NocMig Recordings - Why Shotguns or Parabola Mics (1 Viewer)

[Jon.Bryant]

Probably overthinking this, but why do people favour shotguns or parabolas for NocMig recordings? Does anyone have evidence of what mics work best.

To me the problem with a shotgun, is that it effectively filters out sound off axis and this is most efficient for higher frequencies, which is generally what I want to record. As an example the Sennheiser ME66 would curtail a 4khz sound (say a Redwing) 45 degree off axis by around 2.5db (i.e. almost halving the signal strength. I don't have the ME66, but the longer shotgun (the ME 67), which is even more directional and would further curtail off axis recordings.

Parabolas I think are even worse. I remember reading that parabolas work best within a cone of 15 degrees, but some data sheets show a much narrower focus that that. Even at 15 degrees, it would mean that at an altitude of 500m you are looking at a circle of space with a diameter of only 130m Using a parabola would seem to be lowering the odds that a/ a bird will fly through this cone of space and then b/ happen to call.

Some parabola mics (i.e. the Telinga Twin Science) have two mics, one of which is cardiodal and the other omini directional. Is anyone using this type of kit and if so which mic is picking up the most sounds? - i.e. the omni with the low boost but which is less directional, or the cardiodal, which should have significant boost, but be highly directional.

I presume that if a wide Cardiodal or Omni mic is the best option for NocMig recordings, then the problem is finding a mic with good sensitivity. I imagine that wide angle mics are generally not designed with high sensitivity for picking up distant sounds.

I am going to try the Wildlife Acoustics Recorder SM4, which has reasonably sensitive Omni mics and you can add a fair amount of preamp gain. I will also attach an external mic to the recorder in a homemade parabola housing - hopefully I will get a right channel wide angle recording and a left channel boosted recording of things directly overhead. Has anyone tried the SM4 for this NocMig recordings and was it successful?

 

[wimvdam]

I think it depends on what you are aiming for with your NocMig/NFC recordings. If you want to record as many calls as possible, then, yes, the narrow focus of a shotgun or a parabola would be counterproductive. But, if your goal is to try to ID the calls using their spectrogram, then those setups make a lot of sense. When I started making night recordings with a basic setup I was pleasantly surprised by how many calls I managed to detect. Unfortunately it was very hard to ID them because of the ambient noise and the weakness of the calls themselves. I now use a ME66 and the results are much better and now I have a fighting chance to figure out what species I had. It's still a struggle though so now I'm contemplating shielding my shotgun mic with a large bin. If I had the money I would definitely switch to a parabola to get even cleaner results.

Wim
Solvang, CA

PS: Identifying calls is very time consuming, so I consider a lower hit rate a blessing in disguise.

 

[Jon.Bryant]

As a follow up to my initial post, I set up the Wildlife Acoustics SM4 with an external mic housed in a dish. When testing this setup with a 5 Khz signal, I was amazed at just how narrow the focus of the dish is! The problem is that my recording area does not have an abundance of migrants, so I have failed to get any recordings that were picked up by parabola, but were much weaker on the other omni mic that was not housed in a dish. I am not sure if this means that the benefit of the parabola against a reasonably sensitive omni is very low, or if I just need a greater sample size for the test.

I have now tried switching to having an omni mic mounted on the roof, without the parabola, and am reasonably happy with the results, as there are a reasonable amount of good and identifiable signals.

In summary the omni on the Wildlife Acoustics SM4 seems good enough for the job - although a possible downside of using a omni mic is that terrestrial sounds are also quite loud - the local Robins seem to like singing from around 3am, then there are the foxes and Tawny Owls calling through the night. A directional mic would obviously filter out (reduce the signal strength) of some of this off-axis clutter. That said, it is still fairly easy to spot migrant calls in a sonogram amidst the Robin song.

I will retest the dish against the omni setup when I am next staying at a place with more migration activity. With this in mind does anyone know whether there is a preferred direction to set a parabola for NocMigging? Something says to me that if i set it straight upwards, I will get recordings of migrants at higher altitude, but the target will quickly pass across the focus (and may not call in the process!). If however, I was to point the parabola more towards the horizon in the direction of where birds are like to be coming from, then I may get a longer sequence of calls as the bird approaches, with the bird in the parabola focus for longer.

