how to attenutate signal before mic 2

[electricpete]

I want to record the sound of machinery on my Android phone. The main purpose is just to build a catalogue of sounds we have heard, so we can be more familiar with what's normal and what's not, and what a machine sounded like a year ago. Getting a good quality recording that sounds like the machinery is therefore important.

I have tried several of the programs available on Android, including:
PCM Recorder
Virtual Recorder
Tape Machine

One thing I find is that all my recordings end up with clipping because the environment (inside a machinery building) is too loud.

None of the programs has an input limiter. (Virtual Recorder and Tape Machine imply they do, but they really don't. Closer review of Tape Machine website explains why: the hardware controls to perform input limiting are not available to the Android programs).

So, I would like to try to reduce the signal before it gets to the microphone.

I can't move away from the machine, because that just tends to make the signal/noise ratio go down (I hear more background noise and less of the machine I'm interested).

So I'm thinking maybe I would put something over the mic. A piece of paper? A piece of masking tape? A cotton ball taped over? What would you suggest that might attenuate the signal but hopefully not distort the noise too much in the process?



=====================================
(2B)+(2B)' ?

 

[GregLocock]

That is an excellent question. Presumably you need to attenuate low frequency noise.

The simple theoretical answer is that you need an expansion box, whih would be dead easy to build, and easy to calibrate. ie small big small

I'm thinking a coke can and a couple of bits of 1" waterpipe.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[electricpete]

Thanks. I'm not too familiar with the theory of an expansion box. Is that how an automobile muffler works?

=====================================
(2B)+(2B)' ?

 

[GregLocock]

exactly. The attenuation is governed by the ratio of cross sections small/big, and the tuned frequency is governed by the length of the expansion chamber.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Strong]

It would be crazy to attenuate the sound at the microphone, when all you need is a voltage divider (search for audio Pad circuit) to attenuate the electrical signal about 40-dB to take it from Line Input down to Microphone Input.

Walt

 

[SomptingGuy]

I think the point is that this is being attempted with a fashion accessory rather than engineering equipment.

- Steve

 

[electricpete]

If I had a separate microphone, I would look at electrical attenuation. The microphone is built into the phone.

There is no microphone input jack. There is some kind of multi-purpose headset jack that may suit the purpose if I can figure out the right adapters. If you have a known working solution for Android, I’m all ears.

In the meantime, why is it crazy to attempt to attenuate signal at the mic?


=====================================
(2B)+(2B)' ?

 

[Strong]

This is why: "Getting a good quality recording that sounds like the machinery is therefore important." It is difficult, but certainly not impossible, to place a sound barrier material over the microphone. It would be difficult to maintain the same frequency response (ie. sound quality) with several layers of Gorilla duct tape or closed cell foam over the microphone. It would be by trail and error to observe whether the sound was "the same" on play back as it was direct to the ears.
I am sorry that I read your post too quickly, and I thought that a remote microphone or sound meter was being connected to the phone. As an aside, I have been looking for a new phone and reviewing apps for audio recording and analysis. The Android apps features looked pathetic compared to the iPhone apps. I am still looking.

Walt

 

[electricpete]

I was able to get a pretty darned good (to my ears) recording of a odd ringing sound reported on a single pump running in an isolated quite location by moving the phone far enough away (attached). I have not gotten anything useful in a noisy environment.

I am in agreement that attenuating the sound in front of the microphone is likely to affect sound quality. The things I have tried have not worked so far. Like putting my finger over the mic, putting napkin over mic.... these prevent the clipping but don't give good sound. But you never know… haven’t tried the muffler trick yet (although it would be a challenge to seal the end of it against the phone).

Many of the programs feature fancy recording specs like 44,000khz. But I’m not sure whether this phone has good analog recording circuits to go with that. Some quick recordings in a quiet environment showed quite a bit of static also. Maybe a separate mic would help… I’ll try if I can find one.

It may be that it is just not up to the task. But worth a try to explore the capabilities imo.


=====================================
(2B)+(2B)' ?

 

[electricpete]

One Android App that is interesting is AndroSpectro. It gives you a live waterfall plot showing spectrum vs time. In attached, I was able to isolate frequencies associated with the machine by starting at the machine, walking away, and walking back and watching the changes (you can see it in the powerpoint).

Unfortunately, I’m not sure if those harmonics of running speed shown are real or (more likely) just a figment of the chopping. The recorded sound in this case sounds nothing like the original.

The waterfall format from AndroSpectro will be a very handy tool if only we could get good recording quality. A big if.


=====================================
(2B)+(2B)' ?

 

[GregLocock]

a simple expansion chamber muffler as I described has a pretty horrific transmission loss spectrum, the recorded sound would not be particularly similar to the original. You are on the right track listening to your recordings, that is an excellent first step in an analysis and one that is forgotten far too often. I know someone who spent a day analysing interior roadnoise, when the dominant noise was actually the seatbelt rattling against the B pillar.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[SomptingGuy]

It's been a while since I did real NVH work, but Greg is right. You cannot beat listening to the signals. Similarly, you cannot beat photographing the test set-up. Ears and eyes are subjective tools, but very good ones.

44.1kHz is the standard for CD recordings, so I'm not surprised it's used by the fashion accessories. 48kHz had its day, but I'm not sure if DAT is still alive.


- Steve

 

[electricpete]

“fashion accessories”

Once could be viewed as an attempt at humor, however uninformed and irrelevant the comment is with respect to the current thread.

Twice can only be viewed as a troll.

=====================================
(2B)+(2B)' ?

 

[SomptingGuy]

Not a troll, sorry if it caused offence. Just an aversion I have.

- Steve

 

[electricpete]

OK, no hard feelings. I didn’t mean over-react.

=====================================
(2B)+(2B)' ?

 

[Strong]

Here is a link to an article about audio apps for smart phones and discussion about accuracy and calibraiton:

http://viewer.zmags.com/publication/cdb0dbf9#/cdb0dbf9/14

I keep an eye on phone audio apps, but so far I have not been too impressed. I bought a Sony Linear PCM-M10 digital recorder that is similar size, but thicker, than an iPhone, but it has much better audio quailty.

Walt

 

[electricpete]

Good article. It makes sense we can get better quality with external mic.

The PCM-M10 looks to be around $250. The built-in mic is omni-directional which would not be ideal for most of my recordings in noisy place. Do you use an external mic?... what kind?

I did find this you-tube video with brief explanation of how to hook up external mic to smart-phone:

http://www.youtube.com/watch?v=2IAMgAzufNQ&feature=related

=====================================
(2B)+(2B)' ?

 

[electricpete]

An attractive option is iRig Mic

http://www.ikmultimedia.com/irigmic/features/

$50-$60.
* highly unidirectional mic
* Three levels of gain.
* Then there's this - "The iRig Mic condenser-electret capsule with accurate unidirectional pattern and high-pressure holding ensures high-quality crisp, clean vocal reproduction - even in noisy environments - with no distortion, even with the widest range of volume dynamics."... not sure what high pressure holding is but no distortion over a wide range including noisy sounds good.
* The plug-in connector shown (with the 3 dark bands separating 4 conductors) is the same one that plugs into my Android phone (Samsung Infuse). Certainly will work with iPhone... looks to me like it might work with Android...

=====================================
(2B)+(2B)' ?

 

[Strong]

That's nice, the microphone is larger than the phone!

Walt

 

[Strong]

That's nice, the microphone is larger than the phone! There are other folks who share you interest, but are still struggling to find a solution.

Google Search: "external microphone for android phones"
Agood link is here:

http://android.stackexchange.com/qu...se-a-external-microphone-with-my-galaxy-nexus

Walt