• Welcome to BirdForum, the internet's largest birding community with thousands of members from all over the world. The forums are dedicated to wild birds, birding, binoculars and equipment and all that goes with it.

    Please register for an account to take part in the discussions in the forum, post your pictures in the gallery and more.
Where premium quality meets exceptional value. ZEISS Conquest HDX.

new tiny Zoom F3 recorder (3 Viewers)

Dale Forbes

Swarovski Optik, Austria
Austria
Just seen this news on the Zoom F3 ZOOM F3 - Newsshooter

absolutely tiny
2 XLR inputs
32-bit float
"High-quality mic preamplifier that achieves an ultra-low noise floor of EIN -127dBu or less"
"ZOOM claims that the battery life when using (2x)AA batteries is up to 8 hours."

I love my MixPre 3 II, but if I did not have it, I'd seriously be looking forward to the F3
 
You get the feeling that we may be in for a number 32 bit float announcements... though anyone that had a Tascam Portacapture X8 for sale when I last looked is now back to waiting for another batch, I suspect that the market will not be flooded with this in the short term either.

But this being small, functional, simple and seemingly a good battery life should meet a number of peoples needs.
 
Lithium batteries up to 18 hours, and 192khz sampling rate.

I might need some advice here, in due course, on a compatible microphone for recording bats up to ~96khz, with this.
 
Interesting option. Looks like a good compact solution.

I have been wondering if 32 bit float recorders will open up the opportunity for minimal setups - basically a mic, a short cable and the recorder. One photo in the press relief shows basically what I was thinking of - the recorder attached to boom pole with a shotgun mic, but no headphones in sight!, Could be tricky getting direction spot on with a parabola, but i think a shotgun would be more forgiving. My shotgun has a pistol grip, so just need to think how I could attach on on near the mic without obstructing the grip. I suppose that attaching it to the underside of the wind screen, wouldn't upset the acoustics that much. Then no more getting in knots with cables or sweaty necks or hot hears from your headphones - just turn on at the start, point at whatever appears and hope for the best!

I think the other 32 bit recorders are generally a bit too big to fit to the mic handle or boom pole (and my MixPre definetly is!). The Tentacle track E and the Zoom F2 can of course be used in this manner, but these only have lockable mini jack connections and less gain, as they are designed for lavalier mics. The F3 therefore definitely looks like something to consider further.
 
Interesting option. Looks like a good compact solution.

I have been wondering if 32 bit float recorders will open up the opportunity for minimal setups - basically a mic, a short cable and the recorder. One photo in the press relief shows basically what I was thinking of - the recorder attached to boom pole with a shotgun mic, but no headphones in sight!, Could be tricky getting direction spot on with a parabola, but i think a shotgun would be more forgiving. My shotgun has a pistol grip, so just need to think how I could attach on on near the mic without obstructing the grip. I suppose that attaching it to the underside of the wind screen, wouldn't upset the acoustics that much. Then no more getting in knots with cables or sweaty necks or hot hears from your headphones - just turn on at the start, point at whatever appears and hope for the best!

I think the other 32 bit recorders are generally a bit too big to fit to the mic handle or boom pole (and my MixPre definetly is!). The Tentacle track E and the Zoom F2 can of course be used in this manner, but these only have lockable mini jack connections and less gain, as they are designed for lavalier mics. The F3 therefore definitely looks like something to consider further.
I'm seeing pics with it strapped to wrists. Same wrist as the hand holding the mic presumably!!
 
BHP now are advertising at USD349.99. Normally the USD and GBP prices seem to match 1 for 1, so I suspect this will be the £350 mark in the Uk. Shouldn’t be too long to wait now.

