Taking a stab in the dark...

It may be that the colours are on a 48-bit scale in stead of 24-bit; that is, 16 bits are available for each channel (red, green, and blue).

This is becoming increasingly common in anticipation of 48-bit consumer-level hardware, ie: pretty soon you'll actually need the extra bits, and be glad that they're there.

If this is the case here (but I don't know Tk, so YMMV), the range would be 0..65535; everything else should remain unchanged. Try grabbing $widget->rgb('white') and $widget->rgb('black'), and using their triplets as the limits of your sliders:

something like...

my %colour_min, %colour_max;
@colour_min{ 'red', 'green', 'blue' } = $widget->rgb('black');
@colour_max{ 'red', 'green', 'blue' } = $widget->rgb('white');

# ...

my %triplet;
@triplet{ 'red', 'green', 'blue' } = $widget->rgb($user_entry);
my %slider; # 0..1 range [eg. 50% grey might be (.5,.5,.5) ]
for my $channel ('red', 'green', 'blue')
{
  $slider{$channel} = ( $triplet{$channel} - $colour_min{$channel} ) /
+ ( $colour_max{$channel} / $colour_min{$channel} ) );
}
[download]

That's probably overkill, though, since you can generally assume that black will be (0,0,0), which reduces the final block to

for my $channel ('red', 'green', 'blue')
{
$slider{$channel} = $triplet{$channel} / $colour_max{$channel};
}
[download]

There's also probably a right way to do this that I don't know about... :-)