Another Sectral Analysis -- need Help!

[Giganova]

hey guys & gurlz --

I recorded a test tone from my exact monitoring position after I installed some home-made wall panels (see another thread below). Signal chain was Gefell M930 -> Focusrite Green pre.

Now, how am I supposed to interpret the analyzed recorded test tone (top panel) compared to the test tone (lower panel)?? Are there problems here (maybe between 150-500 Hz?) or does it look OK? Am I in deep trouble?

Needed some advice from guys who have experience!

Thanxxxx!

 

[Ethan Winer]

Gig,

> how am I supposed to interpret the analyzed recorded test tone <

What's to analyze? The data is what it is. This type of skewed response is absolutely typical of untreated rooms. I suspect the real response in your room is even worse than shown. Do you know the resolution (bandwidth) that data is accurate to?

Either way, the solution is acoustic treatment, especially bass traps. Have a look at the Acoustics FAQ, second in the list on my Articles page:

www.ethanwiner.com/articles.html

You'll find other related reading material on the Articles page of my company's web site, linked under my name below.

--Ethan

 

[Giganova]

Hi Ethan --

thanks for stopping by, much appreciated.

> Do you know the resolution (bandwidth) that data is accurate to?

Well, this freebe plugin that I used allowed binning into "channels"; I set it to 1024 but actually don't know what it means in terms of frequency. Binning into 2048 channels clearly smoothes the response curve. If I chose a smaller channel size, the test tone shows some funny sinoidal behavior, so I thought 1024 is a good choice. The frequency range is 20Hz-20kHz, divided by 1024 gives a resolution of 19.5 Hz, I guess.

My problem is that I currently live in a rented apartment and can't really dramatically change the appearance of the room. The idea of this test was more to get a "feeling" about the room acoustics. So it looks pretty bad, does it? Are there any frequency plots out there which show a good room acoustics?

One thing, though: In order to get an idea of the real acoustics of the room, I needed to take the frequency response of the microphone and studio monitors intro acount, right? So the actual room acoustic might be better than what is plotted (although it might be a small effect).

I'll carefully read your articles and see what I can do. Thanks!

 

[Ethan Winer]

Gig,

> The frequency range is 20Hz-20kHz, divided by 1024 gives a resolution of 19.5 Hz, I guess. <

If they did it correctly it will be a logarithmic division, not linear. So 1,024 points across ten octaves (20-20k) is about 1/100 octave. That's fine enough resolution - if that's really what they're doing.

> My problem is that I currently live in a rented apartment and can't really dramatically change the appearance of the room. <

So you build or buy portable acoustic panels. No big deal.

> Are there any frequency plots out there which show a good room acoustics? <

Every room is different. If you could get to +/- 6 dB (hint: you can't on your budget) you'll be doing really well. Achieving +/- 10 dB is probably closer to what you can realistically aim for.

> I needed to take the frequency response of the microphone and studio monitors intro acount, right? <

Yes, but any decent mike and any decent speaker will be far flatter than what your room contributes to the overall response.

--Ethan