Does anyone know any details on the gain level? I can’t see it in the spec sheet. The bumf suggest that the gain is fixed (or perhaps it just means does not need adjustment during recording), but it would be nice to know how much gain is applied (and associated noise). I would like to see that the gain level will be sufficient to result in a strong signal when recording birds with my shotgun or parabola. I know it depends how close, but ideally when hand holding a mic, I would like to see ballistics pretty high and that I was sometimes able to get the sound going over 0dBf. If this is designed for recording louder sounds, then perhaps the gain will be insufficient to see the benefit with 32bit float. If it is designed to be versatile I would have through there would be a pad, so you can change gain for different scenarios. Perhaps questions that will only be resolved when the manuals are available to read.
 
In answer to my own question, I found a Utube video, which shows the device has a zoom factor. This seems to be gain control by a different name. The factor doubles up so 1,2,4,8. The video showed zoom factors up to 1024, with the latter sounding noisy, but not sure how much was mic noise.

I wish gain on devices could just be in dB, as this allows direct comparison and simplify use - you would know if you apply 30dB on one machine you should get substantially similar results if you apply 30dB on another. Now there is a need to play with unknown Zoom factors to work out what works. 3dB is a doubling of sound intensity and if the doubling of zoom factor correlates, then I think you have circa 30dB of gain control (from an unknown base gain at Zoom factor 1)
 
Post at speed - repent at leisure!

Just realised that my concerns above are rubbish! Somehow I was thinking that increasing the levels in post would amplify mic noise, so that 32bit float was only really of benefit for bird recordings to prevent clipping. My rationale then was that gain levels needed to be high enough to create clipping, or it was a waste of time.

I now figure that if a mic produces an analogue voltage and that noise is part of that signal, whether gain levels are high or low makes no difference to this ratio of signal to noise - that is if the bit depth is high enough and noise floor is low enough (which it would be with 32bit float), to accurately record the signal and noise levels. With 16 bit (or even 24), if the gain was very low, there would be potential that the element of mic noise would be lower than the noise floor - in that case when amplified in post, you would amplifying the quietest sound possible with that bit rate - so the the noise floor and not the quieter mic noise, so 16 or 24 bit would be noisier.

So if the actual ratio of signal to noise at the mic is good, then the recorder gain is somewhat irrelevant with recording 32bit float - that is if you ignore the need for monitoring. Obviously if the signal to noise ratio at the mic is poor (the bird is too distant or too quiet or the environment too noisy) then 32bit float can/t save the day.

Anyway, after a little though, I think the Zoom F3 remains an interesting option.
 
Just went on sale today for $349.99 in the U.S. Expected delivery by end of February.

Official product site – F3 Field Recorder

Product demo –

It uses the same preamps and A/D converters as the Zoom F6, so I would expect similar sound quality to the F6.

Intriguing recorder. Hope to see some bird recording reviews after people have spent enough time with it.
 
I am totally new about 32 vs 24 bit recording etc. In bird sound recording, what is the main advantage when using 32 vs 24? I have the Tascam with 24 bit and pleased with the results so far.
 
Hi Horukuru,

The bit depth determines the dynamic range of the recording. In a a nutshell this means it sets the difference between the loudest and quietest thing that you can record. With a 16 bit integer you can have values from 0 to 65,535. The number zero equates to 0dBFs (the loudest sound possible), and 65,535 represents the quietest thing (-96.3dB below full scale). With 24 bits you have more bits to store the number, so can now record values from 0 to 16,777,216. The quietest thing you can then record is therefore much quieter at -144.5dB below full scale - but the loudest is still a value of zero, representing 0dBfs.

32bit float (not just 32 bit) is not an integer representation of a number, but uses floating point format. The big thing here is that floating point numbers can store negative as well as positive values. Providing the sound is not so loud as to create microphone distortion, you can now record a negative value which represents a sound louder than 0dBFs. With 16 or 24 bit recordings this would result in sound distortion due to 'clipping', and ruin the recording. In post production software you can pull back a 32bit float recording so the maximum sound is below 0dBfs, and save as 24 or 16 bit without any sign of clipping. 32 bit float is therefore very good for when you have unexpected loud noises, which would otherwise blow your recording levels.

