Good catch, sorry! here's corrected schematic:

This is a buffer, so gain is unity. To make it into a gain stage, you'd have to insert a divider in the feedback (between "-" and output pins):

Gain formula for non-inverting configuration is: 1 + Rf / Rdiv (in the example, it is 2).
For inverting configuration, it is: Rf / Rin, and Rin sets the input impedance. R4 is added for a minimum value (so, Zin is Rin+R4):

Changing Rin changes gain ratio. You can set Rf lower that maximum Rin value to make it negligible (say, 10K in this example).
In both configs, input capacitor and impedance after it form a high-pass filter, so roll off low frequencies.
Pull-downs are placed outside the input cap. Values must be larger than your input impedance or they will affect it.
Usually I see the resistance in the negative feedback going to the ground, not in the VB. But I don't want to open too much topics!

Thank for your schematic!

Think this is better. Again quick drawing

Thanks for your corrections!

I feel a bit of messy, anyway.
I'm finding different theory about the calculations of the impedance of the two type of schematic inverting input and non-inverting input.
INVERTING INPUT
Looking at the last schematic posted by Kipper4, I can see the inverting input with a dual pot. Usualy the non-inverting input is not used for higher input impedance, but let's see it enyway.
If the impedance is setted just by the "input resistor" I can use a 10k resistor + 1M pot. And to use a fixed resistor in the negative feedback of the op-amp. I guess the input resistor controls even the gain and the 1M pot at max will give me maybe too much attenuation of the volume?
If even the resistor in the negative feedback controls, with the input resistor, the impedance I could use a dual pot, but I guess it's no so easy to get the right range input impedance 10k-1M, because I assume the resistors will work in parallel, etc...
NON-INVERTING INPUT
About the non-inverting input the input impedance is setted by the resistor to the VB, and it should be pretty easy to add a 10k reesistor + 1M pot.
The input impedance could be influenced by a resistor to the ground, in this case the input impedance will be the value between these two resistors in parallel.
I probably need a resistor to the ground as pulldown resistor. Can I put in before the cap to avoid the pop of the switch and avoid the interaction with the resistor in the VB?