Rumours say X-H2 is Xtrans

HatWearingFool

Senior Member
Messages
2,760
Solutions
1
Reaction score
2,280
Location
CA
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.

 
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.
That’s too bad, seems like a really dumb move if this thing is indeed 43MP.
 
Why? I thought it was the processing power that was the hold back. Maybe I’m not on top of its IQ shortcomings - but my T2 is performing wonderfully for me. No shortcomings.
 
Why? I thought it was the processing power that was the hold back. Maybe I’m not on top of its IQ shortcomings - but my T2 is performing wonderfully for me. No shortcomings.
The problem isn't X-trans itself, but that it requires different demosaicing from bayer to get optimal results. The issues aren't at all as big today as they were a few years back, though.
 
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.

https://www.fujirumors.com/exclusiv...-trans-or-bayer-the-answer-now-on-fujirumors/
Not surprising indeed, sadly.

Well, I'm most likely switching to Nikon Z in the future anyway(X-trans being one reason).
I’m sure there are many reasons that some people might see some benefit in moving to Nikon Z. However, I simply can’t see X-Trans alone being enough of an issue to drive a change like that. There are a number of solutions for processing X-Trans images that produce results at least equal to those from Bayer sensors. IMHO, there are so many other factors that could drive a decision to change brands (ergonomics, features, pricing, breadth of product line… etc.) before that would be an issue (for me).

I wouldn’t at all rule out the possibility of a brand change in the future, but X-Trans certainly won’t be any sort of a factor in that decision. Obviously, YMMV. Meanwhile, I’ll anxiously await the X-H2 introduction next year (X-Trans, Bayer, whatever…)..

--
Jerry-Astro
Fuji Forum co-Mod
 
Last edited:
Why? I thought it was the processing power that was the hold back. Maybe I’m not on top of its IQ shortcomings - but my T2 is performing wonderfully for me. No shortcomings.
Having to hassle with a special demosaicing process for optimal results is a big shortcoming in my book, especially for Lightroom users (me). While good results came be obtained with careful processing, at 24MP+ resolution there really isn’t any significant advantage to X-Trans anymore. Bayer sensors no longer use an anti-aliasing filter either so there’s no resolution advantage, and as the pixel density increases, any slight X-Trans advantage in false color moire) rejection dries up too. The significantly more demanding demosaicing processing is a hefty drain on in-camera resources as well that could likely be put to better use (AF, for example). While I’m used to X-Trans and have learned to make the best of it, I’d rather not have to deal with it at all.
 
I’m sure there are many reasons that some people might see some benefit in moving to Nikon Z. However, I simply can’t see X-Trans alone being enough of an issue to drive a change like that. There are a number of solutions for processing X-Trans images that produce results at least equal to those from Bayer sensors. IMHO, there are so many other factors that could drive a decision to change brands (ergonomics, features, pricing, breadth of product line… etc.) before that would be an issue (for me).

I wouldn’t at all rule out the possibility of a brand change in the future, but X-Trans certainly won’t be any sort of a factor in that decision. Obviously, YMMV. Meanwhile, I’ll anxiously await the X-H2 introduction next year (X-Trans, Bayer, whatever…)..
Note that I said ONE reason. So no, it's not the only reason.

I use Capture One and while it does a good job with X-trans files, I still don't find them as detailed as bayer files. The only software with which I've ever achieved equal results to bayer is RawTherapee.
 
Why? I thought it was the processing power that was the hold back. Maybe I’m not on top of its IQ shortcomings - but my T2 is performing wonderfully for me. No shortcomings.
Having to hassle with a special demosaicing process for optimal results is a big shortcoming in my book, especially for Lightroom users (me). While good results came be obtained with careful processing, at 24MP+ resolution there really isn’t any significant advantage to X-Trans anymore. Bayer sensors no longer use an anti-aliasing filter
I could be wrong, but I am pretty sure Canon is still using AA filters on all of their cameras.
either so there’s no resolution advantage, and as the pixel density increases, any slight X-Trans advantage in false color moire) rejection dries up too. The significantly more demanding demosaicing processing is a hefty drain on in-camera resources as well that could likely be put to better use (AF, for example). While I’m used to X-Trans and have learned to make the best of it, I’d rather not have to deal with it at all.
 
I’m sure there are many reasons that some people might see some benefit in moving to Nikon Z. However, I simply can’t see X-Trans alone being enough of an issue to drive a change like that. There are a number of solutions for processing X-Trans images that produce results at least equal to those from Bayer sensors. IMHO, there are so many other factors that could drive a decision to change brands (ergonomics, features, pricing, breadth of product line… etc.) before that would be an issue (for me).