The 32bit float format is used in computing and has good precision - a 16bit float number would have poor precision and therefore is of little if any use in computing and would not be suitable to represent variations in sound levels in a recording. 32bit float has good precision, but is also capable of holding very large numbers, and therefore is capable of storing a dynamic range from +770dB to -758dB either side of 0dBfs. This is a far greater dynamic range than you could ever need, so it could be argued that 32bit float is overkill.

However, the other advantage of 32bit float relates in part to the ridiculous noise floor of -770dB below full scale. Say you recorded a sound at the top threshold of you mic (circa 120dB) and that your mic is relatively quite and has an inherent noise of 10dB - I think this would mean that the signal to noise ratio would be 110dB, so 24 bit would be more than enough in theory. The mic operates independently from the recorded, and merely produces a voltage representing the sound, which is converted and recorder as a digital value. On the recorder the gain would normally be adjusted so the meters peaked close to 0dbFs - i.e. the largest voltage from the mic is set to equate to circa 0dBfs . But what is the gain level was set very low by accident, so that the peak voltage equated to only -50dBFs? Now the noise element of the recording would be very quiet indeed (-160dBFs), and would not be able to be accurately recorded with 16 or 24 bits. If you tried to increase the level of the recording in post production, you would also be applying gain to the noise floor rather than the quieter inherent noise from the mic. The recording would therefore sound noisy. With 32 bit float however, because you can record an extremely large numbers (representing very quiet sounds), the recording cannot really fall below the noise floor. This means that in post processing you can increase the level and recover the actual signal to noise ratio - rather than having the noise floor dominate and ruining the recording.

Hopefully you are still with me, but what the above means is that the actual gain that is applied in the field when making a 32bit float recording is not really of consequence - you can adjust over loud or very quite recordings in post production, and cannot adversely effect the original signal to noise ratio. In fact the Zoom F3 does not seem to even have a gain control! This is great in theory for birds, particularly for short sounds, when you simply don't have the opportunity for a sound check and adjustment of gain levels.

Of course all is not that simple, as you still need to make sure that you are close enough to the bird to get a good signal to noise ratio in the first place - Not even the magic of 32 bit float can solve the problem where the bird vocal is too quite and the signal to noise ratio poor.

The downside of 32bit float is shorter battery life and larger file size - though the latter is less of a problem given the price of memory.

Regards

Jon Bryant
 
Just spoke to a UK supplier and looks like Zoom has made a bit of a mess with the launch date information. The initial date of end of Feb for UK distribution, now looks like it could be mid summer!
 
Hi Horukuru,

The bit depth determines the dynamic range of the recording. In a a nutshell this means it sets the difference between the loudest and quietest thing that you can record. With a 16 bit integer you can have values from 0 to 65,535. The number zero equates to 0dBFs (the loudest sound possible), and 65,535 represents the quietest thing (-96.3dB below full scale). With 24 bits you have more bits to store the number, so can now record values from 0 to 16,777,216. The quietest thing you can then record is therefore much quieter at -144.5dB below full scale - but the loudest is still a value of zero, representing 0dBfs.

32bit float (not just 32 bit) is not an integer representation of a number, but uses floating point format. The big thing here is that floating point numbers can store negative as well as positive values. Providing the sound is not so loud as to create microphone distortion, you can now record a negative value which represents a sound louder than 0dBFs. With 16 or 24 bit recordings this would result in sound distortion due to 'clipping', and ruin the recording. In post production software you can pull back a 32bit float recording so the maximum sound is below 0dBfs, and save as 24 or 16 bit without any sign of clipping. 32 bit float is therefore very good for when you have unexpected loud noises, which would otherwise blow your recording levels.

The 32bit float format is used in computing and has good precision - a 16bit float number would have poor precision and therefore is of little if any use in computing and would not be suitable to represent variations in sound levels in a recording. 32bit float has good precision, but is also capable of holding very large numbers, and therefore is capable of storing a dynamic range from +770dB to -758dB either side of 0dBfs. This is a far greater dynamic range than you could ever need, so it could be argued that 32bit float is overkill.

