Microphone Polar Plots

[Hugh Robjohns]

I'm feeling decidedly chuffed that I managed to beat MS Excel into a grumbling submission and managed to get it to plot a complete set of first-order polar responses... so here, for your enjoyment and edification, is the plot -- (c)Hugh Robjohns 2021 !

[James Perrett]

Very nice - but should there be some numbers on the attenuation scale?

[Mike Stranks]

Good work that man!

Take a bow!

Could you put this somewhere easy to find so that it's easy to retrieve when the next person comes along with a question about the polar sensitivities of different types of mic?

[Hugh Robjohns]

Picky, picky...

Erm... yes... possibly....

I haven't quite figured out yet how to get the underlying to scale to conform to anything useful. Excel really isn't the right tool for this job... and I'm pleased I managed to get it to go this far.

At the moment, those circles are at roughly 1.2dB increments... I'll have another fiddle later when I have some time.

And yes, Mike, I'll copy this to the useful info archive. Good idea.

[James Perrett]

Picky, picky...

I've got the wide cardioid Line Audio CM3's and I was just reading the chart and wondering what the front to back ratio should be with them when I realised that there was no scale.

It is amazing what you can do in Excel. I've just created a suite of macros to handle PRS statements for a friend of mine. They're not pretty and the job could probably be done better with a database but they seem to work. We also started off doing all our post mission data processing for Autosub in Excel before moving to Matlab. It is a great way to get started on something fairly quickly.

[Zukan]

Doing that in Excel makes you proper Jedi!

I'm still on 'text to columns'.

[Hugh Robjohns]

I've got the wide cardioid Line Audio CM3's and I was just reading the chart and wondering what the front to back ratio should be with them when I realised that there was no scale.

Of course -- you're quite right.

The problem I have is that the underlying plot is linear (it's a modified doughnut chart) and I really need it to be log. Not found a way to do that yet without going to the whole faff of calculating and plotting the whole thing from scratch... But it will continue to be a thing I'll tinker with as and when... I quite like playing around in Excel!

The impetus for doing it in the first place was simply to illustrate the relative directivity and rear-tail sizes of the super- and hyper-cardioid paterns... and then I figured I should really add in the other standard patterns too...

[ManFromGlass]

Use bold
Then I can say the plot thickens . . . . . .
did I say that out loud? Inner voice, inner voice

[Hugh Robjohns]

Is this your coat?

[ManFromGlass]

Thanks. How did you know I had to go?

[Drew Stephenson]

The problem I have is that the underlying plot is linear (it's a modified doughnut chart) and I really need it to be log. Not found a way to do that yet without going to the whole faff of calculating and plotting the whole thing from scratch...

Hmm, I think I stumbled across this as an option a few years ago. Let me have a shufti.

[Zukan]

Use bold
Then I can say the plot thickens . . . . . .
did I say that out loud? Inner voice, inner voice

[Drew Stephenson]

The problem I have is that the underlying plot is linear (it's a modified doughnut chart) and I really need it to be log.

Excel will happily play with log scales on the axis, it's a tick box option on the right-click > Format Axis menu.

Hmm. I've got the axes working fine but I must be inputting my data wrong as Omni is ok, fig 8 looks like a reverse cardiod and everything else looks like a variant of a Hypo!

[jimjazzdad]

It all looks a bit like a Psychedelic Shelob in her lair to me...what have I been smoking?

[Hugh Robjohns]

Hmm. I've got the axes working fine but I must be inputting my data wrong as Omni is ok, fig 8 looks like a reverse cardiod and everything else looks like a variant of a Hypo!

Yes, I was aware of the log scale option, but (a) it doesn't work with negative numbers and I have a four-quadrant (bipolar) plot here, and (b) it's geared up for linear XY plots rather than the radial one I've contrived by hacking the doughnut chart option....

Anyway... if I remove the grid the problem goes away, and the comparison of polar patterns becomes even clearer!

[wireman]

I gave up on excel for anything but the most basic chart as it was a neverending source of frustration.

The height of madness was when I wrote macros to turn a spreadsheet of agenda items and times for a multiday meeting into a graphical representation (a drawing).

[forumuser840717]

I thought I was reasonably competent with Excel's charting functions until a few years ago when I came across Tatsuo Horiuchi!

This article about him has a couple of downloadable spreadsheet examples of his work.

[Martin Walker]

This article about him has a couple of downloadable spreadsheet examples of his work.

Wow - just wow that anyone would go to that mathematical amount of effort to create a picture. Most impressive!

Everyone's different I suppose.

Martin

[wireman]

The problem I have is that the underlying plot is linear (it's a modified doughnut chart) and I really need it to be log. .

I had a go at this in python using the standard equations from here:

http://richbreen.info/McGill_Paper/mic_pairs.html

What is not mentioned is that you need to plot the absolute values of the functions.

To get a standard dB plot presumably you don't need a real log scale in Excel, just plot 20 log_10(sens) which will be negative and no comparison is needed as all the equations have a peak of 1.0. This will mean cutting off the minimum value in the plot.

I want to try adding together two cardioid microphones next...

[Drew Stephenson]

I had a go at this in python using the standard equations from here:
http://richbreen.info/McGill_Paper/mic_pairs.html

That's what I was trying to use.

What is not mentioned is that you need to plot the absolute values of the functions.

This is obviously what broke me as I haven't got a clue what it means.
It's slightly worrying that about 25 years ago when I was at University this would all have made perfect sense.

[James Perrett]

Well I've just set our lad a challenge - produce an interactive 3D rendering of the different microphone characteristics using Scratch. He's promised that I can post a link if he manages it.

[Mixedup]

Well I've just set our lad a challenge - produce an interactive 3D rendering of the different microphone characteristics using Scratch. He's promised that I can post a link if he manages it.

I started trying to do that in Blender a couple of months ago... and gave up fairly quickly!

I was hoping for something I could rotate to match the mic angles on a 2D photo and have the lighting similar to in the room. But it turned out to be way beyond my skills/experience!

[James Perrett]

Well I've just set our lad a challenge - produce an interactive 3D rendering of the different microphone characteristics using Scratch. He's promised that I can post a link if he manages it.

I started trying to do that in Blender a couple of months ago... and gave up fairly quickly!

It seems he's given up too - which surprises me because yesterday he was showing me a 3D sphere with a surrounding star cluster that he'd made where you could freely move the camera around and the 3D aspects really worked. He's very into learning about the code behind 3D rendering at the moment but it has to be at his own pace.