The Fujifilm X-H2 High ISO Noise vs the X-T2

Jerry-astro wrote:

Nebell wrote:

I'm too early to the party, it's in 360p. Gotta wait for HD processing.

Anyway, while it's informative, if ISO performance is important, then FF is an option. There are plenty of FF cameras that are at the same price as X-H2.
It's not the same in case of FF cameras. If you want to go up your option is medium format which is both a lot more expensive and limiting.

It's easy for someone to recommend changing or upgrading cameras, and that might solve the problem, but at a significant expense. Before taking that route, I would recommend that the OP try something like Topaz DeNoise (one of a number of good NR solutions out there). I've had really great results in reducing noise while still retaining fine detail. I would certainly recommend taking that route first rather than simply selling existing gear at a loss and then reinvesting in expensive (and likely more bulky) gear. If that doesn't meet the OP's needs, then selling gear and upgrading always remains as an alternative.

If I am going to do available light it will be with my Lecia Q2M.  I would not use an XH2.  After I got my Q2M, I would not even use my Pro3 for available light.  However, up to ISO 1600 to 3200, my wife's XH2 is better than my XPro3 for noise performance.  It is significantly better when I down sample the XH2 file down to the size of the XPro3 in the processed image.  The read noise it todays cameras has practically eliminated by the current state of the art in the CMOS circuits in Sony's chips.  The limiting factor in SNR is shot noise - not read noise.  There are plenty of charts on the Internet that documents this.

While it pains me to admit it - Jerry is absolutely right.    The current smart noise reduction SW be it Topaz or DXO do an incredible job of integrating out noise while maintaining high frequency details.  If I do pick up a XPro4 and I will pick up a version of the latest DXO Pure Raw in case I takes some available light shots in case I want to address noise, although I like my available light shots pretty raw and emotional in which case I might just add grain to the noise.  

One of the issues I see with most of the comparisons is they are apples to oranges. You can't compare a 26 MP at 100% crop on an emissive display with a 100% crop of a 40 MP sensor.  If you do the same scene of the 40 MP sensor is magnified by a factor or 116%.  So there is an enlargement ratio of the 100% crop of the 40 MP over the 26.  That is if the linear length of the 26 MP is 1 unit, the 40 MP is 1.16 units.  In order to make a valid comparison, the 40 MP image should be down sampled to the 26 MP size using a low pass filter.  This of course would not only help eliminate noise it would eliminate the details at spacial frequencies above the capability of the 26 MP sensor to capture.