(to clarify, Tri-X in 35mm, color neg in larger formats, my first post wasn't clear)
Thing is, I'm not going to continually check histograms when I'm shooting. I expect the 109827-pixel metering they advertise to actually work. I don't stop between every shot with my Leica to get a light reading, and Nikon certainly doesn't tell dSLR owners that they should.
When you're shooting film, you have the same issue when the light changes, histograms are a lot easier to deal with than spot meters. But really- you normally don't need to change out the exposure compensation until you get a shift in light that's not linear. Even though the Canon meters bias more toward the high end, the situation is the same- so just dial in some positive exposure compensation. All camera (and off-camera) light meters work essentially the same way- while you get some interesting "database" things going on with 3D Matrix metering, if you're wanting the best exposures, you're going to have to work within the system. It's certainly easier than trying to catch a sunset picture with a 5x7 and a Sekonic.
Underexposure is a commonly reported issue (or 'feature' as some might say) with the Nikon dSLRs of that generation and even into the D200s. Nikon wants to 'protect' the user from blown-out highlights.
Just like setting up for a new film, you need to calibrate the equipment to how you shoot. Changing film types always meant re-calibrating- the sensor is your film- you only have to do it once for the most part, and it's *really* easy to validate with the histogram once you do it.
Shooting for ~2.5 hours outside today in changing light, it took two shots to dial in exposure comp. on the histogram, and all my shots were where I wanted them for the day, despite the ever-changing lighting.
Once again, if you're not doing the right things shooting digital, you're not going to get the best shots from *any* manufacturer's system.
What experience I have had with Canon digital bodies points to it being less of a problem - maybe Canon's more willing to let you blow out highlights if you're not careful.
There are tons of threads on Canon's metering bias, every time I've shot with a Canon, I've used the histogram to adjust exposure compensation- it's generally been lesser number-wise, but it hasn't been negated.
Nikon doesn't offer the same speed and the same features as Canon does in decently-priced prime lenses. In high-end and constant zooms, I'm sure they're pretty much equal.
But my 50/1.4USM has no Nikon analog at its price point, Nikon offers little in the way of fast and wide glass (you have to go to Sigma's 30/1.4 - Canon gives you options in the 24-35 range). The Nikon primes often feature cheaper construction and terrible focus rings (similar to the design of the $65 Canon 50/1.8 - that thin strip of hard plastic perfect for slipping the edge of your finger in the shot).
I've never had a focus problem, or problem from build-quality with any Nikkor I've shot, but then I can say the same for my Mamiya 645 and Pentax 67 lenses- I think the last time I had a lens issue, it was an early-eighties Soligar zoom. If you have a 50/1.4, I don't know why you didn't just stay Canon, but the Nikkor 50/1.4 looks like it's in the same price range.
Less than 1/10000th of 1% of what I shoot would ever need to be both fast and wide, so that's not somewhere I have done a lot of research, but the sales of lenses faster than f/2 has never been a large market.
[qutoe]Why wait? You're obviously unhappy- and you're missing pictures now.
Waiting for some kind of dust protection feature, and I'm quite happy shooting more film now.[/QUOTE]
You may want to read how poorly Canon's anti-dust feature actually performs before "waiting." There's a thread in this board with a link- but the long and short of it is that it actually made things worse.
Personally, if I already owned a D70, I'd be trying to get the most out of it I could. I know I can make perfectly saleable images with a D70, just as I could with a 5D or a 1DSmkII or Rebel 350.
Using sensor swabs really isn't that difficult, and produces much, much better results than any "anti-dust" feature, but Canon's is about the worst out there.