If you play a note really hard, it will often start a hair sharp because of the extra tension while it's vibrating at high amplitude. A gauge tuner probably won't see that as much as a faster-responding strobe or strobe simulator would. So, it's better to wait until the note volume has dropped down to where the pitch stabilizes.
The bigger cause of tuner instability is something else though. When you play a note, you're playing a fundamental and many harmonics. Because of how stringed instruments work, the chances of the fundamental and the harmonics all being in tune is pretty much zero. It's because a real-world string has a stiffness factor that affects a note and it's harmonics differently - the higher harmonics act as if the string it too short and therefore will tend to be sharp. This is why you set your intonation, but that's only a partial fix. If you set intonation at the octave then your fundamental and harmonics will only be in tune around the octave and likewise for wherever you set intonation at.
The tuner hears this stack of out of tune notes and tries to come up with the frequency of it. In most cases, it will end up measuring the loudest harmonic. As a plucked note sustains, the harmonics are dying out at different speeds (in most cases high ones first) and the loudest harmonic is changing. That's what's causing the tuner to give an unstable reading. If you check intonation at a particular note, you may see that that particular note doesn't have very good intonation, even if the instrument is properly set up. Personally, I wouldn't sweat it that much - there's no way you can perfect it across the neck (perfect strings can't be made) and you'll be on to the next note soon anyway!
Your strings will probably be more perfect if you play lighter gauges with thinner more flexible cores. Materials and construction would matter too. Unfortunately at the relatively short scale length of electric basses (vs. something like an upright bass) approaching perfection would probably be a very wimpy sounding string indeed. A guitar low E string at 34 would vibrate much more ideally, but wouldn't sound particularly good.
A true strobe (a Conn Strobotuner or one of the very high end Petersons) is really fast to respond to pitch changes. They way that they work is that there's a disk (or group of disks if you have a really good one) with a black-and-white pattern in concentric rings spinning at a very accurate speed. The audio note you play is amplified and fed to a neon lamp which then blinks at a speed that reflects the frequency of what you're playing. Unlike an incandescent bulb, the neon bulb turns on and off instantly and doesn't slowly dim. The flashing neon light lights the rotating strobe disk, and if the pitch of the note you're playing is exactly correct, the pattern on the strobe disk will appear to be stationary. If the pitch is high or low, you'll recreate the appearance of movement to the left or right.
A regular guitar tuner like the Boss has a little microprocessor in it that tries to count the frequency of what it's hearing, compares it to a reference table and moves the needle to show sharp or flat. The accuracy of the pitch shown depends on the accuracy of this measurement and how fast the measurment is taken. To simplify counting the frequency, most tuners don't listen to the entire spectrum of the instrument - the highs and lows are filtered off to make it easier to count. So it might not be listening to the fundamental of your bass, but a harmonic and trying to measure that. If you filter off too much of the fundamental or high harmonics, you may have a hard time accurately setting intonation.
A true strobe tuner doesn't have this kind of problem, and, in fact can simultaneously show the fundamental and harmonics out of tune with each other. Each of the concentric bands on the strobe display is a different harmonic. So it's definitely a better machine for setting up intonation, if you can get your hands on one.
Boxes like the Peterson Virtual Strobe or StroboStomp are more like the Boss tuner than a true strobe. Instead of wiggling a needle, a VS-1 is drawing a picture of a real strobe display. But what Peterson is doing here is splitting highs, mids, and lows into multiple bands, performing the calculation individually on each band and showing them individually in the simulated strobe display. It's not as fast or accurate as a real strobe, but will show things like the harmonics being out of tune with the fundamental. And they count more accurately too, since they have a more powerful processor.
David Fung