I wouldn’t at all rule out the possibility of a brand change in the future, but X-Trans certainly won’t be any sort of a factor in that decision. Obviously, YMMV. Meanwhile, I’ll anxiously await the X-H2 introduction next year (X-Trans, Bayer, whatever…)..
Note that I said ONE reason. So no, it's not the only reason.

I use Capture One and while it does a good job with X-trans files, I still don't find them as detailed as bayer files. The only software with which I've ever achieved equal results to bayer is RawTherapee.
Well… different strokes for different folks, I guess. I’ve never had any issues rendering fine detail (mostly with LR), and quite a bit of what I shoot (hummingbirds for example) has a fair amount of fine detail. But, then again, I’m not prone to heavyweight pixel peeping either, which might be the reason for the difference in our views. I’ve never perceived a detail issue in large prints (up to 20x30 inches in my case).

--
Jerry-Astro
Fuji Forum co-Mod
 
Last edited:
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.

https://www.fujirumors.com/exclusiv...-trans-or-bayer-the-answer-now-on-fujirumors/
Not surprising indeed, sadly.

Well, I'm most likely switching to Nikon Z in the future anyway(X-trans being one reason).
I’m sure there are many reasons that some people might see some benefit in moving to Nikon Z. However, I simply can’t see X-Trans alone being enough of an issue to drive a change like that. There are a number of solutions for processing X-Trans images that produce results at least equal to those from Bayer sensors. IMHO, there are so many other factors that could drive a decision to change brands (ergonomics, features, pricing, breadth of product line… etc.) before that would be an issue (for me).

I wouldn’t at all rule out the possibility of a brand change in the future, but X-Trans certainly won’t be any sort of a factor in that decision. Obviously, YMMV. Meanwhile, I’ll anxiously await the X-H2 introduction next year (X-Trans, Bayer, whatever…)..
XTrans is not the poison pill for me - though it is close. If Fuji did not have the XPro OVF it would be. A Nikon EVF is most likely no better or worse than a Fuji EVF. For continual use they both suck as any EVF sucks. Neither is how I want to view the world. I will live with XTrans because of the XPro - although at 24 MP up there is zero need for it even if there ever was.

That Capture One and Iridient spent the resources to make XTrans tolerable is one also a reason I still own a Fuji.

It was a marketing gimmick. If the H2 is 43 MP - then why does it require XTrans when the 50 MP GFX doesn't? It doesn't. Fuji has backed themselves in a corner. They over sold XTrans, convinced a few suckers that it was the greatest thing since canned beer and would revolutionize digital photography. They picked up a few sales because of it. Now they are stuck with a difficult time getting out of it.
 
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.

https://www.fujirumors.com/exclusiv...-trans-or-bayer-the-answer-now-on-fujirumors/
Not surprising indeed, sadly.

Well, I'm most likely switching to Nikon Z in the future anyway(X-trans being one reason).
I’m sure there are many reasons that some people might see some benefit in moving to Nikon Z. However, I simply can’t see X-Trans alone being enough of an issue to drive a change like that. There are a number of solutions for processing X-Trans images that produce results at least equal to those from Bayer sensors. IMHO, there are so many other factors that could drive a decision to change brands (ergonomics, features, pricing, breadth of product line… etc.) before that would be an issue (for me).

I wouldn’t at all rule out the possibility of a brand change in the future, but X-Trans certainly won’t be any sort of a factor in that decision. Obviously, YMMV. Meanwhile, I’ll anxiously await the X-H2 introduction next year (X-Trans, Bayer, whatever…)..
XTrans is not the poison pill for me - though it is close. If Fuji did not have the XPro OVF it would be. A Nikon EVF is most likely no better or worse than a Fuji EVF. For continual use they both suck as any EVF sucks. Neither is how I want to view the world. I will live with XTrans because of the XPro - although at 24 MP up there is zero need for it even if there ever was.

That Capture One and Iridient spent the resources to make XTrans tolerable is one also a reason I still own a Fuji.

It was a marketing gimmick. If the H2 is 43 MP - then why does it require XTrans when the 50 MP GFX doesn't? It doesn't. Fuji has backed themselves in a corner. They over sold XTrans, convinced a few suckers that it was the greatest thing since canned beer and would revolutionize digital photography. They picked up a few sales because of it. Now they are stuck with a difficult time getting out of it.
It’s a complete non issue for me. For the few times that it’s gotten in my way, there are easy ways to process an image in a way that makes XTrans moot as an issue when rendering detail. For the rest of the time, it simply is a big nothing burger. No advantage… no disadvantage. For the amount of time expended in this forum discussing it, it’s WAY overplayed as an issue IMHO, and has been for a good long time. But, obviously, I’m biased, or more likely just numb to all the hoo-hah over it. For those of us crazies who insist on sticking with LR, ED all but eliminates the issue with a simple additional PP step.

