Voiceovers - radio filter advice

[baltic_dragon]

Hey guys,

 

I've received questions recently about the filters I use for the campaigns to get the best effects for VOs, so I thought I'd share it with everyone - hope that you'll find it useful for your missions / campaigns. Or if you have a better way of making radio transmissions sound more accurate, please share!

 

For VOs I use Audacity, which is an excellent and free program.

 

The steps I usually follow for the VOs (remember that you can set it up as a "chain"):

 

1) Change track from Stereo to Mono (where applicable)

2) Run High Pass Filter (on 24 dB)

3) Run Low Pass Filter (on 12 dB)

4) Run High Pass Filter again

5) Run Low Pass Filter again

6) Amplify (to 0.0, maximum possible without clipping)

7) Set the volume slider on the left to +12 or +15 dB (important as VOs in game are more quiet than when you listen to them in the Audacity)

8) add (as a separate track) a background white noise and set it - depending on the need - to -5 dB to +5 dB.

9) add ending click to the main track

 

I am attaching some background / click files I usually use for the VOs.

 

Hope this helps!

Sounds.zip

[beppe_goodoldrebel]

I am also intrested in how you handle radio transmissions within the Mission Editor.

 

I suppose you use some sort of script to "give separation" between VOs , am I right?

Can you possibly give us some examples?

[Trubshaw]

Brilliant Baltic...the master at work..!

 

Thanks for sharing..

[baltic_dragon]

I am also intrested in how you handle radio transmissions within the Mission Editor.

 

I suppose you use some sort of script to "give separation" between VOs , am I right?

Can you possibly give us some examples?

 

You add them to a unit as triggered action (Radio transmission) and set the frequency as WP0 action. Then you simply add them as triggers by AI Task Push. Remember to add audio file to each action though, without it it won't work.

 

Also, it is better to add them as triggered actions to a separate units set up directly for this. So you set up one unit for 252.3 UHF, another one for 33.4 FM etc and add all the transmission on the given frequency to them. That makes it more AI-retard proof (sometimes AI did stupid things with too many AI task push actions) and the VOs do not cut each other out.

 

Hope this helps, if you have any further questions hit me up on FB, Discord or with a PM.

[Rudel_chw]

I've received questions recently about the filters I use for the campaigns to get the best effects for VOs, so I thought I'd share it with everyone - hope that you'll find it useful for your missions / campaigns.

 

 

It's really very kind of you to be sharing these, tough voice recording is so much work that I doubt I will ever use that on my own missions :D ... but will bookmark the thread, just in case :)

 

 

Thanks a lot

[beppe_goodoldrebel]

You add them to a unit as triggered action (Radio transmission) and set the frequency as WP0 action. Then you simply add them as triggers by AI Task Push. Remember to add audio file to each action though, without it it won't work.

 

Also, it is better to add them as triggered actions to a separate units set up directly for this. So you set up one unit for 252.3 UHF, another one for 33.4 FM etc and add all the transmission on the given frequency to them. That makes it more AI-retard proof (sometimes AI did stupid things with too many AI task push actions) and the VOs do not cut each other out.

 

Hope this helps, if you have any further questions hit me up on FB, Discord or with a PM.

 

Oh cool! Thank you to share all this info! :thumbup:

[Hardcard]

2) Run High Pass Filter (on 24 dB)

3) Run Low Pass Filter (on 12 dB)

4) Run High Pass Filter again

5) Run Low Pass Filter again

 

I appreciate the fact that you're trying to help people, however...

 

You don't need to filter voice tracks 4 times in order to achieve a radio-like sound.

Just run an FFT filter once (kill all frequencies except those in the 400 - 4000Hz range), then mix that with a static noise track on the background.

 

And yes, you're correct, in order to hear the voice tracks properly in-game, it's a good idea to boost dB (while capping max amplitude, in order to avoid unwanted saturation)

 

EDIT: I've attached a zip file containing a couple of radio message examples from my training mission project (yes, I know, she can be quite cold :cry:)

Cold_Flight_Instructor.rar

Edited December 21, 2018 by Hardcard

[snowsniper]

also in audacity you can manage some sort of script to redo exactly the same filter sequence.

one for Tower, one for player, one for wingman .... etc ... so it's easy once recorded with real voice or TTS like ttsreader, record with audicity apply filter or script, then save as .ogg with minimum size.

[SGT Coyle]

Wow! Thanks for sharing. I've didled with filters a little to achieve a radio quality, but not as good as yours.

 

 

Thanks

[baltic_dragon]

I appreciate the fact that you're trying to help people, however...

