Zoom H6 line out phase inversion problem

[Geoffrey Cox]

I was testing my new Zoom H6 with a Panasonic S5 camera. I connected a microphone to the Zoom input and Zoom line out to the S5 line input so as to have a guide track. I made a recording (i.e. internally on the Zoom and externally via the Zoom line out to the S5) and synched things in post. I worked in Logic so I could drill down right to sample level to get the synch as good as possible.

To my surprise I noticed the audio recording on the S5 video file was clearly phase inverted. If I lined them up exactly I got quite close to complete phase cancellation. I could only assume this was a fault with the input of the S5 or output of the Zoom. I assumed it would be the S5 but tested the Zoom line out on a different device (my laptop using a Scarlett Solo audio interface). The phase was again, inverted.

The only logical conclusion I can come to is that the Zoom H6 line out phase inverts the signal. This should not be happening. It could be a fault with my machine only but that seems unlikely. Could it be a design flaw?

I would be surprised if anyone else had noticed this but any thoughts very welcome.

[Patrick Tracy]

It might be a simple matter of cost, using fewer components in the path to the line out.

[Greg Miller]

If I understand this correctly, the electrical polarity of the Zoom output seems to be reversed with respect to the polarity of a file recorded on the Zoom. I believe you're running 3.5mm TRS output to 3.5mm TRS input, so that pretty much rules out cable wiring issues.

If that's the case, then I can see two possible scenarios. either

(A) the Zoom file polarity is correct with respect to the Zoom input, and the Zoom line output is reverse polarity,

or

B) the Zoom file polarity is reversed with respect to the Zoom input, and the Zoom line output is correct with respect to the Zoom input.

Time to get out the oscilloscope and a signal generator (or reference test source).

[Don Palomaki]

What kind on mic did you connect to the Zoom, and how was it connected?

The 3.5mm zoom output jack and the 3.5mm mic input jacks on both devices appear to be unbalanced stereo.

Phase reversal in one channel with equal audio in the other channel can happen when a balanced source is connected to left and right unbalanced inputs. Without getting into other possible configuration options on the Zoom: an example - a balanced mic connected to the MIC/LINE input jack would likely record reversed phase on one channel.

[Geoffrey Cox]

I tried different mics.

First I tried a mono mic (Rode), then 2x DPA 4061 stereo pair (linked in the H6). The Rode was connected via a 3.5mm TRS to XLR adapter, the DPAs via there own microdot to XLR adapters. So all mics going into the Zoom via XLR.

The results were the same. Mono mic is phase inverted and with the stereo pair, both channels inverted.

Using a phase inverter plugin solved the issue and with the stereo pair I had to invert both channels.

And yes, using a 3,5mm TRS lead.

[Geoffrey Cox]

Originally Posted by Greg Miller

If I understand this correctly, the electrical polarity of the Zoom output seems to be reversed with respect to the polarity of a file recorded on the Zoom. I believe you're running 3.5mm TRS output to 3.5mm TRS input, so that pretty much rules out cable wiring issues.

If that's the case, then I can see two possible scenarios. either

(A) the Zoom file polarity is correct with respect to the Zoom input, and the Zoom line output is reverse polarity,

or

B) the Zoom file polarity is reversed with respect to the Zoom input, and the Zoom line output is correct with respect to the Zoom input.

Time to get out the oscilloscope and a signal generator (or reference test source).

Indeed though I had not thought of scenario B which would be a much more basic fault I would have thought. Unfortunately the oscilloscope test is beyond my capabilities!

[Greg Miller]

I'm glad you seem to have found a solution, but I'm a bit puzzled.

Where did you use the phase inverter? At the mic input of the Zoom? Or somewhere else?

[Geoffrey Cox]

I used a phase inversion plugin on one of the tracks in the audio software (Logic). In fact it is part of a general gain plugin.

In many ways the whole problem is not that important as it is unlikely I would use both the Zoom and S5 audio tracks together and would stick with e better quality, multitrack Zoom audio unless something had gone wrong. I just think it bad that there is this obvious design fault (or whatever it is). Where it is awkward is that it is more tricky to synch to the guide track when it is visually inverted - the phase inversion plugin does not alter it visually as presumably it only affects the outgoing signal.