If there is no obvious answer to the question on parabola direction, I will add it a as another variable to future tests.

Regards

Jon Bryant

 

[Steinn1]

Hi Jon
Don't know what the outcome of this, but Telinga has a small parabola creating a much larger "field-of-view" and not so sharp as the large parabolas. Used it for displaying birds like Jack snipe XC492150 Jack Snipe (Lymnocryptes minimus)

 

[Jon.Bryant]

Hi Steinn,

Unfortunately, I have not had the chance to really test a parabola against an omni mic yet. Where I live migration seems pretty poor, and on recent trips away there has been nowhere to leave a parabola unattended overnight.

I am a bit confused by your statement that Telinga has a 'small parabola creating a much larger field of view'. My understanding is that Telinga only sell at 22 inch rigid or flexible dish, which is practically the same size as the Dodotronics dish (the most popular alternative). To be honest my Telinga dish is large and cumbersome enough, and I wouldn't want to lug around anything bigger. More importantly, I understand that whereas dish diameter influences gain, if the dish is a perfect parabola it shouldn't influence focus. If the dish is not a true parabola it could have a wider focus, but then it would also have less gain - this is a bit like a torch in reverse, where you can either have a reflector that creates a diffuse wide beam or a bright narrow beam (the later created by a reflector with a shape closer to a true parabola).

The mic capsule in the parabola will have a large impact on focus, so for instance if you use a Telinga Twin Science capsule, the outward facing omni mic will pick up reflected and amplified sound, but also non-amplified sound reaching the mic direct. The latter would have a wide focus, but the mic would pickup this sound without the dish in any case. Conversely the inward facing cardiodal mic of the Twin Science, should only really pick up reflected amplified sound, so should have a very narrow focus. The Telinga Stereo mic has two omni mics so should also pick up non-amplified sound reaching either mic directly and hence also have a wide focus. This to me is the big issue with understanding whether parabola are necessary for Nocmigging - I think that just because you are using a dish, you cannot assume that the sound are reflected and amplified. The prove this you would need to either compare the channels recorded with a Twin Science capsule (does the cardiodal channel pick up a significant number of more distant calls), or detach the dish from the Stereo mic and see to what extent it makes a difference.

What was the mic used for the Jack Snipe? I note that the right channel signal is much stronger than the left. So that either you were using the Stereo mic not quite pointed at the bird, or the Twin Science (which creates different gain and I think should probably be isolated as separate recordings - I think with this capsule you are meant to pick the best and isolate it as a mono recording). If you were using a Stereo mic then there does not appear to be any movement from right to left, which suggests that you were tracking the bird in flight - was this the case? I think to really understand the focus of a parabola, the dish would need to be static and the displaying bird would need to fly through the focus - if the focus is narrow, you would then hear a faint call rapidly becoming louder than diminishing again - as is the bird had done a rapid u-turn in front of the mic. Conversely if the focus was wide, the sound would be more natural, passing from left to right channel (or vice versa), and would not increase and diminish so quickly.

The Jack Snipe recording does not seem to have much ambient noise, as if recording was made in a very quite area. This could of course imply that off-axis sound was discarded due to a narrow focus.

As a side issue, I think that parabolas are not great for capturing movement, and that recordists can in theory create some weird effects using stereo parabolas, particularly when there is also a reasonable amount of ambient background noise - you can easily create the effect where the target stands still and the world spins around it. I think that stereo parabolas are probably best for capturing static birds with stereo ambiance, with the dish held in a stationary position. A stereo pair, without paraola, or an ambisonic mic, would be best for capturing a sense of movement in a recording.

