Monday, August 9, 2021

PIXEL ENVY

Submitted by Richard Cooper Why do you need more pixels? Before you run out, and buy a camera with more pixels you should ask yourself, “What do I want to photograph?” and “What is my market?” I hope this blog helps you come to a decision.

Prior to digital photography and camera sensors, photographers used film. Selecting film was an easier choice. If you did not like the film, you purchased a different brand or ISO rating next time. Today, if you choose the wrong format, or pixel count, it is a more expensive proposition to replace a camera. It is important to select the right camera for the type of images you intend to photograph.

Cameras came in different formats. The most common were medium format and full frame/35mm. The ISO was dependant on the size of the silver halide grains. The bigger the grain, the more light it captured, and the film ISO rating was faster. For fine details and documents, there were film speeds of 16 ISO image with capture times that were slow. Therefore, very stable flat-bed cameras were used to photograph detailed documents.

For sport and fast moving subjects, a high speed film was used typically rated at 400 ISO. Faster specialty film was also available. The difference was easy to see. The grain was very fine for 16 ISO, but became larger as the film speed became faster. Large grain had a certain aesthetic, but due to the size of the grain, images were not as sharp as a fine grain film. If you draw a picture using dots with a sharp pencil or a bingo marker, you can see the difference in resolution.

Maybe this is why Henri Cartier-Bresson said “Sharpness is a bourgeois (middle class/ordinary peoples) concept”. Henri Cartier Bresson interview link https://youtu.be/5U63Pf7GS6A

In 1975, Steven Sasson, an engineer at Eastman Kodak, invented and built a self-contained electronic camera. The camera was about the size of a breadbox, and it took 23 seconds to capture a single image. Probably the equivalent to a late 19^th century plate camera in terms of size and speed of capture of an image. Steve Sasson: A Legacy of Innovation link https://www.invent.org/blog/inventors/Legacy-Steve-Sasson

In 1991, Kodak created the first-ever DSLR, the Kodak DCS-100. The camera had a built-in 1.3 megapixel, Kodak CCD to capture images. The camera cost $20,000 equivalent to $39,250 today.

In 1994, Kodak and Associated Press launched a digital SLR designed for photojournalists. The 1.3-megapixel camera had removable memory cards, and up to 1600 ISO. The Vancouver Sun became the world’s first newspaper to convert to all-digital photography. The advantage of digital was that the image could be transmitted in a single operation. Originally priced at $17,950, but discounted to $16,950 for AP members, it was a bargain!

The digital revolution had begun, but are more pixels better? As in the film days, size matters. Instead of grain on film, we talk about pixels on sensors. The larger the sensor, the more pixels it can accommodate, medium format verses 35mm. Typically the larger the pixel, the more sensitive it is to light.

Humans are especially good at recognizing edges, and patterns in what they see. Light captured by pixels is amplified, generating read noise, noise that creates unsharp edges to patterns of light, lines between light and dark, distortion that humans find distracting. Light captured by pixels on a 45 MP sensor requires more amplification than the same quantity of light captured by pixels on a 20MP sensor, thus creating more noise and distortion.

As an analogy. Imagine a sensor filled with pixels and the floor of a room filled with buckets­­­, pixels capture light and buckets capture water. As the pixels fill with light or the buckets fill with water, a pattern is created. A larger bucket can handle a higher volume of water without the water splashing out on the sides of the bucket and a larger pixel can handle more light generating less noise and sharper better defined edges.

As technology advances, light capture and noise factors improve. The question we need to ask is, “Do we want high resolution in bright conditions sacrificing noise in low light?”, Or “do we want less noise in low light, thus sacrificing resolution and colour saturation?” In short, small pixels or large pixels.

In the studio, light can be precisely controlled using strobes, and other artificial light sources so noise is not such a limiting factor compared to wildlife and landscape photography.

SENSOR SIZE AND MEGAPIXELS: As the megapixels increase, the size of the file increases which results in slower processing in camera, plus larger files to process in post processing software including additional storage costs. As the megapixels go up, the cost of the camera increases. APS-C sensors are smaller than full frame sensors, and therefore the MRRP of the camera is generally much less. Remember the size of the sensor means that a 20MP APS-C is smaller and has less light capturing ability, and therefore generates more noise compared to a FULL FRAME.

* FPS - Note: Mechanical FPS is generally less than electronic FPS

COST OF STORAGE: Even though the cost of storage to save your archive library has come down over the years, it costs about CA$0.24 per GB for archive storage. We think that images are almost free compared to film days. However, a 50MP file costs twice as much to store compared to the more common size of 25MP, CAD$0.06 verses CAD $0.12. At present, a conventional HDD is $21 per TB and an SSD is $86 per TB. It is estimated that SSD drive prices will fall and reach parity with conventional hard drives by 2026 at about $15 per TB.

MEMORY CARDS: Another consideration is the size of the memory card you use on your shoot. You can record about 1000 x 20MP full frame images on a 64GB card and less than half of that at 50MP. This may sound like plenty of space until you are taking photographs in extreme conditions, and are unable to download your images. Logically you will need twice the capacity or twice as many cards if you are shooting at 50MP.

PRINT SIZE: This is when the 50MP image comes of age. You can print a 50MP image to a size of 38x21 inches at 250DPI (798 sq inches) compared to 20x16 at 250DPI (320 sq inches). A massive two and a half times larger. However, how many prints do you enlarge greater than 16x20?

Useful Print Aspect Calculator. Link https://www.scantips.com/lights/resize2.html#aspect

What does this all mean? It depends on how you use your camera as a tool.

• Do you take photographs in low light, or good light, or studio light?
• How large do you want to print your images?
• Do you have sufficient memory cards for your assignment?
• Do you need to shoot a fast sequence of images?
• What is your subject?

I hope this helps you make a good decision when selecting your next camera.

Submission by Richard Cooper http://portlightstudio.com/ From a KODAK Brownie 127 to a NIKON Z7. Born in England and living in Canada, Richard Cooper has been involved in photography since his school days. He started developing and printing film with a KODAK Brownie 127. There was a quality of analogue film that is hard to reproduce from digital cameras due to the dynamic range of digital cameras. However, the flexibility and ability to take multiple images in seconds, and store them on a card the size of a single film slide has changed the way photographs are created. These images can now be manipulated in Capture One, Photoshop, and many other imaging editing programs. He now shoots with a silent mirrorless camera which does not disturb wildlife, or distract performers and athletes. It is now possible to take images without the subject knowing. Life for a photographer has changed.