Zaxcom Nomad Auto-Mix

[stacysound]

This would be a complete re-write for Howy, but seems to me the best place to assign what inputs gets AutoMix would be in the output Bus. Add an "A" to the existing "X" and "P" options. If there's an "A" selected, it can be (or not be) effected by AutoMix.

[Jack Norflus]

I need to be able to disable the AutoMix and not have it kill all the audio.

I spoke to Howy about this today and he will be adding this feature shortly.

[fieldmixer]

Quick Fix maybe "disable aMix threshold" on/off.

[Wandering Ear]

You can turn the mic on threshold all the way down so automix turns all the mics on.

[Jack Norflus]

Quick Fix maybe "disable aMix threshold" on/off.

It will be something like this. Sort of an infinity setting.

[Davideo]

I see that in the latest firmware the shortest attack time is now 10ms instead of 1ms.

[Jack Norflus]

Yes it was changed. Before people start to freak out that your mic won't open as quick - let me clarify something.

When a person starts to talk the microphone will open immediately regardless what the attack is set to. The attack time is how long the microphone takes to ramp up to full level.

There was an audible related reason why the fastest attack time was intentionally changed to 10ms - but I don't recall exactly why.

[RScottATL]

IMO, the audible reason is most likely that such a quick attack will just serve to draw attention to the edit. As attacks get shorter, low frequencies can pop if in the middle of a large amplitude wave, usually though we aren't dealing with such as our lows are typically rolled off. Either way, though, if you're losing words and 10ms isn't solving your problem, a threshold adjustment is the better solution than an instantaneous attack.

[Jack Norflus]

I personally think the auto mixer - in most cases - sound pretty seamless running with the attack between 10 and 15ms

[John Blankenship]

I'm sad to say I've found the Nomad AutoMixer less than ideal. With a Dan Dugan (the "gold standard" in automixers) the algorithm is designed to keep the background level constant, even if more than one mic is prevalent at any given time.

Under varying conditions I've found it quite difficult to keep the Nomad "dialed in" to sound consistent.

Not something I'm really happy about since waiting a year and a half for this feature that was one of my reasons for purchasing a Nomad.

I still like the Nomad but feel that AutoMix needs some critical tweaking.

[Christian Spaeth]

John, I used the auto mixer on one project and though I constantly had to change the settings depending on situations (it was doc and sometimes they talked louder and sometimes lower, so I kinda rode the threshold a lot), I found it worked well enough. To get consistent background, setting up a room mic that is always open took care of that for me.

[Jack Norflus]

I don't think anything will ever compare to a Dugan - but given that I think the zaxcom works pretty well. Like Christian said there is no magic one size fits all setting.

I found you need to tweak the threshold, noise reduction and amount of attenuation to find the sweet spot to finess the background noise.

[John Blankenship]

I've used the AutoMix on two different projects, the first was a reality show (a good AutoMix would have been ideal here) but was torpedoed by the garbling gremlin (since fixed).  Another use was a medical conference where I had five lavs and two hand held wireless audience mics.  I tried my best to keep it dialed in but the results were nowhere near consistent.  The panelists varied so much from moment to moment, sometimes in close proximity to each other, sometimes further away, sometimes projecting, sometimes not.  I worked hard to stay on top of it settings-wise, but it didn't give me nearly as satisfactory results as a Lectrosonics or Dugan would (I haven't used the Shure or SD Automixers). 

 

I don't say any of this to knock the feature or the product; I say it in the hopes that further tweaking of the algorithm will lead to better results under more conditions.  It would be neat if Howy could incorporate maybe 3 or 4 different algorithm choices in the menu, that we could try.  That might get us a lot nearer to ideal as we discover what works best under different conditions.

 

I would love this to be a "killer feature," but IMHO, it isn't yet.

  

[James Arnold]

I've had mixed results with the feature (if you'll excuse the pun) though I'm partly convinced that I didn't manage to find the 'sweet' settings. It did put me off using it a bit though. I found it very hard to select an appropriate threshold - 2 of the 4 mics were very present but try as I might I couldn't get the system to catch the other two adequately. I've been a bit reluctant to test it further on important projects without a bit more playing.

[Derek H]

I used the automix on my Maxx once so far and found it really helpful. It was a corporate shoot, medical setting with 4 doctors in a round table arrangement fielding questions as a group. I never knew who would speak first and the room was noisy with a major freeway about 200 yards out the window. The automixer really helped latch on to the speakers quickly and keep the noise floor acceptable. I was really happy to have it!

The settings will need serious thought and tweaking by the user per situation, that's true. I'm not sure there is an "algorithm" at all. All the settings that make the feature function are there for you to adjust.

Also, I used post fader mode and still rode the levels a bit. I think that helped keep it consistent sounding. You wouldn't want to walk away from it and expect good results but it helped me get better results than i could have without it.