It is interesting that Birding Beijing have started NocMigging this autumn, and are using a Wildlife Acoustics SM Mini, which has an omni mic. The recorder has been mounted on the roof of a building, so is away from traffic noise and closer to migrants, but the example recordings seem good. I used the same recorder, but with two omni mics in Cornwall this autumn, and from a first check the recordings seem quite decent. It is quite a nice setup as you can discretely chain it in place and leave it running. I left the recorder in a fairly well known bird-watching spot for 5 days of continuous recording, and don't think anyone noticed it - I definitely have not heard on the recordings any voices saying 'what they hell is that padlocked to the base of that bush?' For NocMigging, I suspect that the ultimate setup would be use something life the Wildlife Acoustics SM4, with the available external mic in a parabola. That way you could record an omni left channel with 360 degree focus for close stuff and a right channel with narrower (circa 15 degree focus) for far off things directly over the mic. This is of course not a cheap option, and the 'ultimate solution' may not be worth the extra cost.

Cheers

Jon

 

[Steinn1]

Hi Jon
I attach a photo of my 36cm/14' Telinga small parabola, it does NOT amplify as much as the large disc and it has a wider field. In the Jack Snipe recording I put the disc on ground and just laid a pair of EM-172 separated by a stick in the disc and covered with fleece for wind protection hoping the snipe would appear through the night (and it did!). Here it is with pro-X and 3.5mm plug PIP. Last summer I had a 1-hour chance to return and managed to get both Jack snipe and Broad-billed sandpiper during those 60 minutes. Here I used a "large" telinga MK2 stereo in the edge of the marsh. XC663127 Jack Snipe (Lymnocryptes minimus)
Stein

 

[Jon.Bryant]

Hi Steinn,

Interesting - I wonder why the dish does not appear on the Telinga webpage?

I still wonder if the parabola was providing any gain in your Jack Snipe recording, and whether if you had detached the dish it would have made a difference. The reason I say this is because I understand that parabolas only provide gain where the wavelength is less than the diameter of the dish, and that the gain increases with frequency. For a 14 inch dish there should be no gain below about 970Hz. As most of your Jack Snipe recording is below 970kHz, I assume that the dish was not really adding anything. I remember pointing a 22 inch parabola and a calling Oriental Turtle Dove and thinking my kit was bust as there was no gain, no matter how accurately i tried to point the mic!

It is interesting that Telinga don't seem to publish the specifications for their mics, but I recall seeing an article where a recordist was using Telinga kit without a dish for nature recordings. I assume that their mics must therefore be fairly sensitive in any case and may try giving this a go to see the difference.

Regards

Jon

 

[Jon.Bryant]

I found this on the net , which was published way back in 2003 by Klaus Strandberg (who I understand is the founder of Telinga)

'A parabola starts amplifying at a the same wavelength as it's diameter. There is no question about this. A Telinga will start amplifying at some 650 Hz.
(340m / 0,53m). Then it gives a further gain by 6db/octave. However, then some other tricky acoustical phenomena occurs. I will not comment on this. (In a week or so, I hope to publish the most extensive scientific paper made on this topic, by Sten Wahlstrom in Audio Engineering for many years ago.)

But: Do not forget that a sound also consists of overtones! If a bird has it's main singing on frequencies below 650 Hz, the dish will still amplify
it's overtones! This is why a parabola "recreates nearness" as distance tends to attenuate overtones by 6db / octave. Roughly....

At 10 kHz, the focus is a "ball" with a diameter of 34 mm. It does not matter where you place the (omni) microphone, as long as it is inside this
ball.

Subjectively, the parabola seem to cause "amplification" also at very low frequencies. First of all this may happen when a pigeon (for example) has
overtones higher than the Telinga 650 Hz. Then there is a "shield off" effect. The parabola becomes a shield, in the way for side and background
sounds. This gives an impression that the parabola amplifies.

Finally you can sometimes hear an effect cause by the ground. Acoustical phenomena, like a 6 - 10 db boost by the ground reflection, interact in
strange ways with the dish and microphone.'

So I was not quite right with my comments before. The 14 inch dish would start to amplify at above 970Hz, but recording with the dish would amplify overtones and create the impression of amplification by blocking noise from the side and back.... and there may also be some effect caused by ground, dish and mic interaction.

