Anyone tried Izotope Neutron? The automatic mixing plugin?

Ok guys. Does this thing work? Has anyone tried it?

I just got handed a project with almost 100 vocal dubs. My assistant/intern is in Nashville for the rest of the week and fucking hell if I’m gonna mix these by hand.

I know how to eq a spoken word voiceover. I just don’t want to. Fingers crossed, gonna download a demo. Lets give this thing a try.

1 Like

Nope. Its shit. Sorry guys.

This thing is dumber than a rock and suggests the same thing for every preset. No, I don’t need “retro harmonics” on every damn vocal. Doesn’t handle speech dialogue very well. Why would they make a 7.1 surround version of this if not intending to use it for ADR and dubbing? It wanted to the do the same thing to Newt Gingrich’s voice as it did Donald Trumps. Dang. Threw a female voice at it, and it changed a little, but not enough to help.

Oh well. Wishful thinking.

1 Like

Yes it is an incredibly dumb plugin. I tried the demo a while back and it was pretty stupid. I don’t remember everything about it since it was a while ago, but I remember it had a feature where it would “listen” to your audio and make suggestions. Problem is you could process it however it suggested, mix that down and analyse the processed audio and it would just suggest the same processing again on it. The actual sound of the audio being analysed had absolutely no bearing at all on what it suggested doing to it. I like ozone though and I’m surprised they put out something as stupid and pointless as neutron.

I would have bought this program in a heartbeat if it actually worked. And I would pay a ton of cash for it, because the amount of time it could have saved is enormous.

Izotope already owns the mixing tools they need to make a plugin like this work. Their problem is purely on the coding and AI end. And OMG!! If they integrated the RX tools (that’s their sound restoration suite) into an AI they’d turn the audio industry upside down.

Imagine having a computer that could fix a distorted clipping microphone by re-drawing and re-syntheizing wave forms based on the wave forms by first analyzing the distortion, then comparing it to the samples it sees before and after. Like a restoring a damaged hard drive from a raid configuration. Guess what? They already have that, and it works VERY VERY well. Now if they could just take that one step further then automate that with an AI, then they’ll own this industry for years.

If I owned a fully functional super smart AI Neutron plugin (that worked with me instead of against me), imagine the competitive advantage!! Holy hell, if this thing could also learn and adapt to your mixing style as Alexa and Siri adapt to your speech style, I could finish days worth of nasty, dry, boring dialogue and SFX editing in hours, then have the rest of the day to smoke cigars and drink scotch…then undercut the living fuck out of other studios that paying an army of entry level nerds to do the same thing.

Hahaha! Neutron, even though its an ass-clown at the moment, is formatted for a 7.1 surround environment. Imagine how many audio engineers it could put out of work, if it could balance and navigate an Atmos object panner and balance a digital summing matrix algorithmically? Wouldn’t that be amazing? :smiley:

I have that plugin. It came with the… what was it called… Music Production Bundle or something.

I don’t use that smart mixing feature at all. (Though I use the plug-in all the time.) I’ve tried it, but I have never gotten good results. It seemed to think that my guitar was a bass more than half the time, and there was no way to tell it that it was, in fact, a guitar.

Sounds like it didn’t work out for you either. Maybe the technology will get there someday!

Mine thought a 40 cal. gunshot was a snare, a bomb explosion was a kick, and a car slamming into a parking garage rail then crashing into a the street was a cymbal. I dunno. The driver hit the breaks and made a screech mark before colliding with the rail, and it thought that was a female vocal. Meh. I had no idea it was so limited.

If I wasn’t having computer problems at the moment (as you can see on that other thread) I’d make a video warning people how dumb this thing actually is. I would probably get a number of views and good laughs, because there are a bunch of people that hate this plugin, but for all the wrong reasons. Most people detest the idea of machine taking their jobs…

…But you’re right. Someday it will happen.

I don’t see how it would give you a competitive advantage if it worked, since then everyone else could just buy it as well. I guess it did have some cool stuff built into it like the eq and what not, but that analysing part pretending to analyse audio when it clearly wasn’t really pissed me off.

I think the main problem with plugins like these is that it’s taking a backwards approach to “smart plugins.” I don’t think it makes sense to try to make an algorithm mix like a person, because there’s too much variability in how that’s done.

I mean, look at photoshop. It’s not like Adobe tried to create a computer version of how a brush works. They took the basic idea of painting/drawing and added all sorts of completely different things to it. That’s what makes it powerful. But nobody, as far as I know, is trying to make smart filters that apply the right settings to white girls in bras. That idea doesn’t even make sense, so why do we try to do it with audio?

To me, that magic of AI when it comes to music won’t be in the mixing department, it will be in the creating and arranging department. For an algorithm to tell me “hey, if you add this chord transition here, it might be cool,” would be way more useful.

Or for an algorithm that listens to the music and automatically creates a drum beat for it, or a bass line, or whatever. Those things are all possible and match the strengths of AI really well.

We already have “dumb” versions of this (meaning not AI, but hard coded performances) with stuff like EZDrummer and EZKeys. But doing an AI version of that would be awesome.

Ah. True. I was thinking an advantage over the people who would resist buying it, much the same as many engineers detested the thought of Pro Tools when it first came out. Now that you mention it, you’re totally right. You’ve pointed out that the advantage at the core level isn’t the software, its the willingness to use technology to beat someone else who won’t.

Again, I hear you there. That thing should have been sold as a metering plugin an nothing more. I thought the ability to see what it thinks are clashing frequencies was pretty cool. I would probably use it if it was available, but right now I’m annoyed enough with that thing to not want to pay for it all together.

Dude. Cool thought. Lets move this thread one out from under the Neutron topic… :smiley:

I have a different take, but it’s not because I disagree with what others have said…

Mainly, the title of this thread, “…Izotope Neutron? The automatic mixing plugin?” I haven’t paid attention to whether they are really touting it as an automatic plugin or not, but you shouldn’t look at it as one.

At best, maybe it gives you some hints at where to start with some eq points on a vocal. I’ve played with that a bit, and don’t find it disastrous. But I consider this a minor side feature that won’t get much use from me.

I gave it a demo and immediately bought it at the $199 intro at the time. But I never even tried the automated stuff—I never expected much from it, nor wanted it.

I find the plugin itself quite useful and quite good. First, the individual components are flexible, and the fact they come as a collection that can be rearranged, modified, and disabled means that you can put together your own presets for specific instruments and sounds easily and conveniently. The EQ is the main thing and the part I’ve worked most with, but I’ve found the exciter and transient shaping to work out very well for the few spots I’ve really needed them. There are so many compressors (I have so many, you have so many) that I won’t comment there, other than, again, it’s nice to have a couple of compressors in that chain. The EQ is really nice, with the compression and side chaining bands to other tracks and all, and very nice metering.

But yeah, forget about the “automatic” idea.

1 Like

Yeah man. Its all fabulous mixing stuff. I use their multi band saturation tool a lot. I’ve gotten really used to advanced 7 for quick mastering touch ups. And their RX suite is killer. Nothing else like it. If you’re using the tools inside Neutron which are the same ones in their standalone sets, then they’re fantastic at doing a great number of things. I was just sorely depressed that I had to mix 150 vocals one at a time by hand lol.

I did spring for an upgrade deal to Music Production Bundle 2. After that I was looking at buying RX6, and found they had a cross grade deal that made it a no-brainer. So I went from no Izotope to leaded pretty quickly. Still much unexplored.

I had to mix 150 vocals one at a time by hand

day-um—I am pretty far on the other end of the spectrum, mainly dealing with me. lol

I mean…some of the people on the video shoot only had like…three or four words. But when the voices are switching from one to another in the sequence of the video, you still have to balance them. Its not like mixing a lead vocal…this stuffs far easier. This session was a little more chaotic than normal.