new tiny Zoom F3 recorder (1 Viewer)

[Jon.Bryant]

Why do you think you need 192kHz?

I don’t think you do. But as memory is cheap, for those who feel the need, it is not a crime to record at a high sample rate and export later to a lower rate. You can always reduce the sample rate in an export, but increasing the sample rate in an export doesn’t add anything.

 

[Joop99]

I don’t think you do. But as memory is cheap, for those who feel the need, it is not a crime to record at a high sample rate and export later to a lower rate. You can always reduce the sample rate in an export, but increasing the sample rate in an export doesn’t add anything.

Downsampling comes at a price, better record with the right sample rate. If you want to distribute it on CD, then this will be 44.1kHz. If this is not possible because you record bat sounds which are in a much higher frequency range, then you have to use a higher sampling rate, but downsampling is then not a big issue.

 

[Jon.Bryant]

Joop, can you explain further?

Firstly, why does down-sampling come at a price - and do you think this depend on the software used? My understanding is that with a WAV file (and presumably a good export engine), there should be no real loss of quality across most of the frequency range if you sample down. There would however obviously be a loss of data at very high frequencies. This brings me to the second clarification - if you use a very high rate to record bats (with an appropriate mic) how can you sample down with no 'big issue' - surely you would lose the high frequency sound you are trying to record. I don't think you can record a bat at 50KHz with a sample rate of 192KHz (circa 4 samples per wave cycle), then sample down to 44.1KHz (less than 1 sample per wave cycle). Or do you mean you could change the pitch (to make the bat sound audible to the human ear) and then save at a lower sample rate - the later makes sense and pitch shift is used by some bat detection devices to make bat sounds audible to humans.

 

[Jon.Bryant]

Answering my own question, it does appear as if the sampling engine (Sampling Rate Conversion algorithms) can have a minor impact on quality. See SRC Comparisons for test of SRC's.

If I look at some of the software I use, then Spectral Layers 9 seems the noisiest, with Audacity, Reaper and Audition all pretty similar (and marginally better) and WaveLab 10 the best. That said the noise of the 1KHz -0.1dBFS test, converting from 96KHz to 48KHz, is a wide frequency but less than -140dBFS. The noise would therefore be practically nothing (and inaudible). The SNR would be drowned out by the SNR of any mic.

The Infinite Wave website info states 'The simplest SRC algorithms change sampling rate by an integer factor. When the sampling rate is reduced by N times, the Nyquist frequency (one half of the sampling rate) is also reduced by N times, reducing the bandwidth. A low-pass filter has to be used in order to prevent aliasing. It will suppress all of the frequency components above the future Nyquist frequency. After the filtering, digital samples are decimated by N times. This operation preserves the spectrum below the new Nyquist frequency.' I have highlighted sections of the text, which state that high frequency sounds will be suppressed, but lower frequency sounds preserved.

 

[S_Man]

Your quote and basic concepts suggest that downsampling by an integer factor eg from 192 KHz to 96 to 48 KHz, is easier to do, on the other hand converting to intermediate sampling rates (eg 192 to 44.1KHz) may add some complexity... (I'm not addressing the problem of the highest frequencies potentially suppressed by downsampling, here!) It is just intuitively "obvious", although I'm not sure anyone could hear the difference, and some editing software may not follow this logic anyway (as your link suggests)...
I personally like 48 KHz, but I see it's now a "minimum" for e-Bird "If your recorder supports it, a sample rate of at least 48 kHz and a bit depth of 24 bits is also recommended." so maybe I should use 96KHz instead... and of course 96-48kHz is not ideal if your goal to make CD's! Not sure about Xeno-Canto requirements/advice about this?
So my idea is that chosing a recording sample rate of 1 x, 2 x or N x the final sample rate could be the best choice, provided there are no other factor/criteria.

 

[Jon.Bryant]

on the other hand converting to intermediate sampling rates (eg 192 to 44.1KHz) may add some complexity

Good point. The same website states;

'To increase the sampling rate by M times, the signal is interpolated (interleaved) by zeroes. This preserves the spectrum below the Nyquist frequency, but creates spectral images (copies) above the Nyquist frequency. These spectral images are filtered out by a low-pass filter...'

... For resampling by a fractional ratio, as in 96 kHz to 44.1 kHz, upsampling and downsampling by integer ratios can be combined (e.g. 44100 = 96000 * M / N = 96000 * 147 / 320).'

