[Ian Stuart Forsyth]

If a zoom lens needs different AF adjustments at different focal lengths

You also have to consider the distance from your subject as this can also require different AF adjustments. Sometimes there are profiles that manufactures use to correct for this, but it is a generic profile that is not specific to a lens and tolerances in lens production can change how AF varies over the range you are using it.

You can also have focus shift while using different ƒ stops so this is another variable you may also want to consider and select your correction for your most used settings.

With some lens manufactures one can profile your own lens for different working distances and zoom range via a dock. this can be very handy for many of the larger zoom and would recommend this if focus is critical for your use

[sebberry]

I'm a little surprised how complex it is. Why can't it just drive the lens until it's in focus?

I guess this is another area where mirrorless with on-sensor PDAF excels, eh?

[Ian Stuart Forsyth]

I guess this is another area where mirrorless with on-sensor PDAF excels, eh?

Not always if the ML does not stop down the lens while focusing it may not account for focus shift as the lens is stopped down and you will see some ML bodies with AF tuning

I'm a little surprised how complex it is. Why can't it just drive the lens until it's in focus?

One of the problems is that AF controls are not done at the image plane of the sensor but at the image plane of the AF sensor. This can introduce errors based on the lens used and how the lens elements move during zooming and focusing. Change what distance you are focusing and that changes how light enters the optics of the AF sensor, and every model of lens reacts differently

[licht96]

May be a new chance/update with the new "Select XS" of the new firmware 1.7
Will try it tomorrow.

[sobkrates]

May be a new chance/update with the new "Select XS" of the new firmware 1.7
Will try it tomorrow.

I'm very interested in the results, because I have the same problem with my 16-50 PLM.

[simon_199]

I tried the select XS mode, which raises a few questions. When using the select S mode, I always assumed (by visual inspection of the VF overlay) that only one focus point was being selected, and that was the maximum "granularity" at which the focusing sensor could operate. However, they apparently implemented a way to use an even smaller focusing area, so which parto of the above assumptions were wrong ? Is that what we see as a single AF points is actually made of smaller (non-selectable) sensitive areas that the firmware handles at a lower level?

Another question, I always assumed that "select S fixed on the center point" should work exactly the same as "Spot" mode. However this new update makes me wonder if the "Spot" mode was already of comparable accuracy to "select XS on center point", therefore different from "Select S on center point". Could it be the case ?

[Kobie]

I tried the select XS mode, which raises a few questions. When using the select S mode, I always assumed (by visual inspection of the VF overlay) that only one focus point was being selected, and that was the maximum "granularity" at which the focusing sensor could operate. However, they apparently implemented a way to use an even smaller focusing area, so which parto of the above assumptions were wrong ? Is that what we see as a single AF points is actually made of smaller (non-selectable) sensitive areas that the firmware handles at a lower level?

Another question, I always assumed that "select S fixed on the center point" should work exactly the same as "Spot" mode. However this new update makes me wonder if the "Spot" mode was already of comparable accuracy to "select XS on center point", therefore different from "Select S on center point". Could it be the case ?

Spot is always the most "narrow" AF point. SEL xs seems to be the same as spot but you can move it around.

[StigVidar]

Another question, I always assumed that "select S fixed on the center point" should work exactly the same as "Spot" mode.

I recommend reading the user manual. Instead of assuming

[simon_199]

I recommend reading the user manual. Instead of assuming

The user manual does not specify that spot is narrower than select S when set to the center point. The manual (I'm reading it right now) states that select S uses "fewer focus points than [select]". Select uses 5 points, as specified in the manual as well as going for what is visible in the viewfinder overlay. Since five in the manual corresponds to five in the overlay, and select S only highlights one single AF point in the overlay...it's rather natural to assume that the description is just vague (not unlike many other sentences in the manual) and that "fewer points" actually means one, as indicated by the overlay. There is no indication anywhere else, before the introduction of this select XS mode, that there is some finer granularity beyond the grid of user-selectable points.

