Archive:Kdenlive/Manual/Useful Information/Surround Sound

From KDE UserBase Wiki

Under Construction

This is a new page, currently under construction!


Editing Surround Sound with Kdenlive

At the time of writing, Kdenlive can only handle stereo (two channel) audio tracks. There is limited support for more than two channels, but in order to edit and create surround sound, some manual steps including external tools are required.

This guide is using a six-channel 5.1 surround sound as example.

External Tools Used Here

  • Audacity - Free Audio Editor and Recorder
  • avconv - A Video and Audio Converter

Note

Kdenlive uses ffmpeg, while on (k)ubuntu, ffmpeg is deprecated and avconv is used instead. So these (and possibly other) distributions already have avconv installed.


Creating New Surround Sound

Kdenlive currently only supports rendering a project with stereo audio. It is not possible to render to more audio channels or to explicitly map audio tracks to channels in the rendered audio.

There are however some relatively simple ways to manually work around this issue, this guide describes just one:

  • Create and edit surround sound with an external application like Audacity


Note

More advanced features such as surround panning (i.e. let a sound move from rear to front) are beyond the capabilities of Kdenlive as well as Audacity - but it is possible to create similar effects manually.


Create and Edit Surround Sound with Audacity

This guide describes how to create a simple 5.1 surround sound with the following example:

  • Some original field recording from the front (stereo)
  • Some voice from the (front) center (mono)
  • Some music from the rear (stereo)

If, like in this example, some original field recording from a video clip is supposed to be used to create the surround sound audio track, it can be easily extracted using Kdenlive with Extract Audio->Wav 48000Hz from the context menu of the clip. This creates a WAV audio file in the same folder as where the video clip is located.

The audio clips to be used in this example are:

  • Field.wav (stereo) for Front L+R
  • Voice.wav (mono) for Center
  • Music.mp3 (stereo) for Surround L+R (rear)

In a new Audacity project, they can be imported in the above order with File->Import->Audio...

The channel mapping for 5.1 surround sound is:

  • 1 - Front Left
  • 2 - Front Right
  • 3 - Center
  • 4 - LFE
  • 5 - Surround Left
  • 6 - Surround Right


Note

LFE (Low Frequency Effects) is often referred to as "subwoofer channel", which is not quite correct. A surround sound speaker setup is perfectly valid without subwoofer, in this case the surround system will redirect the LFE channel to "large" speakers, usually the front speakers.


Editing Existing Surround Sound

When adding a clip with more than two channels to a project, kdenlive creates an audio thumbnail that correctly shows all audio channels:

The clip can be edited and (audio) effects applied to it, and all appears to work just fine - but once rendering the project, it turns out that the audio track in the resulting video file is 2 channels (stereo) only.

The following steps provide a manual workaround for this issue.

Extract and Split the Audio Track

The first step is to extract the audio track from the video clip. This can be done in Kdenlive with Extract Audio->Wav 48000Hz from the context menu of the clip. This creates a WAV audio file in the same folder as where the video clip is located.

The extracted WAV audio file can then be opened in Audacity, it should show all 6 channels, these are:

  • 1 - Front Left
  • 2 - Front Right
  • 3 - Center
  • 4 - LFE
  • 5 - Surround Left
  • 6 - Surround Right


Note

LFE (Low Frequency Effects) is often referred to as "subwoofer channel", which is not quite correct. A surround sound speaker setup is perfectly valid without subwoofer, in this case the surround system will redirect the LFE channel to "large" speakers, usually the front speakers.


The idea now is to split the surround sound into four separate (stereo/mono) audio files that Kdenlive can handle:

  • Front (stereo)
  • Center (mono)
  • LFE (mono)
  • Surround (stereo)


First, Audacity needs to be configured to not always export to stereo audio files: In Edit->Preferences, under Import/Export, select "Use custom mix (for example to export a 5.1 multichannel file)".