The website does not include any info on performance for combined up and down sampling, but given the data on down sampling, I suspect the added noise is still not of real consequence (or really audible).

The website also states

'Sample rate converters (SRCs) are often used when working with digital audio, even when the user isn't aware of it. Almost every ADC or DAC converter, whether it is an expensive external unit or a cheap sound card, is working with so called "oversampling", performing A/D or D/A conversion ..... Software sample rate conversion can happen transparently for the user, inside a sampler or during mixing of audio streams at different sampling rates by the operating system, or at the user's command (i.e. converting a project from 96 kHz to 44.1 kHz during the mastering of an Audio CD.)'

I have underlined the sections that suggest that conversion may be transparent to the user. It seems that what when you think you are listening to a recording on a computer at a certain sampling rate, you may not be!

Xeno Canto only allows upload of MP3, so conversion from Lossless to a Lossy format, which is less ideal. I always export from WaveLab at VBR Highest Quality - not sure this is correct, but the conversions generally sound OK. When I have tried some lower quality conversions, the resultant sound has not always been good.

If EBird require 48KHz and 24bit, I think they are probably over egging it - particularly with the bit rate. I think 24 bit is really for studio recordings, when you may want to be able to hear a pin drop! But as I said previously, if they have the server space (and users don't get bored by the upload time) then why not?

Cheers

Jon

 

[davpen]

Xeno Canto only allows upload of MP3,

Xeno-canto now allows .WAV files

 

[S_Man]

If EBird require 48KHz and 24bit, I think they are probably over egging it - particularly with the bit rate. I think 24 bit is really for studio recordings, when you may want to be able to hear a pin drop! But as I said previously, if they have the server space (and users don't get bored by the upload time) then why not?
...

Agree with all points. E-bird 48KHz 24bit is only a recommendation, maybe they do some editing on their side, so the more, the better (some software/plugins may work better with higher rates/depth, I'm quoting musicians on forums here...)

Are you happy with your F3, any experience with moist environment? Also batteries seem short lived even when not recording (eg idle with 1 x 48V microphone, and following the manual's tips)...

 

[Jon.Bryant]

Are you happy with your F3, any experience with moist environment? Also batteries seem short lived even when not recording (eg idle with 1 x 48V microphone, and following the manual's tips)...

So far so good, but I have not been using the recorder as extensively as a should. Interested to know why you ask about 'moist environments' - from a quick web search I can't see any complaints about the recorder acting up in humid conditions. Do you know of any problems?

I don't like the position of the battery compartmentin the F3. Basically, if you use the rails to mount the recorder on any kind of rig, the compartment is impossible to access. I note that the manual shows that battery life is massively impacted by phantom power, headphones, backlight etc. With everything off Alkaline batteries should give you 8 hours when recording 2 channels at 48KHz 32bit, but is you have headphones and phantom power on this dops to 2 hours. Where I record, it is normally too noisy except for early morning, so it is not too much of an issue - I don't have the option for long recording sessions. But if you want an all-day session, probably best to use Lithium batteries or use a power bank to charge the unit - you sacrifice a bit of potability, but some power banks are quite small. This is not really unique to the F3 - a recorder like the Sound Devices Mix Pre will gobble batteries quickly (but that said there are more power options for that recorder and the batteries are generally accessible in a carry case).

 

[S_Man]

So far so good, but I have not been using the recorder as extensively as a should. Interested to know why you ask about 'moist environments' - from a quick web search I can't see any complaints about the recorder acting up in humid conditions. Do you know of any problems?

I'm just worried about a sudden rain, sometimes it comes like a torrent. Doors for SD card and bluetooth are covered with rubber, but the usb-C and batteries aren't. Also these buttons feel a bit "flimsy". Of course you won't record much sounds when it's raining, but for outdoor use - just like cameras - it's always better to have some protection? Water, sand, there are even tiny bugs that enter devices and grow inside (seriously, they do that)!

I don't like the position of the battery compartmentin the F3. Basically, if you use the rails to mount the recorder on any kind of rig, the compartment is impossible to access.

Same here, I strap mine against a pistol grip, and you also need to remove the strap to access the battery compartment. Design and form factor could be better, although it makes some sense as it is...

