RT60 Rumor

[pepe ferrer]

thanks Arthur,

Yes, the problem of averages with a large window FFT, is time.

I really like using the spectrograph in my classes to explain the delay that occurs just a filter, and see how energy moves in time.

1. Turn down or off the background noise to lower the noise floor of the room. Can be as simple as turning off the air conditioning, telling the custodian to vacuum the room later, telling the lampies to take a break, etc.

A gretting

[MrChris]

Hey thanks for all the responses everybody, I really appreciate it. I had no idea there were more replies, sorry, hope not to seem rude, I don't know what happened, I was subscribed, maybe I need to subscribe again. But check it out, my main confusion for the impulse response was that there never was an impulse, the noise generator never stopped, so I am baffled at how I can trust the impulse response measuring decay, when there is no decay taking place because the noise keeps going.
Update on the female vocal. The problem turned out to be a cordless mic she was using that would compress lower frequencies when female singers get loud. Hooked her up with the sure ksm9, problem solved.
but still have some certain frequencies that have huge decay times.
Im here back checking on the RT60, glad to hear that its being worked on. YAY! Just bought the new usb 2x2 with mic because the church bandit made off with the audix. But thats good because now I don't need to use a mixer channel for reference, and I can bus out a single signal to compare it with the mic signal, Looking forward to the new device this coming weekend. I will try the impulse again and now I know I need lots of time, but I still don't get how decay can be measured while the signal is constantly sounding.

[MrChris]

What you need to understand is that the impulse response is simply a "snapshot" in time of what you consider a continuous event. Essentially with the time window you're telling the impulse recorder to record all events happening in x number of FFT samples or x number of seconds. This includes the direct sound arrival from the loudspeaker, and all reflections happening afterward, along with the background noise of the room. So whether you have a pink noise generator turned on continuously, or a swept sine wave, or a balloon bust or a starter pistol shot, Smaart is simply recording all acoustic events that are happening in a set period of time, determined by how big a time window you set.

HHMMM, I think I can almost understand this. I will try again this weekend, make some adjustments and see if I can see those reflected in the impulse response window.. Thanks! I really appreciate the in depth explanation very much...

[MrChris]

OK I get it. the impulse only records a snippet of sound in time as it travels through time and ignores all new sound introduced later in time. Thats deep but I think I see the light.

[Arthur Skudra]

OK I get it. the impulse only records a snippet of sound in time as it travels through time and ignores all new sound introduced later in time. Thats deep but I think I see the light.

I'd say you're starting to get it.

Part minor clarification of terminology, and part being nit picky: it doesn't *ignore* any events happening beyond the size of the window as if they are completely lopped off from the data being input into the FFT engine, moreso, it weights those events that are outside the window down enough, so you don't see them in your measurement for all practical purposes. Depending on what kind of window you choose, it does so gradually at the beginning and end of the time window. The data is always still there, you're just telling Smaart what to do with it. This is an important concept to understand, particularly when you are looking at a magnitude trace, which essentially uses the same windowing technique, just a different viewpoint of the data being input in to the FFT engine.

For further reading, see this technote:
http://www.rationalacoustics.com/files/technote3.pdf

[gluis]

Hello Chris,

Please remember, Smaart is a dual channel FFT program. What it does is compare your reference signal (i.e. generator output) to your measured signal (i.e. what you pickup through your measurement microphone). The resultant is the difference between one and the other. If you subtract your reference from your measured signal, you are left with audio system and room contribution, and this difference is your impulse response! It does not need an actual impulse. In fact, it would be undesirable because it would be too short to do some of the calculations/averaging.

Please excuse any syntax error as english is not my native tongue.

Regards,

GS

[MrChris]

For further reading, see this technote:
http://www.rationalacoustics.com/files/technote3.pdf

Wow what a read! That also answered other questions I had about the different hamming, etc. windows to choose from. Thanks for the super link. My mind operates on a new level of understanding now. Im enriched! Changed forever! Thanks so much!

[MrChris]

Hello Chris,

If you subtract your reference from your measured signal, you are left with audio system and room contribution, and this difference is your impulse response!

Please excuse any syntax error as english is not my native tongue.

Regards,

GS

Perfect! I will add this to my newly found heap of understanding, fits right in! Thanks so much!
I see no syntax error, I see illumination!

[Adam Black]

Wow what a read! That also answered other questions I had about the different hamming, etc. windows to choose from. Thanks for the super link. My mind operates on a new level of understanding now. Im enriched! Changed forever! Thanks so much!

Awesome! Love those aha moments. Glad the community was able to clear things up for you.

[MrChris]

Thanks! Me Too! Its on this Saturday! Look Out!