 

You don't need to filter voice tracks 4 times in order to achieve a radio-like sound.

Just run an FFT filter once (kill all frequencies except those in the 400 - 4000Hz range), then mix that with a static noise track on the background.

 

And yes, you're correct, in order to hear the voice tracks properly in-game, it's a good idea to boost dB (while capping max amplitude, in order to avoid unwanted saturation)

 

EDIT: I've attached a zip file containing a couple of radio message examples from my training mission project (yes, I know, she can be quite cold :cry:)

 

Thanks happy to check it! I found the filter you refer to (you are talking about the normal radio setting in the equaliser?) not good enough, but I’m curious how your sounds sound (err....).

[Habu_69]

I use the shareware program Goldwave. It has an extensive array of sound processing features and has batch processing that I have set up to perform all the filtering, mixing and amplification in one step.

[Andy1966]

Thank you Baltic Dragon, I am just getting into learning to do VO's, and these tips and files are right on time.

[beppe_goodoldrebel]

You add them to a unit as triggered action (Radio transmission) and set the frequency as WP0 action. Then you simply add them as triggers by AI Task Push. Remember to add audio file to each action though, without it it won't work.

 

Also, it is better to add them as triggered actions to a separate units set up directly for this. So you set up one unit for 252.3 UHF, another one for 33.4 FM etc and add all the transmission on the given frequency to them. That makes it more AI-retard proof (sometimes AI did stupid things with too many AI task push actions) and the VOs do not cut each other out.

 

Hope this helps, if you have any further questions hit me up on FB, Discord or with a PM.

 

It's all good but I still have VOs cut.

 

This is what I've done , a trigger that push all Transmit Message of the radio chatter.

 

Even tweaking the duration (which I suppose is for subtitles) they all start at the same time... I dont quite understand how they supposed to be handled.

Edited December 24, 2018 by beppe_goodoldrebel

[feefifofum]

They all play at once because you're telling them all to play at once. This needs to be three separate triggers.

 

FLAG IS TRUE (4) -> push message 1.

TIME SINCE FLAG (4, whatever) -> push message 2.

TIME SINCE FLAG (4, whatever) -> push message 3

[beppe_goodoldrebel]

They all play at once because you're telling them all to play at once. This needs to be three separate triggers.

 

FLAG IS TRUE (4) -> push message 1.

TIME SINCE FLAG (4, whatever) -> push message 2.

TIME SINCE FLAG (4, whatever) -> push message 3

 

That sucks... a lot! :helpsmilie:

[feefifofum]

How so?

That does exactly what you want, and the file length is displayed when you upload so it's easy to know the proper timing.

[beppe_goodoldrebel]

How so?

That does exactly what you want, and the file length is displayed when you upload so it's easy to know the proper timing.

 

Having X new triggered action every X VOs is very demanding.

It works , yes , but it's very long to do it.

[Sedlo]

Yep, it is a lot of work! :-) But very rewarding when you finish something that is awesome.

 

 

Having X new triggered action every X VOs is very demanding.

It works , yes , but it's very long to do it.

[FragBum]

I would also suggest adding some limiting and some gain to bring the audio well into the limiter. But don't overdo the compression radio systems run voice processing and limiting to improve intelligibility and signal to noise levels and bring lower levels up to a prescribed threshold level.

 

Nominal audio BW 300Hz to 3000Hz SSB maybe to 4000HZ AM/FM

 

Don't forget DSP can easily brick wall audio response which will tend to sound harsh even grating to most people especially using HiFi audio systems so maybe allow a bit of slope on either end of the bandpass filter for the bandwidth filtering.

 

Instead of "clicks" in your DAW make an envelope and use that to modulate bandwidth filtered noise to replicate carrier/signal loss at the end of transmission and the squelch kicking in in the receiver. You could use a shorter envelope for the break squelch at the beginning of transmission as well.

 

I take it mixing "white noise" back in is to give the effect of system noise, this usually should be somewhat down in level compared to the voice over level in radios systems unless your representing marginal signal conditions.

 

Foxtrot Bravo out, schhht. :)

[beppe_goodoldrebel]

Yep, it is a lot of work! :-) But very rewarding when you finish something that is awesome.

 

:thumbup:

 

 

 

I love how TTS voices sound more real after applying some effect.

 

It will be a long and hard road but hopefully it will good in the end.

[beppe_goodoldrebel]

Dont know if that happens also to you guys but sometimes, even with proper separation between "trasmit message" commands, some VOs are cutted.

 

How much separation (in seconds) do you normally use?

Maybe "stop transmission" command before puch message could help?