[Kunzite]

But some are backed up by known pranks

[falconeye]

there are a number of documented test reports of poor AF accuracy of the d800/d4 with f/1.4 lenses on digilloyd.com.

The D800 has its famous accuracy problem with its edge AF sensors, that's true. But it should do fine with its center AF points. Any diverging opinion in http://diglloyd.com/prem/prot/ZF/publish/135APOSonnar-examples-portraits.html ? (I haven't subscribed yet).

Also, the tungsten problem with my D800E is nowhere near as obvious as with the K-5 in low tungsten light. I only care about at very wide apertures (which I don't normally use in the studio) by using a small extra AF fine tune, something which wouldn't have been possible with the K-5 (out of range, luminance-dependent). I use Reikan FoCal to determine these color-temperature dependent values quickly (they differ, but not by that much ...).

The K-5II AF is great, from what I hear. But this now holds true for most competing cameras, esp. a few mirrorless ones. The big area which still does make a huge difference is to focus on fast moving subjects. That's the point of phase detect.

That's the remaining market for a hypothetical K-3 too, assuming it to be more expensive than a K-5II.

High fps, large buffer, locking on fast moving targets, incl. when using the DA 560. An upgrade for those D300s or 7D owners waiting for an upgrade which doesn't emerge from Canikon

[Anvh]

Maybe... if they take the queue from D7100.

I won't say that its a small difference.
Having played with deconvolution sharpening on K30 files, I'd still say that its more practical and convenient not to have the AA-filter.

Then there are those who are very fearful of moire. (even though I've also seen it before on K5 files too)
So everyone seems to have an opinion on it.

See how it pans out.

$50 cheaper?
It should be $100

K30 use 16mp sensor, that's less then 20-24mp sensor i commented on.
If AA filter will blur about 1/2 pixels and the pixels are very small then the blur will also be very small.
ALso moire would be less of an issue with smaller pixels.

So the differnce will be so good as none, even between the K5 mkII versions, when you sharpen the normal version effectively then you wont see the differnce at all in the final print. Some say they can see differnce at 100% but i find it very hard to see a differance.

[Parry]

I hope the K-3 has a full frame Fuji X-Trans sensor should Fuji make one and therefore no AA filter and no moiré, with RAW conversion software that works.

[falconeye]

I hope the K-3 has a full frame Fuji X-Trans sensor should Fuji make one and therefore no AA filter and no moiré, with RAW conversion software that works.

I hope people would stop asking for this. IMHO, X-trans is a dead end.
A good lecture is the 4 part article at ChromaSoft: Demosaicing the Fuji X-Pro1 and its X-Trans CMOS sensor

Most of us tend to forget that even a Bayer image must be reconstructed from a green image of half tesolution and red and blue images of quarter resolution. That's working because even the red and blue spectra are mostly dense up to half the Nyquist freqency, because the human eye is less sensitive to color resolution and because a couple of heuristic cases (e.g., edges) can be treated exactly w/o a loss in resolution. It is kind of a fragile balance between luminous abd colour sampling.

X-trans breaks this, its red and blue sensel positions leave some huge gaps in the spectra. This leads to color artefacts at the reconstructed pixel level. I'd rather have a Bayer-AA filter which can be sharpened.

The raw converters may be able to improve in treating X-trans files. But the same is true with the Bayer pattern. E.g., moirè patterns can be recognized and compensated for (inverted).

UPDATE ...
Just saw Adobe's Fujifilm X-Trans sensor processing tested: Digital Photography Review
This (ACR 7.4RC) provides a much better conversion for X-trans, avoiding many of the artefacts. But ironically, the output now is softer AND shows signs of color moiré (cf. second-last image on the bottom) ...
I think it validates the point I was making above.

[Aristophanes]

I'd rather have a Bayer-AA filter which can be sharpened.

Ironically the industry consensus reached about 15 years ago. Then came Photoshop and PC-based sharpening alternatives, RAW everything, etc. But as less photos are being processed on a home computer as the home computer itself is a threatened beast, we'll be going back in-camera for most processing.

[Digitalis]

I hope people would stop asking for this. IMHO, X-trans is a dead end.

+1 on that, though one has to admit it is a novel approach though.

But as less photos are being processed on a home computer as the home computer itself is a threatened beast, we'll be going back in-camera for most processing.

I don't think there is a camera processor that is anywhere near as capable as the six cored PC processors on the market these days, and don't forget about colour calibration - those panels on the back of our cameras are TN panels, not IPS.

---

Last edited by Digitalis; 02-28-2013 at 04:24 PM.

[gazonk]

I don't think there is a camera processor that is anywhere near as capable of the six cored PC processors on the market these days, and don't forget about colour calibration - those panels on the back of our cameras are TN panels, not IPS.

I think the move away from PCs is more relevant for the pure consumer market so far, as soon as you're a little passionate about photography, you're still going to use a PC.

