Do More Megapixels Mean More Noise? Nikon Z6 II vs Z8 in Long Exposures

Introduction

Do lower-resolution cameras really produce less noise? Does lower pixel density give a sensor an advantage during long exposures? These claims appear regularly in photography discussions, usually in some variation of: “fewer megapixels are better in low light.”

That belief influenced my decision when I bought my first full-frame camera, the 24-megapixel Nikon Z6 II. Its low-light reputation and price both contributed to my choice. After two years of use, it had proved reliable and well suited to my photography.

Some time ago, while experimenting with night seascape photography with Piotr, I noticed that image quality deteriorated as we made a series of long exposures in relatively warm conditions. That experience made me curious about how sensor resolution, exposure time, and heat affect modern cameras. It was only an observation from one session, but it gave me a reason to investigate further.

Recently, I added a Nikon Z8 to my kit. It is a newer model with much higher resolution and considerably greater pixel density. Having both cameras gave me an opportunity to put those familiar internet claims to a practical test.

This is not a laboratory experiment. It is a real-world comparison intended to answer a photographer’s questions, not produce a universal verdict about sensor design. Make yourself comfortable and enjoy the test.

Disclosure: I purchased all equipment used in this article with my own money. I have no commercial relationship with Nikon, Kase, or K&F Concept. No manufacturer reviewed or approved this article, and all opinions are my own.

Why long exposures?

Long-exposure photography is one of my favourite techniques. I first experimented with it 17 years ago, when affordable neutral-density filters were difficult to find, digital cameras offered around 10 megapixels, and ISO 400 could already produce unacceptable noise. To achieve exposures of up to 10 minutes, I used several variations of green welding glass as improvised filters.

Fortunately, those days are over. We now have access to a wide range of ND, variable ND, and CPL + ND combined filters. That makes long-exposure photography much easier and gives me more equipment to experiment with.

Are Lower-Resolution Sensors Really Better?

I currently own two full-frame cameras:

Nikon Z6 II (24.5 MP)
Nikon Z8 (45.7 MP)

Owning both cameras created a good opportunity for a practical comparison. I wanted to see how they behaved at low ISO as exposure time increased, and then at higher ISO with exposures of around two to three minutes.

The cameras differ in age, resolution, and intended market position. The interesting question was whether the Z8’s much higher pixel density would produce a visible difference in this test.

I focused on three questions:

Does either camera produce visibly more noise during long exposures?
Does the higher-resolution Z8 show more hot pixels?
How usable are the files after raising ISO and applying modern denoising?

What Are We Actually Comparing?

At the same sensor size, a lower-resolution camera has larger individual photosites. All else being equal, each photosite receives more light, so a 100% crop may show less noise per pixel. That is the origin of the claim that fewer megapixels are better in low light.

But photographs are viewed and printed as complete images, not as isolated pixels. The Z8 records many more samples of the same scene, and some of its per-pixel noise is averaged out when the image is reduced to the Z6 II’s dimensions. That is why this article considers both 100% crops and images at the same output size. Sensor generation, readout design, processing, and quantum efficiency also matter, so pixel size alone cannot predict the result.

Test setup and limitations

Long exposures are sensitive to changing light, so it can be difficult to repeat a test under exactly the same conditions. I therefore photographed the same scene with both cameras at the same time.

An ideal comparison would use identical lenses and filters, but I did not have two matching sets. Instead, I used similar focal lengths and adjusted the lens-and-filter combinations to obtain comparable exposure times.

For shorter exposures, I used 10-stop ND1000 filters from K&F Concept and Kase. For longer daylight exposures, I used different combinations: an ND1000 stacked with a Kase ND64+CPL on one camera, and a Kase ND100000 on the other. Aperture was adjusted where necessary to equalise exposure.

This matters when interpreting the results. Different lenses, filters, apertures, individual sensors, and sensor temperatures can all affect the image. The test compares my Nikon Z6 II with my Nikon Z8; it does not isolate pixel density as the only variable.

Each camera retained the same body-and-lens combination throughout the test:

Nikon Z6 II + NIKKOR Z 24–120mm f/4 S at 31.5mm
Nikon Z8 + NIKKOR Z 14–30mm f/4 S at 30mm

Long-exposure noise reduction was disabled, and all photographs were recorded as RAW files.

One limitation is especially important when comparing detail: apertures ranged from f/4 to f/22. Diffraction at the smaller apertures can reduce sharpness independently of sensor noise, particularly on the higher-resolution Z8. These photographs can therefore demonstrate practical image quality, but they cannot provide a controlled comparison of fine detail.

Setup of cameras for a field test — That impresive set up was used to perform the actions

Test 1: ISO 100

ISO 100 is the base ISO of the Nikon Z6 II and slightly above the Z8’s base ISO of 64.

The images below are full-frame comparisons resized to a maximum width of 1600 pixels. This represents a normal web-sized output rather than a one-to-one pixel-level comparison.

The changing daylight produced some variation in brightness between frames. Nevertheless, at a maximum width of 1600 pixels, I could not see any meaningful noise problems. Differences became visible only when I examined the files more closely.

