Fujifilm X-T5 Sensor High ISO Performance (vs X-T4)

What’s the difference in the new X-T5 sensor?

I was taken aback when the X-H2 was announced to have a 40.2MP sensor. That didn’t bode well for the X-T5 (for me, at least) because it signaled the X-T5 would house the same sensor.

For one, I thought the megapixel wars were over. Second, I was perfectly content with 26.1MP, just like everyone else was. Third, 40 megapixels crammed into a crop sensor meant its low-light performance would be degraded. Or would it?

A number of photographers proclaimed they would never use the X-T5 because its high pixel count would mean degraded high ISO performance. Proclamations before the X-T5 was even released.

Both the X-T4’s X-Trans CMOS 4 sensor and the X-T5’s X-Trans CMOS 5 HR sensor are back-illuminated. We’re not going to build a sensor here; this just means that the wiring arrangement around the photodiodes allows them to capture more light.

But two and a half years have passed between the two sensors. Are we to assume that technology hasn’t changed? They’re just reducing the pixel size, and that’s it?

Sensors become more efficient with every generation. The base ISO in first-generation X-Trans sensors was 200. Then 160. Now it’s 125 in the X-T5. The pixel structure in this sensor has been improved over the X-T4, and that does help efficiency. Enough to offset the smaller pixels? We’ll see.

In addition to a sensor that gathers light more efficiently, the X-T5 also has a new processor. The fifth-gen X-Processor 5 features a number of improvements over the X-Processor 4 found in the X-T4.

Interactive camera tutorials:

• X-T4 Tutorial Course
• X-T5 Tutorial Course

Fujifilm X-T4 & X-T5 High ISO Noise Comparison

Watch this video on YouTube.

How does the X-T5 sensor perform in low light compared to the X-T4 sensor?

Comparison method

Duplicating conditions is important here, so we’re going to do it in a controlled environment, the good ol’ still life of things found in the house method.

First, we will compare unprocessed RAW files of each camera from ISO 800 to 12800. Each comparison is shown at 100% zoom; since the X-T5 has a higher resolution, it will appear larger.

Lens, focal length, focus distance, Image Quality Setting, and exposure variables are all the same.

Next, we’re going to look at some additional scenes at ISO 12800.

And finally, we’re going to look at some in-camera JPEGs using a HIGH ISO NR (Noise Reduction) setting of -4, which is the minimum amount of noise reduction the camera applies to JPEGs.

You will see some variations in color cast in some of the photos. The X-T5 seems to have a touch more magenta than the X-T4. This is a phenomenon I’ve noticed between all generations of X-Trans sensors. It is irrelevant to these tests, but I just wanted to point that out.

Unprocessed RAW photos

At ISO800 and 100%, I can’t see any noise differences.

Going up to ISO3200, if I strain and search for it, I can see some ever-so-slight increases in color noise in the X-T5 (bottom). That was really searching for it in Adobe RAW, but here on the web it’s nearly impossible to see.

Finally, at ISO6400, we can see some increases in both color and luma noise in the X-T5 (bottom) compared to the X-T4 (top). But it is still something that you really need to look for, apparent in the Polaroid lens and the silver frame next to it, both in the shadows and midtones.

And at everyone’s favorite ISO to be at, 12800, we can now see some more apparent increases in the X-T5 photo (bottom) in the same areas, especially with color noise.

Moving to a new subject, with the X-T4 photos on the left and the X-T5 photos on the right. ISO 800 are nearly identical from a noise perspective.

Additional RAW comparisons

In-camera JPEGs

These photos are straight out of the camera using the HIGH ISO NR setting of -4, the lowest amount of in-camera noise reduction, at ISO12800. It’s pretty darn good!

The most important findings

What’s the most important thing to take away from this comparison?

In the words of Bill Murray’s character in Meatballs, it just doesn’t matter.

First, you can’t really make a pixel-by-pixel comparison here because of the resolution differences. If you view the X-T4 image at 100% next to the X-T5 image at 66% so they “look the same,” the X-T5 image has been resampled by the software. They look pretty darn close as far as noise is concerned.

Second, most differences aren’t really noticeable until ISO 12800. I’m not one to fear 12800, I go there when needed to get the photo, but this is an ISO value that I’m willing to say hardly any of us hang out at for extended periods. Does the noise really look that bad?

But more important, don’t get so wrapped up in this analysis that you can’t see the forest for the trees, as the saying goes.

If you create images that evoke emotion, no one will see or care about “noise.”

Unless you’re looking at the two images side-by-side at 100% and searching for differences in noise levels, no one will notice. Hell, I can barely see differences when I am searching for them. How many other people besides you will be doing that? Exactly zero.

If you don’t need or want 40 megapixels, or you’re always at ISO12800, maybe the X-T5 isn’t for you. If it has features that help you create photos, and you hardly ever go above ISO3200 or 6400, maybe you should consider it.

Just go do photography. Make stuff people will remember. If you take away one thing from this article, I hope that’s it.