Comments on the Pentax SLR forum and various other forums indicate “front focus” problems with Pentax K-R cameras, especially with long wavelength (tungsten) illumination. The purposes of this message are to suggest a theoretical cause of the problem and a proper solution.
The focal length of a refractive optical system depends on the wavelength of illumination. Within the “visible” wavelength region (say from 400 to 700 nm), the focal length of a DSLR lens is essentially constant. Outside this range, for example in the “near infrared” from 700 to 1200 nm, the focal length will be different. Many remember the red line inscribed on lenses to help estimate the focal position shift for use with infrared film. If the wavelength sensitivity range of the phase detection autofocus (PDAF) sensors differs significantly from that of the CMOS image sensor, there will be a focus shift error dependent on illumination wavelength.
It is quite likely that the PDAF sensors on the Pentax K-R have photo-sensitivity extended into the near-infrared (beyond 700 nm) to enhance their low light-level sensitivity. Bare silicon is sensitive to light well into the near-infrared.
Therefore, the proposed theory is that the wavelength response of the Pentax K-R PDAF sensors are not properly matched to the response of the image sensor and this mismatch causes the front focus errors, especially in long wavelength (and near-infrared) illumination from tungsten light sources.
Enabling proper image focus is one of the most fundamentally important functions of a camera. A properly performing (and commercially viable) camera should enable focus accuracy within 1/3rd of the depth-of-field for a given lens and aperture stop. With proper wavelength matching of the PDAF and image sensors, this performance should be readily achieved.
Coping with Pentax K-R autofocus errors by cobbling together compromising firmware patches or requiring the use of Live View mode is not satisfactory. That would be like offering a superb F/1.4 lens but telling the customer to use it only at F/8 and higher.
If the theory is correct, the proper solution to the Pentax K-R front focus problem is to modify the wavelength dependent photo-sensitivity of the autofocus sensors so that they match the CMOS image sensor. This can be accomplished by placing the appropriate spectral filter on the reflex mirror, or the secondary mirror leading to the PDAF sensors, or on the PDAF sensors themselves. It also should be possible to implement this change by means of a field retro-fit. Hoya and Pentax are well known for their expertise in optical filters and coatings so are well-versed in this technology.
Achieving perfect focus should be easy using current technology. It is not complicated. The only tradeoff involves low light sensitivity, which is addressed with the focus-assist illumination. The cost for properly matching the wavelength sensitivities should be minimal if properly designed and manufactured.
Pentax K-R owners would be doing Pentax and Hoya a favor by insisting that this product design defect be corrected. I personally would be willing to pay a reasonable service charge (even though I have a 3 year warranty) to get the problem corrected. If Pentax does not correct the K-R front focus problem properly, I will have to get a different camera, brand uncertain.
Discussion of the front focus problem is notably absent from the recent detailed report on the Pentax K-R by dpreview. I recommend that dpreview add to their standard protocol a test for front focus errors under tungsten illumination.
The focal length of a refractive optical system depends on the wavelength of illumination. Within the “visible” wavelength region (say from 400 to 700 nm), the focal length of a DSLR lens is essentially constant. Outside this range, for example in the “near infrared” from 700 to 1200 nm, the focal length will be different. Many remember the red line inscribed on lenses to help estimate the focal position shift for use with infrared film. If the wavelength sensitivity range of the phase detection autofocus (PDAF) sensors differs significantly from that of the CMOS image sensor, there will be a focus shift error dependent on illumination wavelength.
It is quite likely that the PDAF sensors on the Pentax K-R have photo-sensitivity extended into the near-infrared (beyond 700 nm) to enhance their low light-level sensitivity. Bare silicon is sensitive to light well into the near-infrared.
Therefore, the proposed theory is that the wavelength response of the Pentax K-R PDAF sensors are not properly matched to the response of the image sensor and this mismatch causes the front focus errors, especially in long wavelength (and near-infrared) illumination from tungsten light sources.
Enabling proper image focus is one of the most fundamentally important functions of a camera. A properly performing (and commercially viable) camera should enable focus accuracy within 1/3rd of the depth-of-field for a given lens and aperture stop. With proper wavelength matching of the PDAF and image sensors, this performance should be readily achieved.
Coping with Pentax K-R autofocus errors by cobbling together compromising firmware patches or requiring the use of Live View mode is not satisfactory. That would be like offering a superb F/1.4 lens but telling the customer to use it only at F/8 and higher.
If the theory is correct, the proper solution to the Pentax K-R front focus problem is to modify the wavelength dependent photo-sensitivity of the autofocus sensors so that they match the CMOS image sensor. This can be accomplished by placing the appropriate spectral filter on the reflex mirror, or the secondary mirror leading to the PDAF sensors, or on the PDAF sensors themselves. It also should be possible to implement this change by means of a field retro-fit. Hoya and Pentax are well known for their expertise in optical filters and coatings so are well-versed in this technology.
Achieving perfect focus should be easy using current technology. It is not complicated. The only tradeoff involves low light sensitivity, which is addressed with the focus-assist illumination. The cost for properly matching the wavelength sensitivities should be minimal if properly designed and manufactured.
Pentax K-R owners would be doing Pentax and Hoya a favor by insisting that this product design defect be corrected. I personally would be willing to pay a reasonable service charge (even though I have a 3 year warranty) to get the problem corrected. If Pentax does not correct the K-R front focus problem properly, I will have to get a different camera, brand uncertain.
Discussion of the front focus problem is notably absent from the recent detailed report on the Pentax K-R by dpreview. I recommend that dpreview add to their standard protocol a test for front focus errors under tungsten illumination.
