• 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.

Audiomoth or Song Meter (2 Viewers)

Avetarda

Well-known member
Spain
I am currently using Audiomoth 1.2.0 with a medium gain setting. I get recordings with noticeable background noise. Would Song Meter give me better quality recordings without that annoying background noise and better overall?
Is the microphone better? Both are 16 bytes. Or is it not worth the money?
What is the difference between the Song Meter Mini and the SM4?
 
What is the difference between the Song Meter Mini and the SM4?
I have been a fan of WildLife Acoustics for a long time, and have the SM4, Mini (Mk1) and Micro (Mk 1 and 2) recorders.

The latest comparison information Audio Comparison, has example recordings for each type of recorder and also charts which indicate that the SM4 and Mini 2 have similar characteristics. That said the the FAQ page states

The Song Meter SM4 and Song Meter SM Mini recorders use the same microphone elements for their audible-range recordings, and in many recording scenarios, they will produce recordings of very similar quality. That said, the more expensive SM4 uses more advanced recording circuitry that gives it an edge above the SM Mini in a few different ways:

The analog-to-digital conversion circuitry in the SM4 maintains a consistent signal-to-noise ratio regardless of sample rate, with 16 effective bits of recording depth. The SM Mini’s effective bit depth is still 16 bits when recording on a single channel at a sample rate of 16 kHz or lower, but at higher sample rates, the effective bit depth is reduced. For example, when recording on two channels at 24 kHz or one channel at 48 kHz, the effective bit depth is reduced to 15. This means that when using higher sample rates or recording on both channels at a time, there will be a slight increase in the background noise level. Depending on the noise level of your environment, this electronic noise may or may not be significant when compared to the surrounding acoustic noise.

The SM4 has a very steep anti-aliasing filter. This filter sharply cuts off any sound that is higher in frequency than the SM4 is able to accurately record based on its selected sample rate. Any frequency up to half of the sample rate is passed with next to no alteration. Any higher frequency that would cause aliasing in the finished recording is almost completely eliminated. In comparison, the Mini’s anti-aliasing filter is less steep and less flat. This means that the recorder’s sensitivity will be reduced between ¼ and ½ of the sample rate as the anti-aliasing filter starts to gradually kick in. Additionally, if there is strong noise present in the environment just above half of the selected sample rate, some of that noise will show up in the final recording as aliasing artifacts. In essence, this means that the Mini recordings will tend to exhibit more background noise towards the top of the recording frequency range when compared to an SM4.

For a typical use case of recording animal vocalizations in order to identify species, these differences may not be significant. However, for applications involving measurement of noise levels or where getting the largest possible signal-to-noise ratio is a concern, the SM4’s advantages are more worth considering despite the higher price.


The FAQ section also states that the SM Mini 1 and 2 have the same audio performance, which is odd, because, I am pretty sure that the old comparison charts for the SM Mini 1 used to be different from the SM4. I definitely feel that the SM4 results are better than the SM Mini (1) and I don't feel so concerned about pushing the sampling rate (as you are not going to loose bit rate and hence dynamic range).

The Micro is now relatively cheap, as they have fairly recently decided to sell this at a significant markdown ($249 down to $149) - Some UK retailers have also already followed suit. The mark 2 Micro has definitely changed - the frequency response if now much flatter - the mark 1 was more sensitive at about 4kHz, which was not all bad, as the sensitivity matched the calls of many passerines. But is is nicer to have a flatter frequency response, and to capture the whole environment more authentically.
 
Audiomoth performance testing
Paraphrasing from the link.

Two dramatic patterns should be noted in these results. First, the SongMeter Micro does not have a flat frequency response: a large peak around 7 kHz stands out 15-30 dB above the rest of the frequency response. While this could be problematic, the SongMeter Micro has a much lower noise floor than the AudioMoth, such that even the lowest parts of the frequency response may have plenty of signal to noise ratio compared to the noise floor of the device.

This is text from the comparison of the AudioMoth with the WA Micro mark 1. I think the '7kHz peak' (sorry I said 4kHz in my last post), is a thing of the past, as the mark 2 has a much flatter frequency response - see Mark 1 and Mark 2 Comparison.

