I am searching a solution to use my ecm ms957 (sony). I found a “doesnt-go” thread (and i have no other item (like a mini-disc) than audacity on my laptop) but i cannot imagine that nobody wrote a little utility, though this mic is quite beloved.
I understand that the output are addition and substraction of the channels. So this is at least very simple artihmetic.
Does anybody knows where there is such converter or if not a sony-converter hard?
Thanks in advance.
====
Afterwards, please excuse this thread. I had the idea of downloading audacity 1.3.13 and this works now stereo. Thanks to to community.
I understand that the output are addition and substraction of the channels. So this is at least very simple arithmetic.
Well, yes, but the way Sony implemented the cables isn’t quite so simple. The bottom of the microphone is an XLR5 Male, not XLR3 like a normal microphone, so the first thing you would need is an adapter to split out the Mid and Side into two XLR3s. Then on into two microphone preamplfiers – Windows laptops only have one, Macs don’t have any.
MDs and DATs have two, that’s why they plug together so well.
Then, after you got the two signals amplified so they can managed, then you can apply the two signals to a de-matrixer and line driver and then onto the USB SoundCard like this…
So is this the same “funny” cabling the same on my ECM-MS907, Koz??
It ouputs into a 3.5mm stereo jack, so I’m guessing the wiring and the L/R orientation are all ok - but any insights you have would be useful and interesting.
According to the Instruction/Operation Manual for the ECM-MS957
the microphone has a stereo mini-jack plug and has a switch for selecting between 90 and 120 degree pick-up patterns.
I assume from the presence of this switch that the sum-and-difference matrix for decoding the mid-side stereo is built into the microphone and the output uses “conventional” stereo mini-jack wiring (right / left to tip / ring and common/earth to sleeve).
Although computer sound cards usually have stereo mini-jack connectors for the microphone input, they are not wired as “conventional” stereo mini-jack sockets.
Typical computer microphones are mono and require a power supply, so the computer mic socket supplies power on one contact and uses the other two contacts for signal (mono) and earth.
The ECM-MS957 is battery powered so it does not require the power connector, but instead it requires a stereo input. Basically this microphone is incompatible with a standard computer microphone input.
If the recording sounds like stereo, then the mid-side stereo has been decoded.
Before it is decoded mid-side recording don’t sound like stereo.
Decoding is just a matter of splitting the “side” recording (from a figure of 8 microphone) into left / right channels and inverting one side, then mixing in some of the “mid” (usually a cardioid mic) recording. This gives “Mid + Side” in one channel and “Mid - Side” in the other and (by magic) it sounds like stereo.
Although this is straightforward to do occasionally, it quickly becomes a pain when it has to be done maually for each and every recording. II use a Zoom H2n that produces MS + XY so it’s a lot of track fiddling before the mix can be made.
I was quite surprised to find no plugins to help decoding mid-side tracks. Did I overlook something?
Devices that produce M/S files are quite unusual.
There is a “Channel Mixer” plug-in that can be used to convert M/S into normal stereo. It does not have any presets for M/S stereo because I do not have any M/S stereo recordings to test it on. If you could send me a good “RAW M/S” file then I can probably add some presets.
Some suitable presets, according to litterature, are:
wide ( 150⁰, S = +6dB)
normal ( 120⁰, S = 0 dB )
small ( 60⁰, S = -9 dB )
mono ( only MID )
Would it be possible for the plug-in to select a preset, have the settings copied to the sliders, and then modify the sliders to get the desired effect?
Thanks, I’d have guessed values around those figures, though for use in Audacity it may actually work out better to reduce the mono a bit for the wide settings so as to avoid clipping. Would it be possible for you to upload a good quality M/S recording somewhere on the Internet (for example “sendspace.com”) and post a link to it here on the forum? I’d then be able to test it out rather than relying on the theory.
Unfortunately not. Nyquist plug-ins do not support any kind of interactive control. The plug-in does however display the settings used after processing if you use the “Debug” button rather than the “OK” button. You can then use those figures as a guide.
The values are also the ‘suggested presets’ used by the H2n.
