Thanks so much for this! I’ve been chasing the “boxiness” sound of an amplified double bass and how to get rid of it for years, and a notch @ 220 hz. tames it!
Sorry this is probably TLDR. If your interested, try to get through it.
IMHO the primary "boxy" quality lives in range from approximately 200-400hz.
IMHO, notch filters should be tuned to the most prevalent feedback frequencies...which can vary with the situation and equipment used.
IMHO a single notch filter is totally insufficient. In my experience you can notch between 4 and 6 problem frequencies before you hit the point of diminishing returns. It's best to tune the notch filters each time you set up. Set it and forget it is not optimal.
After notching 4-6 of the most prevalent feedback frequencies, the tone of an upright tends to get a lot more clear and true. However, some gentle broadband tone shaping is also typically required for optimum results. Generally the wider the bandwidth of the filter the less cut you apply.
The corrective EQ I am describing requires a very flexible and powerful EQ. I have had really good results running a 31-band graphic in my amp's Effects Loop. Another option is several bands of fully-parametric EQ. Course, these types of EQ are physically a lot bigger than many of us want in our gig rig. Also many of us don't have interest or possibly the aptitude to learn how to effectively use them.
Background:
Consider the instrument an amalgam of the bass, pickup, amp, cab, spacing between bass and speaker, and room acoustics.
I will refer to this as the "system."
The system's frequency response is not flat or true to the bass's response. Also there are a mix of fixed and variable resonances in the system which modify or distort the system's response.
An example of a fixed resonance is the top plate, tail piece, and afterlengths of an upright bass may combine to form a resonance that is below 100hz. The way I usually solve this resonance is by wedging a rolled up towel under the tail piece. To be honest, I haven't had a frequent problem with this resonance. I had to regularly use the towel trick with one bass that I played long ago. I have experienced this problem on two or three other occasions when a feedback loop formed between the bass and subs on a big festival stage. On those occasions, the PA was notched at the offending frequency.
An example of a variable resonance is the complex interaction between the various elements of the system.
I will try to provide a suitable explanation why these resonances vary.
The entire sound spectrum travels at the same speed. In other words, high frequencies and low frequencies travel at exactly the same speed.
Frequency is cycles per second and the amount of time it takes for one cycle to occur is called the period. Period and Frequency are the inverse of each other (Period = 1/Frequency).
Since all frequencies travel at the same speed, each frequency has a different period and each frequency will have a different wavelength. The formula for wavelength is:
Sound Wavelength (λ) = Sound Velocity (V) / Sound Frequency (F)
FYI, Lambda (λ) is used as the symbol for wavelength.
The electrical signal between the pickup and the amp travels at the speed of light. Then the sound travels at the speed of sound back from the speaker to the bass and pickup. The time it takes the sound to travel the distance is called propagation delay.
Caveat: Let's assume the bass, pickup, amp, and cab have the same absolute polarity, meaning the output of the speaker has the same polarity as the source sound created by the bass.
What we are concerned about is the phase angle between the source sound and the sound that has traveled from the speaker back to the bass and pickup. In other words we are looking at the relationship between wavelength and propagation delay.
If the returning signal is perfectly in phase with the source sound we get summation. If the returning signal is out of phase we get cancellation.
Example of how phase angle between two sine waves effects summing :
Let's summarize some key points and draw a conclusion. Both wavelength and propagation delay determines the resulting phase angle between the original sound of the bass and the sound that has travelled back from the speaker. The resulting peaks and notches are called comb filtering.
Also the amount of propagation delay varies by distance. If you change distance, you also change the propagation delay, and as a result
you change which frequencies sum and which frequencies cancel.
Now consider what happens when a peak in the comb filter occurs at the same frequency as a fixed resonance. For example if the bass is 28 feet from the subs and has a fixed resonance at 80hz. The wavelength for 80hz is about 14', so the subs are two wavelengths away from the bass. This means the signal from subs arrives delayed by two wavelengths and in-phase with the source signal from the bass. So a variable 80hz resonance is potentially set up at the same frequency as the 80hz fixed resonance.
You also get potential variable resonances at 40hz, 160hz, 320hz etc. Explanation: The propagation delay over 28' = 1 wavelength at 40hz, 4 wavelengths at 160hz wavelengths, 8 wavelengths at 320hz.
Note, the frequency centers of these potential variable resonances shift if you flip the polarity somewhere in the signal path.
So potential fixes for this problem include, wedging a towel under the tail piece, applying a notch at 80hz, changing the distance between the bass and subs, or flipping the polarity. You may need to use more than one fix.
When energy is put into the system, it teds to excite the resonances and further distort the frequency response. So the sound becomes progressively worse as volume of the amp is turned up. The notches should be tuned to these resonances. The idea is to minimize the energy at each resonance.
After the system resonances are addressed, gentle broad band cuts are used to massage system frequency response so it more closely resembles the natural acoustic response of the bass.
