Soyuz 013A Ambisonic Microphone

It is certainly easier to drive!

I'll dig around for Wes's paper, sounds like a good read for me at this point. You're right as far as the omni/figure 8 decode, it's just like "real life". As far as the plugin being confusing, I had the opposite reaction (It made a lot of sense to me  ). Especially when you're using it and keeping everything ambisonic like it was intended, it's helped me get what i want quicker. When I did the mix of that bluegrass band around the mic in the first video, I wanted the main singer in the middle to be just a little louder in a few spots. I had to do a couple of weird things to pull that off originally (before I got the ab Imager plugin). It was SO quick and straight forward to pull her up more and make her more present. Pushing the Front/Back control towards more of a cardioid than figure 8 helped a lot, or just raising the omni level. Both great options in stereo.


I've got head tracking working in the Dolby Atmos Renderer for my headphones now and have been messing with an Atmos mix of the bluegrass band (hopefully they'll be up for releasing it). It's really interesting decoding the mic to full Atmos and using the ab Imager to do something like push the singer up in the mix. If I push Front/Back towards a more cardioid pattern to the front towards the singer, I'm able to get her a bit louder as well as physically "push her forward" in the mix towards the front wall. When moving my head around, you get more of an exaggerated sense that the group is in a half circle around your sides and in front of you. If I start to push Omni more, I was able to pull the singer more to the "center" of the environment, almost like she was more in my head than in front of me, while still being able to rotate my head around and locate everyone else around me accurately. Both are really cool, I'm not sure how I'm going to go about it yet. I've done two mixes, one that's more accurate with her located in front of me and one where she's more centered and stays centered more when I move my head around.

As far as upscaling, this is really new to me and I'll need to work with it more before I start to form any valid opinions or thoughts. I have talked with Alejandro a fair amount about upscaling though. Initial thoughts are that if you're just dumping this to stereo ultimately, I have not really found a difference (when listening to the stationary file in stereo on headphones or speakers) between staying in First Order and upscaling. I did one recording where I'm walking around the mic with a shaker. Taking that down to stereo and listening in headphones, I'm not able to perceive any difference or more accuracy. However, if you are going to use fully ambisonic plugins (like some compressors, EQ's and other processors that are able to process each ambisonic channel separately), I am able to get different results and more finely manipulate different areas and parts of the ambisonic signal. I can't tell you if it's going to be practically useful yet, but there is clearly a difference. It's a mixed bag on it being better or worse.

When you start getting head tracking involved (either with the Audio Brewers Ambisonic head tracking, or doing it inside of Dolby Atmos with the external Dolby Renderer), that's when I've heard a clear difference with upscaling, even when taking the First Order tracks from the Soyuz mics and upscaling to Third Order (which is as high as I can go in ProTools Studio...). When I did my very first attempt at the live bluegrass band signal and messing with it in atmos, I kept it in First Order. After doing some aggressive processing with the imaging of the mic, there are definitely times where I would move my head around and when I got to the left side with the mandolin player, all of a sudden he would start to disappear and drop in level a little bit, then pop right back up. It's like there was a tiny dead spot in the image where he phased out. That was totally my fault, I'm learning here and was realizing that I was being a little too aggressive in some of my imaging moves. I dialed it back and the problem went away. When upscaling that ambisonic mic up to third order, then decoding the third order into atmos, it definitely sounds a little different (almost like the players got just a little farther away from me and the image was a bit bigger) but moving my head around felt much smoother with less dead spots.

Once I got head tracking involved on headphones, I started to be more a believe in upscaling. It's still very much a mixed bag, and depending on what you're doing, sometimes staying down in First Order just works better. The only other consideration for upscaling is when you're taking additional microphones and using them to supplement the ambisonic mic (or vise versa). Especially in Dolby Atmos specifically, I'm hearing a difference. You can also take a mono signal, upscale it into an Ambisonic environment with the ab Imager, and move it around inside that ambisonic sphere. Upscaling to a higher order ambisonic environment and placing a mono mic around you sounds different inside of Dolby Atmos (on headphones) than just making the mono source an object or sending it right to the speaker bed and panning it around. I'm not saying it's better or worse, but I'm able to head a difference.

This is opening a whole other can of worms. My initial takeaways are that upscaling and using ambisonics, specifically in atmos, gives you another way to make things sound more spatial than just moving them around in the panner or as an object. Sometimes it sounds better to me, sometimes it doesn't. Moving forward, I'm going to try it a bunch of ways and just think of it as another tool until I get more experience and start to figure out when it works better or worse, and why. There are definitely situations where I've heard upscaling make an audible difference and improvement, but more than likely you're not going to hear that much of a difference when just dumping it down to stereo where everything is remaining stationary. Ask me again what I think in another year or so...