In each image pair below, the Z6 II appears first and the Z8 second.

Exposure settings

ISO	Exposure time	Nikon Z6 II	Nikon Z8
100	1 s	f/4; K&F Concept X-Pro ND1000	f/4; Kase Revolution ND1000
100	8 s	f/11; K&F Concept X-Pro ND1000	f/11; Kase Revolution ND1000
100	15 s	f/16; K&F Concept X-Pro ND1000	f/16; Kase Revolution ND1000
100	30 s	f/22; K&F Concept X-Pro ND1000	f/22; Kase Revolution ND1000
100	120 s	f/4; Kase Revolution ND1000 + ND64+CPL	f/5.6; Kase Revolution ND100000
100	240 s	f/5; Kase Revolution ND1000 + ND64+CPL	f/8; Kase Revolution ND100000
100	720 s	f/8; Kase Revolution ND1000 + ND64+CPL	f/13; Kase Revolution ND100000

ISO 100 / 1 s

ISO 100 / 8 s

ISO 100 / 15 s

ISO 100 / 30 s

ISO 100 / 120 s

ISO 100 / 240 s

ISO 100 / 720 s

Noise and hot pixels

At ISO 100, the noise levels from both cameras were excellent. Some noise was present, but I could not see meaningful degradation from either luminance or colour noise. Both cameras produced completely usable images without software noise reduction at normal viewing sizes.

Hot pixels were more noticeable as exposure time increased. On the Z6 II, I first observed them at 60 seconds. Their number then increased with exposure time, reaching roughly 20 in the 720-second frame. On the Z8, I found one as early as the one-second exposure, but counted only around eight in the 720-second image.

In this particular test, I did not find more visible hot pixels in the higher-resolution Z8. In fact, I counted more in my Z6 II during the longest exposure. This does not prove that pixel density is irrelevant, but it does show that megapixel count alone cannot predict hot-pixel behaviour. The individual sensor, temperature, exposure duration, processing, and the threshold used to identify a hot pixel can all influence the result.

Hot pixels are not specific to either camera in this test, they have existed since the early days of digital photography.

Long-exposure noise reduction can reduce hot pixels by recording a second, dark frame and subtracting fixed-pattern noise. The disadvantage is that the camera needs another exposure of approximately the same duration. After making a 15-minute photograph, I rarely want to wait another 15 minutes before I can use the camera again.

That does not make the feature useless. It can be valuable when image quality matters more than shooting speed, and dark-frame techniques are particularly relevant in astrophotography—unless you enjoy discovering “new stars” in your photographs.

A small constellation of hot pixels near the left edge of the image

What If the ISO Goes Up?

Raising ISO during a long exposure may sound contradictory. If the goal is to extend the exposure, surely ISO should remain as low as possible?

Ideally, yes—but that assumes you have the perfect filter in your bag. In reality, the filter that best suits the scene is often safely stored back in the studio while you are carrying a slightly imperfect alternative.

Using a filter that is too dense may require opening the aperture or raising ISO. The shutter speed is usually a creative choice in long-exposure photography, so shortening it substantially may destroy the effect you wanted.

I began by comparing the photographs at an online output size of 1600 pixels.

In each pair, the Z6 II appears first and the Z8 second.

Exposure settings

ISO	Exposure time	Nikon Z6 II	Nikon Z8
800	180 s	f/5.6; Kase Revolution ND1000 + ND64+CPL	f/9; Kase Revolution ND100000
3200	120 s	f/9; Kase Revolution ND1000 + ND64+CPL	f/14; Kase Revolution ND100000
6400	120 s	f/11; Kase Revolution ND1000 + ND64+CPL	f/18; Kase Revolution ND100000

ISO 800 / 180 s

ISO 3200 / 120 s

ISO 6400 / 120 s

Different Noise Characteristics

At ISO 3200 and 6400, especially with exposure times approaching 120 seconds, noise was clearly visible. In these samples, the Z6 II showed more visible colour noise, while the Z8 sample appeared to contain mostly luminance noise and looked softer at first glance.

That apparent difference in detail should be treated cautiously. The ISO 3200 and ISO 6400 Z8 frames were taken at f/14 and f/18, where diffraction can visibly soften an image. Noise character, lens performance, aperture, and processing are all mixed together in these samples.

At ISO 800, both cameras performed very well.

ISO 800 / 100% crops

ISO 3200 / 100% crops

ISO 6400 / 100% crops

What Happens After Denoising?

I processed every image in the same way using Lightroom Denoise at its default Amount setting of 50, with no other adjustments.

After denoising, I consider ISO 6400 usable from both cameras. In these examples, I cannot see a meaningful difference in final image quality once both files are processed and compared at the same output size. The degradation relative to the low-ISO photographs is still visible, however. The practical question is: how large will the image be displayed or printed, and how closely will it be viewed?

The following images are 100% crops from the denoised files. Remember that the full Z6 II image is 24.5 MP, while the Z8 image is 45.7 MP. A 100% crop from the Z8 therefore represents greater enlargement and is not a direct representation of how equally sized prints will look. For a fair output comparison, both files should also be resized to the same dimensions.

