The 1:0.707 ratio preserves impedances.
This is usually not critical. But German Engineers do it right. At any moment, at any point in this 250 ohm system, they can substitute a different 250 ohm item and get THE expected result.
Overload should not be an issue at mike level. Noise can be. Noise with Neumann mikes is not a serious issue.
If you study "600 match" MS stuff with 1:1 ratios, it is often worked as 600:1200 or so.
> you get 6dB boost.
+6dB voltage, -6dB current.
With zero-Z source and hi-Z load (current is not sucked), then impedance is not critical. 1:1 iron is readily available.
> you get 6dB boost.
In a non-match system, easy to pot-down. Here, or further along the chain. One of the main features of M/S -is- that you can pot-up/down the M and the S as needed.
> a coefficient of 0.5 is safe, 0.7 hazardous, and 1:1 a disaster.
Right. 1+1=2, so two tracks at 99% can make a peak at 198%.
In a nearly random signal (spaced pair), 99%+99% will rarely happen. Possibly not even once per track, and then so briefly that nobody will notice.
In a panpot "stereo" passage where one mike is very dominant, you WILL get 90%+90%= 180%, constant 5dB clipping. If the mix-image engineer drops to pure mono, 6dB. So as Jonte says, You have to have "divide by two" somewhere. For more typical pop-stereo, you might use 4dB and nobody notices the rare 1dB-2dB clip.
> it can't be THIS easy . . . just 4 resistors?
Yes, but:
As mentioned, some port(s) MUST be fully floating. (Not Forced-Balanced.) As drawn, only the S source must float, which is elegant. (However in practice this may upset M-S identity, running S as full-float and M as unbalanced.)
It also directly accounts for "divide by two" somewhere, by throwing-away half your Power. This is acceptable at Line level, may be fine at Neumann level, would be very distresssing at ribbon level.
And if you must add a transformer to gain "fully floating", then you can go ahead and do it with no resistors, just summing powers with series windings. M and S are handled identically, no oddities to the image.
This is usually not critical. But German Engineers do it right. At any moment, at any point in this 250 ohm system, they can substitute a different 250 ohm item and get THE expected result.
Overload should not be an issue at mike level. Noise can be. Noise with Neumann mikes is not a serious issue.
If you study "600 match" MS stuff with 1:1 ratios, it is often worked as 600:1200 or so.
> you get 6dB boost.
+6dB voltage, -6dB current.
With zero-Z source and hi-Z load (current is not sucked), then impedance is not critical. 1:1 iron is readily available.
> you get 6dB boost.
In a non-match system, easy to pot-down. Here, or further along the chain. One of the main features of M/S -is- that you can pot-up/down the M and the S as needed.
> a coefficient of 0.5 is safe, 0.7 hazardous, and 1:1 a disaster.
Right. 1+1=2, so two tracks at 99% can make a peak at 198%.
In a nearly random signal (spaced pair), 99%+99% will rarely happen. Possibly not even once per track, and then so briefly that nobody will notice.
In a panpot "stereo" passage where one mike is very dominant, you WILL get 90%+90%= 180%, constant 5dB clipping. If the mix-image engineer drops to pure mono, 6dB. So as Jonte says, You have to have "divide by two" somewhere. For more typical pop-stereo, you might use 4dB and nobody notices the rare 1dB-2dB clip.
> it can't be THIS easy . . . just 4 resistors?
Yes, but:
As mentioned, some port(s) MUST be fully floating. (Not Forced-Balanced.) As drawn, only the S source must float, which is elegant. (However in practice this may upset M-S identity, running S as full-float and M as unbalanced.)
It also directly accounts for "divide by two" somewhere, by throwing-away half your Power. This is acceptable at Line level, may be fine at Neumann level, would be very distresssing at ribbon level.
And if you must add a transformer to gain "fully floating", then you can go ahead and do it with no resistors, just summing powers with series windings. M and S are handled identically, no oddities to the image.