I note that the manual shows that battery life is massively impacted by phantom power, headphones, backlight etc. With everything off Alkaline batteries should give you 8 hours when recording 2 channels at 48KHz 32bit, but is you have headphones and phantom power on this dops to 2 hours. - I don't have the option for long recording sessions. But if you want an all-day session, probably best to use Lithium batteries or use a power bank to charge the unit - you sacrifice a bit of potability, but some power banks are quite small. This is not really unique to the F3 - a recorder like the Sound Devices Mix Pre will gobble batteries quickly (but that said there are more power options for that recorder and the batteries are generally accessible in a carry case).

I get much shorter times, like 2-3 hours with Alkaline and 1 x 48 V microphone. No lights, mono, 48 KHz. With a pair of AA in reserve it should be enough for most cases but sometimes I'd like to keep it turned on for longer, to catch unexpected sounds with the buffer. From what I've seen, even without recording, batteries are drained quite fast, phantom power most likely a good part of it, but I have only one xlr directional mic, unpowered so I can't compare.

Don't get me wrong, I'm quite happy with this F3, I'm just trying to make sure I have everything "covered" in case I'd use it more extensively.

Where I record, it is normally too noisy except for early morning, so it is not too much of an issue

Noise pollution becomes creepy with this kind of equipment! Our brain is clearly doing a huge filtering work all the time!

 

[Jon.Bryant]

Our brain is clearly doing a huge filtering work all the time!

Yes, quite amazing - we can hear a faint bird call in the cacophony of noise. But on the flip side, we can equally become fixated with the 'unwanted noise'. Once I have heard something that distracts from a recording, I find it virtually impossible to then mentally filter it out!

I take the point on the weather tightness of the F3. Even if you are not recording in the rain, I would not want to get it soaking wet. I think you need to carry a rain cover or bag to place the recorder in just in case.

Are you using rechargeable or single use AA batteries. Rechargeables vary a lot in capacity (I have some cheap ones at 1300 and more expensive at 2500 mAh - so almost double capacity). That said, I think even 2500mAh is less than alkaline. Battery performance is also impacted by temperature (unless you use Lithium, which will tolerate low temperatures without a missive impact on performance).

Cheers

Jon

 

[Joop99]

Mind that AA batteries are 1.5Volt and rechargeables are 1.2Volt. No matter what they invented with chips, you have to go nominal from 3Volt (2 AA) to 48Volt for phantom. There will be some overvoltage, because the regulator eats few volts for regulating. So you need 17 times the 3Volt, you get then 51Volt with spare for the regulator. The maximum power supplied by phantom power is set to 10mA pro mic channel, that is 0.01A. Due to working the 3Volt up to 51Volt you get at the 3Volt a power consumption of 17*0.01A=0.17A or 170mA pro channel. Two channels will be 340mA for phantom power only! That's a heavty load for two AA. It will be likely more because the electronics for making from 3Volt to 51Volt (or so) does also use power (that's not the same circuit as the regulator for 48Volt).

 

[Bartgeier]

To add a data point regarding battery runtime of the Zoom F3, I get about 5,5 - 6,5 hours of constantly pre-recording or recording with this setup:

• one 24V phantom powered microphone with 4x PUI AOM 5024 mic capsules in parallel
• second XLR deactivated
• no headphones
• display light off after 1 minute
• two Eneloop Pro rechargeables with 2500 mAh each

Not great, not bad. Unfortunately, there is no such thing as standby. Phantom power is always on, even when you are listening to your recordings. Some older Zoom devices offered a power saving option that deactivates phantom power when in replay mode.

 

[S_Man]

@Jon.Bryant That brain filter is tricky, true! Right now I'm using my camera strap to wear the mic on its grip with the F3 velcroed on it. But keeping a plastic bag in the pocket may be useful, or buy a new bag... I also found tiny covers for 3.5mm and usb ports, I'm curious to see how long before I lose them.

So far I've only been using Lidl/Aldi batteries which have been the best for the price (used mainly for other devices), but I'll try rechargeables again for the F3, Eneloop are often recommended. From my experience and as described by @Joop99, rechargeables usually didn't last as long as good Alkaline, and I used a lot of them, but hopefully technology is a bit better now than 15 years ago, with 2000+ mAh capacity?

