File this under “just for fun.” I have both a Nikon D4 and Fujifilm X-T1. Both cameras are roughly 16MP resolution, but the Fuji uses an APS-C sensor while the Nikon D4 is a 35mm sensor. I performed this quick ISO comparison simply to see how well Fuji’s sensor stacked up against that of my D4, which is a low-light champ. Before I go on, please understand that these are two very different cameras, for very different uses. I like them both for different reasons!
Methodology
I shot both cameras in RAW, on a tripod. I focused the D4 using Live View. I focused the X-T1 manually using live view focus peaking. I did not go beyond ISO 6400, as the Fuji does not capture raw images above this value. I compared ISO 800, 1600, 3200, and 6400.
I used my 24-70mm f/2.8 Nikkor on the D4, and the Fujinon 23mm f/1.4 prime on the X-T1. I set the 24-70mm lens to 35mm, which roughly matched the focal length on the crop-sensor Fuji. I set each lens to f/4. I processed both images in Lightroom 5 using identical sharpening settings.
Results
Here’s the two shots at ISO 6400, scaled for web size (click to enlarge):
At web size, both images look perfectly good. The main differences are with White Balance (the D4 image looks warmer) and exposure. Interestingly enough, while the D4 exposure was 1/60s, the X-T1 required 1/30s. I’m not sure if that’s due to slight changes in ambient lighting, or a difference in the way each camera meters the scene. In theory, I should have gotten nearly identical exposures using the same ISO value.
Comparisons
Here are side by-side comparisons at 1:1 of each tested ISO. In each case, the X-T1 image is on the left (click to enlarge).
Conclusions
Both the Fujifilm X-T1 and Nikon D4 exhibited luminance noise at ISO 1600 and up. The grain pattern was not what I would call problematic in either camera; subjectively both produced good results. The colors from the Nikon D4 were slightly better (more saturated) at the higher ISO values. I also discovered that in Lightroom, using the default setting (25) for Chrominance NR was too strong for the X-T1 images and created artifacts. Setting Chrominance NR to 8 solved this problem. The D4 images had the appearance of slightly better detail; this could also be due to slight differences in focus. Overall, I was impressed by the quality of the X-T1 images at ISO 6400.
The reason I performed this test was not to see which camera was better, but instead to get a benchmark for the X-T1, which I don’t usually choose as a low-light camera. The Nikon D4 has some significant advantages over the X-T1 with respect to low-light shooting. Not only are the images slightly sharper and have better color, but the D4 can shoot raw files above ISO 6400; something the X-T1 cannot do. Plus, the D4 has better ergonomics, 10fps shooting, and a killer flash system. Nevertheless, the X-T1 image quality was pretty darned good for a crop-sensor (APS-C) camera. When I want to go light, I choose the X-T1. When I need maximum performance, I go with the D4. Considering the price difference between these two cameras, it’s nice to see this kind of ISO performance in a $1500 camera.
Order a Fujifilm X-T1 or Nikon D4s from B&H Photo and support this site!
Jason, your comments are the latest of several tests I have read where the X-T1 has required approximately 1 more stop of exposure to equal exposures of other branded cameras. The speculation is that Fujifilm’s ISO settings are artificially high, thus appearing that the image files have a lower digital noise level. I don’t know, but a trend seems to be emerging.
Maybe a more accurate test would be to use the same shutter/aperture/ISO settings, adjust the brightness of the Fujifilm image image to match the D4, then check for luminance/chrominance noise. In other words, remove the variables and don’t rely on the manufacturer accurately lableing the ISO.
Dennis-
I did some more testing and it looks like the culprit is the “standard/matrix” metering mode. When I put the camera into spot metering or into full-frame metering, I get exposure values that are nearly identical to my Nikons. It seems that blank walls (uniform subjects) are the worst offenders. Go figure!
It’s pretty well known that Fuji overstates their ISO values. I love my Fuji’s, but it’s not apples to apples when using same exposure values.
Metering shouldn’t be part of the equation when exposing manually. The Fuji’s always need close to a full stop of exposure to match other cameras. Thus comparing ISO 6400 on the Fuji is more like comparing to ISO 3200 on the Nikon or Canon for that matter.