The scheme available on Ricoh websites does not suggest that the center focus point has anything special:

Google Image Result for https://www.ricoh-imaging.co.jp/english/products/k-3-3/feature/img/03-af-img-01.png

The little square is a bit smaller, in the overlay, in Spot mode, but I would not necessarily give it any practical meaning it could just be a visual cue to indicate the selected mode.

[StigVidar]

The user manual does not specify that spot is narrower than select S when set to the center point. The manual (I'm reading it right now) states that select S uses "fewer focus points than [select]". Select uses 5 points, as specified in the manual as well as going for what is visible in the viewfinder overlay. Since five in the manual corresponds to five in the overlay, and select S only highlights one single AF point in the overlay...it's rather natural to assume that the description is just vague (not unlike many other sentences in the manual) and that "fewer points" actually means one, as indicated by the overlay. There is no indication anywhere else, before the introduction of this select XS mode, that there is some finer granularity beyond the grid of user-selectable points.

The scheme available on Ricoh websites does not suggest that the center focus point has anything special:

Google Image Result for https://www.ricoh-imaging.co.jp/english/products/k-3-3/feature/img/03-af-img-01.png

The little square is a bit smaller, in the overlay, in Spot mode, but I would not necessarily give it any practical meaning it could just be a visual cue to indicate the selected mode.

I have to apologize. I cannot find it in the manual either.

But it is mentioned here: How to optimize focusing accuracy with large-aperture lenses / explore | RICOH IMAGING

Highlight: The advantage of Spot setting in AF mode
The Spot setting is an autofocus mode that detects the in-focus point using a single AF sensor positioned at the center of the image field.
For PENTAX digital SLR cameras marketed after the PENTAX K-3, the Spot setting detects an area approximately 30 percent narrower than that of the Select setting, which uses a set of middle AF sensors in autofocus operation.

[simon_199]

I have to apologize. I cannot find it in the manual either.

But it is mentioned here: How to optimize focusing accuracy with large-aperture lenses / explore | RICOH IMAGING

The last link is interesting, the wording is peculiar as it is stated that the select setting uses a number of "middle AF points in operation"...as if there is really some sort of denser array of points that the camera can utilize at a low level.


(Or maybe it's a software thing, the points are always the same but for select mode they use several and apply some sort of weighted average with a radius that is not zero even for sel S that apparently, visually select a single point...)

Guess we'll never know for sure. The new mode suggests that there indeed could be a difference between spot and sel S, and that Sel XS is a sort of movable Spot mode. If so, I expect it could make a difference in terms of accuracy. Difficult to test scientifically though.....

[sebberry]

My initial impression after a couple of days with the DA* 50 1.4 is that using the wider focus points are producing better results. I wonder if the 16-50 PLM is any better in this regard than the original 16-50? I'd like a faster focusing lens for the new family puppy and 50 on APS-C is a little tight for indoor action shots of an energetic puppy.

[licht96]

Ha... simelar here... but really exactly the same combo as you, seberry. The DFA 50 1.4 has a better hit rate now ( I think it is better)
But the 16-50 PLM is still not really better at 2.8. Although the focus is not so tight it has still lots of front and back focus inconsistent over the area.
I will still have to use only the center.
May be it's better in some "through a branch" focusing... but not for my use, to but a person eye NOT in the center of the picture.
For that I still use recompose ore Live view. ***
By the way: it would be interesting if now the XS center is more or less the SPOT focus.
It (Spot) was always mentioned to be better/more thin than the Select center. (as mentioned here or in a parallel thread)
That could be one atvantages for me at least.

***by the way 2: I did some smal testing at 1.4 in live view.... and the face detect is really more or less an eye detect (focusing) ... sure, the whole face is highlighted... but even person's with long nose, have a sharp eye an a blurred nose tip ;-)
So quite nice to use at this aperture
Normally I use the smallest AF area in Live view...but that works quite good, too