On Sat, 14 May 2016 08:55:12 +0200 fred <f.rech(a)yahoo.fr> wrote: Thanks for the links. None, simply because there are no Linux plugins, as I know of, that have this option. So I thought nevertheless to give it a try to see how it might sound. I would use EQ. As Paul mentioned, it seems that the new plugin pin management feature would allow something like M/S, perhaps in an easier way.

On 05/14/2016 05:35 PM, Paul Davis wrote: And with older versions of Ardour: 1. you have a stereo track already. disconnect it from the master. 2. create a mono "M" bus- 3. create a stereo "S" bus (yes, see below). 4. route both outs of the stereo track to "M". 5. route the L out of the stereo track to the first input of "S". 6. route the R out of the stereo track to the second input of "S". 7. invert the polarity of the second input of "S". 8. now add your desired processing (usually EQ) to M and S (use mono plugins for S despite the bus being stereo). 9. set the M panner to center. 10. set the S panner to center and maximum width. 11. set the M and S faders to -6dB to begin with, adjust S level to taste. But: M/S processing is really only useful when you are remastering a stereo mix without access to the individual components, or maybe if you are dealing with stereo mic recordings as part of a larger mix. If you are creating a multitrack mix from individual channels, M/S buys you exactly nothing that couldn't be done better and more precisely in the individual channels. -- Jörn Nettingsmeier Lortzingstr. 11, 45128 Essen, Tel. +49 177 7937487 Meister für Veranstaltungstechnik (Bühne/Studio) Tonmeister VDT http://stackingdwarves.net

On 05/18/2016 09:04 PM, Emanuele Rusconi wrote: My apologies, you are completely right. I typed this on a train somewhere without having access to an Ardour system, and I jumped ahead of myself. What's missing is one step: create an S' channel just like the S channel, with right polarity inverted. S -> S' -> Master. S must have the panner width set to zero, and S' to full width. The nice thing about an M/S matrix is that it is its own inverse. That means you can use the exact same operation twice. That's what the two S busses basically do: M = L + R (happens because you route both chs of your stereo track to M) S = L - R (happens because you have one side of S inverted and then sum in the panner) L = M + S (happens by feeding M and S' left to Master left) R = M - S (happens by feeding M and S' right inverted to to Master right) Of course, you could also use a little plugin to do the matrix operation. -- Jörn Nettingsmeier Lortzingstr. 11, 45128 Essen, Tel. +49 177 7937487 Meister für Veranstaltungstechnik (Bühne/Studio) Tonmeister VDT http://stackingdwarves.net

On Thu, 19 May 2016 16:18:44 +0200, William Light wrote: In more than 30 years of audio recording I _never_ used M/S processing. M/S microphone technique is something else and is useful, because it does cause a mono-compatible signal, due to not being based on travel time, but instead it's based on intensity. To "exaggerate" a signal by intensity with a mixing console, just pan pots are required. IOW Jörn is right, if you have full access to the individual components, you could use pan pots instead (more than one channel completely to the left and the other completely to the right is impossible), resp. you could use more enhanced techniques to widen or "shape the stereo image subtly". "Natural" stereo hearing is much based on travel time as well as intensity. Regards, Ralf

On Fri, 20 May 2016 01:41:15 +0200, William Light wrote: We explained that this technique due to it's nature is used on stereo signals, assumed you don't have access to the individual original signals and that it doesn't provide something useful, if you have full access to the individual channels. It depends on the kind of music and wanted result and at what stage you start taking care about the wanted result. I would stay away from M/S processing to spread a stereo chorus or reverb of a mix, I would chose another chorus or adjust the reverb instead. However, since "natural" means that travel time is important, you could add delay without feedback to one side of a single to spread the audio signal or better instead of doing this, you could add early reflections by aux send to all channels by different intensity. If you're making Moog alike synth music you could add flanging, phasing that is not dominating as effect, since it's masked by the synth sound filter effects, so it quasi only makes a mono sound stereo or spreads a less wide stereo signal. Instead of using 2 oscillators by a mono synth that have a different fine tuning, you could use 1 oscillator and a chorus or two of this mono synth with one oscillator with different fine tunings. When playing two synth of that kind by the same source (MIDI track or manually connected by MIDI), you also could use a little bit different portamento to the left and right synth (to make it sound similar to two synth played individually). An auto-panning effect could be useful for several tasks, one is related to the stereo perception. Anyway, less is more, most of the times good EQing is all that is needed to have much impact on _all_ aspects of a mix. You might have noticed a difference between most modern and vintage delays. Oldish delays often are ping pong delays, while modern delays usually try to avoid real, complete ping pong. The more stereo, the harder it gets to keep mono-compatibility. If you're making "natural" music with violins, guitars etc., adjust microphone positions, adjust pan pots and EQs, usage of a reverb with early reflections might be ok too. Regards, Ralf