Been there done that.
I have found that when using full-frame or spot metering modes, I get virtually identical exposure values as compared with my Nikons (± 0.3EV).
Hi Jason.
Very interesting comparison, though I’m not generally a “low light shooter”.
At the risk of being off-topic, I am highly considering the XT-1 with their new 18-135 WR Zoom, for a shoot in Oregon, since it’s cold and sometimes wet up there, and I wouldn’t have to worry about weather with that kit. But I don’t want to sacrifice image quality (particularly sharpness/detail) over my Nikon D5300. I have read many stellar reviews of the Fuji glass, such as the 35, 56 and 60 macro lenses.
Have you made any large (poster size) prints yet from the XT-1? I wonder how the detail holds up from that sensor when printed large (30×40).
Thanks again.
Regards,
Steve
Hi Steve-
I haven’t done any printing from my X-T1 yet, but honestly the detail I get is similar (at lower ISO) to what my 16MP Nikon D4 produces. Given that I’ve made wonderful large prints from my 12MP Nikons, I wouldn’t worry too much. However, at 30×40, you are definitely going to lose some detail and you’ll want to sharpen for print appropriately.
I agree with the other comments. I have first hand experience with Fuji’s overstating their ISO values by 2/3 to 1 stop. I am convinced at this point that this is intentional as the jepg and raw ouput is also slightly softer (meaning less contrast and sharpness). This further enhances the high ISO performance at the expense of detail (saturation remains good, however). Fujis perform well, but no better than the top of the line APS-C competition when matching shutter speeds, aperture, sharpness and contrast settings.
Which, IMO is fine for my needs.
FYI, please check your exif data. The XT-1 is shot at 1/30, while the D4 was set at 1/60. Both cameras were at IS0 6400 ISO and at F4. There’s your one stop overstated ISO.
Based on the shutter speed and f-stop combination, the XT-1 was actually shooting at ISO 3200, while the D4 was at ISO 6400. That’s the problem with Fuji.
Unfortunately, your test and conclusions fail to take this into account .
To properly compare cameras and their low light performance, you must have the same light levels, while shooting at the same ISO setting, shutter speed and f-stop. When using different sensor sizes this will result in a difference in depth of field, but we are just looking at high ISO performance here, so that is not a relevant variable for this test.
Try it again, this time with matching parameters, and then tell us what you conclude. I think you will find that the XT-1 still performs well in low light, but its performance will more closely match its direct competition…not full frame sensors.
I just noticed that you did in fact realize that the shutter speeds were different. However, your interpretation baffles me. If the shots were taken inside under the same conditions, the light levels should have been the same and exposure values should have been similar.
Are you going to believe me or your lying cheating eyes (homage to Groucho Marx) ? The exposures look to be roughly the same. I would have expected at most a 1/3 of a stop difference, if any. The exif data makes no sense when looking at those two pictures. A one stop exposure difference in low light images is readily apparent when viewing two images side by side. Unfortunately, this confirms that the ISO value could not have been the same for both cameras.
Hi Jason,
I own both the XT-1 and the D4s as well. I would love to know exactly how you get great results like that from the XT-1 at ISO 6400.
Thanks,
Todd
I didn’t do anything special, they were simply tripod-mounted test shots. I don’t use the X-T1 for high-ISO work, because it requires a JPEG workflow.
I am a Nikon convert (D4) moving to the X-T1 over the past year. I find that the D4 is far superior to the X-T1 in low light action shots (example – indoor Basketball). For normal daylight landscape shots the X-T1 (with the better Fuji lenses) gives very similar results to the D4. Add in the smaller kit size and the lighter weight, my X-T1 is my normal choice.
Something to consider is the proper conversion from DX to FX.
Shot on DX at 23mm, f/2.8, ISO 1600 from 2.3 meter should be compared to FX at 35mm, f/4, ISO 3200 from 3.5 meter (both taken at the same shutter speed).
This conversion results in identical field of view, exposure and Depth of Field. Moreover, the performance from the same generation of DX/FX sensors trends to be about the same (i.e. one stop difference in exposure, e.g. ISO 1600 vs 3200).
So everything reduces to availability of fast glass.