Question about connectors parabolic mic (1 Viewer)

[Avetarda]

I'm looking for information before buying a parabolic mic. I've doubts. Is an xlr or jack connection better? Does the sound of a jack connection lose quality compared to an xlr? I've two recorders, Zoom H5 and Olympus LS P4.

And another question:
Does the number of capsules (2 or 4) influence? Which one is recommended?
I'm interested in a stereo Dodotronic, but you can choose the type of connector (xlr or jack) and the number of capsules (2 or 4)

 

[Jon.Bryant]

I am not sure that the connection differs much in sound quality, especially as you will only be using a short cable. I find however, that mini jack connections are prone to movement and bad connections, especially if the connection is being stressed by movement. If I was using the mic hand held and stalking a bird, I think XLR would be my choice.

From what I understand the Dodotronics 4 capsule stereo setup is more sensitive than the two capsule stereo setup. The following link gives pros and cons of different options https://www.dodotronic.com/hi-sound-comparison/?v=2a47ad90f2ae. I understand that the term saturation refers to the risk that the recording is clipped. Obviously you can adjust the gain on the recorder to prevent clipping, but then I think you need to consider how you will actually use the parabola and either of your recorders - neither of your recorders fit comfortably in a bag, and can be operated easily with one hand. This means than you may have a bit of a problem if you want to be mobile and hand hold the mic, but still be able to adjust the recording level.

Another thing to consider may be to invest in a 32bit float recorder. With this option, you do not need to alter the recording level and cannot really clip the recording, so the downside with the four capsule module or the problem of having to monitor and adjust levels go away. At the moment there are only a few 32bit float options, but recorders with mini jack or XLR connection already are on sale.

Regards

Jon Bryant

 

[Borjam]

Balanced connections have a higher signal to noise ratio but the advantage is irrelevant with a very short cable. As Jon Bryant says, XLR connectors are more robust which is a plus especially outdoors.

High quality recorders use XLR connectors because it's the standard professional connector. So you will find that the best preamps in the market have XLR connectors. Another advantage is phantom power which would be problematic to implement with a jack, even if balanced.

The number of capsules is trickier to understand than it seems, let's see, it's quite technical, and it comes down to what happens when you add two signals.

Imagine that you record two microphones side by side, capturing the same sound source. You can say that each microphone is picking up two different signals.

First one is sound. Both the intended signal (say, a bird song) and whatever ambient noise that happens to be picked up by the microphones.

The second one is electronic noise in the microphone itself, its electronics, the preamplifier and the A/D converter.

While the sound picked up by both microphones will be the same (yes, depending on relative microphone positions there will be a phase difference but we can disregard it) the electronic noises generated by each component are unique.

What matters here is: if you add two "equal" signals (technically the term is "correlated") you double the output. But adding uncorrelated signals does not work in the same way. So, you gain some signal to noise ratio by using several capsules.

Sound Devices uses this trick in the MixPre recorders for example. Each channel is sampled by two A/D converters and they average the samples as far as I know. It gives an increase of 3 dB in S/N ratio.

As for 32 bit, remember that the ultimate limit to dynamic range is the actual dynamic range of the microphone and preamp. A preamp with a low signal to noise ratio will anyway limit usable dynamics a lot.

 

[Jon.Bryant]

Borjam raises an Interesting point about summing the two signals and the impact on noise.

With regard to 32 bit float, I really raised this as Dodotronics advise that the 4 mic array is very sensitive, which may lead to problems with 'saturation'. A couple of weeks ago I asked Dodotronics whether there is an issue with saturation with the DR, which also has 4 capsules. I also asked what they meant by the term - my simple logic being that I could not see how the number of capsules could impact the sound pressure in the dish and hence the potential issue with mic distortion. I understand that the issue is that the combined signal from the two capsules can be too high, resulting in clipping.

