REW for Audio Visualization

[ForensicAudio]

Hello All,

Is it possible to visualize or record a wav file in REW but simply look at the wav file for what it is without any filters or analysis? I really love the waterfall visualization and although I can use the spectrograph in Audacity, I would love to see something like that for the full spectrum of a wav file.

Cheers!

 

[John Mulcahy]

Yes, but only for WAV files that aren't too long - only the first minute or so will be loaded. Use File - Import audio data or drag the WAV file onto REW while holding the Ctrl key.

 

[Keith_W]

You're better off using something like Audacity. Like REW, it's a free tool. But you can certainly open a .WAV file in REW:

That's the first few seconds of Billie Jean. REW has a limit of 101 seconds. I'm not sure how you can increase this limit.

If you want to record, I would use Audacity and not REW. And you will need an interface. This is how:

1. In your player software, set output to an unused output. Let's say AN1/AN2 OUT.
2. In your interface, set up a loopback from AN1/AN2 OUT to AN1/AN2 IN. Some interfaces support software loopback. Some require hardware loopback, so you need an actual cable.
3. In Audacity, set input to AN1/AN2 IN.
4. Start recording in Audacity.
5. Start playback.

 

[ForensicAudio]

I will try this in REW. The waterfall visualization is what I am after! The recordings will only be about 10 seconds long.

Having said that, I will look at Audacity for recordings with the UMIK 1. I believe it's possible to use the calibration file to set a corrected EQ to give a flat recording.

Basically, I want to compare the sound recorded by a device like a security camera with the UMIK 1 and then compare the two to see what is the frequency cut off and response of the device.

Much appreciated for the advice.

 

[Keith_W]

Basically, I want to compare the sound recorded by a device like a security camera with the UMIK 1 and then compare the two to see what is the frequency cut off and response of the device.

If that's what you want to do, then you're better off taking a proper logarithmic sweep rather than recording random noise? Also bear in mind that the microphone in the security camera and the UMIK-1 are not comparable. The UMIK-1 is omnidirectional and will be much better at capturing ambient sound - which is what you want for an acoustic measurement, but not what you want in a security camera.

 

[ForensicAudio]

Yes, I want to run a sine sweep and capture that using REW but record the security camera at the same time. The issue is then how do I compare them? If I could get the WAV recording from REW, then I could bring both into Audacity and make some plots showing differences but I was hoping I could do something like this inside of REW directly.

What I am thinking is playing a sine sweep from REW and recording with Audacity and the UMIK-1 on a different laptop. I then can just import the audio from the security camera and align them in Audacity. What I want to show is how a security camera differs greatly in frequency response to a much better microphone. Some of these cameras are capped at 4-5 kHz since they mostly care about voice.

For reference, I have the JBL 305P MKII, the UMIK-1, Audacity, REW and Adobe Audition, and Sonic Visualizer amongst a few other smaller "toys".

If you can think of an alternative or better way, let me know...I am relatively new to this so trying to figure out what tools do what.

Thank you.

 

[Keith_W]

I am sorry, I misunderstood you. When you said "recording", I thought you meant recording music or using REW to analyse music files. I now realize that you meant recording a test signal for the purpose of analysis. Re-reading your initial posts, I don't think it's entirely my fault that I was mislead

So i'll state what I understand of your request: you want to compare the performance of the microphone in your security camera to an objective benchmark?

Everything in the process is straightforward, EXCEPT extracting audio from the security camera microphone. It is best to do this entirely within REW. Try this:

1. Connect the security camera to your PC and download the appropriate driver.
2. Open REW. Under "Preferences - Soundcard" look for the security camera microphone under "Input Devices". Select it.
3. Press "check levels". You will hear pink noise, and you will see the metering bars start to dance. Keep an eye on the middle bar, "In". If it starts to dance, then your mic can hear the sound and is successfully sending it to REW.

If it can do that, the rest of the procedure is easy, and i'll be happy to talk you through it.

 

[ForensicAudio]

Hey Keith, much appreciated on the reply. Yes, I would like an objective comparison between the UMIK-1 and the security camera. I am not trying to calibrate the security camera, but rather want to understand its limitations compared to a calibrated mic. Maybe I am thinking of this the wrong way but some of what I see in REW seems to make a lot of sense from a standardized test method.

One issue that I have is that these security cameras might be wireless doorbell cameras or other types of cameras that don't have a way to hook them up to a PC. They are stand alone devices and you have to get the video files through the internet or through an app.

Let me know if you see a pathway forward here, with or without REW.

 

[Keith_W]

Unfortunately, I don't see a way forward. The problem with those cameras is the processing that is done to the audio once it is recorded - it may be compressed, or recorded at a different sample rate. Some of them even have noise cancellation. Goodness knows what that will do to a test signal!

I suppose it won't hurt you to try. Here are two ways that MIGHT work. No guarantees!!!

Method 1: take an offline measurement
1. Go to Generator - Sweeps - Measurement
2. Create a sweep from 20Hz - 21kHz, length 1M, Mono, 44.1kHz, 32 bit float, and "add timing ref". Save .WAV to file.
3. Play the .WAV file on good quality calibrated speakers and take a recording with the security camera
4. Repeat with the UMIK-1 in the same position using REW's offline measurement procedure
5. Load the .WAV file from the security camera and compare

Method 2: fool REW's RTA
REW's Real Time Analyzer (RTA) is usually used to take real-time recordings of pink noise. Here we are going to play back recorded noise via a loopback and lie to REW.

1. Go to Generator - Noise - Pink random
2. Create a pink noise file. Full range, Mono, 44.1kHz, 30s duration. Save to file.
3. Follow steps 3 and 4 in the previous method
4. Use music playback software (let's say Foobar). Connect the player output to REW's input*
5. Open REW's RTA analyser. Press "play" on Foobar, then press "record" in REW.

* this is the critical step and how you do this is going to be OS dependent. The easiest way to do this is to use an interface with loopback. If you don't have an interface, you could try using a virtual audio cable, e.g. VB-Audio, or VAC if you are on Windows. These things aren't easy to use. I can't speak for other OS'es since I don't use Mac or Linux.