Experiment: Level Matching Frequencies


For fun, from my listening position, I experimented with a Radio Shack SPL meter and a demo disc. Using the ROON equalizer, I level matched frequencies at 80dB (50Hz, 63Hz, 80Hz, 100Hz, 125Hz…etc.) This took a few hours and the results were not good.

Of interest, I corrected a peak of +12dB @ 80Hz, a dip of -8dB @ 100Hz,  and a peak of +6dB between 160-200Hz. Other frequencies up to 10kHz needed minor adjustment of +/- 1 to 3dB.

With adjustments, music sounded terrible. For example it was very thin in the mids, compared to no EQ. This experiment was a waste; I expected and hoped these adjustments would improve the sound. Please note: I am a recreational listener, not and EE nor majored in music which means this experiment was ad hoc and  unscientific.

Have any of you tried this or have thoughts as to why level matching frequencies this way had poor results?

C7e2e6f0 47c8 425b ba49 b015511fec814hannons
Possible problem with your experiment.           
First off, how were you able to adjust the 'Q' of the areas you manipulated? The 'Q' of the actual sound may vary a great deal from the 'Q' of your EQ device! And usually the EQ device is the one failing to have as tight a 'Q' as the real world speaker. So (at least guessing) when you adjusted the  system you were not just dropping the actual peaks, but dragging down a broad range of frequencies nearby.      
That might explain the general leaness of your results.       
One way to alleviate would be boost around the peak instead of dropping the peaks. leave the peaks up, and bring up the rest of the sound. That way you are not as caught by the failure to limit 'Q' as much . (at least I think it might work better)                
(If anyone is confused by what 'Q' is, the Q is the curve of the slope. A tighter Q is a narrower and thinner slope to the peak of the change. A low Q is a broader wider range across the frequency spectrum.Failure to control Q is probably the worst feature of many equalizers.

Thanks Elizabeth, forgot to mention that I adjusted the Q as necessary, for example the peak at 80Hz and at 100Hz both had a Q of 1 (very thin slope), the Q at 160-200Hz had a Q of 6 (or 8?) at 185Hz to cover the range.

Quite possible I didn’t have tight enough Q; perhaps I should have used 0.5 Q in the above examples.

Your suggestion of boosting around the peaks might have merit, it’s certainly something I’ll give some thought.

I am no audio pro, but I recall in my youth buying an equalizer to correct sound imperfections. It, like your experience, was one of disappointment.
One of the problems is that most speakers have crossovers and other such innards that will react, sometimes negatively, to external changes in frequency levels.- I am probably not phrasing this correctly, but I hope you get the idea.
A better way to affect sound deficiencies would be room correction. At least to my simple mind.
Bob

Thank you, Bob.

IIRC the EQ function of ROON has received praise for not harming the signal. The more I think about it the more this experiment may become a laypersons journey into optimizing the ROON equalizer.

Can any other  ROON users chime in? Or anyone with experience in EQ? Perhaps this will morph into a  discussion of tips, successes and failures using the EQ function of ROON.

Um, don’t do that. :)

First, if you need a good, cheap calibrated mic, use the Dayton imm-6. The RS SPL meter is severly constrained at both ends.


https://amzn.to/2Ly6G65


If you have an Android, Audio Tools works perfectly.

If you have a new iPhone you will need an adapter and I don’t know what software.

Next, flat is not actually ideal. A descending frequency range from around 100Hz is. Among others, look for the Bruel & Kjaer ideal speaker curves, or look at the documentation for something like Dirac. Whether you like Dirac or whatever, it shows a descending, not flat curve.

There is a lot written on why a flat EQ is not actually desirable in a room, that’s what Google / Bing / DuckDuckGo are for, but we use flat for anechoic, and quasi-anechoic measurements. That's why you'll see those terms in a lot of speaker measurements in reviews.
Eric, I recall that you build speakers. Right?
Do you think my assessment regarding crossover/wiring affecting hound level balancing correct?
Bob
IT is true, plenty of cheap microphones have far far worse frequency response than speakers!
@gdnrbob2
Not really, assuming you have a decent SS amplifier and reasonable speakers, the crossover and EQ won't interact very much, but your point that EQ should be used sparingly is not wrong.

Using EQ well is an art. In the bass, you want surgical precision. In the midrange and treble  you want to be delicate.

You can only adjust a speaker so far, and the idea of using an EQ to make up for it's deficiencies never works out well. What EQ does well is in the mastering process, or to help correct for certain room problems.

Also, using them as tone controls, with grouped, and gentle adjustment is a good thing.

Best,
E