ISO 6400 / 120 s / denoised 100% crops

Nikon Z 6 II / ISO6400 / CROP / Denoised

A Real-World Edit

Finally, I gave both photographs a quick editorial-style treatment: cropping, white-balance and exposure corrections, tonal and contrast adjustments, denoising, and sharpening. Judge for yourself whether the resulting quality is sufficient for your intended use.

ISO 6400 / 120 s / final edits

Stray Light in the Stacked Filter Setup

Noise was not the only issue revealed by this test. In the longest Z6 II exposures, a magenta cast appeared in the upper-left corner of the frame. The sun was behind the camera, and the cast appeared when I used the stacked square-filter setup.

The most likely explanation is stray light entering around the filter holder or reflecting between the filters. A multi-minute exposure can record light that would be insignificant at a normal shutter speed. Stacking filters also adds more glass-to-air surfaces, increasing the possibility of internal reflections, flare, and uneven colour.

This does not show that square filters are inherently unsuitable for long exposures. Holder design, sealing, filter placement, and lighting conditions can all affect the result, so the complete setup should be checked carefully:

Make sure the filter is seated correctly and its sealing gasket faces the holder.
Shield the holder from direct sunlight, even when the sun is behind or beside the camera.
Cover the viewfinder if the camera design allows light to enter through it.
Avoid stacking filters unless the combination is necessary.
Make a shorter test exposure and inspect the frame before committing to several minutes.

When possible, I prefer one strong filter instead of several stacked filters. Using fewer optical surfaces can reduce possible paths for stray light and internal reflections. Square systems remain useful, particularly for graduated filters and positional adjustments, but sealing and shading deserve attention during extreme long exposures.

Most importantly, the magenta cast most likely came from the optical setup rather than sensor noise or pixel density. Without recognising that distinction, it would be easy to blame the camera for a problem that probably originated in front of it.

Choosing a Filter System

After several years of using neutral-density filters, I have learned that there is no universally best format. The right choice depends on what you photograph and how you prefer to work. Colour neutrality, coating quality, density accuracy, and sealing can vary between individual products, regardless of format.

Square and Rectangular Filters

Square systems offer the greatest flexibility. Filters can be stacked easily, and graduated ND filters can be moved up or down to match the position of the horizon. This makes them particularly useful for landscape photography.

The tradeoffs include size, the need for a dedicated holder, and greater exposure to wind. A large filter can vibrate during an exposure, while gaps around the holder may allow stray light to enter. These factors become more relevant during multi-minute exposures.

Screw-In Round Filters

Screw-in filters are often more compact than square systems and are available across a wide range of prices and optical specifications. Once attached, they form a direct connection with the lens that can help limit paths for stray light.

Changing them can take longer, and each filter must match the lens diameter or be used with step-up rings. Thick filter frames and stacked filters can also cause vignetting on ultra-wide lenses. Round graduated filters offer less control over the position of the transition line relative to the horizon.

Magnetic Round Filters

Magnetic filters share most of the advantages of screw-in filters but are much faster to attach, remove, and stack. They connect to a magnetic adapter ring that remains screwed onto the lens.

Compatibility varies between magnetic systems, so filters and adapters from different product lines may not work together. You may also need adapter rings for different lens diameters, and some standard lens hoods may not fit while the system is attached. Stacking several magnetic filters can still cause vignetting at very wide focal lengths.

My Choice

I prefer magnetic round filters because speed and convenience matter most to me in the field. I bought adapter rings for each of my lenses so that one filter set can be used across the entire kit. The compromise is that I usually cannot use the original lens hoods while the adapters and filters are fitted.

That preference is personal rather than universal. I would choose a square system if I regularly needed graduated filters, a screw-in system for a compact and affordable setup, and a magnetic system when fast filter changes were the priority.

Conclusions

This test did not reveal a meaningful long-exposure noise disadvantage for the higher-resolution Z8 at equal output sizes. It also did not show that a denser sensor necessarily produces more hot pixels: I counted fewer in the longest Z8 exposure than in the equivalent Z6 II frame.

Those findings apply only to these two camera bodies and this shooting session. Because the lenses, filters, apertures, and base ISO settings were not identical—and because sensor temperature was not measured—the results should not be treated as a controlled sensor benchmark.

The practical lessons are still useful:

At ISO 100, both cameras produced clean files even after 12-minute exposures.
Hot pixels increased with exposure time, but their number did not follow sensor resolution in any simple way.
ISO 800 remained very usable on both cameras.
ISO 3200 and 6400 required more processing, but modern denoising produced usable results at normal output sizes.
A 100% crop is not a fair final-output comparison between cameras with very different resolutions.
Aperture and diffraction can affect apparent detail as much as noise, so they must be considered when interpreting these samples.
During extreme long exposures, filter sealing and stray-light control can matter more than the difference between the two sensors.

My honest conclusion is that megapixel count should not decide which of these cameras you choose for long-exposure photography. The Z8’s higher resolution did not create the dramatic noise penalty that internet wisdom led me to expect. Camera features, handling, lenses, filters, and the final use of the image matter more.