On Sat, 21 May 2016 07:38:10 +0200 (CEST), Tim Goetze wrote: Hi Tim, please explain what it gains. Summarized: On Sat, 14 May 2016 08:57:59 +0000, Fons Adriaensen wrote: Fons didn't mention what those rare cases are. On Thu, 19 May 2016 17:07:46 +0200, Ralf Mardorf wrote: Jörn explained what those rare cases are. Wiliam disagreed with Jörn's explanation, but didn't mentioned why he doesn't agree with Jörn. I pointed out, that if you have full access to all components of a mix, the left/right intensity could be controlled by the pan pots (in combination with the faders and EQs ;). In my reply to Wiliam's question what kind of enhanced techniques are available, I mentioned that if e.g. an effect shouldn't spread the stereo signal wide enough, choosing another effect gives you better control. So again, what exactly is M/S processing useful for daily mixing music and what's wrong with Jörn's points and where I'm mistaken? There is a karaoke effect, the voice canceler, that tries to cancel out the vocals, by eliminating mid-range sounds panned to the centre. This might be useful, if you don't have access to the original multi-track recording, but as soon as you have got access to the original multi-track recording, muting the vocal tracks is the easier way to do it and the result is much better. IOW, if not doing karaoke, but instead mixing music in a studio, using voice canceling processing gains you nothing compared to muting the vocal tracks. What is M/S processing good for, that can't be done easier and better, apart from what Jörn mentioned? Please explain why we are mistaken? Regards, Ralf

On Sat, May 21, 2016 at 09:47:45AM +0200, Ralf Mardorf wrote: Yes. When you have to modify a stereo mix, or a stereo signal that is part of a mix. Do you claim that such cases don't exist ? In theory anything that can be done in the L/R domain can be done as well in the M/S domain and vice versa. But the simple fact is that some things are much easier in one domain than in the other. In particular: * Anything that affects L and R in an assymetric way (e.g. modifying only L or R) is easy in L/R and difficult in M/S. * Anything that affects L and R in a symmetric way (e.g. modifying stereo width) is much more easily done in M/S. That remain true even if you have access to the components of a stereo mix. And certainly if doing the same thing in L/R would require you panpots to go beyond their range limits, or be frequency dependent. Maybe you're not familiar which such techniques. In that case just stop commenting. Ciao, -- FA A world of exhaustive, reliable metadata would be an utopia. It's also a pipe-dream, founded on self-delusion, nerd hubris and hysterically inflated market opportunities. (Cory Doctorow)

On Sat, 21 May 2016 11:08:22 +0200, Ralf Mardorf wrote: *lol* Replacing microphones by other microphones could be useful too, but actually I wanted to say, that changing the position of the microphones, e.g. using another microphony technique, could replace complicated processing.