@Bartgeier 5.5 hours is great, keeping a 2nd pair in the pocket should cover any daily needs. My microphone needs 48V (user manual) so I should expect half of that, which is exactly what I get with Alkaline (same settings, although I sometimes forget to disable the 2nd channel).

A 24V microphone or one with its own power would be more appropriate for the F3, but I'm happy with its specs, I even get some sounds slightly above 20 KHz (edi: not sure they come from birds though, maybe just happened synchronously)...

 

[Joop99]

@S_Man It has nothing to do with the capacity of a battery. You have to go to 48Volt or if available 24Volt if your microphone can work on the lower voltage. If you have only 2AA then you have with batteries 3Volt or with rechargeables 2.4Volt. So you have to go from let's say 3Volt all the way up to 24V. You need some more because the 24V needs to be regulated. So this will be 9 times 3Volt with 3Volt for the regulator. The standard phantom power is 48V or 24V at 10mA. This is (24+3) * 10mA = 0.27Watt / 3Volt = 90mA. So your 2AA sees a load from only 1 channel phantom power of 90mA. If your battery pack has a capacity of 2500mA(h) then you could in theory run more then 27 hours, this will be far less because it is impossible that you only use phantom power without recorder. The recorder F3 uses about 225mA (1900mAh / 8.5hours - page 104 manual). So with recorder and 1 channel phantom on 24V you use 225mA + 90mA = 315mA. With 1900mAh this means a continous recording of about 6hours at best, you have also to calculate with temperature and as I don't know what the system uses exactly it will be probably less then the calculated 6 hours. The calculated time is in accordance with the manual on page 104.

 

[S_Man]

To add a data point regarding battery runtime of the Zoom F3, I get about 5,5 - 6,5 hours of constantly pre-recording or recording with this setup:

• one 24V phantom powered microphone with 4x PUI AOM 5024 mic capsules in parallel
• second XLR deactivated
• no headphones
• display light off after 1 minute
• two Eneloop Pro rechargeables with 2500 mAh each

Not great, not bad. Unfortunately, there is no such thing as standby. Phantom power is always on, even when you are listening to your recordings. Some older Zoom devices offered a power saving option that deactivates phantom power when in replay mode.

Trying with rechargeables this time (eneloop 1900 mAh) same settings as yours, but 1 x 48 V, 2 mA microphone, I get at least 5 hours of mixed recording/stand-by, at which point I stopped as it was quite cold...
I was a bit worried after my first test, but that sounds good enough for me, although eneloop Pro and their higher capacity should last longer, of course.

NiMH are supposed to be less sensible to temperatures than Alkaline, maybe that's the reason for the longer times I get now, despite the fact that the Alkaline I used are supposed to provide higher voltage and capacity (1.5V 2000mAh vs 1.2V 1900mAh for my eneloops), it was cold when I used both types.

Then you have the microphones, some are more power thirsty than others, I'm being told.

The sample rate may also be a factor, since they mention it in the operation manual...

That's good enough for me. A 15-20Ah powerbank should provide at least 24 hours of continuous field recording, for those who practice that kind of thing...

 

[Avetarda]

Would you recommend this recorder for use with a parabolic mic? The difference is evident?

 

[Bartgeier]

Would you recommend this recorder for use with a parabolic mic? The difference is evident?

Yes and no - it certainly is a good recorder. 32bit wavs are useful as is the 6s pre-record option. Also it seems completely immune to RFI.
Sound-wise I must say I was expecting more. There is hardly any difference compared to a Olympus LS-3 with PIP in a similar parabolic setup which also has the added advantage of better battery life and smaller size. I do prefer the F3's XLR-plugs over 3,5mm jacks though.

 

[Avetarda]

I've the zoom h5 with 3,5mm jacks. Forgetting to use gain can be great. And if I've better quality of the recordings, volume... better. For convenience, I'm considering buying it.

 

[Jon.Bryant]

Would you recommend this recorder for use with a parabolic mic? The difference is evident?

I think it is good with a parabola, however, I don’t think you will see an amazing sound difference - THD, signal to noise ratio etc are pretty good on most modern recorders.

For me the benefits are easy of use - 32 bit float (so no need to monitor), 2 No. XLR inputs (obviously if you have a parabola with PiP mics the F3 is no good) and small size meaning you can mount it on the mic handle. See Small (But Good?) Part 5 where I did an initial review of using the recorder mounted on a Telinga parabola.