However, the other advantage of 32bit float relates in part to the ridiculous noise floor of -770dB below full scale. Say you recorded a sound at the top threshold of you mic (circa 120dB) and that your mic is relatively quite and has an inherent noise of 10dB - I think this would mean that the signal to noise ratio would be 110dB, so 24 bit would be more than enough in theory. The mic operates independently from the recorded, and merely produces a voltage representing the sound, which is converted and recorder as a digital value. On the recorder the gain would normally be adjusted so the meters peaked close to 0dbFs - i.e. the largest voltage from the mic is set to equate to circa 0dBfs . But what is the gain level was set very low by accident, so that the peak voltage equated to only -50dBFs? Now the noise element of the recording would be very quiet indeed (-160dBFs), and would not be able to be accurately recorded with 16 or 24 bits. If you tried to increase the level of the recording in post production, you would also be applying gain to the noise floor rather than the quieter inherent noise from the mic. The recording would therefore sound noisy. With 32 bit float however, because you can record an extremely large numbers (representing very quiet sounds), the recording cannot really fall below the noise floor. This means that in post processing you can increase the level and recover the actual signal to noise ratio - rather than having the noise floor dominate and ruining the recording.

Hopefully you are still with me, but what the above means is that the actual gain that is applied in the field when making a 32bit float recording is not really of consequence - you can adjust over loud or very quite recordings in post production, and cannot adversely effect the original signal to noise ratio. In fact the Zoom F3 does not seem to even have a gain control! This is great in theory for birds, particularly for short sounds, when you simply don't have the opportunity for a sound check and adjustment of gain levels.

Of course all is not that simple, as you still need to make sure that you are close enough to the bird to get a good signal to noise ratio in the first place - Not even the magic of 32 bit float can solve the problem where the bird vocal is too quite and the signal to noise ratio poor.

The downside of 32bit float is shorter battery life and larger file size - though the latter is less of a problem given the price of memory.

Regards

Hi Horukuru,

The bit depth determines the dynamic range of the recording. In a a nutshell this means it sets the difference between the loudest and quietest thing that you can record. With a 16 bit integer you can have values from 0 to 65,535. The number zero equates to 0dBFs (the loudest sound possible), and 65,535 represents the quietest thing (-96.3dB below full scale). With 24 bits you have more bits to store the number, so can now record values from 0 to 16,777,216. The quietest thing you can then record is therefore much quieter at -144.5dB below full scale - but the loudest is still a value of zero, representing 0dBfs.

32bit float (not just 32 bit) is not an integer representation of a number, but uses floating point format. The big thing here is that floating point numbers can store negative as well as positive values. Providing the sound is not so loud as to create microphone distortion, you can now record a negative value which represents a sound louder than 0dBFs. With 16 or 24 bit recordings this would result in sound distortion due to 'clipping', and ruin the recording. In post production software you can pull back a 32bit float recording so the maximum sound is below 0dBfs, and save as 24 or 16 bit without any sign of clipping. 32 bit float is therefore very good for when you have unexpected loud noises, which would otherwise blow your recording levels.

The 32bit float format is used in computing and has good precision - a 16bit float number would have poor precision and therefore is of little if any use in computing and would not be suitable to represent variations in sound levels in a recording. 32bit float has good precision, but is also capable of holding very large numbers, and therefore is capable of storing a dynamic range from +770dB to -758dB either side of 0dBfs. This is a far greater dynamic range than you could ever need, so it could be argued that 32bit float is overkill.

