Zoom F6 Portable Field Recorder Review

[Blumlein 88]

Maybe I'm just a brutal moron. But when you say split the feed. Like what does that mean, how would I do that? Like, do I need to have two channels recording the same thing with multiple mics, or one mic that feeds into multiple channels with a splitter? And then somehow take that recording and.. do what exactly in software? Like, I'm wondering what does it mean literally from a directions point of view (what's actually happening when you're doing this goes over my head, I'm wondering specifically what would I be doing with options in a DAW that would be doing this sort of "combining" process that yields better results).

Also, this adds bit-depth as you say, but not reduction in noise. What use is such if noise hits before your theoretical bit-depth? Or lets say your noise is super low (amir posted a video once of some company recording insects eating and a caterpillar walking), could you then do this technique you speak about to achieve something that company was able to with their mics?

I assume he was thinking of recording audio interfaces which may have 8 or more channels. You could split the input to more than one of them I suppose for a tiny benefit. It really won't matter as room noise is what will get you in the end. Unless you are making totally electronic music that is.

 

[Tks]

I assume he was thinking of recording audio interfaces which may have 8 or more channels. You could split the input to more than one of them I suppose for a tiny benefit. It really won't matter as room noise is what will get you in the end. Unless you are making totally electronic music that is.

Just so I can comprehend how much of a benefit this is. Let's say you are making electronic music. And you have something like a 32 channel interface. What benefit would that yield? And also, how is this "combination" done. Like what exactly are you doing? Is this like a function you can choose from in software like audacity perhaps? Where you have tracks that you can merge or something? I'm just lost as to what exactly is going on.

 

[PeteL]

Just so I can comprehend how much of a benefit this is. Let's say you are making electronic music. And you have something like a 32 channel interface. What benefit would that yield? And also, how is this "combination" done. Like what exactly are you doing? Is this like a function you can choose from in software like audacity perhaps? Where you have tracks that you can merge or something? I'm just lost as to what exactly is going on.

this article explain the theory behind this: https://www.analog.com/en/technical-articles/get-a-3db-snr-boost-using-a-dual-adc.html# (simplistic)
More torough:
https://www.electronicdesign.com/te...-1-3-db-averaging-adc-channels-to-improve-nsd

 

[Blumlein 88]

Just so I can comprehend how much of a benefit this is. Let's say you are making electronic music. And you have something like a 32 channel interface. What benefit would that yield? And also, how is this "combination" done. Like what exactly are you doing? Is this like a function you can choose from in software like audacity perhaps? Where you have tracks that you can merge or something? I'm just lost as to what exactly is going on.

Well if its electronic music you aren't going to benefit. The max you'll benefit is 3 db reduction in noise from doubling channels. And often you don't quite get that in practice. If we just split a cable and send it I don't think you will get any benefit. I think you'll get a 3db increase in noise.

Here is an explanation of how adding ADC channels reduces noise. It only does so if the noise is not correlated. Sampling noise, quantization noise, and thermal noise are all involved.

Here is an in depth examination of the issues in a paper you can download here.
https://drum.lib.umd.edu/handle/1903/9608

Well PeteL gave a much simpler page to explain it. Doubling doubles the signal 6 db, and doubling the noise only raises it 3 db for an improvement of 3 db Signal vs noise.

Also, remember, 3db is per doubling. 2 ADC's is 3 db. 4 ADCs is 6 db. 8 ADCs is 9 db. So to get large improvements you'll quickly get into large ADC arrays being necessary.

 

[DLS79]

As far as software compatibility more DAWS/etc need to support it though (and recording with it directly) https://www.sounddevices.com/32-bit-float-applications/

I'm sure support will come, the industry isn't exactly fast moving when it comes to adopting new technology.

 

[bennetng]

I'm sure support will come, the industry isn't exactly fast moving when it comes to adopting new technology.

Most DAWs, including the "doesn't look like a DAW" Audacity support import/export and processing in 32-bit float. The $60 Reaper supports 64-bit float.

Also, I've been using wavpack, a lossless codec with 32-bit float support for two decades, actually, earlier than the first time I started to use flac (I used APE before that).
http://www.wavpack.com/

 

[DLS79]

Most DAWs, including the "doesn't look like a DAW" Audacity support import/export and processing in 32-bit float. The $60 Reaper supports 64-bit float.

I was referring more for video, as that is who the 32 bit feature is geared towards.

for example see:
https://www.sounddevices.com/32-bit-float-applications/

 

[GTAXL]

@amirm I have a Zoom F4 and a Zoom H4n Pro. The H4n Pro only does mic and instrument level though. The Zoom F4, unlike the F6 has a real line level input, direct to the ADC and bypasses the preamp.

I also have a DAP, Fiio X5 III as well. I don't know if you'd be interested in testing any of these, but I do use the Zoom F4 often for videos as well as use the Fiio X5 III regularly. I'd like a turnaround of 1 or 2 weeks if not faster.