The text states that the SM Micro 1 has a lower noise floor than AudioMoth. I understand the Mark 2 has a better signal to noise ratio, and that the Mini 2 is better again than the Micro 2. The SM4 only really comes into it's own when recording audio at 16 bit 48Khz (slightly above CD quality), when 16bits (and the noise floor) are maintained, whereas the Mini drops to 15bits with a slight increase in the noise floor.

Attached, as an example is a Greenish Warbler call taken in Qinghai, China with the SM Mini. The recording has been normalised to -3dB - if the bird had been a smidgen closer the noise would be reduced, but too close and it would have clipped. It would be fantastic if WA produced a 32bit float version to avoid such issues.
View attachment Greenish Warbler.wav
 
Soy fan de WildLife Acoustics desde hace mucho tiempo, y tengo las grabadoras SM4, Mini (Mk1) y Micro (Mk 1 y 2).

La última información comparativa Audio Comparison, tiene ejemplos de grabaciones para cada tipo de grabadora y también gráficos que indican que el SM4 y el Mini 2 tienen características similares. Dicho esto, la página de preguntas frecuentes dice:

Las grabadoras Song Meter SM4 y Song Meter SM Mini utilizan los mismos elementos de micrófono para sus grabaciones de rango audible y, en muchos escenarios de grabación, producirán grabaciones de calidad muy similar. Dicho esto, el SM4 más caro utiliza circuitos de grabación más avanzados que le dan una ventaja sobre el SM Mini en algunos aspectos diferentes:

Los circuitos de conversión de analógico a digital en el SM4 mantienen una relación señal-ruido constante independientemente de la frecuencia de muestreo, con 16 bits efectivos de profundidad de grabación. La profundidad de bits efectiva del SM Mini sigue siendo de 16 bits cuando se graba en un solo canal a una frecuencia de muestreo de 16 kHz o inferior, pero a frecuencias de muestreo más altas, la profundidad de bits efectiva se reduce. Por ejemplo, cuando se graba en dos canales a 24 kHz o en un canal a 48 kHz, la profundidad de bits efectiva se reduce a 15. Esto significa que cuando se utilizan frecuencias de muestreo más altas o se graba en ambos canales a la vez, habrá un ligero aumento en el nivel de ruido de fondo. Dependiendo del nivel de ruido de su entorno, este ruido electrónico puede o no ser significativo en comparación con el ruido acústico circundante.

El SM4 tiene un filtro anti-aliasing muy pronunciado. Este filtro corta bruscamente cualquier sonido que tenga una frecuencia más alta de la que el SM4 puede grabar con precisión en función de la frecuencia de muestreo seleccionada. Cualquier frecuencia de hasta la mitad de la frecuencia de muestreo se pasa casi sin alteración. Cualquier frecuencia más alta que pueda causar solapamiento en la grabación terminada se elimina casi por completo. En comparación, el filtro anti-aliasing del Mini es menos empinado y menos plano. Esto significa que la sensibilidad de la grabadora se reducirá entre 1/4 y 1/2 de la frecuencia de muestreo a medida que el filtro anti-aliasing comience a activarse gradualmente. Además, si hay mucho ruido presente en el entorno justo por encima de la mitad de la frecuencia de muestreo seleccionada, parte de ese ruido aparecerá en la grabación final como artefactos de aliasing. En esencia, esto significa que las grabaciones Mini tenderán a mostrar más ruido de fondo hacia la parte superior del rango de frecuencia de grabación en comparación con un SM4.

Para un caso de uso típico de grabación de vocalizaciones de animales con el fin de identificar especies, estas diferencias pueden no ser significativas. Sin embargo, para aplicaciones que involucran la medición de niveles de ruido o donde obtener la mayor relación señal-ruido posible es una preocupación, las ventajas del SM4 son más dignas de consideración a pesar del precio más alto.


La sección de preguntas frecuentes también indica que el SM Mini 1 y el SM 2 tienen el mismo rendimiento de audio, lo cual es extraño, porque estoy bastante seguro de que las antiguas tablas comparativas del SM Mini 1 solían ser diferentes de las del SM4. Definitivamente creo que los resultados del SM4 son mejores que los del SM Mini (1) y no me preocupa tanto aumentar la frecuencia de muestreo (ya que no va a perder la tasa de bits y, por lo tanto, el rango dinámico).

