Recabinet 3 - Opinions so far?

[khanyz]

How are you guys getting on with the project? Do you see what I'm trying to do?

[Amon1973]

Hi Nigel, I checked the project a couple times and I found it extremely interesting.

It gives a big, spacious feel to the guitars... curious to test this setup on some metal stuff. I guess I'll have to handle the concept properly, prior to do whatever

[khanyz]

^ The Cabinet IRs you use are critical, but it's not that difficult. If, when you listen to them, you classify them as bass/mid/treble type and then just pick 2 of different types to try. It's best to start with the lower frequency one in the centre, but you don't have to end up with it there. Just remember the usual bass frequencies in a stereo image issues/limitations.

Did you notice how single notes are in a different stereo position than chords. This is because of the more limited frequency range. It can be handy that solos automatically move so you can put a support track in the "hole" created. Or you can centre the solo and have the chords move to one side.

Also, it's not just for Cabinet IRs. I've started using with Microphone IRs and Vocals. It's not really for lead vocals as they need focus, but for backing, it's nice.

Cheers,
Nigel

[Dewdman42]

I read this thread with some interest, but since i don't have some of the plugins needed to check this out in Reaper, its a bit confusing to me what you did. I admit I have to learn about mid-side processing also. It would be great if you could back up and explain this technique generally, instead of as it relates to recabinet

[hibidy]

I like it.

[Dewdman42]

Ok I read up on mid-side process. I have used M/S mic techniques in the past, so I was already somewhat familiar.

If I understand you correctly you have essentially been wanting to use more airy sounding cabs for the sides and the meat in the middle. Also potentially applying more FX on the sides. Of course you don't have to use different cabs, you can simply create the Mid-side streams and apply different EQ to the sides, etc.

The thing is, this technique requires the original X/Y signal to be stereo to begin with. Pulling it into M/S allows you to operate on the middle and sides separately, but in order to realize the result you have to convert it back to X/Y again. The side channel has the far left and right encoded into a single channel by virtue of the polarity.

if you use two different cabs, and somehow could force a mono incoming signal to be split to hard pan left and right for the two cabs, then you would definitely end up with a lopsided sound. Sending to MSED and back to stereo would still be lopsided. I am missing something, but what?

I guess I don't understand what you're doing with Recabinet. Recabinet has two cabs, I guess in more or less a mono mode. What I like is that this provides a way to create one virtual mono cabinet. I get what you're trying to accomplish, but I am still not sure how you will get the airy sounding stuff to both the left and the right sides, stereoized, when the input signal is mono to begin with.

I guess recabinet could have a mode where it simply outputs the two cabinets on seperate channels instead of mixing them, which would let you do whatever you want to do with it, but I still don't understand exactly what you're doing, and I would like to understand because I'm always looking for better tone.

[khanyz]

Ok I read up on mid-side process. I have used M/S mic techniques in the past, so I was already somewhat familiar.

If I understand you correctly you have essentially been wanting to use more airy sounding cabs for the sides and the meat in the middle. Also potentially applying more FX on the sides. Of course you don't have to use different cabs, you can simply create the Mid-side streams and apply different EQ to the sides, etc.

This is an artificial effect/technique that I started using with Bass in the 90s. I wanted to blend the cab (single close mic) and direct sounds (both mono), and have a bit more width to the usual mono bass. I found that by restricting the low response cabs (15 or 18 inch) to the center and the hi response cabs (10 or 8 inch) to the side, along with the same band split component from the DI, I got what I wanted. It was an artificial stereo image and so did require balancing. All in all, a thorough pain for an engineer (so I would have to mix it down to a stereo bass track) and not much better with the early software. Luckily, I never really used much FX with bass so it didn't get too complicated. All it really needed was HP and LP filters on each sound, and independent (mono) compressors on the mid and side, to make it sit properly. You then fed the M/S to L/R and balanced with gain.

When I started recording guitar, I tried the technique and quickly dropped the DI. Instead, I replaced it with different FX chains. I basically had a solid state amp with a 12" (Peavey) and a tube amp with an 8" (Crate). The 12" was distorted (with the help of an Ibanez TS-9) and close mic'd (mono). The 8" was used with a delay pedal (Dod Digital one) and again close mic'd (mono). You then put the 12" in the mid, the 8" in the side, sat with HP/LP filters and balanced to get a nice airy sound.

The thing is, this technique requires the original X/Y signal to be stereo to begin with. Pulling it into M/S allows you to operate on the middle and sides separately, but in order to realize the result you have to convert it back to X/Y again. The side channel has the far left and right encoded into a single channel by virtue of the polarity.

if you use two different cabs, and somehow could force a mono incoming signal to be split to hard pan left and right for the two cabs, then you would definitely end up with a lopsided sound. Sending to MSED and back to stereo would still be lopsided. I am missing something, but what?

What you are missing is that there are no original stereo, or mid/side recordings. It is entirely an artificial construct. You are using differences in frequency responses of the same (mono) signal through different cabinets/DI EQ to act like comb filters (ish, this is more an educational comparison than technical). It's a bit like how stereoizers work, but not really. There's also a lot of phase/delay interplay going on.

I guess I don't understand what you're doing with Recabinet. Recabinet has two cabs, I guess in more or less a mono mode. What I like is that this provides a way to create one virtual mono cabinet. I get what you're trying to accomplish, but I am still not sure how you will get the airy sounding stuff to both the left and the right sides, stereoized, when the input signal is mono to begin with.