Cheers to zaxcom for including it on the Maxx too.

[studiomprd]

Derick: " I'm not sure there is an "algorithm" at all. "

huh ??   of course there is.... big time!

[John Blankenship]

The algorithm is the internal mathematical programming that is the "brains" of an automixer. Contrary to TV ads, there aren't little elves in there pulling levers.

[James Arnold]

I used the automix on my Maxx once so far and found it really helpful. It was a corporate shoot, medical setting with 4 doctors in a round table arrangement fielding questions as a group. I never knew who would speak first and the room was noisy with a major freeway about 200 yards out the window. The automixer really helped latch on to the speakers quickly and keep the noise floor acceptable. I was really happy to have it!

The settings will need serious thought and tweaking by the user per situation, that's true. I'm not sure there is an "algorithm" at all. All the settings that make the feature function are there for you to adjust.

Also, I used post fader mode and still rode the levels a bit. I think that helped keep it consistent sounding. You wouldn't want to walk away from it and expect good results but it helped me get better results than i could have without it.

Cheers to zaxcom for including it on the Maxx too.

Thanks Derek. I've not tried it in post fade mode. I think that would have helped. I'll try that on my next round table.

[Jack Norflus]

I always use the auto-mix as post fade. I have tried pre fade and have gotten mixed results.

Also keep in mind that no auto-mixer is 100% foolproof - since there is no way for it to determine the difference between wanted and unwanted sound. This is also true for a speakers bleed onto another mic - this is where the adjustable threshold comes in handy.

[James Arnold]

I always use the auto-mix as post fade. I have tried pre fade and have gotten mixed results.

Also keep in mind that no auto-mixer is 100% foolproof - since there is no way for it to determine the difference between wanted and unwanted sound. This is also true for a speakers bleed onto another mic - this is where the adjustable threshold comes in handy.

Cheers Jack. Shame that the pre fade seems not quite there. In the cases I've used it I opted to use it pre fade and send my own mix to camera as requested by production. The environments were quiet so it was differing voice levels and the difficulty of setting the correct threshold that boggled me a touch.

[berniebeaudry]

I would think that the majority of the time you'd want to use it post fade because after all aren't you using it to create a better mix?  I know the Dugan is the gold standard and I've not used the Lectro version.  I have used the Shure auto mixers and they're really not bad.  We once used three 8 channel ones on a huge discussion group.  24 lavs.  It helped tremendously!  You do have to keep an eye on the levels and adjust as needed when someone talks too softly or coughs.  I saw at NAB last year that Shure has improved the auto mixers with dsp.  The demos were impressive.  Maybe Zaxcom can talk Dan Dugan into creating a software module for the Maxx and Nomad.

[John Blankenship]

If I'm not mistaken, the Lectrosonics automixers use some Dugan magic (except for the earliest ones).

[Christian Spaeth]

I'm not quite sure what a different algorithm could improve on the automix function. I see the main problem being different levels on the inputs, also the s/n ratio differing from speaker to speaker, depending on how loud or dynamical they talk. My impression is that if you had identical speakers speaking at identical levels, with identical wireless systems, the Nomad automix would be flawless. Since I never used another automixer, I wonder: can a different algorithm handle varying input levels and dynamics in a better way? To me Automix seems to be all about S/N, and that is something that greatly relies on the human element, which is hardly controllable. Maybe the way SD call theirs Mix Assist is a more accurate term? Again, I'm curious if the Dugan or other systems really do all the magic and you don't have to worry about "the human element"?

[Paul Isaacs]

Maybe the way SD call theirs Mix Assist is a more accurate term?

Maybe this will answer your question ....

[berniebeaudry]

I'm not quite sure what a different algorithm could improve on the automix function. I see the main problem being different levels on the inputs, also the s/n ratio differing from speaker to speaker, depending on how loud or dynamical they talk. My impression is that if you had identical speakers speaking at identical levels, with identical wireless systems, the Nomad automix would be flawless. Since I never used another automixer, I wonder: can a different algorithm handle varying input levels and dynamics in a better way? To me Automix seems to be all about S/N, and that is something that greatly relies on the human element, which is hardly controllable. Maybe the way SD call theirs Mix Assist is a more accurate term? Again, I'm curious if the Dugan or other systems really do all the magic and you don't have to worry about "the human element"?

Its not just about the noise floor.  When you have people miced that are in close proximity you get phasing and comb filtering if the levels aren't set just right and you don't duck the mics that aren't being used.  That's were the mixing comes in when multiple mics are in play. How good you are at it has an impact on the mix.  I don't think auto mixers in general are designed to do all the work, just help do a better mix.  Haven't used the Dugan but saw it at NAB last year and that might be the exception.