OK ALL YOU full frame camera fanatics

I absolutely agree.

Yes!

There are many important properties to the sensor, and sensor size is one of the biggest ones (pardon the pun).  Other important attributes of the sensor are:

• QE (Quantum Efficiency -- the proportion of light falling on the sensor)
• Read Noise (noise added by the sensor and supporting hardware)
• Pixel Count
• Microlens Efficiency
• AA filter
• CFA (Color Filter Array)

Please note that higher pixel counts do not adversely affect noise performance.  For example, a 24 MP sensor will produce photos with the same noise as a 12 MP sensor, so long as the QE and read noise of sensors is the same.

Indeed, what we usually find is that the newer sensors with more pixels are more efficient than the older sensors with less pixels, and thus produce photos with more pixels and less noise.

This is true.  That depends on the lens, specifically, the aperture, as you noted above.  For example, a 4/3 lens at 50mm f/2 has an aperture diameter of 50mm / 2 = 25mm.  A FF lens at 100mm f/2 has an aperture diameter of 100mm / 2 = 50mm.

Thus, the aperture area at 100mm f/2 on FF is four times the aperture area at 50mm f/2 on 4/3, so four times as much light (two stops more light) will fall on the sensor at 100mm f/2 on FF than for 50mm f/2 on 4/3, for a given shutter speed (the DOF will also be half as much, as well).

For equivalent settings, 50mm f/2 on 4/3 and 100mm f/4 on FF, for example, the aperture diameters will be the same (50mm / 2 = 100mm / 4 = 25mm), and, consequently, the same total amount of light will fall on the sensor for a given shutter speed, as well as having the same DOF.

Exposure is a function of three, and only three, parameters:

• The scene luminance
• The t-stop of the lens (closely approximated by the f-ratio, in most situations)
• The time of the exposure (shutter speed)

Note that ISO does not play a role in exposure at all.  All the ISO setting on the camera does is apply an amplification (either digital or analog) to the signal.  For example, 50mm f/2 1/200 ISO 100 and 50mm f/2 1/200 ISO 400 have the same exposure for a given scene.  The difference is that the ISO 400 setting has amplified the exposure by two stops.

So, why not shoot at base ISO all the time and push the exposure in PP?  Good question -- I'm glad you asked!

First of all, the ISO setting you chose on the camera will affect what f-ratio, shutter speed, and/or flash power the camera chooses in an AE (Auto Exposure) mode.

Secondly, even if you shot the scene in M mode, neither the camera's image playback feature nor the in-camera jpg engine are not linked to the camera's meter, so the playback and in-camera jpgs would look dark for lower light photos (an easy firmware fix, by the way).

Lastly, for many sensors, such as Olympus and Canon sensors, the analog amplification of a higher ISO setting results in less read noise.  For example, if f/2 1/200 ISO 400 gave the desired image brightness, then f/2 1/200 ISO 100 pushed two stops would be more noisy.  However, the ISO 400 setting would push more of the highlights outside the bit depth of the capture file, resulting in more of the photo being blown.

OK, now that exposure and the role of the camera's ISO setting has been cleared up, we can address the f/4 1/25 ISO 3200 on 4/3 scenario vs FF.  The options for the FF photographer are:

• f/4 1/25 ISO 3200 (Half the DOF and half the noise than the 4/3 photo)
• f/8 1/25 ISO 3200 (same DOF and same noise as the 4/3 photo)
• f/8 1/6 ISO 3200 (same DOF and half the noise as the 4/3 photo, but greater motion blur if scene is not static)

All the comments about noise imply equally efficient sensors.  The system with the more efficient sensor will either extend its noise advantage, or have its noise advantage eaten into, depending on the scenario.

Lastly, we have to give IS a mention, since most Olympus DSLRs have sensor IS, and not all FF lenses have IS.  So, for a static scene when the FF photographer is not using an IS lens or tripod, the 4/3 photographer may enjoy a substantial advantage.