However, the other advantage of 32bit float relates in part to the ridiculous noise floor of -770dB below full scale. Say you recorded a sound at the top threshold of you mic (circa 120dB) and that your mic is relatively quite and has an inherent noise of 10dB - I think this would mean that the signal to noise ratio would be 110dB, so 24 bit would be more than enough in theory. The mic operates independently from the recorded, and merely produces a voltage representing the sound, which is converted and recorder as a digital value. On the recorder the gain would normally be adjusted so the meters peaked close to 0dbFs - i.e. the largest voltage from the mic is set to equate to circa 0dBfs . But what is the gain level was set very low by accident, so that the peak voltage equated to only -50dBFs? Now the noise element of the recording would be very quiet indeed (-160dBFs), and would not be able to be accurately recorded with 16 or 24 bits. If you tried to increase the level of the recording in post production, you would also be applying gain to the noise floor rather than the quieter inherent noise from the mic. The recording would therefore sound noisy. With 32 bit float however, because you can record an extremely large numbers (representing very quiet sounds), the recording cannot really fall below the noise floor. This means that in post processing you can increase the level and recover the actual signal to noise ratio - rather than having the noise floor dominate and ruining the recording.

Hopefully you are still with me, but what the above means is that the actual gain that is applied in the field when making a 32bit float recording is not really of consequence - you can adjust over loud or very quite recordings in post production, and cannot adversely effect the original signal to noise ratio. In fact the Zoom F3 does not seem to even have a gain control! This is great in theory for birds, particularly for short sounds, when you simply don't have the opportunity for a sound check and adjustment of gain levels.

Of course all is not that simple, as you still need to make sure that you are close enough to the bird to get a good signal to noise ratio in the first place - Not even the magic of 32 bit float can solve the problem where the bird vocal is too quite and the signal to noise ratio poor.

The downside of 32bit float is shorter battery life and larger file size - though the latter is less of a problem given the price of memory.

Regards

Jon Bryant

With 32 bit float however, because you can record an extremely large numbers (representing very quiet sounds), the recording cannot really fall below the noise floor. This means that in post processing you can increase the level and recover the actual signal to noise ratio - rather than having the noise floor dominate and ruining the recording. ---- This is exciting!!!
 
This week, I received my Zoom F3 to complete my first audio field recording kit:
  • Sennheiser MKE 600 microphone
  • Audio-Technica ATH-M50X closed-back headphone for audio monitoring
  • Rycote pistol grip
  • Zoom F3
It has been raining for 3 days here. So, I didn't have fully tested it yet. I'll probably write a small review in few weeks or months. However, I didn't have any experience with audio recording and 24 bits recorders.
 
Recently I purchased a Zoom F3 and have had some time to work with it in the field. It is a very nice little field recorder for capturing birds. I am using it with a Sennheiser MKH-416 shotgun mic, and use an Auray wind softie on the mic.

My previous setup was connecting the MKH-416 to an iPhone using IK Multimedia's iRig Pre2 mic preamp. The iRig Pre2 provided 48v phantom power for the mic. I used Voice Record Pro to record the audio set to 96 kHz sample rate, and both 24-bit and 32-bit (fixed). This setup worked well, was compact, light, and easy to carry around. The downsides were the self noise was higher than I cared for, and I was always having to reset the gain on Voice Record Pro every time I restarted the app.

The Zoom F3 only records in 32-bit float. There is no option to change to, or use, 16-bit, 24-bit, or 32-bit fixed bit depths. Because of the huge dynamic range capability of 32-bit float, there is no gain control on the F3. Setting and worrying about gain control is gone. At first it seems a bit strange not to have gain control, but in practice it works great. You cannot blow out a recording due ambient noises. This makes the F3 an incredibly easy recorder to use in the field. You set the sample rate, and off you go.

The preamps on the F3 are extremely quiet. Any noise floor issues with recordings will be due mic selection and the self noise level of the mic. Mic selection is very important to consider.

The switch to 32-bit float recording requires one to rethink what they have learned about input recording levels. Just like the move from analog to digital required people to get used to lower peak recording levels (resulting in more headroom), the switch to 32-bit float allows for what appears to be recording levels that shouldn't work. But they do because the amount of dynamic range and the amount of headroom one has to work with are so large. This is one reason a mic with low self noise is going to become even more important as part of the recording chain.

Because there is no need to worry about gain, the device's screen is essentially not used during recording. There is no recording information on the screen that one needs to monitor or watch. As a result the screen is a small LCD used to change settings basic settings such as sample rate, battery chemistry type, format the microSD card, activate and set phantom power, etc.

