The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Fujifilm GFX100 Pixelshift Firmware is out

DO NOT USE 2s intervals!Screen Shot 2020-12-23 at 9.43.31 PM.jpg

I'm getting obsessed with this... but what if the problem is the firmware?

I did tonight a test using
  • GF120mm (IS off)
  • f/8, ISO 100, 1/125
  • White balance daylight
  • manual focus
  • Manfrotto Mini-Salon column...very heavy
  • continuos light LED
the target was a page form a book, super flat with very small details
all images have been left As-Is, with no sharpness or curves fixed.

here's my conclusion: the shortest interval is the best, along with the 15s interval shots.
what I'm shocked is the 2s is the absolute worst, totally unusable shot.

Which interestingly enough it coincides with my initial test done with Profoto strobes: I was lamenting that photos taken with flash are terrible...because I was keeping a 2s interval to let the generator generously recharge...

damn... now I want to do a test with strobes and the shortest interval... I have the feeling the photos may come up very well!

Screen Shot 2020-12-23 at 9.37.08 PM.jpg
 

f8orbust

Member
It would be interesting to see a side-by-side comparison of pixel-shifted images from the Olympus (OMD EM1 MK II/III/X), Leica (SL2) and the Fuji GFX100 (I know DPReview has some side-by-side images, but they're not very good). Surely somebody on GetDPI must own all three :ROFLMAO:
 

gerald.d

Active member
So I thought I'd have another go with the pixel shift on the S1R. A while back I took a shot that made me suspect that possibly the initial combing artifacts had been addressed in a firmware update at some point, but I didn't get around to examining the issue in detail.

I have just done so, and can confirm that without question the latest firmware of the S1R does NOT exhibit the bad combing artifacts that I originally was experiencing, the same artifacts that are evident in every shot from the Fuji shared to date from all sources.

This gives me some encouragement that Fuji will be able to address this issue in firmware at some point in time. There can be no way that they are unaware of the problem.

First image is 100% crop, the second an 800% screenshot.

Nothing in my technique has changed. Exact same camera, lens, camera mount, subject mount, environment, etc etc.

IMG_0474-100pc.jpg

IMG_0474-800pc.jpg
 

Paul2660

Well-known member
It would be nice to hear from Fuji on this issue, so far they have been moot. Not even Fuji Rumors seems to have picked up on this issue. As for a firmware update, tough call. Fuji does tend to address more issues than other camera makers with firmware, however considering it took almost 2 years to get what we have now, not holding my breath that it will be address anytime soon, especially with all the fuss over the 100S rollover.

I personally wish they would create something with less output resolution, 200MP if possible, which would take less than the 16 frames. Or work on a Dual exposure solution similar to what Phase One did, (same sensor as IQ4 just smaller). In reality, 16 frames is too many for me, the software they came out is is very limited in scope and it can't be used in any outdoor scene I can imagine, beside Monument Valley. And if the solution is wait 15 seconds between each shoot, then simple movement of the earth would cause misalignment.

Paul C
 

gerald.d

Active member
It would be nice to hear from Fuji on this issue, so far they have been moot. Not even Fuji Rumors seems to have picked up on this issue....

Paul C
It's not just Fuji - not a single reviewer/Fuji ambassador/dealer that I am aware of has addressed the issue, yet I don't doubt for one moment that they have all seen what the problem is. I've had owners of the camera pass off the criticism as being only visible at 400%, and that they can't see the problem on their monitors at 100%, so what's to complain about?

Well, if they can't see the problem at 100% on their monitors, that just goes to show that they can't see the individual pixels at 100% on their monitors, because it is patently obvious this is an issue at the individual pixel level. And Fuji are marketing this as being great for archival work - the very work where every individual pixel needs to be accurately recorded.

There's been plenty of time now for an objective response to the criticism from those who have seen it, and yet complete silence.

At the end of the day, I can only assume that all those individuals and companies have too much of a vested interest in making money off people buying Fuji that they will not address the problem forthrightly and honestly. A very sorry state of affairs.

Kind regards,


Gerald.
 

bab

Member
Gerald
why not contact your Fuji dealer and ask to bowser another camera and lens to test then if results are the same he could possibly send a letter to Fuji requesting a firm response!
 

gerald.d

Active member
a new version of the combiner software is out, but no mentioning of improved results
Will be interesting to see if they have addressed it. I certainly hope so, since I've just put down a deposit for the GFX 100S.

Seems that the release date here in Thailand is the 25th, so I'll get to test it out for myself in a week!

Kind regards,


Gerald.
 

gerald.d

Active member
Will be interesting to see if they have addressed it. I certainly hope so, since I've just put down a deposit for the GFX 100S.

Seems that the release date here in Thailand is the 25th, so I'll get to test it out for myself in a week!

Kind regards,


Gerald.
Confirmation from a number of owners that the issue has not been addressed.

On a positive note, one of the main Fuji ambassadors (Jonas Rask) has now confirmed that he has spoken with Fuji about the issue, and they acknowledge it exists.
 
Top