El Micro es ahora relativamente barato, ya que recientemente han decidido venderlo a un precio significativo (de 249 dólares a 149 dólares) - Algunos minoristas del Reino Unido también han seguido su ejemplo. El mark 2 Micro ha cambiado definitivamente -la respuesta de frecuencia es ahora mucho más plana-, el mark 1 era más sensible a unos 4kHz, lo que no era del todo malo, ya que la sensibilidad coincidía con las llamadas de muchos paseantes. Pero es mejor tener una respuesta de frecuencia más plana y capturar todo el entorno de manera más auténtica.
Thank you so much for the extensive information, John.
 
Hello,

If you are using audiomoth for birds, try using good quality external mics.

I use two PUI AOM 5024 L HD connected in parallel, it works better than with one.
Very interesting. And how do you connect 2 microphones? I have one that could be useful, a Primo EM258, but I'm not good at soldering, I could try it. On AliExpress they sell 3.5 mm jack connectors, I don't know if they can be useful or if you have to buy a specific one. On the Open Acoustic website they indicate that it must be SJ2-35464A-SMT-TR but I can't find much on the web
 
Paraphrasing from the link.

Two dramatic patterns should be noted in these results. First, the SongMeter Micro does not have a flat frequency response: a large peak around 7 kHz stands out 15-30 dB above the rest of the frequency response. While this could be problematic, the SongMeter Micro has a much lower noise floor than the AudioMoth, such that even the lowest parts of the frequency response may have plenty of signal to noise ratio compared to the noise floor of the device.

This is text from the comparison of the AudioMoth with the WA Micro mark 1. I think the '7kHz peak' (sorry I said 4kHz in my last post), is a thing of the past, as the mark 2 has a much flatter frequency response - see Mark 1 and Mark 2 Comparison.

The text states that the SM Micro 1 has a lower noise floor than AudioMoth. I understand the Mark 2 has a better signal to noise ratio, and that the Mini 2 is better again than the Micro 2. The SM4 only really comes into it's own when recording audio at 16 bit 48Khz (slightly above CD quality), when 16bits (and the noise floor) are maintained, whereas the Mini drops to 15bits with a slight increase in the noise floor.

Attached, as an example is a Greenish Warbler call taken in Qinghai, China with the SM Mini. The recording has been normalised to -3dB - if the bird had been a smidgen closer the noise would be reduced, but too close and it would have clipped. It would be fantastic if WA produced a 32bit float version to avoid such issues.
View attachment 1612491
I like!
 
Very interesting. And how do you connect 2 microphones? I have one that could be useful, a Primo EM258, but I'm not good at soldering, I could try it. On AliExpress they sell 3.5 mm jack connectors, I don't know if they can be useful or if you have to buy a specific one. On the Open Acoustic website they indicate that it must be SJ2-35464A-SMT-TR but I can't find much on the web

To connect 2 microphones in parallel: positive with positive and negative with negative.

TRRS conector -- smartphone
TRS -- recorder
Audiomoth -- Audiomoth with external mic

Primo EM 258 is a good alternative.
 
Last edited:
I am currently using Audiomoth 1.2.0 with a medium gain setting. I get recordings with noticeable background noise. Would Song Meter give me better quality recordings without that annoying background noise and better overall?
Is the microphone better? Both are 16 bytes. Or is it not worth the money?
What is the difference between the Song Meter Mini and the SM4?
Hi, I used for a 60-day audio monitoring of migrating birds the WA SM Micro 2 and am very fond of it. Very easy to set up thanks to its configurator app, it gave me a lot of peace of mind and reliability throughout the period without showing any problems. The batteries have shown a surprisingly long life. My previous experiences with the Audiomoth had not been so positive for me
 
I think I'll buy the WA Mini with 2 microphones. I'm ruling out the SM4 because of its large size and how difficult it will be to hide, and the little improvement in recording quality compared to the Mini. What do you think?
 

Users who are viewing this thread

Back
Top