BUT... there are some hybrid PCs now where the screen is a detachable tablet which contains the actual processor (which may be a (yuck!) Atom, but also Core i5 for some models) and actually also wacom-licensed pressure sensitive touch hardware in the screen. The better of these PCs can definitely run Photoshop (my daughter is going to art school and we were looking at these but found out that she's better off with her mac and a Wacom Intuous for a while still). I imagine that in the not too distant future, real image processing software will also run on dedicated tablet OSes, and that we will see large and advanced tablet computers with pen input being used by photographers (as well as artists) to a much larger degree than now.

[Digitalis]

I think the move away from PCs is more relevant for the pure consumer market so far, as soon as you're a little passionate about photography, you're still going to use a PC.

BUT... there are some hybrid PCs now where the screen is a detachable tablet which contains the actual processor (which may be a (yuck!) Atom, but also Core i5 for some models) and actually also wacom-licensed pressure sensitive touch hardware in the screen. The better of these PCs can definitely run Photoshop (my daughter is going to art school and we were looking at these but found out that she's better off with her mac and a Wacom Intuous for a while still). I imagine that in the not too distant future, real image processing software will also run on dedicated tablet OSes, and that we will see large and advanced tablet computers with pen input being used by photographers (as well as artists) to a much larger degree than now.

I happen to use a tablet PC - ad as far as i'm concerned I would really hate to have to rely on it for post processing. The current version of Adobe Photoshop touch for the android OS has less than half the relevant filters and tools needed for my work. While I agree that tablet PC systems hold great promise there are still some significant hardware hurdles to overcome - more powerful processors eat up batteries, large IPS-S based touch screens with high DPI and 2048 levels of pressure sensitivity and a high polling rate will eat up batteries, yet consumers always want the next product to be thinner than the next*. Manufacturers can only work within the limits of their materials and as far as I can see battery technology has some way to go. Current solid state storage, processors and display technologies are pretty energy efficient - it is just carrying enough battery power required to run all this stuff that is the hard part.

*if you have ever seen a teardown of an Apple Ipad you will see that 90% of the internal space behind the screen of the tablet is taken up by the battery.

[mee]

The problem I'm seeing in this thread is people are basing the current technology on what will happen in the future.. much like the 'kitchen of the future' back in the 1950s.. hilarious today because it looks like a comic book 1950s kitchen and nothing modern.

PCs aren't going anywhere for many years. Too many niches that require heavy floating point performance (3d artists, audio production, film / effects, serious gamers etc) that current portable processors cannot and will not be able to match for a longg time. Intel can shrink the die a few more times I think on their x86 lineup with currently known methods. Expect slower release cycles and higher prices though.. We won't know what we are using in 10 years, however, until we get closer to there.

[Medium FormatPro]

MAINGEAR Nomad 17 [HM77] - Ultimate

Here's a laptop PC that can well excess using any version of programs such as Adobe CS also with numerous plug-ins. It can also run that on battery power for over two hours - fully tasked. If one runs close to running out of battery power (unlike Apple/Mac) simply switch out the battery. But also there are plenty of other laptops (some of less than 1k in price) that can also run circles with CS on battery power for well over three hours.

In less than one years time the world will see how many laptop makers will also be offering 6 and even 8 core laptop processors.

[Alizarine]

In less than one years time the world will see how many laptop makers will also be offering 6 and even 8 core laptop processors.

Manufacturers can only work within the limits of their materials and as far as I can see battery technology has some way to go. Current solid state storage, processors and display technologies are pretty energy efficient - it is just carrying enough battery power required to run all this stuff that is the hard part.

I agree to these. So far developments in the battery industry are the ones progressing slower than the processor and other semiconductor components, though there's the interesting development of a very thin battery which may come soon for those "foldable" smartphones (and probably into mirrorless cameras in the near future):

Thin film rechargeable lithium battery - Wikipedia, the free encyclopedia


it's old news, and that's quite the situation... where has this technology progressed yet?

[gazonk]

I happen to sue a tablet PC

Sue? Just a normal typo or a Freudian slip? ;-)

The current version of Adobe Photoshop touch for the android OS

Yes, I was thinking in terms of real operating systems (if you can call Win 8 that, which is what the newest products come with) and real Photoshop. My daughter, for instance, wouldn't have use for anything which doesn't work well drawing with a variety of brushes on at least 4048px wide canvases in Photoshop.

One example which seems to be a step in the right direction but not yet quite there, is the Samsung "ATIV Smart PC Pro 700%": http://www.bhphotovideo.com/c/product/891266-REG/Samsung_xe700t1c_a01us_Inte...Windows_8.html

[Digitalis]

Sue? Just a normal typo or a Freudian slip? ;-)

Suing ASUS would be a fools errand, have you read one of their badly translated manuals? they are almost as impenetrable as some academic theses. I have some old camera manuals that use genderised instructions E.G: "grab her crank"

[Ash]

Clearly, he meant 'use' not sue.