--
Jerry-Astro
Fuji Forum co-Mod
 
Last edited:
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.
That’s too bad, seems like a really dumb move if this thing is indeed 43MP.
I cant see this thing being 40+ mp.

Ive really no idea who the xh1 xh2 are for anyway.
People with adult sized hands.

I'm a little disappointed with it being xtrans because that means no sensor shift high res photos and no raw video. It also likely means that AF isn't going to catch up if my understanding is correct of xtrans requiring more processor power. It's one thing to rewrite the algorithm (which remains to be seen if they've actually done), it's another to have to have that algorithm be so good that it overcomes the additional processing handicap to catch up to other manufacturers.
 
I agree with Truman. If it makes no difference to you, Jerry, then there should be no issue in going back to Bayer.

I agree with Truman that at this point Fujifilm is digging in their heels. To abandon X-Trans on the X-H2 would be some sort of admission on Fujifilm’s part.

if the X-H2 were to come out at 43MP it would have a far greater pixel pitch than a GFX 100. If there were real tangible benefits to X-trans then why aren’t the GFX bodies using it? They are the premium IQ flagship bodies.
 
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.
That’s too bad, seems like a really dumb move if this thing is indeed 43MP.
I cant see this thing being 40+ mp.

Ive really no idea who the xh1 xh2 are for anyway.
People with adult sized hands.

I'm a little disappointed with it being xtrans because that means no sensor shift high res photos and no raw video. It also likely means that AF isn't going to catch up if my understanding is correct of xtrans requiring more processor power. It's one thing to rewrite the algorithm (which remains to be seen if they've actually done), it's another to have to have that algorithm be so good that it overcomes the additional processing handicap to catch up to other manufacturers.
You could still have sensor shift it would just need to move differently.
 
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.
That’s too bad, seems like a really dumb move if this thing is indeed 43MP.
I cant see this thing being 40+ mp.

Ive really no idea who the xh1 xh2 are for anyway.
People with adult sized hands.

I'm a little disappointed with it being xtrans because that means no sensor shift high res photos
Is there really a point to sensor shift high res when you already have a 40+ megapixel crop sensor?
and no raw video. It also likely means that AF isn't going to catch up if my understanding is correct of xtrans requiring more processor power. It's one thing to rewrite the algorithm (which remains to be seen if they've actually done), it's another to have to have that algorithm be so good that it overcomes the additional processing handicap to catch up to other manufacturers.
The phase detect pixels have no color filters and are completely independent of the X-Trans array. While you do need to demosaic for subject recognition, you only need to demosaic a small fraction of the sensor (subsampling). This can be done by skipping alternating lines and rows on the sensor. Depending on how the data is subsampled, the demosaic may be no more complicated than a Bayer pattern. This subsampling is also how the feed for the EVF is created as the entire sensor can not be read out at 100fps. Even for the X-T4 EVF, you only need a 1280x960 resolution feed.
 
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.
That’s too bad, seems like a really dumb move if this thing is indeed 43MP.
I cant see this thing being 40+ mp.

Ive really no idea who the xh1 xh2 are for anyway.
People with adult sized hands.

I'm a little disappointed with it being xtrans because that means no sensor shift high res photos and no raw video. It also likely means that AF isn't going to catch up if my understanding is correct of xtrans requiring more processor power. It's one thing to rewrite the algorithm (which remains to be seen if they've actually done), it's another to have to have that algorithm be so good that it overcomes the additional processing handicap to catch up to other manufacturers.
You could still have sensor shift it would just need to move differently.
You wouldn’t be able to have sensor shift raw files or sensor shift full RGB raw files.
 
Not really surprising, but the site says more info on the X-H2 is coming soon which is exciting.
That’s too bad, seems like a really dumb move if this thing is indeed 43MP.
I cant see this thing being 40+ mp.

Ive really no idea who the xh1 xh2 are for anyway.
People with adult sized hands.

I'm a little disappointed with it being xtrans because that means no sensor shift high res photos and no raw video. It also likely means that AF isn't going to catch up if my understanding is correct of xtrans requiring more processor power. It's one thing to rewrite the algorithm (which remains to be seen if they've actually done), it's another to have to have that algorithm be so good that it overcomes the additional processing handicap to catch up to other manufacturers.
You could still have sensor shift it would just need to move differently.
You wouldn’t be able to have sensor shift raw files or sensor shift full RGB raw files.
Why not? You’d be able to get rgb at each pixel with the same number of moves.
 

Keyboard shortcuts

Back
Top