why does xt4 reviews focus on the video so much?

Started 5 months ago | Discussions thread
nostatic Regular Member • Posts: 399
Re: why does xt4 reviews focus on the video so much?

Truman Prevatt wrote:

KaanG wrote:

Truman Prevatt wrote:

KaanG wrote:

jjz2 wrote:

KaanG wrote:

Thats how they position the product. All those "independent" reviewers are clearly using the same script from the company.

We have already reached the potential of the APSC system and the only direction you can push forward is the videography features.

So would you say XT3/XT30 specs/sensor is already peak aps-c performance in terms of IQ/ISO and available lenses? And to get more you'd need to jump to FF?

I do think the jump to 16 to 24-26 mp was pretty good.

I would respectfully argue with that statement. I had the gen 1 (x100) xtrans 2 (xt10) and currently the xtrans 3 (xt2)

Yes the jump in pixels result in improvement with croping but the quality of the image is on the same level.

The last gen produces more clarity in imagea but with the expense of lost in the smooth mojo from the original x100

In practical terms and day to day use i would not expact regular people to recognize the difference between gen 3 and gen 4

Sure the lastest gen is faster but is it better ?

With the 16 MP Xtrans I can print at 300 dpi (which is the min dpi I like to print at) at 10x16. With the 24 MP Xtrans I can print at 300 dpi at 13x20. The 26 MP is just a nit above the 24 and I'd still stick to 13x20 or so.

So sounds to me like unless you want to crop a little bit - the 16 would work just fine in most cases.

If the XH2 uses the new Sony 43 MP chip like rumored than one could print at 300 dpi to 17x26 which wold be a major step up from the first gen XTrans but not a huge jump for the 24/26 MP generation.

Face it - the improvements today are incremental in resolution and certainly not in dynamic range as more pixels on the same size sensor equals smaller detectors. The improvements is the multilayer of DRAM with increases speed with a side benefit for faster AF, faster read out and better video. To double the resolution one needs 4 times the number of detectors and technology is not taking those leaps in the same size sensor since the early 2000's.

For a stills only camera where video is not of any interest whatsoever, the 16 MP XTrans is not a bad option at all.

The increase from 16 to 24 mpx brings another benefit with the improved cropping.

I can now carry two prime bodies, the GR3 with 28mm and XT2 with 50mm eqv. field of view.

When you crop the 28mm to 35 and the 50mm to 85mm, you end up with very usable and clean 16 mpx files. This helps bring the practicality of the zooms and image quality of the primes together in one single setup.

BTW I really liked your earlier comment on the topic of AI in this post. It's not directly related to photography but very good insights

Yes, the jump from 16 to 24 does allow for more cropping so more flexibility.

As far as AI, we tend to loose perspective on "new technologies. For example Apple's "Seri" is far from new. It was first developed funded by DARPA by SRI International during the late 1970's. DARPA wanted to work to remove the second man from the cockpit of military aircraft by allowing the pilot to talk to the plane and the plane talk to him to do the task that was done by the second man.

Again the drawback was not the mathematics but the signal processing horsepower. Now it sprung on the commercial market 5 or 6 years ago as "brand new" with Amazon following on with their smart speakers. Most technologies in fact have been around and are fairly mature when they do become commercial. The digital camera for example has been in use by the US military since the late 1970's with the first digital camera flown in space in 1976 - an ECL CCD designed and built by Fairchild Camera and Imaging Fairchild Semiconductor division. The break through in most technology for commercial use comes when it can be produced cheaply enough to be commercially viable - that is cheap enough that a large portion of the population will buy it.

That is usually years after it has been developed. The first digital computer that was Turing complete was put into operations in 1945, known as the ENIAC.



It was 30 years later in 1975 until the first machine resembling a PC came to market and that was a kit for nerds to build. It wasn't until about 1980 when Steve Jobs pushed H/W S/W design and a Bill Gates produced a descent disk operating system to power the IBM H/W that the PC's really became commercially feasible.

But you can be assured that there were plenty of computers around between 1945 and 1975. Technology evolution is just that evolution - one small step at a time. We often loose sight of that.

The next big impact on the commercial world will be computer vision as digital sensors become less and less expensive with more and more capability and computational horse power grows - thank you Moore's law - so robots can navigate their environment and make decisions because they can see. As a young fledging algebraic geometer I never gave one though to how such an abstract part of mathematics could have any sibilance of applications (except to particle physics) but it turns out to be the foundations of computer vision as the processing of sensor data to simulate sight one has to work in projective space (since all light from a ray is on the same point in a sensor hence the sensor has to be viewed as a plane in projective space and not traditional three space).

Now there is even a field in applied mathematics called - wait for it - computational algebraic geometry. Now our cars stop for us in an emergency, warn us of cars on the side, park the car for us all because of this "new" technology. The ideal of projective geometry goes back to the third century AD and the projective transform was defined and studied starting in the 1400's. In the mid 1950's, W. L. Chow, my co-adviser, and Oscar Zariski wrapped the new developments in Algebra around the subject that by that time had run out of tools to make any more progress to put the geometry of projective space onto a firm logical mathematical footing - kicking off rapid development in the field. Who knew that would lead to a car that can drive itself or a robot that can actually do something.

A glimpse of the future.



This one my favorite


There is really little new under the sun and we do stand on the shoulders of giants.

Another take on the future - this was made a number of years ago (Matsuda is an architect by training). Most of the capabilities are already in play. I've used clips of this in talks and as now as framing for policy issues.


-- hide signature --
 nostatic's gear list:nostatic's gear list
Canon EOS 6D Sony a7 III Fujifilm X-T3 Canon EF 16-35mm F2.8L II USM Canon EF 24-105mm f/4L IS USM +5 more
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow