Micro 4/3 Despite lust towards FF and DSLR dinosaurs, I'm just drifting towards m4/3...

In good light, mu43 can hold it’s own. Low light became an issue. The 61 mp Sony FF sensor has a 26 mp APS-C crop. Mu43 is about half the size of APS-C, so figure the 12 mp sensors have similar pixel sizes. It’s interesting that mu43 never tried a newer tech 12 mp sensor, which might have had much better low light behavior. Sony did that with the A7S series for 12 MP FF sensors. Mu43 could have had 12 and 20 mp models.
It wouldn't make that much of a difference at the image level.

Larger megapixel counts mean fewer photons per pixel, but once you downsample the larger mp image down to the smaller mp count - or display at the same absolute size - to make for a fair comparison, the only photons you've lost are the ones that bounce off the "walls" between the pixels. BSI technology (which current consumer m43 cameras don't have) and microlenses on the sensor can help further reduce the number of bouncing photons.

The reason video centric sensors lile the GH5S use lower pixel counts are 1) because they don't need any more given a video resolution standard (HD, 4K, 8K etc), and not having to downsample your sensor output saves processing power, heat etc; and 2) fewer pixels makes for faster readout and less rolling shutter with electronic shutter, which is important for video.

The Sony a7S really solidified the thought in peoples minds that the lower resolution allowed for much better low light results at image level (as opposed to "just" at pixel level). But that insane low light capability was just because it was Sony's first Dual Gain sensor. The high res a7R series also got Dual Gain sensors from the Mk II onwards, and perform about as well as their a7S contemporaries in low light - simply because the total number of captured photons is what matters, not the photons per pixel.
 
I should revisit Bryan Peterson composition books next.

His shots are all Nikon but he is all about "story telling apertures" around f/16. In short, adapt his gear and advice to fit the M4/3 framework perfectly.

The notion of a zoom and a super zoom at that, still intrigues me. I have shot some normal zooms enough to realize I'm really more of a prime guy but they still interest me. Perhaps a 28-300 superzoom is a different thing than a 24-120. And there's no better platform for zoomin' than M4/3.
 
GX80 has a pretty low headroom for highlights but Panasonic G9 just like Pen-F has extra reserves. G9 and Pen-F probably share the sensor chip but obviously things like signal processing and clipping are solely at the mercy of the camera's processor. Wonderful to see that G9 chooses to leave a hefty headroom for easy correction of exposure mistakes.

I set up Zebras for 105% and I took a sample shot of white mugs with plenty of positive EC so that they mostly bathed in zebra patterns. Still the raw file opened in Darktable shows zero clipping in highlights. G9's own live histogram also showed no clipping so it appears that the histogram is tracking raw data and not processed JPEG histogram? Nice.

Much like Pen-F shots, the raw data and the JPEG processing has a big discrepancy in exposure and for this I will probably set darktable to automatically apply an extra stop of exposure for each G9 photo automatically during opening. This will get me very close to what I saw in the viewfinder during taking the photo and I can always reduce the push without detriments if things look too hot.
 
Back
Top