Multimeter readings

[rhdwave]

Hey all, I was trying to debug an axis fuzz that i was building for a friend.  The original i built some time back works great, has all the right transistor voltages etc. This one is off.  I was thinking that there might be something wrong with some of the resistors around the transistors, so i took out the mulitimeter to try and get readings.  With the one that's supposed to be 680k it goes up to like 380k and then .45M and then real real slowly continues to go up.  This also happened with the 820K resistor.  This doesn't happen with anything else on the board.  I checked the batteries on the meter.  Has anyone else encountered this?  The other thing is, i actually like the way this fuzz works, it's just missing some of the anticipated boost and mayhem with the drive control that i have on mine.  But i like the pure distortion sound of this build better, it seems smoother somehow.  I was wondering also if the resistors are lower than they should be, i'm assuming this would change the type of distortion. 
Any feedback would be greatly appreciated...much thanks...Here's a link to the schematic:

http://www.diystompboxes.com/pedals/axisfuzz.gif

[R.G.]

Measuring high value resistors soldered into a board is hard to do accurately. To do a good job, you have to unsolder one end.

This is because the rest of the circuit is in parallel with the resistor, and changes the reading. In your case, it sounds like you're charging a capacitor somewhere.

An ohmmeter works by feeding in a small DC current and reading the DC voltage this makes. The good ones use a voltage less than a silicon diode drop so they can measure resistance in-circuit without turning on a junction. The less-good ones use a larger voltage. In either case, paralleled R-C networks can give false readings.

[rhdwave]

Okay, thanks RG that makes a lot of sense.  Now, these are the readings i was getting for the transistors:

Q1 E= 3.44
     B= 2.85
     C= 3.39
Q2 E= 2.78
     B= 3.38
     C= 5.22

My original build was much closer to what someone had posted the actual Axis Fuzz readings to be:

Q1 E= 4.32
     B= 3.63
     C= 3.21
Q2 E= 2.64
     B= 3.21
     C= 5.87

Q1 is 2N3906   and Q2 is 2N3904

I tried swapping out the transistors and got different readings, but still nowhere near as close as what they should be.  The thing that's also strange is that i like the sound of the new pedal better, it seems warmer, not quite as harsh on the distortion.  But it doesn't have the full drive and level boost that i was expecting.  I wish i knew what was causing the change in sound so i could duplicate it in the future.