I have personally not experienced clipping when recording with a parabola, but I use a Telinga Twin Science, and don't know how this compares with the Dodotronics systems - to be honest Telinga are shocking silent on the specs of their mics. The only time I have really experienced clipping is with static omini directional mics (without parabola) when a bird has approached too close - Cetti's Warblers are particularly good at creeping up, blowing a recording then flitting away!

Assuming the 4 capsule stereo system is super sensitive (and that is what you want), then 32 bit float could be beneficial to overcome any possible problems with clipping.

As Borjam states, 32bit float has limitations. In particular I don't think it is really that great for increasing levels for most outdoor recordings. If you have a super quiet recording with very limited noise, then perhaps. However, I think that for most wildlife recordings the ambient noise will be above the noise floor of even a 16 bit recording. I can see that potentially the 32bit signal has enough data to bring up the levels of very quiet recordings, without loosing fidelity, but boost the recording and you must be boosting recorded noise too.

 

[Borjam]

I am not sure how they have implemented the four microphone arrray but I misused the term "summing". Actually it is averaging, which in this case would mean summing them and adding a certain attenuation. That should increase the S/N ratio. If they just sum the microphone signals it means you have to reduce gain and you will achieve the same effect.

Parabolas have gain anyway. Unlike shotgun microhones whose strength is in attenuating sounds coming from unwanted directions, parabolas really amplify sound. So you can get away with a somewhat noisy microphone. Of course when recording sound it is still desirable to achieve the best S/N ratio in every step. For the same reason, it is always better to achieve a proper recording level turning the 32 bit feature into a safety net instead of depending on it.

And this is especially important now that 32 bit can become commonplace. Analog electronics make a big difference and "bitness will not avoid bitterness"

 

[Avetarda]

Thanks!
for now I want to continue with my recorders (zoom and olympus). So, I can understand the conclusion from your comments, that 2 capsules can be better than 4 and less problems?

 

[Borjam]

for now I want to continue with my recorders (zoom and olympus). So, I can understand the conclusion from your comments, that 2 capsules can be better than 4 and less problems?

If you are using an assembly from a given manufacturer such as Dodotronic I guess four capsules will get you a better S/N ratio.

The distance between capsules will be very small, so the combing filtering should be negligible.

Which is the exact model you are considering? Unfortunately Dodotronic don't give many technical details :/

 

[Avetarda]

From this page: https://www.dodotronic.com/hi-sound-comparison/?v=2a47ad90f2ae
The model 2 or 3 (stereo with 2 or 4 capsules), between those two I doubt.

 

[Borjam]

You see that the noise floor is 3 db better in the dual capsule version, while the sensitivity is quoted as 3 dB better, for a net gain of 6 dB in signal to noise ratio, which is a lot.

These are theoretical values, I don't think they have measured them. But that said, there will be a gain in S/N ratio for sure.

I guess saturation shoudln't be much of an issue. It´s just a 3 dB increase in sensitivity after all. If you record too close make sure to have your limiter engaged

 

[Avetarda]

Thanks Borj. Then the option of 4 capsules sounds good...😄

 

[Jon.Bryant]

Hi Borjam,

I think I am being a dummy, as I am struggling with the 3dB + 3dB calculation - I would be really grateful if you could explain a bit more.

My confusion is that I normally see noise stated as dB(A) , which I understand is the noise (in the main human hearing range) in relation to a 94dB source. What I don't understand is that this is normally a positive figure, so +10dB(A) would be a a good figure for a mic. I can't recollect seeing a negative dB figure for self noise - are you reading the Dodotronics as +14dB(A) and +11dB(A)? In this case I can follow your calc. If it is really meant to be a negative value, I am a bit baffled - can you advise whether a negative figure makes sense to you, and if so how this is defined and what values are best? I presume that the sensitivity is dB re 1 V/Pa, so in this case -21 is 3dB better than -24.

