vocal megaphone effect

[Rob Strand]

Haven't you heard? the iPhone 12 won't have a charger in the purchased box.

I didn't hear that but after reading about the reasons it really shows they are a bunch of petty money grabbers.

[SmileyNK]

Way back in high school, we used a cheap microphone with a 1/4" plug into a Boss PW-2 into a little practice amp, which we then miked for vocals. That sounded very much like a really bad bullhorn. Check out Tim Escobedos Lofo Mofo for a simpler solution.

Andy

Honestly the Lofo Mofo does this job perfectly for me. It's just the right amount of cut and crackle. I have one in a box with a bitcrusher/sample reducer for a one stop analog/digital style lofi pedal.

[amptramp]

I tried looking up "Winchester Cathedral" by the New Vaudeville Band to see if they had any whiz-bang technology but it looks like they used a real megaphone.  This might be a spot where all those little transistor radio transformers you have lying around might come in handy.  On some radios with tiny transformers and 2 inch speakers, you really couldn't tell there was anything different about the vocals on this song.

[iainpunk]

how about you get 2 1inch speakers and glue the diaphragms together, drive one with a 386 and use the other as a mic, just add a Jfet gain stage and output jack.

cheers, Iain

[amptramp]

how about you get 2 1inch speakers and glue the diaphragms together, drive one with a 386 and use the other as a mic, just add a Jfet gain stage and output jack.

cheers, Iain

If your drummer is in the middle of a 110 db drum exercise, this might get into the speaker/mic.  Transistor radio transformers offer all the bandlimiting of this combination with small size and no sensitivity to ambient noise.  But I agree that 1 inch speakers would do the job.

[ThermionicScott]

I tried looking up "Winchester Cathedral" by the New Vaudeville Band to see if they had any whiz-bang technology but it looks like they used a real megaphone.  This might be a spot where all those little transistor radio transformers you have lying around might come in handy.  On some radios with tiny transformers and 2 inch speakers, you really couldn't tell there was anything different about the vocals on this song.

I wonder if just buying a little megaphone and reverse-engineering it might give the most insight.  The frequency response of that horn speaker may play a bigger role in the final sound than just the circuit.

[amptramp]

I tried looking up "Winchester Cathedral" by the New Vaudeville Band to see if they had any whiz-bang technology but it looks like they used a real megaphone.  This might be a spot where all those little transistor radio transformers you have lying around might come in handy.  On some radios with tiny transformers and 2 inch speakers, you really couldn't tell there was anything different about the vocals on this song.

I wonder if just buying a little megaphone and reverse-engineering it might give the most insight.  The frequency response of that horn speaker may play a bigger role in the final sound than just the circuit.

Their megaphone was just a passive metal cone - no electronics involved.  But it was brass, so it was suitable for a musical instrument.

[idy]

Thanks for the enthusiasm. I tried a 2ndf order high pass like this:


Sorry add image has changed... the top of this page:
https://cepd.com/calculators/highpass_ana.htm][url]https://cepd.com/calculators/highpass_ana.htm[/url]

The values I saw for 1k were all caps 22n, the resistor to ground 5.36k, the resistor in the fdbk 15.8k. I used the nearest value I had and liked the overall sound with the vocal (headset) mic. I have two next steps in mind:one to experiment with dirt diodes, two to make this into a stand alone preamp with some gain and a mix to bring in a slightly more full range (but probably still bass cut) signal path.

Or: is there an easy way to make this filter variable?

[idy]

Add image is working!

[ElectricDruid]

Or: is there an easy way to make this filter variable?

Short answer: Perhaps..

Slightly longer answer: It's an equal-capacitors version of the multiple feedback highpass filter, like you can derive on this calculator here:

http://sim.okawa-denshi.jp/en/OPtazyuHikeisan.htm

From what I read, you can tune the cutoff "to some extent" by changing R1. Changing R2 will affect Q, but change the cutoff too.

Try it, either on the breadboard or in a simulation. You'll soon see whether it's enough to do what you want.

good luck!

[PRR]

...Their megaphone was just a passive metal cone - no electronics involved....

Rudy Vallée was known for a megaphone.
https://en.wikipedia.org/wiki/Rudy_Vall%C3%A9e#Music
https://youtu.be/-ubiw7MXmTE?t=430

His seems to be fiber:
https://upload.wikimedia.org/wikipedia/commons/1/1e/Rudy_Vallee_Megaphone.JPG

[Rob Strand]

. I tried a 2ndf order high pass like this:

The values you gave imply a Bessel filter (Q=0.5774) which has a fairly slow roll-off.

Butterworth Q=0.7071 is a good starting point and has a faster rolloff.
Q's of 1.0 would be even higher roll-off and add a bit of peaking.