[Patrick Tracy]

Interesting. In Vegas Pro, which is a complete DAW as well as an NLE, there's a polarity invert switch right on the channel control, and it does flip the waveform. There's no need for a plugin. I'm kind of shocked that Logic doesn't have that kind of functionality.

[Geoffrey Cox]

Originally Posted by Patrick Tracy

Interesting. In Vegas Pro, which is a complete DAW as well as an NLE, there's a polarity invert switch right on the channel control, and it does flip the waveform. There's no need for a plugin. I'm kind of shocked that Logic doesn't have that kind of functionality.

It's a good point. I will double check but don't think there is a simple reverse polarity switch.

[Paul R Johnson]

Can you just flip it in the audio editor? When aligning audio tracks for sync, I never get close enough for this to ever be a problem. I assume that when you look at the zoomed in waveform you are seeing one waveform go positive on it's first transient leading edge and the other one goes negative? My guess is it's just a design feature. Have you actually tested the E-E of the Zoom, as in the input to output to see if the zoom is the same polarity in as it is out. I'm not sure it's that critical , but if the output is reversed to the input that is poor design. If the in to out is the same, I doubt it matters too much.

[Greg Miller]

Paul, I suggested testing the Zoom in/out polarity in post #3. The OP said, in post #6, that he didn't have that capability.

[Paul R Johnson]

I was thinking more of checking a file recorded on the card, ingested into a computer to see if the first data was positive going compared to that card via the audio out or usb out. Ideally they'd all be the same, but probably not.

Thinking about this, I'm not sure it makes any actually difference practically as I don't think I've ever zoomed in close enough to have ever noticed. After all - the camera audio won't ever cancel completely as the mics are in different places? Plus - there's no point being more accurate than the frame rate - so minute shifts which are possible in an audio editor aren't usually possible in the video editor, locked to frames,

I edit in Premiere, and it's simple to pop into Audition to do individual sample shifts, but I've never had to?

[Geoffrey Cox]

Originally Posted by Paul R Johnson

I was thinking more of checking a file recorded on the card, ingested into a computer to see if the first data was positive going compared to that card via the audio out or usb out. Ideally they'd all be the same, but probably not.

Thinking about this, I'm not sure it makes any actually difference practically as I don't think I've ever zoomed in close enough to have ever noticed. After all - the camera audio won't ever cancel completely as the mics are in different places? Plus - there's no point being more accurate than the frame rate - so minute shifts which are possible in an audio editor aren't usually possible in the video editor, locked to frames,

I edit in Premiere, and it's simple to pop into Audition to do individual sample shifts, but I've never had to?

If I understand you correctly this is what I already described above. I compared the recording made straight to the card in the Zoom device (ingested into a computer via the card - my laptop has a card reader built in) to the same simultaneous recording fed out via the line out of the Zoom. First I tried the line out plugged into my camera and then noticed the phase inversion so assumed the camera input would be at fault but tried a second test with the Zoom line out going into my laptop via an audio interface. The result was the same - an inverted signal so I had to assume that it was the Zoom line out that was the problem. The actual phase cancelation that resulted makes it clear this is not just some issue with the waveform depiction but an actual inversion.

When it comes to how much one needs that kind of close synch, it is a good point but with dialogue it does need to be pretty close and I have sometimes found frame only accuracy is not enough so I would finalise things in an audio editor (Logic). I do a lot of fine-tuned audio work anyway and find the limitations of such work in a video NLE (premiere in my case) frustrating.

[Don Palomaki]

On audio sync to video. Light travels much faster than sound, about 186,000 miles per second vs. ~1000 ft/sec. So audio lagging video slightly corresponds to standing back a bit further from the source. Our brains process it. However, audio leading video is unnatural

FWIW: Phase inversion and delays in one channel of stereo sound compared to the other channel can effect the spatial perception of the sound. You can demonstrate it with loud speakers by reversing the +/- connections on one side and listen. With proper phase you can generally pinpoint instruments, with one side flipped phase the same instrument may seem to wonder in space.