I think for a good mix the difference between the mid and side channels matter. Just adding 6dB of side would easily lead to clipping.
I’m at a concert today and I will try to make some MS recordings to play with. PM me your mail address and I’ll send you a link when I finish uploading (expect them no sooner than 21.00 UTC).
Thanks for the M/S recording snippets sciurius.
I’ve added 4 new presets to the Channel Mixer plug-in:
M/S Decode Wide
M/S Decode Normal
M/S Decode Narrow
M/S Encode
The first three will decode M/S stereo recordings, assuming that the Mid channel is the upper (left) channel of the stereo pair.
The fourth preset will encode a normal stereo recording into M/S stereo with the Mid channel at the top.
That brings the number of presets up to 18 (plus the custom option) which is probably about the most that we would want. As different operating systems handle long menus differently, please can you check that the full list of presets is accessible on your machine. The first in the list is “Use Custom” and the final one in the list is “M/S Encode”.
With M/S decoding it is not possible to know in advance what the exact peak level of the decoded audio will be until it has been decoded. In an extreme case a 0 dB side mix could produce an output with peaks up to 6 dB higher than either of the M/S channels, though that would be highly unusual. I’ve adjusted the presets so that hopefully the peak output level will usually be about the same as the peak level of the original audio, though depending on the phase differences between the channels it could actually be a bit higher or lower.
Please could you check on as wide a range of M/S recordings as possible that the output levels are reasonable and not excessively high or low. I can tweak the presets if necessary.
Do you think that the “Wide / Normal / Narrow” presets cover useful ranges? (not too limited and not too extreme?)
If you want “mono” then there is already a preset higher up the list “Both Left” which will do that. channel-mixer.ny (2.88 KB)
Please could you check on as wide a range of M/S recordings as possible that the output levels are reasonable and not excessively high or low.
No problems so far.
However, I was wondering whether it is possible to calculate the exact mix factors. For example, if the MID channel has a max level of 0.7 and the SIDE channel has a max of 0.6, the resulting channels would be limited to 0.7 (the max of both maxes) and the mix factors adjusted accordingly.
Do you think that the “Wide / Normal / Narrow” presets cover useful ranges?
This will require some more experimenting…
In earlier experiments with the channel mixer, I got satisfactory results with:
If you want “mono” then there is already a preset higher up the list “Both Left” which will do that.
Yes, although for the sake of completeness I’d suggest adding it to the presets nevertheless (but I can see why you want to keep the number of presets down…).
The peak value of the output for a given M/S file cannot be calculated without first applying the decoder and then measuring the output, but the problem with that approach is that, for Nyquist plug-ins, the measuring can only be done in RAM, so for long recordings the memory usage could be greater than the amount of physical RAM available in the computer, which would then cause Audacity to freeze or crash.
I think that in most cases where a mono track was wanted, it would be better to simply split the M/S track (click on the track name then select “Split to Mono”) and then delete the “Side” track (click on the [X] in the top left corner of the track).
I’ve been thinking that it may be better to have a separate Mid-Side decoder plug-in rather than just tagging it onto the Channel Mixer.
Which do you prefer for decoding M/S, the channel mixer or this one? (below)
Do you think that it needs the option for selecting whether the Middle channel is the “left” or “right” track channel? I’ve not seen a lot of M/S encoded files, but those I have seen have either been as two single channel files, or as a single channel Middle and a 2 channel Side file, or as a 2 channel file with Mid in the first (left) channel (like your files). I don’t recall ever seeing a M/S 2 channel file with Mid in the second (“right”) channel. MidSide.ny (2.94 KB)
Do you think that it needs the option for selecting whether the Middle channel is the “left” or “right” track channel? I’ve not seen a lot of M/S encoded files, but those I have seen have either been as two single channel files, or as a single channel Middle and a 2 channel Side file, or as a 2 channel file with Mid in the first (left) channel (like your files). I don’t recall ever seeing a M/S 2 channel file with Mid in the second (“right”) channel.
I haven’t seen am M/S with mid right (S/M ?) either so as far as I’m concerned this option can be removed. OTOH, it doesn’t really get in the way of anything.