On Sat, 21 May 2016 09:43:35 +0000, Fons Adriaensen wrote: I don't have a problem with this thread, I just disagree with Wiliam, because Wiliam disagrees with Jörn. This is true for a stereo signal, but from where do you get a stereo signal? In origin there can not be more than left only and right only, there is nothing beyond those extremes. Regarding a spatial impression of a stereo signal, were you can't change the origin, you're right. But from were do come stereo signals in the field? Actually I experienced such cases, but then there wasn't the gear available to do M/S processing and under better circumstances there likely would have been the needed gear available, so that M/S processing anyway wouldn't have been needed. Yes, that's what Jörn already mentioned and I agree with this. I also don't say something against his work flow. I only pointed out that Jörn isn't mistaken. So regarding the OP's request "What are your thoughts on M/S EQ processing ?" Jörn was the first one who has given a reasonable explanation. If somebody likes another work flow is something else, but actually everything Jörn mentioned is correct. If you have access to all components, M/S processing doesn't make sense. How can one disagree with this? If you e.g. use the sampled gated drums, then you already have no access to the components. If you use sampled drums that aren't gated and you gate them, you have much better control. I neither would use a samples of gated drums that don't fit my stereo requirements, nor would I use a sample of normal drums and do the gating myself, I simply would chose other gated drums samples. To disagree with Jörn's explanation is wrong, because he doesn't pointed out something that is wrong. He pointed out that you gain nothing by using M/S processing, if you have access to all components, He did not mention that you couldn't use it, even if you have access to all components. If you e.g. wish to use M/S processing for the stereo sum of the final mastering, you are free to do so, but an experienced engineer usually makes live easier, by mixing the individual channels as wanted in the first place. The result with M/S processing could be perfect, but it unlikely makes it easier to get the wanted result, more likely it would be harder to get the wanted result. If you don't exactly know what you want as the result, experimenting with what ever tool is ok, but to precisely get what you want, it's not a good work-flow. Regards, Ralf

On Sat, 21 May 2016 12:35:25 +0000, Fons Adriaensen wrote: What are examples in the filed? Yes, that already was pointed out by Jörn's explanation. I'm right there with you. However, if you're using double sided large membrane mics you could stepless chose directionality. So you won't use pan pots for spatial impression at all and you unlikely need M/S processing. If you're using headphones as microphones, you'll experience issues that can't be fixed, what ever technique you'll use. Then give an example. Again, there are no doubts that it could be useful if room or mics are good, but not ideal for the wanted recording. Apart from the examples given by Jörn, what are the examples in the filed, that Wiliam and you disagree with Jörn's claim? Please, why don't you provide an example? Regards, Ralf

On Sat, 21 May 2016 15:50:49 +0000, Fons Adriaensen wrote: This is not the contentious issue. That is why microphony was given as an example when M/S processing could provide something, that can't be done without it. No, you only pointed out an example that isn't a contentious issue. *IF* you're doing microphony or *IF* you don't have access to individual components of any other stereo mix, then M/S processing could be helpful. That's beyond all question. Anyway, it can't do magic, the more components are part of a stereo signal, the less you can adjust with S/M processing. *IF* you've got access to all components, then there is no natural stereo source, excepted of microphony. Other sources could be - Stereo synth that don't provide access to their stereo components, but usually a synth allows to do this, exceptions are sounds such as gated drum samples, of which all kind of samples are available, so you just need to select the one that fits. - Stereo effects that don't allow adjustments, don't provide the needed kind of effect. This are absolutely rare exceptions. Assumed you need to pay for what you need, then you better pay for a good reverb, a selection of good effects and synth that allow to edit their stereo options, instead of paying for M/S processing, to fix issues that shouldn't exist. *IF* you are doing a multi-track recording and no complex microphony is involved, you unlikely ever need M/S processing. "Natural" and "intuitive" spreading of a stereo signal isn't if you use the M/S processing to improve the stereo master of a multi-track recording or even parts (sub groups), excepted of stereo microphony or assumed you shouldn't own a usable reverb, other usable effects or you only own cheapest stereo sample players. Regards, Ralf

On Sat, 21 May 2016 20:32:45 +0200, Jörn Nettingsmeier wrote: All kinds of stereo enhancer, exciter, intrusive recording technique are "in the air tonight" and nobody will listen to this the morning after.

On Sun, 22 May 2016 08:10:59 +1200, cbannister(a)slingshot.co.nz wrote: https://en.wikipedia.org/wiki/In_the_Air_Tonight#Drum_sound And then we had an era of Linn, Sequential Circuits TOM, E-mu Drumulator, Simmons and similar drum sounds. Not necessarily the first choice of sounds we want, if we want sound samples or emulations of artificial vintage drum sounds ;).

On Sat, 14 May 2016 08:57:59 +0000 Fons Adriaensen <fons(a)linuxaudio.org> wrote: What would you consider such a case that would eventually call for M/S processing ? Cheers.