First time here? Check out the FAQ!

Record Timeline to 16 Channels

+1 vote
Hi everyone,

I would like to record a Timeline to a circle of 16 speakers and I tried to do this by running the sound two times and changing the placement of the speakers between the two runs. I basically tried to render all the odd speakers in the first run (1,3,5...) and all the even speakers (2,4,6...) in the second run. The placement of the speakers are in a circle with 22,5 degrees between each speaker. So my speaker one is at: 56,25 degrees, speaker 2 at 78,75 degrees, speaker 3 at 101,25 degrees and so on.

The result I get, is that it seems to me (looks and sound) like the files rendered for speaker 1 and 2 are identical, 3 and 4 are identical and so on.

Has anyone tried this and experienced the same behaviour or have any ideas on things that I might have done wrong or not thought of?

All the best,

asked Mar 27, 2017 in Using Kyma by mathias-josefson (170 points)
Hi Mathias, are they precisely identical or is it possible that they are just very similar?  In the speaker placement preference, there is a lot of overlap between adjacent speakers, so it would not be unexpected for speakers 1 and 2 to be very similar in a 16-channel pan.

You could try experimenting with the Radius control: the larger the radius, the more difference between each speaker.  You could try a radius of 1.5 or even 2 to see whether you get more differentiation between the speakers. (With a radius of 2, you might have to boost the overall signal, since it will be attenuated as if the source signal is further away).

Try testing it with super simple test signal at first (something like pulsed noise slowly panning) so you can hear when the amplitude dips in each channel as the virtual source leaves that position.

thank you so much for your suggestions! I have now inspected the recordings made using the SoundField Vectorscope from iZotope and the files are identical. The vectorscope shows a straight line in the middle when checking the files from speaker 1 and 2. Checking the files between speaker 2 and 3 they are totally uncorrelated. The angular distance between speaker 1 and 2 is the same as between speaker 2 and 3. I made a new record to disk after changing the Radius to 2 and the result is the same, i.e files from speaker 1 and 2 are identical.

I then tested, following your advice, with pulsed noise panning a full circle in one minute. And for this test the result is different, this works as expected. When looking at the vectorscope of the files from speaker 1 and 2 I can see a line moving with the panning (looks like a monophonic source being panned). Same for files 2 and 3. It is also audible the panning in the files. So the test works differently from my patch.

What I am doing in my patch is that I am playing stereo files using DiskPlayer and rotate them in the same/similar way as in the test patch using the Angle Automation. For all cases the sound has been placed on a radius of 1.  

I then decided to do another test. Instead of using stereo files for the Disk Player I tested with mono files (patch otherwise identical). This worked!  Just like the test patch with the pulsed noise I see if I look on files 1 and 2 the monophonic line moving in the vectorscope space. So it seems that the behavior with identical correlated files happens when a stereo file is played by the DiskPlayer, but not when a mono file is played. I will continue to test to split up the stereo file into two mono files.

Thank you and all the best,

1 Answer

0 votes
Yes, the best approach is to use mono sources in the Timeline.

The Timeline considers each channel of a multichannel source to be in the same position as the corresponding speaker in the Speaker Placement Preferences. This means that a 2 channel source will be separated by the same angle as separates speakers 1 and 2.

For mono sources, the pan position places the single source at the desired location. For multichannel sources, each source channel is placed at the speaker location and then rotated by the pan angle.
answered Mar 30, 2017 by ssc (Savant) (127,180 points)
Thank you so much for the additional information!

All the best,
Hi everyone,

just a follow up on this topic, if anyone wants to do the same. I split up the stereo file into two mono files and separated them of what the angle of the speakers will be when I play from all 16 speakers (i.e 11.25 degrees) and then followed the process described above, and that worked well!

All the best, Mathias
I know I'm like 5 years late to the party here, but I'm now working on a piece for 15 speakers in a linear setup, and this is all super-helpful. I'm parsing through all this now, and I'll probably have some follow-up questions myself!
Hi Simon! That is awesome!