[bwDraco]

Update: Details on the f/2.8 high-precision point can be found in post #6 of this thread.

---

The press release reads the following:

The new PENTAX K-5 II / K-5 IIs offer a compact, solid body that delivers high quality image reproduction, especially when working in low light conditions. The camera’s new, highly sensitive AF sensor enables photographers to work with a luminous flux based on F2.8 levels in addition to F5.6, which increases the AF accuracy when using fast lenses. The AF sensor also includes an upgraded AF algorithm as well as the Select-Area Expansion feature, which automatically tracks the subject when it deviates from a pre-assigned selection point, by assessing data from neighboring sensors.

It is clear that the SAFOX X system inherits the AF-algorithm improvements from SAFOX IXi+. However, are there any more specifics on how the f/2.8 high-precision AF is implemented, such as the number of high-precision points and whether the corresponding sensors are oriented vertically, horizontally, or diagonally?

--DragonLord

---

Last edited by bwDraco; 10-14-2012 at 12:08 AM.

[blackcloudbrew]

I think I'd like to know more about this too. I'm sure it will come out in time but ... er .... well, this was a problem I wasn't aware of. I assumed / presumed that AF worked the same for all f-stops.

[rawr]

It is interesting.

At a minimum, I guess they threw all the AF goodies of the K-30 (new lens assembly, better sensitivity under tungsten, tracking feature improvements etc) into the new K-5, and added some extras into the pot - like making all 9 of the existing AF cross-points f2.8 sensitive, intead of the current single 2.8 sensor in the centre. (?)

They should release a white paper on it or something. That would be cool.

[snostorm]

AFAIK, Pentax has never claimed any level of precision for any of its AF sensors before like Canon and Nikon have.

All I know is that all of the SAFOX variations in their DSLRs so far have focused my f2.8 and faster lenses very well in center point mode. Low light sensitivity threshold (and max aperture tolerance) has improved through the models, but they've all worked as expected AFAIC.

Scott

[bwDraco]

At a minimum, I guess they threw all the AF goodies of the K-30 (new lens assembly, better sensitivity under tungsten, tracking feature improvements etc) into the new K-5, and added some extras into the pot - like making all 9 of the existing AF cross-points f2.8 sensitive, intead of the current single 2.8 sensor in the centre. (?)

Pentax has never had f/2.8 high-precision points before.

For reference, here's some info on how the older SAFOX IX works: https://www.pentaxforums.com/forums/pentax-dslr-discussion/181885-analysis-p...provement.html

--DragonLord

[bwDraco]

Update: Only the center point is f/2.8 high precision. The high-precision sensor is vertical line-sensitive (see attachment). The center 9 points are sensitive to EV -3; the outer two horizontal line-sensitive points are excluded. See the Google translation of the Pentax K-5 II / K-5 IIs feature page. Note that it is not clear whether the horizontalvertical line-sensitive portion of the central AF point will work with slower lenses as well.

--DragonLord

[Class A]

I assumed / presumed that AF worked the same for all f-stops.

It does.

The "f/2.8" vs "f/5.6" AF sensitivities mentioned relate to the size of the AF baseline that is used to measure the phase difference.

An f/2.8 AF sensor has better accuracy because it "sees" a stronger variation of phase shifts but it stops working with slower lenses that have f-ratios of f/4 (e.g. DA 15/4) or even f/6.3 (e.g., DA 18-250/3.5-6.3 at the long end) wide open.

The f/5.6 AF sensors have the advantages that they still work with slower lenses, but are less accurate.

---

Last edited by Class A; 10-14-2012 at 08:37 PM.

[JinDesu]

Update: Only the center point is f/2.8 high precision. The high-precision sensor is vertical line-sensitive (see attachment). The center 9 points are sensitive to EV -3; the outer two horizontal line-sensitive points are excluded. See the Google translation of the Pentax K-5 II / K-5 IIs feature page. Note that it is not clear whether the horizontal line-sensitive portion of the central AF point will work with slower lenses as well.

--DragonLord

So the center point is sensitive to -3EV, and is highly accurate (due to being the F2.8 sensor), but the remaining center 8 points are also -3EV sensitive?

[bwDraco]

So the center point is sensitive to -3EV, and is highly accurate (due to being the F2.8 sensor), but the remaining center 8 points are also -3EV sensitive?

Well, that's what it says in the translation.

--DragonLord

[Alizarine]

Nice find here. it's good to learn more info even before the actual units come

[froeschle]

To cite noodlelow on this aspect: What is the Exact Layout of the f/2.8 AF Sensor(s) in the K-5II/s?

In general, as per each AF "point" that is consisted of two crossed linear sensors (which is generally known as a cross sensor by people), if a f/2.8 sensor is included, there are several possibilities as follows:-

1. Modifiying either the horizontal or the vertical sensor to a f/2.8 one. Indeed, this is the easiest and cheapest way to go. What it needs to do is to replace the pair of image projection lenses as seen in the above diagram. The purpose of this is to measure light from the outer part of the glass, instead of measuring the AF closer to the centre. There is simply no need to change anything of the AF sensor array as it is only about the AF projection lens pair that all matters!

Now with the f/2.8 sensor, if with proper software logic, higher AF accuracy can be achieved for fast lenses. What I mean for the "proper" software logic refers to the camera should choose the better measurement value that is closer matched to the actual f-number that is used. But this implementation has one major drawback, that is, with slower lenses than f/2.8, one of the crossed sensors will become blind and virtually the cross sensor will be reverted back to linear! [...]

[JinDesu]

So this implementation means, k-5II users should be aware that with slower lenses, they should be looking for vertical lines for contrast, while faster lenses should look for horizontal lines.

[froeschle]

This could e.g. mean (not knowing all facts) that the central AF point

- would be more accurate for certain structures (arranged in parallel to one of the picture frames - shooting horizontally or vertically) with fast lenses (f2.8)
- could be insensitive to the same structures with slow lenses (!)
- would perform no better for structures, which are arranged in right angle to these structures - irrespective of the lens

Knowing the problems of a non-cross type sensor (from e.g. the Z-1P), this could be a strong argument against a K-5 II(s), if one often uses slow lenses and the central AF point.

[soalle]

Update: Only the center point is f/2.8 high precision. The high-precision sensor is vertical line-sensitive (see attachment). The center 9 points are sensitive to EV -3; the outer two horizontal line-sensitive points are excluded. See the Google translation of the Pentax K-5 II / K-5 IIs feature page. Note that it is not clear whether the horizontalvertical line-sensitive portion of the central AF point will work with slower lenses as well.

--DragonLord

A curiosity... do you know if the k30 has any f/2.8 precision point?

[Class A]

Knowing the problems of a non-cross type sensor (from e.g. the Z-1P), this could be a strong argument against a K-5 II(s), if one often uses slow lenses and the central AF point.

One needs to bear in mind, though, that the f/2.8 sensor should still work for f/4 lenses. There aren't that many lenses that are slower than this.

Also, given the relatively large size of the sensor areas, in practice they should be able to pick up some structure with contrast. Artificial patterns, as used on some focus charts, could be problematic but most real world targets should not.

Anyhow, a lot of speculation (as you also acknowledge) based on a marketing diagram which may or may not reflect the actual design. The AF areas are a lot larger than they are marked in the marketing illustration, for example.