November 26, 2022

Nikon’s Z9 and computational photo capability

Recently Nikon Announced Its new main camera, the Z9. This camera represents Nikon’s first mirrorless camera. A great specification list Designed for use in demanding fields such as photo journalism, sports, landscape, birds, and other utility cases that call the camera to shoot at incredible speeds in high resolution. The $ 5,500 Z9 is Nikon’s first camera, avoiding the traditional mechanical shutter, allowing speed and autofocus performance to reach new levels.

Speed ​​is great, especially for sports photographers. But it is interesting to think about where this technology could be used to capture traditional style cameras in the future. This may be the first step towards larger design cameras that adopt computational smartphones that smartphone cameras have been embracing for years.

The Z9 features a multi-angle 3.2-inch LCD that tilts in both directions but cannot face forward.
Image: Nikon

Nikon does not mention the rotating buffer that smartphones make for computational photography for HDR-style photos or for capturing up to nine or 10 frames at a time and attaching them at every pressure of the shutter button. But the new 45.7-megapixel full-frame rear-illuminated stacked CMOS sensor has not been far from phones for years, at least in terms of core design. This type of construction uses a sandwich configuration of sensor, logic board and dedicated RAM – providing incredibly fast reading speeds.

Today, the Z9 allows you to use a full-time electronic shutter with a shutter speed of 1/32,000 in a second and achieve incredibly fast burst shooting. It can capture up to 20 frames per second in RAW / JPG at full resolution or at 11 megapixels at 120fps without producing audible sounds (dummy shutter sounds can be turned on for audible targeting if desired). The new Xpeed 7 processor and dual CFexpress / XQD card slots provide the Z9 with 1,000-shot buffer at full resolution in high-performance compressed RAW, but the stacked sensor’s fast reading speed is key to the calculation. Photography Puzzle.

Being the first of the major camera manufacturers to push the mechanical shutter, Nikon puts itself ahead of its competitors in the race towards computational photography. Sony’s A1 And A9 The lines already use stacked sensors for faster reading speeds, enabling electronic shutters for full-time duty, and Canon’s upcoming R3 Will use the same technology. Switching to a full electronic shutter is the logical next evolution of cameras, although the responsibility of proving its electronic shutter now depends on the day-to-day work and demands of pro photographers.

The Z9’s stacked CMOS sensor is fast enough to shoot without blackouts when writing images to the card.
GIF: Nikon

To date, efforts to implement computational photography from camera manufacturers have been limited to features such as Olympus. Direct ND And Panasonic Post focus And In-camera focus stocking. Simple features, yes, but these are side views compared to the paradigm shift, which is the full computational photo that will one day be activated at every pressure of the shutter. OM System, the newly renamed Olympus, Recently promised Its next camera to use computational photography technology, but we have to see if it focuses on the main or another aspect on the side.

The object detection of the new Z9 was used before some capabilities by Deep Learning, Olympus, Panasonic and Canon used in the autofocus system. This helps to improve autofocus tracking performance, but in the end, a glassless camera captures an image defined by the sensor’s dynamic range.

The primary barrier to blocking cameras such as the Z9 may be the data capacity and image processing pipeline of other pro- or interest-level glassless cameras with fully computable stacked sensors. Ten frames taken simultaneously from a 45-megapixel full-frame sensor and combined into one file can be exponentially larger than a set of images of the same size taken from a smartphone sensor.

Additionally, the rotation buffer is essential for continuing to write and rewrite images to the camera’s buffer in the background before you press the shutter. Even the new processor of the Z9 may not do these tasks. In place of the smartphone, CPUs are designed to be more suited to this process, Sometimes uses specialized hardware, But the cameras are not made identical. More innovations at the CPU level are still needed from camera manufacturers.

Processing power may be the last technical barrier to computational photography on full-size cameras.
Image: Nikon

There are some obvious advantages to using computational photography. Most any modern smartphone can create symmetrical expression, they glow well, visible details and clouds all over the shadows – all in one frame. Improvements like Night Sight and Night Mots allow you to do the hardest things to achieve with a standard camera, while Google continues New computational tricks to keep subjects sharp in motion, And Apple allows too RAW files with computational data.

On the other hand, photography taken even with the most advanced mirrorless camera today – elevated in sharpness and clarity – requires some sacrifice. Achieving the same look of most smartphones requires at least a bit of processing and editing, which must be exported from a RAW file to JPG or some other global format. The computational photo coming to exclusive camera systems could re-energize the camera market, although camera makers are finally scaring away connected Wi-Fi apps – admittedly, they can find another tall line.

Cameras like the Z9 can bridge that path, and can be appreciated even by professional photographers who can edit in less time to achieve the look many of their clients are looking for. It’s “What is a movie?” While further blurring the lines, full-size cameras can be a little exciting again.