Channel designation for 5.1

WaveLab
#1

Most modern receivers and components label the channel outputs as:

FL
FR
Surround Left
Surround Right
Center
Sub
Surround Back Left
Surround Back Right

Where the last two are typically reserved for 7.1 configs (or upmixing to a 7.1).

RIFF wav file headers must have any channels that are present in the order specified here:
http://www.microsoft.com/whdc/device/audio/multichaud.mspx under the section “Default Channel Ordering”. No all channels need be present but any ones there must be sequential in that order. The actual mapping of where these channels are sent is specified in the dwChannelMask word as specified in that article.

My question is the 5.1 specification. There are two different ways to specify a nominal “5.1” config. in the wave file (from the KsMedia.h VisualStudio header file):

#define KSAUDIO_SPEAKER_5POINT1 (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT |
SPEAKER_FRONT_CENTER | SPEAKER_LOW_FREQUENCY |
SPEAKER_BACK_LEFT | SPEAKER_BACK_RIGHT)

#define KSAUDIO_SPEAKER_5POINT1_SURROUND (SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT |
SPEAKER_FRONT_CENTER | SPEAKER_LOW_FREQUENCY |
SPEAKER_SIDE_LEFT | SPEAKER_SIDE_RIGHT)

Which is the mapping used in most commercial 5.1 authoring applications or does it vary? My tests with my Pioneer VSX-9140txh receiver uses the SurroundLeft/Right speakers when I use the SIDE_LEFT/SIDE_RIGHT version (5POINT1_SURROUND mapping).

#2

There are more than 2 ways.
The commonly used L,R,C,LFE,Ls,Rs is ITU standard.
L,R,C,LFE,Lss,Rss,Ls,Rs is a 7.1 configuration.
Running 5.1 from an amp outputting for 7.1 requires the manual to be looked at - most use the 2 side channels for 5.1 but some use the 2 rearmost (rare).

To make things interesting, you may get interleaved files in L,C,R,Ls,Rs,LFE config, or even L,R,Ls,Rs,C,LFE.
we love our standards - it’swhy we have so many of them, I think :smiley:

#3

Thanks Neil. It’s dizzying!

I’m not sure if the following will help but I created a “complete” set of 28 audio multichannel test wav files (14.9 Mb expanding to 240 Mb:
http://www.jensign.com/testaudio/OmniWave.zip

The set includes 14 files with 16bit/44.1 kHz and 14 files with 24bit/96kHz. Files are:
mono, Stereo, 2.1, 3.0, 3.1, 4.0, 4.1, 5.0, 5.1, 5.1s, 6.0, 6.1, 7.0, 7.1
All files are -20dBfs (1/10 full digital amplitude) in each channel and consist of multi-tones.

The “examples.txt” file in the archive lists the actual tones used in each of the wav files.

Some interesting info here also on multichannel and receiver designations matrices:
http://en.wikipedia.org/wiki/Surround_sound

#4

The fileset looks like it may be interesting - thanks for this. I’ll grab it & take a look.
Can be very useful adding test signals to discs…

Not sure about Wiki is a source though, especially when the head of the linked article states it needs citations & verification…

#5

It will be interesting to get some feedback on this … particularly the 5.1 versus 5.1s files.
Specifically, if you test these two files, observe which of the speaker channels are used. In both cases, the FL, FR, C and LFE should be used. The difference will be in the receiver designated channels SL, SR and SBL, SBR usage (Surround Left/Right and Surround Back Left/Right).

The 5.1 wave channel mask specifies: Back Left and Back Right in the channel mask (0x003F)
The 5.1s wave channel mask specifies: Side Left and Side Right in the channel mask (0x060F)

My experience is that specifying “Back Left and Back Right” in the channel mask usually maps to the Surround Left/Right receiver speaker channels (but not always!).

(The channel locations and their relation to the “channelmask” in a multichannel wav file are shown here:
http://msdn.microsoft.com/en-us/windows/hardware/gg463006.aspx#EMLAC

My plan for the next release of this archive is to add (or replace) each constant tone in each channel with a sequenced recorded voice-over in each channel for easier identification (created with, of course, a montage in WL6)