I also found a formula for a parabola beam width, which is (Angular Proportionality Factor x Wavelength)/Dish Diameter. The Angular Proportionality Factor is a constant for any dish, as is the Diameter. This means that the beam width narrows as frequency increases and wavelength reduces, but you are right that a smaller dish will have a wider beam width.

Working with a Angular Proportionality of 70 degrees (as stated in the article), the beam width of 14 and 22 inch parabolas would be as follows

Frequency (Hz)	Wavelength	Beam Width 14 inch (degrees)	Beam Width 22 inch (degrees)
1000​	0.343​	67.52​	42.97​
2000​	0.172​	33.76​	21.48​
3000​	0.114​	22.51​	14.32​
4000​	0.086​	16.88​	10.74​
5000​	0.069​	13.50​	8.59​
6000​	0.057​	11.25​	7.16​
7000​	0.049​	9.65​	6.14​
8000​	0.043​	8.44​	5.37​
9000​	0.038​	7.50​	4.77​
10000​	0.034​	6.75​	4.30​
11000​	0.031​	6.14​	3.91​
12000​	0.029​	5.63​	3.58​
13000​	0.026​	5.19​	3.31​
14000​	0.025​	4.82​	3.07​
15000​	0.023​	4.50​	2.86​

At 1000Hz (as per your Jack Snipe recording), the theoretical beam width is pretty wide, but up at say 4000Hz (as per the flight call of a migrating Redwing) the beam width has narrowed considerable. Even the 14 inch dish would only have a beam width of circa 17 degrees for a sound at 4kHz. For a call of a Goldcrest at 9000Hz both beam widths are very narrow.

Regards

Jon Bryant

 

[Steinn1]

Hi again Jon
This is the best explanation on dish-size and amplification seen on net so far! Thanks a lot!
Can I publish it on my site?? https://www.fuglesang-troms.net/english/
https://www.fuglesang-troms.net/lydopptak-sound-recording/

 

[Jon.Bryant]

Hi Steinn,

No problem with using the information. I think that theoretical equations are standard stuff, but the source is used was https://www.montana.edu/rmaher/publications/maher_aac_0805.pdf.

The article is a bit technical, but gives a theoretical equation for parabolic mic gain in DB. Figure 1 in the article shows theoretical gain for different dish sizes (although I don't think it excludes frequencies where the dish is smaller than the wavelength). I think using the formal and the assumed values in the paper, you could easily create a plot of gain for your 14 inch parabola compared to a 22 inch dish (from a rough calculation the difference seems to be over 3dB).

The table I created in the previous post is actually the 'half power beam width', which I understand is a standard measure for a parabola 'field of view'. The half beam width has a rather clunky definition in the paper (i.e. 'the angular separation between the half power points of the antenna directional pattern, i.e., where the gain is one half the maximum on-axis value.') So in the image below theta would be half the value calculated in the table.

One of the arguments for using a parabola is that the higher the frequency the more sound attenuates over distance, so the fact that a parabola creates more gain with increasing frequency is counteracting this process. For some purists this is a cause of debate - are recording what a human hears (i.e. a sound with higher frequencies attenuated) or what the bird would sound like if you stood directly next to it.

Regards

Jon Bryant

 

[Steinn1]

Hi again Jon,
For your last comment, for me who has passed 60 and start losing the higher freqs it is perfect with amplification in these range, I've seen the discussions on this aspect - but then they can lower this part of the frequency range if needed. To see this information in a table creates a clear picture of the behavior of the dish-sizes, really great stuff for me being a non-mathematician The parabola is a great tool in field and I will try to bring it with me where-ever I go as it really makes a difference in the stuff brought home.
Your Stein

 

[Joop99]

Hi Steinn,

Interesting - I wonder why the dish does not appear on the Telinga webpage?

Try this: Products - Telinga Parabolic Microphones

 

[Joop99]

The Sennheiser microphones with K6 (Me 6x series) are no longer available from Sennheiser. They discontinued this series.