EQ melodica

Rewind · 09-26-2012, 05:53 PM

If you were going to do some shelving for a melodica, what would you do? Would you cut it, or boost it? I'm guessing you'd want it kind of high medium, low treble... I might be wrong about that. At what frequency would you boost or cut. Maybe 11kHz, or 12k?

Freebase Dali · 09-26-2012, 09:15 PM

Honestly, you should probably utilize a frequency analyzer to show you the frequencies the melodica is occupying. When it comes to keyed instruments, it's not as cut and dry when EQing as it is with instruments whose fundamental frequencies are actually embodied in the instrument itself.
With keyed instruments, like piano, for example, there's the sound of the piano's shell, and then there's the sound of the piano strings themselves, which will extend into a large frequency range, unlike the piano itself as a whole. Generally in those situations, you'll find yourself needing to EQ the overall character, versus particular frequencies, and you'll also be mostly focusing on any characteristics that are negatively impacted by the recording space.

I've never actually worked with a melodica, but I know what it is and the principle is the same. Basically, you don't simply EQ because you can. You EQ because you need to. Using a proper pair of studio speakers in a proper listening environment, listen to the recording. What stands out? If you have a really shrill sound on certain notes, then to reduce them, you'd simply apply an EQ that gives you Q, freq and Decibel control, and with a very narrow Q, raise the decibels on it so that it's basically a spike in that particular frequency.
Then sweep it back and forth through the frequencies until that particular shrill sound is exaggerated. Once you land on the spot, then simply reduce the decibel level so that it is a cut, rather than a boost. The amount by which you do this is purely to taste, but I would not recommend completely removing problematic frequencies, as they are a natural part of the instrument, regardless of what it is. Simply reduce it so that it does not stand out irregularly.
Follow the same approach for any other errant frequencies.

Also, if recording in an acoustic environment, the microphone itself, as well as the room you're recording in, will place a sonic imprint on the recording itself. That can be a good thing or a bad thing. If your microphone and room isn't so good, there will be a frequency correlation.
When dealing with mic and room characteristics, you'll want to do frequency sweeps, with mostly thin Qs, boost and sweep to find frequencies that resonate with a lot of notes, instead of just a few. If you find a frequency where everything is ringing to hell, even when the notes themselves are changing with a variety, chances are, you've identified a common resonant frequency that can be reduced. Once you've found those frequencies, simply cut there, but not completely. Just do it enough for the sound to be balanced, and constantly bypass your EQ program to see the difference from the original sound.

Also, remember to put a high-pass shape regardless. When you're dealing with a melodica, especially one recorded with a microphone, there really isn't any need for very low frequencies to be in play, so just roll of at below where the lowest frequency of the audio "lives". You'll either need to use an analyzer to know where that is, or just experiment with where the highpass rolls off. Either way, it's a standard operation to ensure that low-end rumbles aren't eating up headroom and stepping on things. Always better clean and safe, than muddy and sorry.

Obviously, this doesn't take into account the fact that you should be mixing in context with the mix itself.. meaning, you should be making these sorts of decisions based on what it sounds like with the rest of the music. Provided you're not just mixing a single recording of a melodica, then make your decisions based on how your changes affect the mix as a whole, not how it affects the melodica.

SATCHMO · 09-27-2012, 12:17 AM

Just to give you a head start on the great advice that Freebase has given you, (as someone who has sound checked a handful of melodicas) they're notoriously hot in the 4k-6k region.

09-26-2012, 05:53 PM	#1 (permalink)
Rewind Groupie Join Date: Sep 2012 Posts: 18	EQ melodica If you were going to do some shelving for a melodica, what would you do? Would you cut it, or boost it? I'm guessing you'd want it kind of high medium, low treble... I might be wrong about that. At what frequency would you boost or cut. Maybe 11kHz, or 12k?

09-26-2012, 09:15 PM	#2 (permalink)
Freebase Dali Partying on the inside Join Date: Mar 2009 Posts: 5,584	Honestly, you should probably utilize a frequency analyzer to show you the frequencies the melodica is occupying. When it comes to keyed instruments, it's not as cut and dry when EQing as it is with instruments whose fundamental frequencies are actually embodied in the instrument itself. With keyed instruments, like piano, for example, there's the sound of the piano's shell, and then there's the sound of the piano strings themselves, which will extend into a large frequency range, unlike the piano itself as a whole. Generally in those situations, you'll find yourself needing to EQ the overall character, versus particular frequencies, and you'll also be mostly focusing on any characteristics that are negatively impacted by the recording space. I've never actually worked with a melodica, but I know what it is and the principle is the same. Basically, you don't simply EQ because you can. You EQ because you need to. Using a proper pair of studio speakers in a proper listening environment, listen to the recording. What stands out? If you have a really shrill sound on certain notes, then to reduce them, you'd simply apply an EQ that gives you Q, freq and Decibel control, and with a very narrow Q, raise the decibels on it so that it's basically a spike in that particular frequency. Then sweep it back and forth through the frequencies until that particular shrill sound is exaggerated. Once you land on the spot, then simply reduce the decibel level so that it is a cut, rather than a boost. The amount by which you do this is purely to taste, but I would not recommend completely removing problematic frequencies, as they are a natural part of the instrument, regardless of what it is. Simply reduce it so that it does not stand out irregularly. Follow the same approach for any other errant frequencies. Also, if recording in an acoustic environment, the microphone itself, as well as the room you're recording in, will place a sonic imprint on the recording itself. That can be a good thing or a bad thing. If your microphone and room isn't so good, there will be a frequency correlation. When dealing with mic and room characteristics, you'll want to do frequency sweeps, with mostly thin Qs, boost and sweep to find frequencies that resonate with a lot of notes, instead of just a few. If you find a frequency where everything is ringing to hell, even when the notes themselves are changing with a variety, chances are, you've identified a common resonant frequency that can be reduced. Once you've found those frequencies, simply cut there, but not completely. Just do it enough for the sound to be balanced, and constantly bypass your EQ program to see the difference from the original sound. Also, remember to put a high-pass shape regardless. When you're dealing with a melodica, especially one recorded with a microphone, there really isn't any need for very low frequencies to be in play, so just roll of at below where the lowest frequency of the audio "lives". You'll either need to use an analyzer to know where that is, or just experiment with where the highpass rolls off. Either way, it's a standard operation to ensure that low-end rumbles aren't eating up headroom and stepping on things. Always better clean and safe, than muddy and sorry. Obviously, this doesn't take into account the fact that you should be mixing in context with the mix itself.. meaning, you should be making these sorts of decisions based on what it sounds like with the rest of the music. Provided you're not just mixing a single recording of a melodica, then make your decisions based on how your changes affect the mix as a whole, not how it affects the melodica.

09-27-2012, 12:17 AM	#3 (permalink)
SATCHMO Al Dente Join Date: Mar 2005 Location: Texas Posts: 4,708	Just to give you a head start on the great advice that Freebase has given you, (as someone who has sound checked a handful of melodicas) they're notoriously hot in the 4k-6k region.