One thought on the low light: while poor illumination and contrast are likely the main issue for AF, if the camera is using DSS/SENS UP/other slow-shutter tricks, that could confuse the AF as well. I have an old analog PTZ on my bench that goes up 64X SENS UP and while it looks great at night with no motion, even a person walking across the field of view leaves a huge streak... and accordingly, and attempts to AF cause major blurring, which prevents the camera from focusing.
Testing Autofocus Accuracy
************************.
** ***** *** **** *** ** getting ******* ***** ********, ********** ********* for **** ** ***** *******. *******, some ******** *** ******** ** ************ versus ****** ***** * ***** *** carefully ********* *** ****** **** *****.
Autofocus ***********
** ****** * ******* ** ******** scenes, ******** ******* ** **** ***** (~160 ***), **** ** *** **** (~2 ***), *** **** ******* ** the *******. *** ******* **** *** testing ****:
- ******** *.*-**-** (****)
- **** ***** (*.***)
- ***** ***-**** (****)
- ***** ***-**** (*****)
- ******* ***-**** (****)
- **** ***-***** (****)
- ********* *** ********* ******** ** **** lit ******, ********** ** **** *** in *** *** ** *** ****, indoors ** ***. **** *** ****** had ****** ******** ******* * ***** wall, *** **** ***-*****.
- *********** * **** ***** ** *** middle ** *** *** ****** *** Sony ******'* ***** ******. ***** ******* were *********.
- ** *** ***** (~* ***) ** an ***** *****, ******** ******* *** severe ***** ******. ****** * ***** improved ********* ***********, ******** *** ******* to ***** **********. ***** **** ******* automatically ******* **** ********* ** ********** mode, ***** ****** ** ******* ** double ***** ***** ** *****.
- ********** ** *** ******* ****** *** not ****** *** ******* "*******" *** "focus" ** ** *****, *** * camera **** ****** ********** **** ** easier ** ******* ******* ** *** ** ***** *** in ***** **** *** ***** ***.
- ****** ***** *** ** ************ ********* to ********* *** ** **** *********** in ********* ******* ******.
- **** ***** ******* ***** ************* ******* after ********* **** *** ** ***** mode, ***** *********** **** *** ****** ******** ***********. This ** ** *** *** ******** risk **** ***** ********* *******, *** should ** ************ *******.
- ********* ** * ******** ********, *** as ****, ****** ****** ******* ************* and **** ****** ****** * ****** line. ** **** ********************** ************ ****** *************, ***** ***** in ******** *******, ** ***** ****** expect ***** *********** *** ****.
**** *** *** *** ******** **** this ****:
***************
** *******, ********* ******** **** *** may ** *******, *********** ****** *********** time ** ************. ** **** ***** this ** ** **** ********** ** what ******* *** ** *** ***** (subjects, **** ******, ***.).
*******, ***** *** *** *** ******* to ****:
Test ***********
******* ***** ****
*****, ** ******* ******* ** *** conference **** ***** ** ** **** light (~*** ***) **** ******* ******* at * ***** ****. ***** ** a ******* *** ***** ** *** scene.
*** **** ***-***** *** *** **** camera **** *** *** ***** ***** properly, *** **** *** ** **** by ******* ** *** ***** ** the ***** **** ** *** *****.
**** **** ***** ** *****
****, ** ****** ********* **** ** eye ***** ** *** ****** ** the *** ** *** ** ********* ** ** ****** ** *** center ** *** *** ******* ***** performance.
**** *** ***** ******** ** *** FOV, ** ******* *** *** ******* focusing, ********* *** **** *****, ***** had ********** **** *** ***** ****.
Indoor ****
** *** **** ***** ** *** full ***** ****** *******, ** ****** the ******* ** * ***-***** (~* lux) ***********. *** ******* **** *** to ***** ****, **** * */** exposure. ** *** *** ******* *** cameras *** ***** ***********, **** *** well **** *** ******** ** * low ***** *****.
***** ****
**** *** *** ******* ***** *** camera ************, **** ******* ** *** FOV.
*** ******** *.*-**-**, ***** ***-****, *** ******* SNB-5004 ******* **** *** ***** ********** in **** *****, ****** ** ******* a ******, ******* *****, *** ******** multiple ****** ******** ******** ** *** the ******* *****. ***** ** ** image ******* ***** ***** ******* ***** autofocusing.
**** ***** ** *****
**** *** ***** ** *** *****, all ******* **** **** ** ******* a *****, ******* *****.
*******
** ****** ** ** ******* ******* lot ** *** **** ******* * more ******* ***** **** * **** field ** **** ***, **** ******** objects ** *** ********** *** **********, including ****** ****.
** ******* **** ** ****** ** defined ***** ******, *** ****** *** the ******* ********* ** *** *****. The *** ** *** ***** ** presented *****.
*** ******* **** ********* **** ** focus ********** ** **** *****. *******, we then **** *** **** ***** **** the **** ****** ** **** **** inside. ***** ** *** ***.
******* ***** ****** ** *** ***** did *** ****** *** *******' ********* performance ** *** ***. *******, ***** inspection ** *** *** ***** ******* the ******* ** ********* ** *******' apparent *****.
*** *******, ****** *** ***** ** the ******* ***** *** ** ***** optimal ***** *****, ******* ****** ** sharpness **** *** ***** ****** **** different, ** **** ** **** *** softest ***** *** **** *** ** focus ** ****. ***** ****** ** careful ** ******* *** **** **** focusing *** ********* *** ********** ******* soft *** *** ** ***** *******.
Comments (2)
MI
Matt Ion
Dec 18, 2013JH
John Honovich
Dec 27, 2013We test with shutter at 1/30s so that would not impact our test runs. However, I agree that it could be for those using slow shutter presuming there's no prominent fixed objects to focus on.
I do wonder if the camera manufacturer would not automatically adjust the shutter while refocusing to avoid this potential problem and then return it to whatever slow shutter afterwards. Of course, if the image becomes pitch black with the slow shutter temporarily turned off, then that's an equally problematic issue for focusing....