olympus e30 or canon 40d?

[Anbesol]

I don't know how they stack up statistically, but there is FAR more anecdotal evidence around about sample variation in Sigma lenses than about any other brand, Tamron would be next worst. BTW, what does that have to do with 4/3 sensors being "bad"? They're not, they just have a roughly 50% higher pixel density than comparable resolution APS sensors - which lessens the size of the photosensitive parts of the pixel, making them somewhat less light sensitive.

True, also demanding better lens to achieve those good results. 10 megapixels on an APS sensor vs 10 on 4/3, put an identical lens on each and the lens problems will be more noticeable on the 4/3 than on the APS.

Thanks for the info on the crop factor.

[Sushigaijin]

True, also demanding better lens to achieve those good results. 10 megapixels on an APS sensor vs 10 on 4/3, put an identical lens on each and the lens problems will be more noticeable on the 4/3 than on the APS.

Thanks for the info on the crop factor.

The opposite, actually.

A 4/3 sensor would just use the "sweet spot" of an APS lens, while an APS sensor would record the imperfections at the edges like vignetting and softness. Since the center of the lens is closest to "on plane," it is the sharpest and most accurate part of the lens.

Quality diminishes towards the edges, which wouldn't even be visible on a 4/3 sensor.

Full frame cameras are the most susceptible to lens flaws. The smaller the sensor, the cheaper and easier it is to manufacture a crackin' lens.

This is why the (sigma, rebadged by olympus Zuiko) 70-300 lens is sharper on 4/3 than any of the other mounts it comes in: the 4/3 sensor just uses the good parts of the image circle and discards the rest. It's just like cropping the center out of an APS shot and blowing it up to equal size, without losing any IQ due to enlargement.

[Anbesol]

The opposite, actually.

A 4/3 sensor would just use the "sweet spot" of an APS lens, while an APS sensor would record the imperfections at the edges like vignetting and softness. Since the center of the lens is closest to "on plane," it is the sharpest and most accurate part of the lens.

Quality diminishes towards the edges, which wouldn't even be visible on a 4/3 sensor.

Full frame cameras are the most susceptible to lens flaws. The smaller the sensor, the cheaper and easier it is to manufacture a crackin' lens.

This is why the (sigma, rebadged by olympus Zuiko) 70-300 lens is sharper on 4/3 than any of the other mounts it comes in: the 4/3 sensor just uses the good parts of the image circle and discards the rest. It's just like cropping the center out of an APS shot and blowing it up to equal size, without losing any IQ due to enlargement.

No, that is only one aspect of lens quality. Higher pixel density means things like diffraction will be more severe. High pixel densities can bring out the best things in a lens, but it also makes lens problems worse.

[Sushigaijin]

No, that is only one aspect of lens quality. Higher pixel density means things like diffraction will be more severe. High pixel densities can bring out the best things in a lens, but it also makes lens problems worse.

True, but APS pixel density is higher than 4/3 - the 50d's 15 mp is higher than the density of it's closest 4/3 competition, the 10mp E3.

You can rightly say that those are two very different image sizes, but those are the best fit for a camera comparison. based on exact same image sizes, 4/3 will always be more dense.

My point isn't to split hairs but to provide a real-world working example...my point is that you are MORE likely to notice edge distortion on a large sensor than 4/3 pixel density issues in the sweet spot of a lens.

[Anbesol]

True, but APS pixel density is higher than 4/3 - the 50d's 15 mp is higher than the density of it's closest 4/3 competition, the 10mp E3.

You can rightly say that those are two very different image sizes, but those are the best fit for a camera comparison. based on exact same image sizes, 4/3 will always be more dense.

My point isn't to split hairs but to provide a real-world working example...my point is that you are MORE likely to notice edge distortion on a large sensor than 4/3 pixel density issues in the sweet spot of a lens.

Another point here - most people shooting APS are also using lens in which they pull through the sweet spot. EF-S lens aren't anywhere near as common as the standard EF, and the same with Sony's DT lens. Nikons the only exception with their popular DX lens type.

And well, pixel density can also be a very good thing, my original point was simply that greater pixel density makes the camera much pickier with lens, a very expensive catch 22.

[Dougjgreen]

True, but APS pixel density is higher than 4/3 - the 50d's 15 mp is higher than the density of it's closest 4/3 competition, the 10mp E3.

No it's not. The Canon has 15.5 million actual pixels (15.1 million effective), on a sensor that is 329 mm square, for a total density of 47.1K pixels per square mm.

The Olympus has 11.8 million actual pixels (with 10.1 million effective) on a sensor that is 225 mm square, for a total density of 52.4K pixels per square mm.

Incidentally, the reason Olympus has so many more actual pixels vs. effective pixels relates to their live view implementation, which basically consumes around 1/8th of the total sensor resolution.

That being said, those two density numbers are pretty darn close, so there isn't any fundamental reason that a 16 MP Canon APS-C sensor would be any more sensitive than a 10 MP 4/3 sensor of the same generation. Other than the fact that Canon has been making their own sensors for quite some time, and has lots of experience at high ISO sensitivity. Of course, the 50d's sensor is actually a generation newer than that of the E-3. The more direct competitor from a silicon generation standpoint is the E-30's sensor.

In any case, BOTH an APS-C sensor and a 4/3 sensor get the benefit of the sweet-spot effect when only using the center portion of a lens that's designed for the full 35mm frame. Most good lenses have no problem resolving that density at the center, at least when stopped down 1-2 stops from the max, but the further away one gets from the center point, the more it becomes an issue.

[Sushigaijin]

Oh yeah, got me on actual vs. effective, i hadn't even thought about that. I just cranked the math on a calculator without checking specs...but still too close to call IMO. Of course the new 12 mp 4/3 sensors are going to be denser still.

But to address the original topic, the pixel density is unlikely to make much difference in APS vs. 4/3 format in regard to "sweet spot" sharpness. Sensor and processor manufacturing is more likely to effect output - some companies are just plain better at making sensors than others, but it's really a marginal improvement and only applicable to the minority of photographers.

In reality the 4/3 format and the APS formats are very very similar in image quality, especially at normal sensitivities.

As far as lenses go, the 4/3 system is FAR cheaper than APS systems for very high quality glass. 4/3 just doesn't have the huge selection of primes that are found on the larger systems. No matter, the ZD zooms are as sharp as the other system's primes

Where olympus really gouges you is the pro-level cameras. The entry level models are probably the best deals on the market, but the E3 and E30 are priced by a crazy person - they are both easily 1/3 higher priced than they should be. The E3 is 18 months old and it is just now going for the price it should have MSRP'd for. The fortunate thing is that you can go from 24-400mm for $1400 (new) in two lenses that easily compete with the "L" series in quality (and weathersealed, too).