Avertade,

I understand that 3dB is roughly a doubling of sound intensity, so the difference between the 2 and 4 capsule systems should be readily apparent. I think occasional saturation should be able to be easily avoided by adjusting gain, but I think to do this with a Zoom H5 (or Olympus LS-P4) you need to think about how you will use and in particularly hold the recorder and parabola. I think it will be tricky to hand hold the parabola and make any necessary adjustments of gain with one hand - unless a shoulder bag exists that holds the recorder securely in place, but leaves the controls exposed, allowing the gain to be readily adjusted. Perhaps having the recorder and parabola mounted on the same tripod would be best - then one hand to aim the parabola and the other avialable to adjust gain as required.

Regards

Jon Bryant

 

[Avetarda]

Do you think I'll need both hands to hold the parabola? The weighs about 700 grams. Can't you hold it with just one hand?
Currently I hold a Sennheiser microphone and zepelin with one hand.

 

[Jon.Bryant]

Can you adjust gain on the recorder while holding the Senheisser mic? Is so not a problem to operate the parabola in the same way. The main recorder I have is a Mix-Pre which fits in a shoulder bag with the controls facing up towards you. This means you can hold the parabola in one hand and adjust the gain level with the other. I haven’t seen a bag like this for the Zoom H5. If you hold the parabola in one hand, but then have to take the recorder out of you pocket or from a bag to operate, you then have to be able to hold and adjust the recorder with just the one hand (or put the parabola down). Is it possible to hold and adjust gain on the H5 with one hand? You can always hope that the preset gain is ok (and you may have to do this for short non repeated sounds such as calls), but this will be trial and error and you could loose some recordings to clipping - particularly as the parabola is so sensitive - which brings us back to 32 bit float!

 

[Avetarda]

Well, that's how I do it now, Sennheiser and a zeppelin in one hand and a recorder in the other. The H5 has very comfortable spinner-shaped gain controls.

 

[Jon.Bryant]

Ok. Sorry to make an issue out of nothing. On a different forum the issue of operating ‘all-in-one’ type recorders was discussed. My two all-in-ones are not that easy to operate in the way you describe. Glad the H5 is different.

 

[Borjam]

I was a bit thick (my mind was somewhere else) so I didn't quote units properly. Anyway there is an advantage to the microphone array.

I am not a signal processor expert myself, but remember that adding signals of two different microphones, ideally close and aiming at the same target, means you are adding the same sound signal twice (assume that both microphones are picking up the same) which means that signal will gain 3 dB, but the microphone noise will be different. That what is called uncorrelated.

So, if we call the sound signal S, it is obvious that S + S = 2 * S, that´s the 3 dB gain.

What about the sum of uncorrelated N1 and N2 noise signals? Well, statistically they average to the same. You could try an experiment even. Record, exactly at the same level, two samples of white noise (something similar to noise in a blank portion of the FM broadcast radio frequencies) and add them together. How much will the level meter climb?

The gain in S/N ratio would be 3 dB or less, which anyway is very good. Sorry about the confusion.

 

[Jon.Bryant]

Thanks Borjam - wasn’t trying to catch you out - I am far from knowledgeable about signal processing, and noise and THD figures often have me completely stumped.

Averrarde - not wanting to bang on, but… I was looking at a thread on the new Tascam 32bit recorder. One comment was that XLR cables impact the ability to hand hold recorders. The arguement was that the cable weight makes handheld recorders unbalanced, so you hold the recorder tighter and could introducing unwanted handling noise - I suppose this depends on the mic to cable connection. My shotgun shock mount includes a thin cable connection from the back of the XLR in the mic to an XLR connector in the handle, meaning any cable handling noise should be isolated from the mic. I am not sure if parabolas have the same type of system to isolate cable noise.

Not sure if you have used your shotgun with XLR cables, but this may be worth a try to get the feel of handling the recorder with the thicker cables attached.

If it doesn’t feel right, and you think the mini jack parabola solution will be better suited to your recorder(s), then taping up the connections may be a work around - I found this substantially improved issues with broken contact noise. This would only be required if you don’t have locking minijack inputs on your recorder.

Regards

Jon