1 Topic by Martin86 (edited by Martin86 2021-10-23 18:50:49)

  • Registered: 2021-10-23
  • Posts: 7

Topic: Mono/stereo mathematical relationship

I have just bought the ADI2, wonderful piece of equipment which I am very pleased with. Last night I had occasion to switch it to mono, exceptionally bad stereo mix of a mono original, and the perceived sound level dropped which prompts the question, how have RME applied this?

Years ago when I was in BBC Engineering, M=A+B-3dB, the IBA used M=A+B-6dB which the BBC has now adopted as their standard also. Apparently the original was too difficult for non BBC types to cope with. I also believe the EBU possibly used M=A+B-4.5dB. THis was all done so that the same signal heard in mono on a single loudspeaker and in (stereo) on a pair of loudspeakers would be at the same subjective level, hence the varying subtractions to achieve this.

I appreciate within the ADI2 a monoising adds A and B and sends it to both L and R outputs so I would expect the sound level to either increase or remain the same, it seems sensible to ask though what is actually happening.
Regards
Martin

2 Reply by KaiS (edited by KaiS 2021-10-23 23:18:11)

  • KaiS
  • DAW Mastermind
  • Offline
  • From: Germany
  • Registered: 2020-02-05
  • Posts: 3,154

Re: Mono/stereo mathematical relationship

ADI-2 Pro:
Mono = (Left + Right) / 2
Just tested with a sinewave.

Remark:
-6 dB isn‘t exactly 1/2
-6,0205999133 dB = 1/2
Bit-shifting one bit down is 1/2

KaiS's Website

  • Registered: 2021-10-23
  • Posts: 7

Re: Mono/stereo mathematical relationship

Kai
Thanks very much for doing that, I now know what to expect, which should be no actual change in level if using tone, programme will vary depending on material. My ears cannot detect a change of 0.0205etc dB, so 6 will do. When your as old as me and brought up firmly in the analogue era, nearest dB will do,  which is probably adequate enough most often even today.
Thanks
Martin

4 Reply by KaiS (edited by KaiS 2021-10-26 00:12:26)

  • KaiS
  • DAW Mastermind
  • Offline
  • From: Germany
  • Registered: 2020-02-05
  • Posts: 3,154

Re: Mono/stereo mathematical relationship

Martin86 wrote:

My ears cannot detect a change of 0.0205etc dB, so 6 will do.

Nobody can.

Most soft- and hardware I know does in fact do -6.0205999133 dB (exact 1/2), even if their label shows straight -6.0 dB.
As mentioned, in the digital domain the calculation is a simple bit-shift.

So does ADI-2.

Mathematically the figure -6.0 dB is correctly rounded.
Practically generating mono by (L+R)/2 makes sense too, this way there is no level change when switching a centered signal (L=R) to mono.
It even avoids overload problems, as the level can’t get higher in mono than in stereo setting.

KaiS's Website