Recabinet 3 - Opinions so far?

[Dewdman42]

Another thing you might do for now to help you a bit setup your mushroom cloud each time is check out plogue bidule. With that you can setup a more complicated patch of stuff with your amp sim, recabinet and MSED and then save that as a single preset easily inserted on a track.

It would still require two instances of recabinet unfortunately, which means more CPU then if the panning feature was there, not to mention some cpu for plogue bidule too. But on the positive side, you could combine two cabs for the mid and two cabs for the sides, and plogue bidule may also open some other cool ideas for patching up interesting guitar setups.

Code: Select all

             ----recabinet mid----
           /                       \
AMP SIM---< (split)                 MSED -> stereo out
           \                       /
             ----recabinet side---

Also kind of think that for now you could simply create one recabinet with the main mid sound, then split the signal, apply EQ to what will become the side input for MSED. This could give you some of the mushroom cloud without having to use two instances of recabinet. Personally I think putting a very slight delay and/or pitch change on the side (id, doubling), would help to enhance the effect.

Code: Select all

                         ------ mid --------
                       /                     \
AMP SIM---recabinet---< (split)               MSED--> stereo out
                       \                     /
                         ---EQ + FX--side---

[Dewdman42]

As a side note, I want to also suggest some concerns I have about potential phasing problems with that trick. I have absolutely nothing against unorthodox use of FX to achieve creative ends. That being said, in this case the MSED is being used to take a non-encoded side channel and turn it into some kind of stereo. To me this is very unpredictable what it will end up creating, but there is not necessarily anything wrong with unpredictability, with the exception that this trick relies on whatever polarity of signal exists in the source side channel.

In my mind this could result in all manner of weird stuff happening from side to side in the resulting stereo image. Its also taking a sound which starts out mono in the middle, and puts stuff from one polar extreme on the right and the other polar extreme on the left. If you solo just the left or the right result of that, it would be interesting to hear what you get, for one thing. If you combine them back to mono again, it theoretically should sound exactly like the original signal I think(but not sure). However, put them in a room with speakers X feet apart from each other, and I predict possible phasing problems.

So the net result is that this might be a cool trick, or it might be a mastering engineer's nightmare. I'm not entirely sure about that, but just expressing my concern about phasing issues

[khanyz]

I've used Energy XT but it's a hassle. DDMF's Meta looks a better bet.

If you read a bit further back. I mentioned the stereo issues such as single notes being in a different place to chords. Also it is a total pain for engineers, but you have to challenge them occasionally. The easiest way, as I said, is to bounce it down before you sit it in the mix.

Any stereo image/phase issues are obvious in meters, as they are so extreme. You can correct them with automation for the bounce.

I think you should try it a few times. It isn't that bad to do.

[Dewdman42]

DDMF is certainly free. I had problems with it when I tried it on my mac.

[khanyz]

One way to think of it (and how I do) is this.

In Mid/Side recording, you have 2, possibly different, microphones recording the same source. Their relative positions with the primary source, and any secondary sources/reflections, create the phase and delay differences to stereo image encoded. However, the different responses of the microphones (even if they are matched) will also create further differences (sometimes intended as character/colour).

It it's simplest form, you are sort of recording the guitar in Mid/Side mode with the guitar cabinets as the microphones. With this in mind, you can compensate for the phase differences being unrealistic.

Things are complicated when you don't use the same source through each cab. It takes practice to learn the limitations and how far you can compensate to end up with a usable stereo sound.

BTW: Meta is $39. Last time I tried it, it wasn't stable enough to use, but I'll be trying it again soon.

[Dewdman42]

hmm....

I think there may be some misconceptions here.

in mid/side recording, you use a cardoid microphone facing the subject and a bi-directional ( figure-8 ) microphone placed in the same location as the cardoid but facing side to side. The cardoid gives you the "mid" as a mono direct signal of the soure. The bi-directional picks up the L and R artifacts and put them onto one single channel in such a way that polarization of signal is present, that's the side. If you try to listen to that side as is, it will not tell you much. But due to the way the signal is polarized on that channel, decoding it will create actual track separation.

So the first thing, the mics MUST be different, not maybe. And their position is not arbitrary. They must be in the same place and configured the way I just described. There is no phase issue as of yet. Its not phase differences that create the encoding, as you suggested, its the bi-directional microphone and the MS matrix that will be used with it.

In order to create a stereo XY signal from the M/S matrix, you take the side and split it to a third channel, flipping the phase switch 180degrees on one of the sides. Then you combine SIDE1 with mid for the resulting stereo left, and flipped-side2 with mid for resulting stereo right.

In the above the negative polarity in each side cancels out from the mid, while the positive polarity adds to it, creating a resulting stereo image with L and R channels.

Now back to what you are doing with recabinet. You are not really recording in M/S mode at all, you are recording in X/Y mode, where you have one mic on one cab and another mic on the other cab. In recabinet's normal mode of operation these are completely isolated from each other, there is no concern about phasing between mics. The resulting sound is combined as you wish to result in a mono phase-free combined tone.

