Blackmagic Forum

Hiya There.

I've been using timecode recently and I'll never ever go back to syncing by waveforms automatically or manually, I've been sending LTC through my cameras audio directly from my recorder and Resolve works amazing with it and makes my workflow an absolute dream especially with Polywavs files and the way Resolve combines them all into the clips with all tracks separated, I love it so much!

The problem is that resolve will ONLY see my timecode through audio if I plug a cable directly into the camera from my recorder or use a proper timebox box (Tentacle) etc etc

I've tried sending timecode audio via wireless audio packs (Sony UWP-D11 as I don't have a TC box) and while yes it's absolutely not as good/reliable as a proper timecode box etc but it should still work but the problem is no matter my audio level low or high Davinci will not see and update the timecode from the audio as I guess it's not clean enough?

However, TENTACLE TIMECODE TOOL does see the timecode in the audio instantly leading me to assume Davinci Resolves Audio to Timecode is quite simple and is not quite as adaptive way to seek out the timecode in audio? I also hear the GH5 and timecode has issues to the audio prerecord on the camera?

Honestly would LOOOOVE to see this feature Audio to LTC updated with more adaptability and be able to read audio even better.

Tentacle Seeing the TC in the audio
https://imgur.com/xeXoYoU

Resolve won't see the TC in the audio
https://imgur.com/m4ZandU

Would also like to hear your experiences is there anything I can do in resolve to help?

You said "no matter my audio level low or high" - but did you try actually boosting the level in the file itself, ie saving a new copy of the file with boosted audio levels? As opposed to just adjusting the mixer slider in Resolve itself.

I have zero experience of TC, LTC, etc. But that's what I'd try - boost the volume level in the audio track and save it back to a new file, and try that new file. Maybe Resolve's TC detection ignores slider levels and just reads the raw audio, so it would require actually baking in higher audio levels to the file for Resolve to see a hotter signal.

Failing that, if you listen to the audio can you notice anything that seems out of place? Like pops or crackles, noise that doesn't sound like the encoding. I don't know exactly what the encoding sounds like, but I found a YouTube video called "5 minutes of EBU LTC" which I guess is similar, even if not the same format. It sounds a bit like loading from a Commodore 64 / Spectrum data cassette tape (if you're old enough to know what that is!) . It's quite a regular and consistent sound.

So if your audio tracks has anything that seems quite different to that, like sudden clicks, pops, etc, then maybe you could try some very gentle and targetted noise reduction/anti-click filters to remove that.

Likewise if there are variances in the volume level - I'm guessing that in a correct signal, volume should be fairly flat - then you could try smoothing that out.

Then bake all that back to a new file, cross your fingers, and try Resolve again.

Just to emphasise again, I have no experience of this and have absolutely no idea if any of that can help and it's quite possible it could make the signal even less usable - especially the noise reduction part, as it could easily detect parts of the encoded data as 'noise'. But it's what I'd try, in the absence of better information.

Here's that YouTube video I found, for reference. It's what all the kids are listening to these days!

Yes, you do need a clean signal. LTC produces an asymmetrical waveform so any noise in the signal can cause problems. It brings back memories of working with LTC and VITC back in the 1980s and 90s.

TheBloke wrote:You said "no matter my audio level low or high" - but did you try actually boosting the level in the file itself, ie saving a new copy of the file with boosted audio levels? As opposed to just adjusting the mixer slider in Resolve itself.

I have zero experience of TC, LTC, etc. But that's what I'd try - boost the volume level in the audio track and save it back to a new file, and try that new file. Maybe Resolve's TC detection ignores slider levels and just reads the raw audio, so it would require actually baking in higher audio levels to the file for Resolve to see a hotter signal.

Failing that, if you listen to the audio can you notice anything that seems out of place? Like pops or crackles, noise that doesn't sound like the encoding. I don't know exactly what the encoding sounds like, but I found a YouTube video called "5 minutes of EBU LTC" which I guess is similar, even if not the same format. It sounds a bit like loading from a Commodore 64 / Spectrum data cassette tape (if you're old enough to know what that is!) . It's quite a regular and consistent sound.

So if your audio tracks has anything that seems quite different to that, like sudden clicks, pops, etc, then maybe you could try some very gentle and targetted noise reduction/anti-click filters to remove that.

Likewise if there are variances in the volume level - I'm guessing that in a correct signal, volume should be fairly flat - then you could try smoothing that out.

Then bake all that back to a new file, cross your fingers, and try Resolve again.

Just to emphasise again, I have no experience of this and have absolutely no idea if any of that can help and it's quite possible it could make the signal even less usable - especially the noise reduction part, as it could easily detect parts of the encoded data as 'noise'. But it's what I'd try, in the absence of better information.

Here's that YouTube video I found, for reference. It's what all the kids are listening to these days!

Thanks for you're responses but none of the following worked, There are no clicks and Pops and the LTC signal is obviously not perfect there is like a small noisefloor as it was sent through a Sony UWP-D11 pack.

I'm not surprised that resolve did not see the Timecode through a wireless transmitter that wasn't a ultra clean signal, however again Tenticles timecode software found the Audio Timecode instantly and correctly unlike resolve so it would be cool to see resolve update and improve there Audio to LTC because I honestly think it's better than Sync via Audio.

Turns out that Tenticles software also easily allows you to export repacked files with the timecode embedded so with just a 5 more mins I can be ready to sync in Resolve.

I've had all sorts of issues with LTC. For example, if the LTC starts later than the 1st frame, you will get an offset. The actual file metadata timecode may not reflect the LTC time code.

I have solved this by creating a bunch of utilities in a docker container that will read the LTC from an audio file, correct any offset, and apply it to the actual audio file metadata such that when you read the file into Resolve, the timecode is right without having to even interpret the LTC. You can apply the timecode from one audiofile to another in the utility.


If this is of interest, let me know and I will look at making it available,

MadPanic wrote:If this is of interest, let me know and I will look at making it available,

This is exactly what I'm looking for! If you could toss your Dockerfile and related files up on GitHub, that would be great!

MadPanic wrote:I've had all sorts of issues with LTC. For example, if the LTC starts later than the 1st frame, you will get an offset. The actual file metadata timecode may not reflect the LTC time code.

I have solved this by creating a bunch of utilities in a docker container that will read the LTC from an audio file, correct any offset, and apply it to the actual audio file metadata such that when you read the file into Resolve, the timecode is right without having to even interpret the LTC. You can apply the timecode from one audiofile to another in the utility.


If this is of interest, let me know and I will look at making it available,

I'm Having the same needs for this utility tools. Could you please share the dockerfile or github repo/reference. Thanks beforehand

exarva1 wrote:I'm Having the same needs for this utility tools. Could you please share the dockerfile or github repo/reference. Thanks beforehand

Hi,

I have uploaded the docker image to the following:

https://hub.docker.com/r/madpanic/tcutil

I have added some basic documentation. It's been a while since I looked at this project and there is a later version partially done. For now, the version posted needs a mapped in volume containing your audio files. One of these being a ltc track which can be used to extract the timecode and apply it to other recorded tracks. It includes the following utilities:

ltcdump (extracts the timecode)
ffmpeg - Generaly video / audio manupilation.
bwfmetadataedit - For editing metadata.

computerguru wrote:
This is exactly what I'm looking for! If you could toss your Dockerfile and related files up on GitHub, that would be great!

Very sorry for response 3 years late!!