I guess recabinet could have a mode where it simply outputs the two cabinets on seperate channels instead of mixing them, which would let you do whatever you want to do with it, but I still don't understand exactly what you're doing, and I would like to understand because I'm always looking for better tone.

What I want to do with Recabinet is only use one instance, select the lower response and higher response cabs, and have the output to different channels (hard panned to L & R). I could then simply put Voxengo's MSED after it and do the balancing easily. Right now I need to use 2 instances on different channels and use Reaper's Routing to get them to MSED.

I hope that explains everything. If not, ask away .
Cheers.

[khanyz]

I forgot to mention. I'm in the process of encapsulating this technique in a guitar pedal. It's called the "HiMo LoJo" and is basically a dual-band distortion pedal where you can pass the bands through a M/S decoder to get a stereo signal. The gain, band-split and distortion stages are on the breadboard and are final. The auto balance is still a work in progress but I'm getting close. I'll release a schematic once I have something that works well enough. I've got Dean Aka Nekro and insonicbloom lined up to sanity check it at that point (though Paul doesn't quite know it yet).

As part of this, I've also designed a Mid/Side FX Loop pedal called "Lil' Ms Loopy" which is the M/S Decoder with Auto Balance. Actually, that will probably be ready as a prototype first as Dean got rather excited when I mentioned it to him.

Cheers

[Dewdman42]

no I get you are trying to create fake stereo, but I still don't quite understand how you get balanced stereo.

if you have low freq cab on the left and high freq cab on the right and send that to MSED, you will get the low+high mono mix in the mid channel. It should be a pretty even mix of the two channels. However the side channel will not be just the highs, it will also be the lows and the highs, except the way that M/S works is that it uses the polarity to get both the L and R signal into a mono channel, such that when you decode it back to stereo again it retains everything on the left and right. Normally if you are working with a stereo track to start out with, then M/S will definitely put in the mid channel the meat of what is in the center and the side channel will tend to have the content from BOTH sides, but encoded by the polarity so that when you go back to L/R you still have the stereo image.

However in this case you are talking about having the lows on the left and the highs on the right. So the mono mid will have the blend of both(what recabinet already delivers now), and the sides will also be a blend of both, except that it will be encoded with polarity such that the lows and the highs are out of phase but on a single mono channel, until you decode it back to stereo again, at which time you will end up with lows on the left and highs on the right.

So I still don't understand how you are isolating the highs from the lows and creating a usable stereo image in the end. MSED does not create stereo, it can tweak an existing stereo image, it doesn't create one.

[Dewdman42]

or maybe I am still missing something...

[khanyz]

However in this case you are talking about having the lows on the left and the highs on the right.

I think is where you are misunderstanding. In the routing, yes the Highs and Lows are labelled as Left/Right, but they are treated as Mid/Side. That is what MSED does. It treats it's left/right inputs directly as mid/side, it does not covert them. So MSED is creating a stereo image.

So Mid (Left + Right) is the low response cab and the Side (Left - Right) is the high response. From a basic RMS level understanding and Left = Mid + Side; Right = Mid − Side, the Right will usually be at a lower level and so will need a gain boost to balance the stereo signal. This is simplistic, because of all the phase and frequency dependency you are intentionally introducing, but is a good starting point. The phase coherence with Recabinet, is a particular feature I wanted to see simplfy this.

[Dewdman42]

So you are taking the left and right signals from recabinet and treating them AS IF they were already M/S encoded and then using MSED to decode back to stereo from that?

Under that presumption, your left low channel you are treating as the mid and your right high channel you are treating as your sides, but its not really encoded properly to begin with, which makes the results a bit unpredictable, but interesting to think about.

In order for it to be more of a valid decoding, you would need to have that right channel from the high freq cab to have been previously encoded with L and R information from somewhere. Since you're starting with a mono-non-encoded channel, MSED will try to interpret that however it does I guess, Looking for polarity in the signal to try to decide what goes to the left and what goes to the right.

Now, that being said, I can see how simply running it through the MSED would definitely do SOMETHING to the material, mangle it up, based on the content, turning it into some kind of L/R stereo separation.....so....maybe I'm getting that you are making it sound more spacious that way and doing so only with the high freq cab.....while leaving the meat of the middle cab alone. Its an interesting technique. I will have to play with it sometime.

I will say, you could probably do the very same thing simply by taking the mono output from recabinet and split the signal off to MSED, but through an EQ first. I'm not sure its really so necessary that recabinet be able to give you different cabs on seperate channels in order to do that interesting production trick. But hey, if its not hard for him to add that feature to recabinet and it won't complicate usage, then why not. A lot of people apparently do ask to have the seperate outs. Some people might simply want to apply different FX to the different cabs before mixing them.

[khanyz]

I've done proper Mid/Side technique, and this is anything but proper or valid.

I started out doing this by patching cables so the software version is already more "accurate". The reaper project is a simplified example to just show the basics. I include a lot more phase processing when I do this in production. It's one of the big reasons I've stuck with samplitude.

[Dewdman42]

yea man I actually really want to try it sometime, its interesting...

[khanyz]

yea man I actually really want to try it sometime, its interesting...

Just think "Mushroom Cloud of Sound" whilst you do it, and you'll get it .