Audio dropouts in render

I can't get clean audio exported from Resolve. The 2-minute render has at random intervals brief audio dropouts that I can't explain. They shouldn't be in there at all. I'm getting them randomly during playback also, but in playback I get an audio glitch, back up to play it back, and usually end up with it being clean... implying that it's buggy.

I've tried DNxHR and h.264 in QuickTime containers, using the native as well as nVidia encoders for h.264 and also tried both encoders with mp4 containers. All unusable.

This is in 16.1 beta 3, I'm going to try this in the latest release version also in the hopes that I can get a clean output from it. It's bad. Unusably bad.

The system that I'm using presently is a review model from HP:
Z6, 2x12 core Xeon, 96GB, Quadro RTX 6000.

I'm uploading one of the mp4 renders here:
https://1drv.ms/u/s!AnkPSq9AIf_fgfsIldT ... w?e=UUcjbg

(Might take a couple of hours.)

Audio export issues, which can include crackle, don't appear to be present in 16.0. Somebody suggested reducing the render rate to something other than "maximum" in the Deliver page (for the 16.1 betas), but can't firm or deny.

Reducing the render rate to 100 fps worked for my crackle issues.

I tried 25, that didn't help.

Same problem in v16, release version. I also dropped the render speed to 10fps, still have dropouts.

It's completely unusable.

Try using a different clip. That you're having playback issues even before export suggests a problem with that clip.

Same problem in v16, release version. I also dropped the render speed to 10fps, still have dropouts.

It's completely unusable.

Possible to know which format are the sources and if in color page there are nodes with "special" settings or only basic color corrections?

The video is Redcode, the production audio is 24 bit bwav recorded on a Sound Devices recorder.

There's nothing non-standard, only Fairlight VSTs... This should work fine. This is after all core functionality...

Sent from my H1T1000 using Tapatalk

As a stop-gap, did you try rendering the audio, with the cache option? The VST thing could be the problem.

I just tried that... same result (v16).

Are dropouts still present in deliverables with the Fairlight AFX bypassed?

On the Fairlight page > Timeline > Bounce Mix to Track.
Are dropouts still present selecting the bounced tracks as Output Tracks on the Deliver page.

Please check the other topic on the audio issues. It's very likely the NVidia driver causing the encoding problems. Can you give a spin with the software render or going back earlier NVidia driver?

Please check the other topic on the audio issues. It's very likely the NVidia driver causing the encoding problems. Can you give a spin with the software render or going back earlier NVidia driver?

Maybe. Maybe it's an incompatibility problem between DR and Nvidia drivers. But I render files with PPro and Vegas Pro with zero crackles, using GPU, CPU or any other combination. The problem is in DR and they have to fix it ASAP.

I render files with PPro with zero crackles, using GPU

Premiere Pro doesn't use the GPU for encoding purposes. If hardware encoding is selected, it's using the Intel QuickSync feature for that.

Somebody suggested reducing the render rate to something other than "maximum" in the Deliver page

I can confirm that this does NOT work for me. Still lots of crackles.

I'm currently downloading 16.0 so I can deliver to my client. It's now 10pm and I've been troubleshooting this ******** bug all day long. I'm NOT happy.

Are dropouts still present in deliverables with the Fairlight AFX bypassed?

On the Fairlight page > Timeline > Bounce Mix to Track.
Are dropouts still present selecting the bounced tracks as Output Tracks on the Deliver page.

This is the only way I can get clean audio; bounce mix to track, then use the generated WAV file to encode externally.

Are dropouts still present in deliverables with the Fairlight AFX bypassed?

On the Fairlight page > Timeline > Bounce Mix to Track.
Are dropouts still present selecting the bounced tracks as Output Tracks on the Deliver page.

This is the only way I can get clean audio; bounce mix to track, then use the generated WAV file to encode externally.

Can you explain this more? I am seeing the same issue. But not sure I get your steps to fix.

Sent from my SM-G975U using Tapatalk

This issue still exists in 16.1.2 - had this now for two days. My workaround for deliveray was to export audio as a .wav, then import again and export the whole clip. But this means always duplicating the timeline, importing audio, add audio to timeline, export again and return to your "normal" edit.

To be honest, exporting proper h.264 with video and audio from timeline without any issues shoulb be one of the main jobs that an "pro" editing suite should be capable doing. So everyone @blackmagic. When will this issue going to be fixed? Seems to be present already for couple of month.

This issue is still present and just bit me on a 48 Hour Film Project. Exports sounded totally fine a day before. This feels like a garbage nightmare. The rest of the film holds up so well, watched multiple times on TV no issue but in headphones is completely unuseable. What a damn shame.

This issue remains in Studio ver. 16.2.6. I may have traced the problem to the NVIDIA encoder. The Native render seems to work ok.
Problem: The render leaves gaps in the audio about every 7 seconds. The gaps are about 5 frames wide.
Hardware: I9, 32GB memory, GTX 2070 Super, 1GB Nvme SSD
OS: Windows 10, 1909
Settings:
Video: Captured using VirtualDub2, Cineform compression, PCM audio, converted to deinterlace to 60fps, 57 minutes in length.
Timeline Resolution 640x480 59.94 fps
Playback framerate: 59.94 fps
Video format: NTSC
Deliver Settings:
Quicktime, x.264, Native, 640x480, 59.94fps, Restrict to 8192, un-check "Bypass re-encode when possible"
I deleted all render cache.

The problem occurs when the encoder is NVIDIA. (HOWEVER - rendering a 5 minute In/Out with NVIDIA encodes OK.)

I hope this helps. (And yes, this an "in-retirement" project. I have about 40 VHS and 8mm video tapes to digitize.)

I am having audio dropouts in 17 both with preview and render in same spots. I have tried to fix it both with audio caching on and off to no avail. My clips with problems are all short few second dialog (for animation) WAVs at 32 floating point bit depth/ 9600 sample rate. If it wasn't a render problem I would just think it was a system resource playback problem. The problem came up overnight which is distressing because it could come up the day of a due date. I have a Nvidia card but I am using the free DVR version, which I didn't think had GPU rendering. Regardless, how do I change my audio renderer to CPU?
Are people still having this problem in 17 or only me?

WAVs at 32 floating point bit depth/ 9600 sample rate.

I guess you mean 96K?

Have you tried converting the wav files to 48K to see if that makes a difference?

I was able to get rid of the audio dropouts by setting the render time down to 10.