$m->method(); vs $m->$method(); benchmark

.pl

#!/usr/bin/env perl
use strict;
use warnings;

use Benchmark;
use lib map { "$ENV{HOME}/sandbox/$_/lib" } qw(Music-FretboardDiagram);
use Music::FretboardDiagram ();

my $count = shift || 100_000_000;

my $m = Music::FretboardDiagram->new(chord => 'x02220');

timethese($count, {
    raw_method => \&raw_method,
    var_method => \&var_method,
});

sub raw_method {
  my $x = $m->spec_to_notes('x02220');
}

sub var_method {
  my $method = 'spec_to_notes';
  my $x = $m->$method('x02220');
}

.txt

Benchmark: timing 100000000 iterations of raw_method, var_method...
raw_method: 535 wallclock secs (535.25 usr +  0.02 sys = 535.27 CPU) @ 186821.60/s (n=100000000)
var_method: 544 wallclock secs (543.73 usr +  0.08 sys = 543.81 CPU) @ 183887.75/s (n=100000000)