I've been lucky enough to try out the new Pettersson M500 detector over the past few months. This is a small hand-held device that allows you to both view and record full spectrum bat echolocation calls onto a windows laptop or tablet. The principle is that rather than record echolocation calls onto the device, or onto a separate recorder and analyse them later, you can record the calls directly onto your laptop/tablet via a USB cable. Not only that, but you can actually see what you are recording at the same time as well.
While this seems like a new idea, in fact Pettersson Electronik produced a special version of Batsound some years ago that allowed you to record onto a laptop via a PCMCIA data acquisition card. The limited computing power and battery life of laptops at the time, plus the cost of the data acquisition card made this something of a specialist item however, though the quality of recordings made via this method was outstanding. Now, with better faster electronics, and a price below £400, we have the M500.
What do you get?The initial package is surprisingly small, giving some hint as to the size of the device, but it's only when you unwrap it from the plain box that the tiny size of the device is brought home.
It's a tiny little thing, very light but it feels solid and robust and seems to be mostly made from aluminium, so should take a fair bit of field use. There are with no buttons, just a small translucent spot behind which is a red LED to show you when it's operating, but more of that later.
The little cone on the front protects the microphone, which is the same electret microphone used in the D500x static logger. The cone also enhances the directionality of the microphone and its on-axis sensitivity, and can be removed if you want to make the detector more omnidirectional.
To remove the cone you simply gently twist and pull. Doing this does leave the microphone rather exposed, and while these microphones are reasonably robust, they can be damaged by water, so unless you have a real need to do this, I would suggest leaving the cone on.
At the other end is the USB connection. Mine came with a curly UBS lead that extended out to about 60cm, but as there is a highly standardised USB mini-B socket on the bottom, any decent quality cable should do. I would be cautious about leads over 2m as due to the high data throughput, there might be signal loss along the way. The device is powered via the USB socket as well, so no need for batteries.
And that basically is that, a small box with a microphone and a lead, and I assume some wild electronics inside.
The box itself can sample at up to 500 kHz at 16 bit resolution. This is quite a remarkable feat, and one that should ensure very high quality recordings. The electret microphone, while arguably not as sensitive at higher frequencies as the capacitance microphone of the D1000x, is certainly very good and also much more robust and resistant to humidity. Recordings using the same microphone in the D500x show that it is a very capable performer indeed. The box also appears to have a built in anti-aliasing filter set at 190 kHz meaning that all bats with the exception of a few tropical species, can be recorded.
The bundled software allows you to record and view files and is compatible with Windows Vista, 7 and 8, both 32 and 64 bit versions. I have installed it on a range of machines and had a few minor issues that were quickly resolved.
There are separate installation files for Windows Vista and 7 and for Windows 8, so you need to select the correct ones. The problems I encountered were that either installation failed or on starting the software I received an error message. This turned out to be having outdated .NET frameworks and an older Visual Studio C++ re-distributable. These are components of the Windows operating system, so if you computer is up to date, you should be ok, otherwise you may need to install them from the Microsoft website. As the software is 32 bit, if you are using a 64 bit machine (most Windows 8 machines are) you need to make sure you have the 32 bit Visual Studio C++ redistributable installed as well as the 64 bit version.
A very important thing to note is that you should install the software BEFORE plugging the M500 into a USB port, otherwise the device drivers may not install properly.
The M500 will also be supported in the new version of Batsound due out later this year.
If all has gone well, you should now be running BatMicRecorder and should be looking at something similar to that below.
The very minimal look is both for the limited screen space on small tablets and also to reduce the load on the computer's cpu due to the vast amount of data about to be streamed through it.
The fun begins when you click the 'Start Sampling' button, and get the following...
The software - BatMicRecorder
The software has really been designed for automatic logging. This puzzled me at first as I was expecting a big button on the microphone marked 'go' so it took some time to get used to the fact that you need to set up the software to record automatically. There are a lot of functions to customise how the software works, and it does take some time to find your way around the various options, so I will only deal with the major features here. It's also important to point out that while the software allows you to see what you are recording, it doesn't allow any analysis after capture. For that you'll need Batsound or some other package. Fortunately as the files are standard .wav files and are time and date stamped in the file name, this can be anything you are used to, though some freeware packages might not like the high sampling rate. Batscan, while reporting the correct sample rate, doesn't seem to want to display the frequency correctly. One work-around is to open the files in Audacity (freeware) and either view them there or convert the sample rate down to 50 kHz using the 'Set Rate' command which changes the sampling rate in the file header and then view the files as 10x time expansion ones.
Capturing the signal
Firstly, and most importantly, you have two modes to capture the signal. These are 'Automatic' and 'Circular'. By default, the recorder starts in Automatic mode. Note that there is a rather confusing tick box also labelled 'Auto' this does not switch the recorder to automatic mode, rather it enables the 'Auto arm' function that allows the recorder to automatically reset itself so that after capturing one sequence it then waits for the next trigger to capture the next sequence. Without this check box enabled, the recorder will record one sequence and then stop and you have to manually push the 'record' button to start it again.
For the automatic record mode, you can select how long to record for in steps of 0.3, 0.5, 1, 3, 5, 10, 20 and 30 seconds, 1 minute or manual. In manual mode, it carries on recording until you press 'abort'. You can also set a pre-trigger length of 0 seconds (off), 0.1, 0.3, 0.5, 0.7 or 1 second. The pre-trigger allows the software to go back through the recording to before the trigger point. You total recoding time them becomes Recording length + pre-trigger.