Now, in the first and the fifth track, click on the small triangle besides the name and select Make Stereo Track - you should end up with 4 tracks, two stereo and two mono.

Next, rename the tracks to "Front", "Center", "LFE" and "Surround" starting from the top, using Name... from the context menu of each track.

The tracks should now look like this:

After all this hard work, exporting the four tracks to four separate audio files is easy with File->Export Multiple.... Use "WAV" as "Export format", the rest of the settings should be already okay: "Split files based on: Tracks" and "Name files: Using Label/Track name".

The "Edit metadata" dialog might pop up for each track, it is fine to just say "OK". At the end there should be a confirmation dialog, and four audio files should have been exported: "Front.wav", "Center.wav", "LFE.wav" and "Surround.wav".

Import Audio Tracks into Kdenlive

The previously created audio files can now be added to the Kdenlive project using Project->Add Clip.

SInce there are only two audio tracks in a project by default, it is necessary to add two more using Project->Tracks->Insert Track before adding the four audio tracks to the timeline.

The next thing to do is to group the four audio tracks with the video clip by selecting all of them and then choosing Timeline->Group Clips.


Note

Don't forget to mute the original audio track in the video clip if necessary!


The Kdenlive project should now be ready for the usual editing, like cutting clips and adding effects, and should look something like this:

Rendering the Project

The pity now is, that it is still not possible to render the project with a surround sound audio track. Several manual steps are necessary to accomplish that.

First, the video track needs to be rendered without audio. This is simply done by rendering the project as it would normally be done, but without audio, by deselecting the "Export audio" checkbox.

Then, each of the four surround sound audio tracks "Front.wav", "Center.wav", "LFE.wav" and "Surround.wav" need to be rendered into a separate audio file. For each of them, do the following:

  • Mute all other audio tracks
  • Enter a respective file name for "Output file"
  • Select "Audio only" as "Destination"
  • Select profile "WAV 48000 KHz"
  • Check "Export audio"


Note

Unfortunately, the mono tracks "Center.wav" and "LFE.wav" are rendered as stereo tracks, and there seems to be no way to avoid this. But this can be handled later in Audacity.


Compose a Surround Sound Audio File

Now the separate audio tracks rendered by Kdenlive need to be "merged" into a single multichannel 5.1 surround sound audio file. This is again done in Audacity:

  • Import "Front.wav", "Center.wav", "LFE.wav" and "Surround.wav" (in this order!) using File->Import->Audio...


"Center" and "LFE" are now stereo, which is not what is needed. This can be fixed by selecting Split Stereo to Mono from the context menu of each track, and deleting one of the two resulting mono tracks.

Eventually, there should be four tracks in the Audacity project:

  • Front (stereo)
  • Center (mono)
  • LFE (mono)
  • Surround (stereo)


The project can now be exported into a 5.1 surround sound audio file:

  • Select File->Export...
  • Provide a name for "Name:" and select "AC3 Files (FFmpeg)"
  • Click "Options..." and choose "512 kbps" as "Bit Rate:"


The "Advanced Mixing Options" dialog should show up. The number of "Output Channels" should be 6 and the channel mapping should already be correct:

The result of the export should be an *.ac3 file which is playable with i.e. VLC or Dragon Player.

Muxing Video and Audio Together

Since video and audio was rendered separately, both need to be multiplexed into a single file containing both the video and audio stream.

Note

When muxing audio and video files into one file, the actual streams are just copied, and not transcoded. So there is no quality loss to neither the audio nor the video streams. Also, because the streams are just copied, muxing is very fast.


Assuming the video track was rendered to "Video.mkv" and the surround sound was exported to "5.1.ac3" the command to mux both to "Video-5.1.mkv" with avconv would be:

avconv -i Video.mkv -i 5.1.ac3 -c copy -map 0:0 -map 1:0 Video-5.1.mkv

The result should be an MKV video containing a Dolby Digital 5.1 surround sound audio track.