As far as varying the cut-off with the resistors you need to scale the values up and down in proportion.   If you want an adjustable filter using a dual-gang pot then you need a filter with equal R values.   The MFB filter you gave doesn't let that happen so easily.  A simple one-opamp adjustable high-pass filter is the Sallen and Key filter with gain.  The down side of the simple circuit is the gain and Q are linked but for initial testing it is probably OK.

Here's a Butterworth version of the MFB and the adjustable Sallen and Key type.

[ThermionicScott]

...Their megaphone was just a passive metal cone - no electronics involved....

Rudy Vallée was known for a megaphone.
https://en.wikipedia.org/wiki/Rudy_Vall%C3%A9e#Music
https://youtu.be/-ubiw7MXmTE?t=430

His seems to be fiber:
https://upload.wikimedia.org/wikipedia/commons/1/1e/Rudy_Vallee_Megaphone.JPG

Any fans of "Mr Show" here?  My mind went straight to this bit:

[idy]

Mr Strand: I have set up the filter you suggested, Butterworth (or Sallen Key? difference is variable gain?) It seems about right.

The 10k (to VB on single supply?) at the inverting input can be varied to control gain, right? That's what I have been fooling with. Non inverting op-amp the gain will always be 1 or more? And as gain goes up the Q gets.... I'm guessing wider? (low q is wide...)

[Rob Strand]

The 10k (to VB on single supply?) at the inverting input can be varied to control gain, right? That's what I have been fooling with. Non inverting op-amp the gain will always be 1 or more? And as gain goes up the Q gets.... I'm guessing wider? (low q is wide...)

Yes, some tweaks are required for single supply
- R8 (10k) to VB is correct 
- You also need to connect R6 (7. 234k) to VB.

The Q and gain should increase as you increase the opamp gain.

Off-hand:
- Buffer (no R8=open):  Q = 0.5, (high freq) gain = 1.0  ; I should check that one.
- Gain of 3:  Q = infinite, (high freq) gain = 3.0   ; probably want to keep below that say 2.5 max.
- For gains above the Butterworth values shown on the schematic, the response peaks just above the cut-off frequency, the gain at the peak is higher than that set by the opamp gain.

You can see the idea here,



Don't worry about zeta symbol (not many pics showed up for high-pass).  The thing is as the Q increases it gets more peaky and at the Butterworth value  (Q=0.7071) it's flat as possible without peaks.   At Q = 0.5 there's no peaks but the roll-off is slower than Butterworth.    Also as the peaks get taller they get thinner near the peak.

One other thing to notice on the pic is the -3dB shifts a bit with the change in Q, slightly lowering as Q increases.    The filter "f0" doesn't change but the f3 does.  Not usually a big deal.

[garcho]

Butterworth (or Sallen Key? difference is variable gain?)

Sallen-key is a topology, Butterworth is a response, or "type" of filter. Butterworth filters can be made with Sallen-Key topology. The terminology is a bit confusing.

[FUZZZZzzzz]

https://www.pedalpcb.com/product/lofinator/

[Rob Strand]

Sallen-key is a topology, Butterworth is a response,

Maybe,
High-pass is a type of filter response.

Butterworth is a type of filter response shape.  (Simple view is frequency response. Pedantic view (time) impulse response.)
Bessel is a type of filter response shape

Sallen & Key is a type of filter circuit (= filter circuit topology).
MFB (multiple feedback) is also a type of filter circuit.

A "filter" to an electronics person is a circuit but it's really a filter circuit.    However, a filter can be just the response without a circuit.   There's no filter circuit required for a "filter" in sound processing software.

You could go nuts playing with words.

[iainpunk]

how about you get 2 1inch speakers and glue the diaphragms together, drive one with a 386 and use the other as a mic, just add a Jfet gain stage and output jack.

cheers, Iain

If your drummer is in the middle of a 110 db drum exercise, this might get into the speaker/mic.  Transistor radio transformers offer all the bandlimiting of this combination with small size and no sensitivity to ambient noise.  But I agree that 1 inch speakers would do the job.

i fully know radio transformers sound lofi, but the speaker thing can be mechanically over driven with only 3Vpp input for decent crunch, if you throw the whole 9V it gives a nasty distorted mess!
the outside noise isn't picked up well because i have it sandwiched in between yoga-mat foam and pillow stuffing, which is housed in a cigar box. remember that the back of these small shitty speakers are sealed and thus no sound can reach the diaphragms to pick up drums or ambient, you have to put it IN the kick drum to hear it.

cheers, Iain

[amptramp]

No problem with that.  Try it out and see if it works.