Comparing noise and depth of field- full frame versus Micro 4/3
Disclaimer- The following information is not intended to bash Micro 4/3rds, but rather to foster an honest understanding of the effects of shooting a sensor 1/4 the size of full frame. I very happily shoot BOTH a Panasonic G9 and a Sony A7R III. These are both top tier cameras within their respective brands. Why would I ever opt for the G9 when I can shoot the Sony? Because there are many unique features to the Panasonic that makes it a superior choice in some scenarios. The amazing in body image stabilization on the Panasonic is easily a couple of stops superior to what is in the Sony. So when shooting static images the Panasonic can deliver great images at a low enough ISO to be competitive with the Sony. Additionally, In body focus stacking is awesome for Macro shots. The compact nature and light weight of the telephoto lenses is also a real bonus. If compared to a traditional DSLR then there are many other "mirrorless" advantages, however for my shooting scenarios both the Sony and the Panasonic are mirrorless and offer amazing electronic viewfinders, options for silent shooting etc.
But for anyone accustomed to Full Frame and contemplating Micro 4/3rds, there truly are a couple of differences that you should fully understand.
The first of these matters is to understand the difference in noise performance for current generation M43 cameras when compared to full frame. For these two sample images- both were shot on a tripod with the focus point manually selected to be on the "11" on the ruler. For the Micro 4/3rds I shot the Panasonic G9 set at 25mm on the Leica 12-60mm lens at F/4, with an import into Lightroom and then an export at 2048 pixels on the long edge. For the full frame image I shot the Sony A7R III with the Sony 24-105 lens set at 50 mm with an import into Lightroom and then an export at 2048 pixels on the long edge. No additional noise reduction was applied to any of the images.
The images can be viewed in detail at https://photos.app.goo.gl/fwMdaYQSMZyfyZpy7
By virtue of the laws of physics there are two things that happen when you comparably frame the same subject on a Micro 4/3rds camera as compared to a full frame camera:
(1) The fully frame sensor gathers 4x as much total light. This equates out to two stops, and as the physics would predict, we see about two stops worth of performance difference when comparing an image shot at 1600 ISO on Micro 4/3rds and one shot at ISO 1600 on a full frame camera. To be clear, the intensity of light arriving at each individual photosite is the same- so the exposures hold constant. But in total the full frame sensor gathers more light from the scene, and this is reflected in the relative noise performance when the images are then magnified to a comparable display size.
For my test I shot the Panasonic at ISO 1600 and the Sony at 6400. Both images were downsized to an image size of 2048 pixels on the long side. As you can see, the noise is nearly identical between the images. So if you KNOW you will frequently be shooting in low light venues that necessitate higher ISOs, then simply be aware that if your comfort level for noise on full frame is at say ISO 3200, then you will need to shoot the Micro 4/3rds at ISO 800 to have comparable noise levels.
Now is this a big deal? As my mentor Joe Edelman is fond of saying- digital cameras have passed the threshold of being "good enough" for quite some time now. If 90% of your images are being shared to Instagram or Facebook then you definitely should not worry about the difference in noise levels. By the time you downsize your images and apply a touch of noise reduction, no one is going to notice a difference. Or if you a studio shooter, and can always manage your light levels, then this difference in ISO performance simply does not matter. Micro 4/3rds at base ISO (200) delivers an extremely clean file. If however you frequently shoot in available light, and find yourself often shooting above 1600 ISO, then the difference in noise becomes noticeable when comparing the two platforms. If you can embrace a little grain, then no worries. Otherwise you are probably better off to leverage a full frame solution, particularly if you have a penchant for pixel peeping.
(2) The other fundamental difference to understand is how the Micro 4/3rds sensor impacts depth of field in your images. To be clear- (given roughly comparable pixel density) a lens of a given focal length and aperture will ALWAYS maintain the same depth of field, regardless of what sensor is sitting behind it. For example, a 35mm F/2 lens at 6 feet from your subject on a full frame camera will deliver a depth of field of 1' 3/4". Keep the lens position identical, but put a 4/3rds sensor behind the lens, and the depth of field does not change. What DOES change is your field of view.
So while technically it is correct to say the sensor does NOT affect the depth of field, from a practical standpoint it DOES, because you are forced to step further away from your subject to maintain a comparable framing. With Micro 4/3rds the CROP factor is 2X, so to maintain an identical framing (Say a 3/4 length portrait) you will have to stand about 12 feet away instead of six, and this of course changes your depth of field to a bit over four feet.
Is it bad to have a deeper field of view when using a comparable field of view? If you are shooting groups of people then it can actually be an advantage, since you can use an F stop that is two stops lower than on a full frame camera while still maintaining the same depth of field. A full frame camera might well need to be stopped down to F/5.6 of F/8 to get a small group shot in focus. Meanwhile the Micro 4/3rds user can shoot the same group at F4 and use either a lower ISO or higher shutter speed.
But what if you are a portrait shooter and your reason for being is shallow depth of field? Well simply remember that any lens you mount on Micro 4/3rds is going to yield two stops deeper depth of field than a full frame lens that delivers the same field of view. So lets say you are a member of the Fedora nation and see no reason to ever shoot your 85mm 1.4 lens at anything other than wide open. Lets imagine you are taking a shot from 8' away. Your depth of field is going to be 2 and 21/32 of an inch. If you shoot Olympus or Panasonic you will be shooting one of the F/1.2 45mm lens if you are a fan of shallow depth of field. The Olympus 45mm 1.2 shot at 1.2 from 8' away gives you a depth of field of 4 and 9/32" of an inch. So we are only talking about a difference of a couple of inches of depth of field between these competing portrait lenses for Full frame and Micro 4/3rds. If your subject is at anything less than a straight on view to your camera then the full frame camera will not be rendering both eyes in focus on the full frame shot at 1.4, but for the Micro 4/3rds shooter you have a bit more wiggle room even when shooting at 1.2. If you ever want to know what F stop on Micro 4/3rds is comparable to full frame, then simply multiply by 2. For example, when the field of view of each lens is identical, An F/1.2 lens on Micro 4/3rds will deliver the same depth of field as full frame lens at F/2.4.
So bottom line- you can definitely still have a shallow depth of field on Micro 4/3rds. You just need to invest in the fast F/1.2 glass if what you are accustomed to is full frame fast lenses shot wide open.
As always- Happy shooting!