Computing Life

中文版

I was recently investigating whether I should invest in a full frame or medium format camera. And here are some thoughts to share.

Full frame sensor is the sensor that has the same size as a 35mm regular film. And medium format sensors refer to all the sensors that have a larger size, but smaller than a big format (4’’x6’’) sensor (which is always the case for consumer cameras). And the common sensor size for popular armature level DSLRs are APS-C size, which is about half the size of a full frame sensor. The definition of the medium format itself shows it’s not a very rigid concept, and there is indeed huge size difference among medium format sensors. For example, the medium format sensor I’m interested in is 68% larger than a full frame sensor.

Then the next question is, why medium format? What advantages can it bring us? There are mainly four advantages.

• First, larger sensor naturally bring more pixels, which are essential for bigger prints. So for commercial prints that are several-floor high, a large sensor such as medium format sensor will have to be used.
• Second, medium format sensors usually also have bigger pixels than the full frames. The pixel size is important in the Signal Noise Ratio (i.e. how much sensor noise). And bigger pixels will bring less noise.
• Third, the depth of field is correlated with sensor size. So it’s easier for larger sensors to get shallower depth of field.
• Fourth, diffraction is an enemy to the image quality. And its impact also has negative correlation on the pixel size. Usually we say the impact of diffraction is observable when the size of the Airy Disc is larger than a pixel. And it’s obviously harder with larger pixels. So the impact of diffraction is less pronounced.

Traditionally, medium format cameras are usually the choice for professional studios for several reasons. First, commercial photography sometimes asks for very large prints, and has little tolerance on the imperfection on the photo quality. This makes large sensors a hard requirement. At the same time, they also don’t need portability in the studio, this also makes the major drawback of medium format sensors less important here. Third, commercial photography has relatively more time in adjusting the parameters and focusing, and thus more tolerant to the slower speed of medium format system (the processing chip needs to process much more information).

But recently, Hasselblad, which has been famous for its medium format camera, released a new compact medium format camera, x1d. Its unique advantage is to make the camera (not the sensor) to have the same size as full frame cameras, and make “regular people” be able to carry medium format cameras around. I’m very curious about the product, and was wondering whether I can utilize the advantages of medium format sensors in the real life. So I rent a camera body and three native lenses, to get a feeling of what it’s like to use a medium format camera.

Unfortunately, the conclusion so far is, I don’t see any perceivable advantages.

One key factor is, although a big sensor will bring better image quality, the lens and the glasses inside also become larger, which also makes the optical design more sophisticated. The direct result of this design difficulty is, the aperture is usually smaller than the full frame lens. For example, for the three prime lenses of x1d, the largest f number is only 3.2. This especially causes a problem in dark environment. For example, comparing with a 50mm f/1.4 full frame lens, the light it collects is (3.2 / 1.4) ^ 2 = 5.2 times of the f/3.2 medium format lens. Then for the same focal length, say, the full frame camera needs ISO 400 to make a safe shutter speed, x1d needs an ISO of 2000 to maintain the same shutter speed. This is OK for studio users because the surrounding is usually very bright, and x1d’s image quality if superb in such environments. But my observation is, it hasn’t reached the stage that you can easily tell it’s from a medium format sensor (like you will for DSLRs vs iPhones).

Another drawback is, because the sensor has so many pixels, the processor chip of the camera system has a hard time to crunch all the data. This is more a problem of x1d rather than general medium format systems. For example, x1d needs about 8 seconds to boot. And you need to wait about half a second between pressing the button and hearing the unique sound from the leaf shutter. After the shutter releases, you cannot do anything in about 3 seconds, during which period even the EVF is totally blacked out. To make things even worse, x1d becomes quite warm after a while, even when it’s not used for shooting, but just standing by there. I heard Fujifilm GSX S50 doesn’t have problems like this. Hope it’s the case.

Back to the image quality. One advantage of x1d against full frame cameras is it has 16 bit color depth. So theoretically it has richer color. But someone online plot a histogram of the pixel values from the x1d’s raw file, and found 2 out the 16 bits are pure noise, so essentially it’s still a 14-bit sensor. I compared the image quality between Hasselblad x1d and Leica SL, and didn’t find much perceivable differences, except that x1d has much more pixels and can crop to a much smaller area.

X1d or typical medium format cameras have a selling point is leaf shutter. It brings very short flash sync time, to about 1/2000 seconds. So there will be no need to use various tricks to sync the flash and the (focal plane) shutter. But for an armature user like me, this doesn’t matter too much.

In summary, I don’t find it especially attractive to buy a slower, and pricier compact medium format camera. This thing is definitely useful and has its market. But currently it hasn’t reached a mature stage in terms of technology and user experience. X1d has a lot of small problems, possibly GFX S50 is better, but it’s also considerably bigger than x1d. Currently I still haven’t found a nice candidate in this niche product line. For non-professional (or non-studio) users, full frame cameras are a much more reasonable choice.