The F3 is powered by two AA batteries. I've tried both alkaline and rechargeable 2700 mha NiMH batteries. When using 48v phantom power to drive a mic, the power does go down. I've been on bird walks several hours long and have not depleted a set of batteries beyond 50%. If one is out for a full day, and is doing a lot of recording, then bring an couple extra sets of batteries. Using rechargeable batteries make sense with this setup. The F3 also has a USB-C input that can be used for both data transfer and powering the unit with 5v DC.

I've used the F3 and MKH-416 in some very windy conditions and am very impressed with how 32-bit float handles sudden gusts. I don't have a full blimp setup, just a softie. With the application of some highpass filter at 150 Hz or lower, the wind goes away and useable bird audio emerges. Because of how quiet the preamps are, one can then boost the volume quite a bit during editing. What I have been left with is the volume and pitch of the wind through the foliage competing against the volume of the recorded target bird. Of course there are times when the wind gusts are too much for just a softie, but the F3 seems to capture more useable audio in windy conditions than my previous setup.

In terms of audio editing software... Before you consider buying a Zoom F3, look at the website and review the list of 32-bit float compatible audio editing software for each type of computer and operating system.

For example, 32-bit float file support in Audacity is supported for both import and export under OS X (Mac) (requires OS X 10.15.1 or higher), but is not supported for either import or export under Windows 10.

I have a Mac and have successfully imported the F3's audio files into Audacity, edited them, exported them out as 32-bit float WAV files, and uploaded them to eBird.

Hope this helps provide some more information for folks thinking about the Zoom F3.
 
Last edited:
Recently I purchased a Zoom F3 and have had some time to work with it in the field. It is a very nice little field recorder for capturing birds. I am using it with a Sennheiser MKH-416 shotgun mic, and use an Auray wind softie on the mic.

My previous setup was connecting the MKH-416 to an iPhone using IK Multimedia's iRig Pre2 mic preamp. The iRig Pre2 provided 48v phantom power for the mic. I used Voice Record Pro to record the audio set to 96 kHz sample rate, and both 24-bit and 32-bit (fixed). This setup worked well, was compact, light, and easy to carry around. The downsides were the self noise was higher than I cared for, and I was always having to reset the gain on Voice Record Pro every time I restarted the app.

The Zoom F3 only records in 32-bit float. There is no option to change to, or use, 16-bit, 24-bit, or 32-bit fixed bit depths. Because of the huge dynamic range capability of 32-bit float, there is no gain control on the F3. Setting and worrying about gain control is gone. At first it seems a bit strange not to have gain control, but in practice it works great. You cannot blow out a recording due ambient noises. This makes the F3 an incredibly easy recorder to use in the field. You set the sample rate, and off you go.

The preamps on the F3 are extremely quiet. Any noise floor issues with recordings will be due mic selection and the self noise level of the mic. Mic selection is very important to consider.

The switch to 32-bit float recording requires one to rethink what they have learned about input recording levels. Just like the move from analog to digital required people to get used to lower peak recording levels (resulting in more headroom), the switch to 32-bit float allows for what appears to be recording levels that shouldn't work. But they do because the amount of dynamic range and the amount of headroom one has to work with are so large. This is one reason a mic with low self noise is going to become even more important as part of the recording chain.

Because there is no need to worry about gain, the device's screen is essentially not used during recording. There is no recording information on the screen that one needs to monitor or watch. As a result the screen is a small LCD used to change settings basic settings such as sample rate, battery chemistry type, format the microSD card, activate and set phantom power, etc.

The F3 is powered by two AA batteries. I've tried both alkaline and rechargeable 2700 mha NiMH batteries. When using 48v phantom power to drive a mic, the power does go down. I've been on bird walks several hours long and have not depleted a set of batteries beyond 50%. If one is out for a full day, and is doing a lot of recording, then bring an couple extra sets of batteries. Using rechargeable batteries make sense with this setup. The F3 also has a USB-C input that can be used for both data transfer and powering the unit with 5v DC.

