alexisgreat wrote:
After some more testing with the camera, I've found some interesting quirks, some of whom I've also found in reviews for previous cameras.
1) Using spot metering (not sure about other metering modes) in Manual exposure mode and AUTO ISO, the camera selects a -0.3 EV lower exposure than in Shutter or Aperture Priority Mode. This manifests itself in a slightly lower ISO being selected (for example ISO 4000 vs ISO 5000 or ISO 200 vs ISO 250.)
2) When saving MySets for some reason when I save Shutter Priority mode MySet it always saves it as a +1.3 EC regardless of what the current Shutter Priority setting is and always reverts to the +1.3 EC when I turn the camera on in that mode. So I've decided to save SP MySet in Movie mode on the mode dial (since I can access movies in any mode by pressing the red button) and have left S without a MySet but keep that at +0.3 EC. I've also programmed a MySet into Manual exposure mode so that it reverts back to AUTO ISO when I turn the camera back on even if I program ISO into one of the function buttons and use it to manually select the ISO and leave it there when I turn the camera off.
3) This might be a more serious issue than the above. When zooming the kit lens I noticed it doesn't take into account the changing fnumber early on in the zoom range. For example, at 14mm the maximum fstop is f/3.5 and at 22mm it's f/4.1 But when setting AUTO ISO in both Shutter Priority and Manual exposure mode at both focal lengths and maximum fstop, it selects the same ISO even though the fstop is different! It does adjust the ISO when you move further up the zoom range (for example when the maximum fstop if f/5.6) but not earlier on. More experimentation is necessary. I've only used spot metering and there is still the issue of Manual exposure mode always selecting a 1/3 stop less ISO than the Priority modes.
4) Another possibly serious issue which needs to be cleared up is that does increasing the ISO only boost the light that the camera has already captured or does it also do what increasing the shutter speed does, also reveal light sources that a lower ISO or shutter speed wont? For example assuming the same aperture (f/3.5) my old E-520 at 15 second exposure and ISO 1600 showed way more stars than my E-PL6 at 2 second exposure and ISO 12,800. Now the E-PL6 image was just as bright (if anything perhaps brighter), and the bright stars were just as bright or brighter, but the dim stars the E-520 captured were nowhere to be found in the E-PL6 image, even though the exposure was the same. Again spot metering was used on both cameras and both Shutter Priority and Manual exposure modes (with the E-PL6 once again defaulting to 1/3 stop lower ISO in the latter mode- so I had to manually select the ISO in that mode to match SP mode- which brings me to another point which could explain what's going on here).
5) When manually selecting the ISO in Manual exposure mode on the camera with ISO programmed into one of the function keys, I saw that anything above ISO 5000 is labeled as "extended mode." What does this mean? Could it be the reason that I can detect fainter stars with the E-520 with a lower ISO despite equivalent exposures? Do other Olympus mirrorless 4/3 cameras also display "extended mode" when using ISO 6400 or higher? I found this review for the previous generation model (E-PL5) which agreed with my findings:
http://www.thephoblographer.com/2012/12/03/review-olympus-epl5/#.VsHl-yASymw
At the moment of publishing this piece, there is some confusion as to what the real highest ISO setting should be. What do I mean by that? Well, I received what is called a Champion sample of the camera. These are some of the first off of the production line. The camera’s highest native ISO setting should be 6400. However, mine and my Olympus rep’s are actually 5000 until it starts labelling 6400 as the extended range. With that in mind, the highest ISO setting I shot at with this camera was 3200.
More about this "extended ISO" issue on this page- it seems to be what I thought about not showing more light, rather just boosting the signal that is already present. My question is do other Olympus cameras also go into "extended" mode starting at ISO 6400 and is it better to stick to ISO 5000 or less or is there some reason to only use full stop ISO and not go higher than 3200 like the other author suggested?
http://www.stevehuffphoto.com/2015/02/23/mirrorless-battle-micro-43-vs-aps-c-vs-full-frame/
The ISO with digital sensors is electronically or digitally made. Where with film it was just purely about silver halide crystals dispersed in gelatin and the size of those crystals was main element in the film light sensitivity, others were the chemicals, emulsion and finally temperature etc. All kind chemical reactions. You couldn’t know what kind details you really got because those crystal sizes were random, their placement was random and many other random features when it came to photons and chemicals in different environmental situations (time, temperature, colors etc).
