| |
For directional microphone work I have identified three excellent microphones that seem both very sensitive and very low noise as well as highly directional (See Microphone Specification Table p51). The Wildtronics Pro Mono 22" parabolic dish, and the Rode NTG8 (22 in) and Rode NTG5 (8 in).
I try to look at and compare the polar plots of the three mics but find it difficult as Wildtronics and Rode have chosen different dB scales. Where Wildtronics places the origin at -40dB and maximum gain at 0dB, Rode places the origin at -25dB with maximum gains of +2 to +5dB depending on the mic. I made graphic for the Wildtronics for 1kHz and traced the Rode mic plots simply scaled so all max at 0dB, at least you can compare the plot shape this way.
To do this I digitized the data and first plotted it on cartesian coordinates. This way I can review the actual data. I again scaled the amplitudes such that all three max at 0dB at 0 degrees (center axis). Eventually I learned how to plot the data on polar coordinate charts which are tricky.
From the following chart comparisons I must seriously question the directionality for the Rode NTG5, better than a simple cardioid, but by how much? The Rode NT1 is a side-address cardioid that is more sensitive and has lower self noise and should be easier to carry about in the field, but is not really directional.
The Wildtronics comes out as a clear winner in terms of directionality, sensitivity, and self noise. It is also less expensive than the NTG8.
Comparison at 1000 Hz
Comparison at 4000 Hz
For the previous plots comparing the Wildtronics Pro Mono with Rode NTG5 and NTG8 I have had to manage things digitally as the Rode polar plots contain much deceptive mischief. If you look at the published Rode polars, you should note some oddities. First for the NTG8 they include data for three frequencies only, and none below 500 hZ. This gives the impression of near total side and rear sound rejection. Secondly, both plots at zero azimuth have the curves in positive values instead of a standard maximum of 0 dB. This gives the false impression of forward gain which is most certainly NOT the case. Finally, the Rode plots have an origin of -25 dB, (slightly more given the positive forward). This chops out all measured data below that amplitude, especially on the sides and rear and so further giving the false impression of near total rejection of side/rear sound.
To mitigate these slights of hand I had to digitize the data in order to manipulate it for a standard presentation comparable with the honest Wildtronics presentation. In order to digitize the curves I utilized "PlotDigitizer" (plotdigitizer.com/app). It is a very nice app that allows digitizing values from various types of charts, I highly recommend it! With the data in digital format, it was straightforward to shift the Rode data back to 0 dB forward. Plotting polar data is another thing. I eventually found "Plotly" (chart-studio.plotly.com/create/) that does an excellent job.
The plots I show (for 1 kHz and 4 kHz) include a standard X/Y chart and one polar chart. Note that the Wildtronics data is for 3150 Hz so it is quite pessimistic withy respect to the Rode data. [Note: in the next post I suggest that the Wildtronics at 4000 Hz will look much the same as the Telinga at 4000 Hz. If so then the Wildtronics is significantly more directional than the NTG8]. For all data plotted I have scaled the origin to -40 dB and the Rode data to 0 dB forward. These charts give a direct comparison of the three microphones considered, Wildtronics Pro Mono, Rode NTG8 and Rode NTG5. With the X/Y charts it becomes obvious where Rode cuts off the data, their curves "flatline". There is almost certainly missing data that was removed with this technique. Its less obvious on the polar charts but noticeable once you know what to look for. Rode has wiped out all the smaller side lobes.
Finally, it is apparent that the NTG5 looks fairly Cardioid rather than highly directional. The Wildtronics and NTG8 appear fairly similar overall. One thing to note though, unlike the long Rode NTG8, the Wildtronics 22" dish DOES give noise-free gain. From the gain equation (G = 20Log(3.25DE/W) where G = dish gain dB, D = Diameter in inches, W = wavelength in inches, and E = Efficiency from 0 to l). Efficient dish gain is: @ 1000 Hz = 14 dB dish gain, and @ 4000 Hz = 27 dB dish gain. As you will see, the dish is NOT 100% efficient.
Lets therefore take another look at the 1 kHz comparison plot between the Wildtronics 22” dish, the Rode NTG5 and NTG8. This time I have made an important modification to the polar chart data. The normal polar presentation starts with 0 dB at the 0 degree central axis. Then all points around the plot represent decreased sensitivity in the off-axis directions. BUT, not all microphones have the same sensitivity at the 0 degree location. The reported sensitivity in the spec sheets is just this, the sensitivity at 0 degrees.
For this plot then, I added the spec sensitivity to the data, shifting the curves to their true relative positions. Thus you can see that each curve intersects the 0 degree line at its respective sensitivity and things go downhill as you get increasingly off-axis. Visually the Wildtronics dish ballons out somewhat although I tried to compensate by plotting the chart origin at -45 dB.
Here I assume the data was taken with microphone array and dish combined. This should thus reflect the dish gain at the three frequencies reported. The reported sensitivity values when plotted on the dish gain chart (see p 85) are well below that predicted for a perfectly efficient dish. My best match for the data to theory is with a dish efficiency of just 11%.
Wildtronics sensitivity: -4 dB @ 1kHz, +6 dB @ 3k Hz, +14 dB @ 10kHz
Gain equation @ E = 0.11 -4.7 dB, +4.85 dB, and +15.3 dB
OK, to finish the comparison I did the same slight of hand to the comparison at 4 kHz. Here the dish gain overwhelms the two Rode microphones. Its all about the dish! For the Wildtronics I used a sensitivity of +5.27 dB from the Gain equation at 3.15 kHz and an efficiency of 11%. It would be 7.35 dB 4 kHz. The two Rode microphones retain their same sensitivity as they have a flat response with no forward dish gain.
Interpreting polar chart may not be for the faint of heart. I make two presentations with the EXACT same data but with different scales to demonstrate. Below is the plot scaled for all the data, +10 to -50 dB. The Wildtronics data for 3150 Hz now appears rather distorted, but this is not the case.
Even though the Wildtronics curve “looks” puffed out and distorted, it is unchanged and should be interpreted the same way as before. To prove the point, I plotted the data (below) with the scale at +10 to -35 dB so that it is clear the Wildtronics curve has the same “look” as in the original published report. It should be more straightforward to interpret and compare the off-axis rejection now.
Of course in plotting the true sensitivities with this scale, the Rode curves look itty bitty and some of their data has been lost in the singularity at the center of the chart. Which microphone do you want to own?
KJS 04/2025