I've used the F3 and MKH-416 in some very windy conditions and am very impressed with how 32-bit float handles sudden gusts. I don't have a full blimp setup, just a softie. With the application of some highpass filter at 150 Hz or lower, the wind goes away and useable bird audio emerges. Because of how quiet the preamps are, one can then boost the volume quite a bit during editing. What I have been left with is the volume and pitch of the wind through the foliage competing against the volume of the recorded target bird. Of course there are times when the wind gusts are too much for just a softie, but the F3 seems to capture more useable audio in windy conditions than my previous setup.

In terms of audio editing software... Before you consider buying a Zoom F3, look at the website and review the list of 32-bit float compatible audio editing software for each type of computer and operating system.

For example, 32-bit float file support in Audacity is supported for both import and export under OS X (Mac) (requires OS X 10.15.1 or higher), but is not supported for either import or export under Windows 10.

I have a Mac and have successfully imported the F3's audio files into Audacity, edited them, exported them out as 32-bit float WAV files, and uploaded them to eBird.

Hope this helps provide some more information for folks thinking about the Zoom F3.

When you do the recording, do you put the mic on the tripod or handheld?
 
When you do the recording, do you put the mic on the tripod or handheld?
I handhold the mic.

Some of the great benefits of the Zoom F3 (aside from 32-bit float) are the small size, light weight, and ease of use. For me adding a tripod would be a hindrance. I could not start recording as fast, get recordings in confined areas, etc. The only place a tripod mounted mic would be a benefit for me is if I was doing a long timeframe recording (10-60 minutes) of ambient sounds rather than targeting a specific bird - and that is not how I record.

Handholding does not induce any noise into the mic if you don't adjust your grip. Don't let the mic cord rub against anything, control your breathing, stand still, and you will not add any sounds.

For how we bird the recording kit needs to be small, lightweight, directional, and have good quality. It is something I carry with me all the time because I enjoy getting recordings, but also so that if a rarity is spotted it can be documented with either sound, a photo, or both.
 
Last edited:
I have a Mac and have successfully imported the F3's audio files into Audacity, edited them, exported them out as 32-bit float WAV files, and uploaded them to eBird.
I think that 32 bit float is really all about post production editing - it allows a signal above 0dBF to be pulled back and also allows a low signal to be normalised to -3dBF (as per Xeno Canto guidance) without adding significant noise due to the low the noise floor.

Post production and editing, you will probably not be able to tell the difference between 16 bit, 24 bit and 32 bit float outputs - after all audio CD's are only 16 bit, and we have been happy with the quality of music CD's for over two decades now.

I would therefore save as a 16 or 24 bit final version and then upload to EBird - quality will be substantially the same and it will save their server space,

Regards

Jon Bryant
 
Because there is no need to worry about gain, the device's screen is essentially not used during recording. There is no recording information on the screen that one needs to monitor or watch. As a result the screen is a small LCD used to change settings basic settings such as sample rate, battery chemistry type, format the microSD card, activate and set phantom power, etc.
I would agree that this is a real beauty of 32bit float, and have opted at times not to monitor recordings at all - just wing it and adjust in post. Are you using headphones at all to monitor the recording, or just relying on your judgement of whether the signal to noise (i.e. how loud the target is compared with the environment) will be OK? I have sometimes opted not to use headphones at all, although headphones can be useful to ensure directional mics (particularly parabolas) are aimed properly at the target.

I take the point on mic specifications and low noise, but I would caution this with the fact that very low noise mics are really for studio work (i.e. in a silent environment). Out in the field, I think that generally environmental noise is likely to drown out mic noise. 32bit float can't obviously solve the problem of a low volume recording in a noisy environment (and you may wish to monitor to avoid this), but the magic of not having to worry about gain or clipping is great.

Regards

Jon Bryant
 

Users who are viewing this thread

Back
Top