With digital sensors we have always the exact pixel layout, exact pixel size etc.
Basically it goes that sensor is in dark, as empty in DSLR or collecting photons all the time but just passing trough when comes to Live View (mirrorless etc).
The electrons the pixel collects, is passed to ADC, that is “Analog-Digital Conversion” processor. A digitalizer component that reads the electronic current and translates it to digital 1’s and 0’s, based wanted mathematical algorithm.
If we want to raise the sensitivity to collect more light, meaning we want to raise ISO, we do not change the pixel size like we did with film. We do not even change the depth (pixels are like silos). Instead we read the electronic current that pixel is feeding when flushing, and we add more electrons with amplifier, drawn from camera battery where those electrons are stored.
Now we are electronically amplifying the signal, changing the SNR. We mathematically change the ratio in the SNR, Widening its range, hence the name “Dynamic Range” because it is dynamic by math with electronics. It is just values we manipulate. There is no “Dynamic Range” value that can be compared between sensors, as most those sensors are using totally different designs and other electronic elements that has nothing to do with sensors itself. We can use exactly same sensor and different electronic components and algorithms and get totally different results.
A dynamic range is just a mathematical scale that is… dynamic. Not comparable to anything else than itself.
So we feed more electrons to electrons the pixel flush to ADC, and now the ADC needs to do the conversion. It needs to detect and identify every sub-pixel color current and combine it with the other sub-pixels to get the correct digital value.
Now comes the reason why we have two kinds of ISO settings. We have so called “native ISO” and then so called “extension ISO”. The difference between these to is that “native ISO” is this electronical boosting, adding electrons to signal before processing, widening the sensor dynamic range, a one actually collected photon transforms to multiple electrons. Multiple collected photons becomes folded amount of electrons. And now we have good estimation of values in much wider dynamic range. But the sensor SNR has not changed a bit, while dynamic range has!
Now we hit to limits in electronics some point. We can’t feed more electrons to widen the dynamic range because the SNR is already so small. We get problem where we can’t just anymore resolve the correct information from the fed information.
So instead trying to make things complex in electronics, we use digital technology.
Now we use “extended ISO”. We take a value given by electronics, convert those to digital value and then we take that value and we mathematically multiply those digital values to something else they are.
Extended ISO is like pressing the film. Like on film era we could press film, we had just ASA 200 (ISO 200) film but we needed ASA 800 or even 1600 but those were expensive ones or we didn’t get those. So, we loaded the ASA 200 film in camera, we rotate the camera ASA dial to ASA 800 or 1600 like we would have loaded ASA 800 or 1600 film. Then we went and we photographed with the exposure values the camera lightmeter guided us.
Then when developing the film (or sending the film to be developed), we simply developed the film like it was ASA 800 or 1600. The chemical process did the trick. We really didn’t get more information, but we made the available information more visible. We had usable film, with often colors little off etc. But we had usable film. We had usable photo as well, because we could use faster shutter speed to freeze motion or use smaller aperture to get deeper depth of field.
Looks like I have my answer, even the EM-1's ISO 6400 is "extended"
http://webcache.googleusercontent.com/search?q=cache:q0ocFs-ga24J:www.cnet.com/products/olympus-om-d-e-m1-body-only/+&cd=10&hl=en&ct=clnk&gl=us
The E-M1 is just awesome camera. JPEG is best what is out there, even that the article claims otherwise. It is small, almost perfect controls what just "melts" to your hand so you don't even think camera anymore at all as you just shoot. ISO 3200 is awesome (and you can even use extended 6400 with JPEG if wanted) and available lenses are just best ones and you have wider range than what you can ever get to any FF camera.