The circular mode is essentially a manual record mode where all the data is pre-trigger. You can select from a record length of 0.3, 0.5, 1, 3 and 5 seconds. On pressing 'record' the device continually acquires data into a circular memory buffer. If you selected a record length of 10 seconds, then once the buffer is filled up after ten seconds, the data acquired from seconds 10-11 overwrites the data acquired from seconds 0-1 and so on. Once you see or hear and bat and press the 'stop' button, all the data acquired before that point (10 seconds worth) is saved to file. Any trigger settings you have input have no effect in this mode as you control the recording manually and should be greyed out.
How you set the trigger is very important to ensure that you don't miss bats but also don't acquire too much non-bat noise (such as the ever-rustling menace of the 'transect trousers'.). You have two options here, either 'Level' or 'Frequency'. Level is the simplest, in that you adjust the slider between 0% and 100% of the amplitude of the signal. If you have the live oscillogram enabled (Go to Show>Oscillogram) AND the check box marked 'Show' is clicked then you can see white lines that show the trigger level. If the signal goes above or below these lines then the system will trigger. There is a risk that other non-bat noise might trigger the system unless you have the trigger set very high.
A more sophisticated method is to use a frequency trigger. Enabling this allows you to adjust two frequency sliders which will define the bandwidth you are interested in. At its simplest you might select 20 kHz in one and 250 kHz in the second, which will look at anything ultrasonic and ignore anything in the normal sonic range. A more sophisticated way might be if you are interested in pipistrelles, set on to 40 kHz and the other to 60 kHz, or even more specific, lesser horseshoes, set one to 105 kHz and the other to 120 kHz. You then set the slider for the intensity (dB) that the signal has to reach before triggering the system. The lower the value, the more sensitive it is. Note that if you set a very low value the system may trigger itself from its own internal electronic noise and continually record.
The final setting in the trigger section.is the 'Single' check box. This means that once a signal exceeds the trigger level, the sonogram view and the oscillogram view are frozen so that you can see the signal in more detail. If you have the 'Auto' option enabled, then once a signal triggers the device it is shown for a few seconds before the system arms itself waiting for the next trigger.
The settings option
This confused me a little at first as you get a series of options for controlling the horizontal axis time scale which is simple enough, but also a series of settings for recorder settings, such as oscillogram, spectrogram, grid and audio. What these do is to turn off those displays/audio while the recording is taking place. This can stop breaks in the recording due to the pc not being able to keep up with the rate at which data is being acquired. You can check this by looking at the throughput written at the bottom of the window 9see bottom of image below. This shows the acquisition rate currently being used. It's important to check this during a few test recordings of about 10 seconds or more. If it drops below 500,000 (goes red) there is a risk that some data may not be recorded. The occasional drop into red is not an issue as there is a data buffer for the data so it may be able to catch up later on. If it stays below red, disable the spectrogram view from the setings>recorder settings menu to free up cpu resources.
The set of options at the bottom allow you to configure how the software looks. You have a range of options for controlling the spectrogram colours, threshold, FFT size and spectrogram and oscillogram scaling. Just play with them until you get the view you like, they don't affect the recordings at all. If you can't see them, go to the 'Show' menu and click on the 'Advanced' box. The show menu also lets you display various parts of the display, useful for saving space on small screens or cpu resources.
The enable audio check box allows you to listen to the signals as they are being acquired. It uses a process of under-sampling which is to sample the audio at a lower rate than that needed to preserve the signal. in effect you are aliasing the signal to make it lower in frequency. The output sounds a little like a frequency divided output. The slider controls the volume. It's useful for listening for bats - turning your M500 and laptop into a kind of frequency-division detector, but it's best to listen through headphones or you may find that the audio output from your laptop gets re-recorded by the M500 and then output again and re-recorded creating both horrible recordings and nasty feedback. The audio option also uses up cpu resources, so if you have a lower-specified machine I would suggest turning it off if the system struggles.
This is the tricky part, what do you use it for? It's a lovely thing, but what would you use it for in the field? Every consultant has their favourite way of working, some like to take out the best equipment going, others like to go minimalist. Personally I like a detector that has a 'go' button so that I can just watch bats and not equipment. For that reason I actually prefer a D240x with a recorder, small, not too many settings and good quality recordings. I probably wouldn't take a laptop/tablet in the field as I like to watch bats and not screens, but I can see that as an automatic recorder, you could set the trigger settings, set up the laptop/tablet to stay on when you closed the lid, put it in a rucksack and set off on your transect. Essentially you get the quality of a D500x at a fraction of the cost assuming you already have a laptop. Ditto for remote logging if you can leave the laptop in a safe place, such as in lofts or barn. It would however be perfect for education use to both see and hear echolocation calls as they happen.
Like most new products, especially software, there are bound to be a few issues. The most troubling was the installation of the C++ redistributable packages. Once this was sorted out, everything else went fine. Occasionally menu items don't work properly, or can't be selected, but this has always been solved by closing down the program and restarting. One issue I have had a few times is that if you leave the M500 attached, close the software and restart, an error is reported. Simply close the software, unplug the M500, plug it back in and restart the software. I'm sure software updates will gradually sort these minor problems out.
A few examples of some bats to show you the screen outputs. I haven't particularly optimised the views in terms of spectrogram levels or time axes, but it gives you an idea of different species.
This is really an amazing piece of kit, which for the price is even more amazing. You do have to ask yourself what you would use it for. This approach of screens in the field may not work for some and it's worth having a good long think about how you would use it. The quality of recordings is second to none save perhaps those from a D1000x at ten times the price. The size, sensitivity, versatility and quality make it very attractive for the price, though you will need a laptop or a tablet and some form of analysis software as well. Essentially it turns your laptop into a real-time bat analyser and data-logger, so perhaps the limitation is me for not being inventive enough to think of new ways of using it.