If you take those phase-free signals separately and send to MSED, then what MSED does is the M/S matrix I described above. It takes the side channel and splits it off to a third channel, flips the phase on one of them and then combines it back with the mid. Since you are starting out with a non-encoded side channel, that means one side will have the real high freq cab sound, mixed with the low freq cab, while the other side will have the reverse phase of that same high freq cab, mixed with the mid.

In headphones it will sound fine, though I question how "stereo" it will sound. However, mixed to mono the high frequency content should cancel out entirely. In a room with speakers what you end up hearing will depend entirely on the placement of the speakers and position of the listener, but since you have introduced a lot of reverse phase content on each side, its going to be difficult to predict or fix until listening time. iIts hidden there, but also its the entire basis of your stereoification.

This is very different from a typical X/Y recording session where you place two mics on a cab, say a near mic and a far mic on the same cab and try to blend those two signals of the cab. That can and will create phase issues depending upon how far each mic is from the cab. Engineers try to avoid that problem, but they can always adjust the phase of one mic to move them virtually closer to the same distance from the cab when such problems exist so that phase cancelation will be reduced.

But such situations have nothing to do with M/S arrays, and since you are starting out with non-encoded side channel, the M/S matrix is introducing out of phase sound into your mix between left and right.

[khanyz]

If it takes you off track, then forget what I said. It's just an abstraction I use to keep things straight in my head. My other metaphor is Bose speakers but that needs beer to work.

I see that it may be confusing that an essentially single source, which in practise would only produce a mid signal, somehow gets side content. However, if you done M/S recording badly, or with dodgy microphones, or even with 3 (in a triangle with one phase inverted), you'll see some side content. Play with that stuff and it gets fun.

Have you done any M/S recording? The side does have to be figure 8, but the mid can be either cardioid, or a figure 8. I have used matched figure 8s because I preferred the sound. It helped to maintain consistent levels/settings.

I think you are putting way too much thought into this. Just try and see if it works for you. I've got the background to go into all the physics and math, but sometimes I just like a sound and leave it at that. If you would like me to go through the theory properly, just let me know the level I can pitch it at. It might be interesting to see if you can predict some results.

[Dewdman42]

the mid could be omni too, but that's not really relevant here

[khanyz]

the mid could be omni too, but that's not really relevant here

Although, technically you can use omni, in practice I never do. With multiple microphone setups, I always try to minimize the overlapping of polar patterns. It makes it a lot easier to tune the environment with screens/blocks/etc. With Mid/Side I like to keep strict perpendicular, independent patterns. So for the mid, I don't really mind rear reflections, but don't want anything from the sides.

I have never been formally trained (just educated in the theory and shouted at by many, many drummers, guitarists and singers), so I've picked up methods from doing them, and them working (or not and not repeating them). Hence my methods are quirky, but produce good results.

Anyway, I think I'll leave it at that and stick whilst I'm behind. Enjoy the method and let me know how it works out for you.

Cheers.

[hibidy]

You guys are CRAZY into this......love it!

[khanyz]

You guys are CRAZY into this......love it!

It starts out as poverty and having to get the best out of the equipment you can afford.

It gets crazy when you've got too many options. Luckily, I'm over the hump and stick mostly with my fave ways. Well, until sub-space microphones come along. Then we can do Mid/Side/Yesterday recordings . Yesterday's sessions (that you didn't record at the time) are always better.

[Dewdman42]

hey like I said, I have nothing against creativity, knock yourself out. I'm just pointing out the phase issues.

[Dewdman42]

Another idea that came to me for you to try since you're outside the box already.

1. Run guitar through pedals-amps, but not cab yet.
2. stereoize that signal with something, could be a lot or a little. A simple 11ms delay will do. But there are all manner of ways to stereoize it, but for this example, do NOT use MSED for this, some delay, pitch, modulation, etc, as subtle or drastic as you want. Making it very subtle to the point of almost not even hearing it is still valuable for what is to come next.
3. Now that you have a stereo sound, run into MSED to encode to M/S
4. Now finally run the Mid and side to recabinet, feed the mid to the low freq cab and the sides to the high freq cab.
5. Apply any other fx or EQ you want to use with mid-side
6. decode back to stereo with MSED

I the believe the above would give you some of the mushroom cloud you are after without the phase issue I described earlier.

cheers

[khanyz]

Or just use Voxengo Sound Delay (pre cab) and PHA-979 (post cab). I tried a stereo chorus, on a mild setting, whilst recording but it's too restrictive.

You can also record the part twice and hard pan (doubling). Schwa CMX also works nicely to do this artificially, as does Martin Eastwood Duet (on the Computer Music DVD).

The phases "issues" you have mentioned are part of the sound so you do need some of them. IMO it is better to record the parts as independently and as dry as possible so you don't commit too early. You can do all the processing in the DAW and it only takes practise to know at what point in the chain you need to start "fixing" things.

Like I said before, what I posted were the very basics. There's no point me giving out a full project as not everyone will have the same plugs, or want the same sound.

[khanyz]

Forgot to mention that you can use PHA-979 pre cab, it's just my preference to use it post. Pre, it will let you know of potential issues and let you remove them. Post, it shows you the conflicts and lets you fix them.