Fellow FR-reader Sven is specialized in IR and Astro conversions, and he recently did a fullspectrum conversion on his Fujifilm X-H2.
Here some excerpts regarding the teardown:
- After a deep look under the hood, the design of this camera can be summarized as follows: Heat Management and Dissipation.
- mainboard of the X-H2 reveals countless SMD components. With the amount of features packed into such a small camera, this is hardly surprising
- the processor is covered with a heat conduction pad
- The entire shutter mechanism is supported by small springs all around and has no direct contact with the body
- this also explains why the camera’s shutter is relatively quiet and has this very special muted sound
- The heat of the X-Trans V sensor is tapped directly on the back of the sensor via a delicate heat sink and conducted to a copper plate on the stainless steel chassis
As for the IR performance, they write:
- Some modern sensors can show fine, mostly horizontal lines in the images. This is probably caused by PDAF pixels (“AF pixels”) on the sensor. These occur more often with strong IR filters and with very high-contrast processing. How does the X-H2 perform with an 830 nm infrared filter? With overly high-contrast processing, no lines or other artifacts are visible at 100% view (aside from noise, of course, induced by the high-contrast processing). The new sensor is a clear recommendation and can be used without restrictions.
As well as a firmware request:
- Unfortunately, the X-H2 shares the same handicap as all X-Trans cameras. The manual white balance is always set without any problems, but the available range is not quite sufficient. […] Unfortunately, a warm magenta color cast remains in the camera and has to be corrected on the PC.
- Fujifilm can actually do that, with their Bayer sensor cameras (X-A series and X-Txxx) it works without any problems
- The X-H2 seems to work a bit better than the X-T3, but you can’t get rid of the color cast completely
- In the unlikely event that a Fujifilm firmware designer reads this: Please expand the limits in both color axes in which manual white balance can be set. A little bit wider and the cameras internal white balance will work for infrared as well. Such a change can be distributed in a future firmware update in my eyes and should neither harm anyone nor cause too much work in programming.
Read the full article and see all the images at irrecams here.
