I have several test CDs and a pink noise generator built into my EQ.

Problem is, all the CDs and the EQ's pink noise all sound different....

I know it depends on the band filtering and coherent/non-coherent (incoherent? sic) but what is 'true' pink noise, and where does one get it from?

on a light note, you can make it with your lips just say shhhhhhhhh

it plays every frequency between 1hz and say 25000hz , they are all recorded individually and played back at the same output level.
Well thats my assumption.
That way if a certain band is to low it will show on the graph as a low point.

Don't compare pink noise tracks with your ears. Pink noise should be 20 - 20kHz of equal amplitude noise and when played back on a system will reveal frequency response anomalies - but you need to use an RTA to compare the tracks themselves to see any differences. ONce you have an RA you can plug your CD tuner's RCA directly into it and play each track to see what the true content is. Pink noise will give a flat response line.

White noise is equal power per frequency, pink noise is equal power per octave (or white noise with a 3db/oct low pass filter applied) which more closely matches how we hear sounds.

volenti is correct
but if u have a spare 1000 friends, my lecturer said that if u had 1000 people all in a room, all having different conversations at the same time, that is an EXTREMELY close representation of pink noise
got that many friends?

not that many friends - I am too tied up listening to pink noise.

What I was trying to get at -
1. How can pink noise be the same if it sounds different?

2. Will I get different RTA results if I use different sounding pink noise?

i've played pink noise, and i find that the RTA reading varies as the track plays!!
so i use a slower setting...

some RTA's have a pink noise generator, i'd use that one if you can!

yeah, thought so