 

[amirm]

I also have a DAP, Fiio X5 III as well. I don't know if you'd be interested in testing any of these, but I do use the Zoom F4 often for videos as well as use the Fiio X5 III regularly. I'd like a turnaround of 1 or 2 weeks if not faster.

Thanks a lot for the offer but I just can't turn around review items that fast.

 

[amirm]

There were requests to test the F6 in stand-alone mode. I initially started testing it by recording content but found a simpler pass through mode from Line In to Line out/Headphone Out. Note that there are NO computer connections this time. The F6 is operating on batteries alone and is being fed using balanced XLR:

As you see, that hugely rising low frequency noise remains. It is a shame as distortion is not too bad at -8 dB or so.

 

[Blumlein 88]

There were requests to test the F6 in stand-alone mode. I initially started testing it by recording content but found a simpler pass through mode from Line In to Line out/Headphone Out. Note that there are NO computer connections this time. The F6 is operating on batteries alone and is being fed using balanced XLR:

View attachment 81286

As you see, that hugely rising low frequency noise remains. It is a shame as distortion is not too bad at -8 dB or so.

Any way we can get you to test in 32 bit float the input to the SD card. Then export the file and see if it really can't be clipped. Not that I expect it will improve the noise situation or get close to 25 bit dynamic range. Just wondering if it will scale the various input levels so it works not to clip any signal as claimed. That is the reason anyone who purchases this recorder would choose it over another Zoom model or other brand recorder.

@amirm

And boy it really is looking bad on batteries. That isn't really sufficient to be usable that I see.

 

[bennetng]

The key to clipping prevention for a product like this is not really floating point, since under the hood it is still running behind two or more integer based ADCs, and in physical and analog domain you simply can't have 770dB headroom. It's more like how the auto-gain math is done before exporting to the destination format. The fact is you can have something like 48-56 bit accumulator in fixed-point ICs, and reduced to 24/32 bits integer to storage devices so there would be no clipping as well.

But 32-bit float storage format is a nice choice since the DAWs we have today mostly support at least 32-bit float, if storing in 32-bit fixed, when you open it with a 32-bit float DAW, it can only see 25 bits, not all DAWs support 64-bit float.

 

[amirm]

The key to clipping prevention for a product like this is not really floating point, since under the hood it is still running behind two or more integer based ADCs, and in physical and analog domain you simply can't have 770dB headroom.

Indeed in pass through mode I tested it would easily clip if you turned up the gain.

 

[Blumlein 88]

Indeed in pass through mode I tested it would easily clip if you turned up the gain.

You have to send it to the SD card for their 32 bit float. Unfortunately there is no pass thru for that on the F6.

 

[Blumlein 88]

The first 5 minutes of the video review posted here earlier in the thread goes over using the recorder in 32 bit mode.

https://www.audiosciencereview.com/...table-field-recorder-review.15668/post-500962

 

[bennetng]

https://www.audiosciencereview.com/...y-of-software-volume-control.5922/post-172865
A simple experiment everyone can do without having a field recorder, to experience audible differences with floating point processing.

 

[GTAXL]

Thanks a lot for the offer but I just can't turn around review items that fast.

Thanks for responding. Say I don't send them all at once, I just send the Zoom F4 for now. How long do you think the turnaround would be? I'm very curious how it performs as Zoom doesn't list any specifications, just dynamic range, ADC dynamic range is 120dB or greater.

 

[GTAXL]

There were requests to test the F6 in stand-alone mode. I initially started testing it by recording content but found a simpler pass through mode from Line In to Line out/Headphone Out. Note that there are NO computer connections this time.

You really should test field recorders by recording a file to the SD card and analyzing said file. Otherwise you'll be testing the performance of the headphone amplifier instead, which usually on these Zoom products isn't implemented as well as the ADC. Most field recorders you should record 24-bit 48kHz as that's what most people for video production use. Since this one specialized 32-bit float I would use that if your analyzer can support it at 48kHz. The F6 does have a 3.5mm line out, but oddly Zoom shows the headphone amp as having better DAC dynamic range. 108dB vs 95dB. Still I would analyze files recorded to the SD card. You said you initially started but later switched to the headphone output, was performance about the same in your initial tests??!

 

[bennetng]

Here is my Sony MZ-R3 (1995) MD walkman's line-in directed to toslink output [the opposite: toslink input directed to line-out] without ATRAC compression. Obviously everything is 16-bit but still better than what I see in this review.

 

[GTAXL]

Another thing to note is the line out isn't a direct rail after the pre-amp, the line out will be fed from a DAC so you are able to hear any digital processing you did and also playback files. So testing the line out, you effectively test the DAC implementation, not the ADC. Test the headphone output you test the headphone amplifier. Both of these are most used simply for monitoring purposes only during recording to make sure levels are right, etc. I'm not sure if the F6 has a block diagram but my F4 did so I understand how it's laid out internally. The XLR inputs are always mic-level, and the TRS inputs are line-level, but really just a ~20dB pad in front of the preamp on most of their recorders. The F4 has a real line input, direct to the ADC.