consistently reproduces it for you?
Also, a few questions:
- Are you able to reproduce it in both Studio (i..e Sandbox) and UE?
- Do you receive any warnings or errors in the Studio or UE log when the issue occurs?
- What version of UE are you using?
It does not happen inside Sandbox, only inside Unreal Engine and in the built game and there are no warnings or errors (my logging is only on error level, I could turn on extra logging to check it further). Lastly, I’m using Unreal Engine 5.4.
As for steps, it’s simply:
- I use a skill in the game that spawns an instance of an event that has a 700ms fade-out
- I cancel the animation of the skill, trigerring the AHDSR
- I use the same skill again within the 700ms fade-out and it’s volume is set to whatever point of the release the first sound was in. It plays normally if it’s after the 700ms.
If it’s relevant, the instance of the first play is just being explicitly released with the stop call unlike normal one-shots since this one we want to be able to manually stop but I don’t believe both the first and second play are sharing the same instance, at least they shouldn’t be.
Thanks for the extra info and repro steps! I’ll take a look and get back to you on this soon.
Just adding some further information here. I’ve been able to profile the issue and confirm that the sounds really are in two separate instances so they shouldn’t (in theory) be affecting one another.
Sorry for the delayed response!
Thanks for the more detailed repro steps - while I wasn’t able to exactly able to repro the issue, I was able to find an issue with AHDSR that may be related. To confirm this is the cause, I have a few more questions for you:
- Can I get you to elaborate on how the animation skill event is set up? For example, any event logic/transitions/loop regions, whether you’re using any event instruments, where the AHDSR is placed, what clock source the AHDSR is using, etc.
- Did you migrate your project from 2.02 to 2.03? Or was it created in 2.03?
- Are there any AHDSRs or other modulation present downstream of the event in question i.e. in the mixer, or in a parent event if the skill SFX event is being played nested?
Some screenshots or a short video demonstrating how the event and UE animation are set up would be great. You’re welcome to upload them here, or to your FMOD user profile, though your user profile is better if confidentiality is a concern.
Thanks for the answer! Let me answer each of your questions:
- While I haven’t tested if this is happening with more complex events (because they’re not often spammed) the events I can confirm are being affected by this are as simple as they come, the only “complex” thing they might have is a parameter sheet to potentially select different sounds but it also happens with just an action with a multi-instrument. The AHDSRs are using the clock source “Instance (Default)” and in-game they’re triggering using code (not attached to timelines or BPs) and it just creates instance, stops it if needed and releases the instance after stopping.
My AHDSRs are always in the master of the events and there are no instances of AHDSRs downstream, they only pass through a subtle distortion and compression at bus level (as well as reverb sends).
I’m attaching a screenshot of one of the affected events with name of the event and parameters removed.
- Yes, we did migrate the project from 2.02 to 2.03 so the current project was initially created in 2.02 and then we went to 2.03.03, to 2.03.05 and finally after it left EA we updated to 2.03.09.
The events being affected were both created before the transition to 2.03 and after.
I can send you more info if you need but like I said, the setup is as simple as it comes with the exception of the parameter sheet sometimes being used to trigger different sounds.
Unfortunately I still haven’t been able to reproduce the new event’s volume being affected by the AHDSR of the previous. If possible, could I get you to upload a stripped-down UE & FMOD project in which I can reproduce the issue to your FMOD user profile?