Donzo98 wrote: j4nu wrote:
I might not be in the BIF crowd, but I just don't think this is how eyeAF is supposed to work on a modern Sony camera:
I actually did some test shots on 1.30, where I noticed an improvement. After upgrading to 1.31, unintentionally this time I noticed some regression pretty quickly...
How do you unintentionally upgrade your FW??
It’s not that simple…
Unfortunately the upgrade part was intentional . I didn't intend to do any test shots regarding eyeAF accuracy, I just noticed it when taking casual pictures after the upgrade...
---------------------------------------------
AGeoJO wrote: j4nu wrote:
I might not be in the BIF crowd, but I just don't think this is how eyeAF is supposed to work on a modern Sony camera:
I actually did some test shots on 1.30, where I noticed an improvement. After upgrading to 1.31, unintentionally this time I noticed some regression pretty quickly...
Jan, no question that both crops show front focusing. I have seen plenty of your images and I am not questioning your photography. But the face, let alone the eye is tiny relative to the entire frame on both. The AF can be tricked by something contrasty in the front of the target. Plus, the green square is just an overlay. While I could see that the camera may have a tough time trying to focus on the eye of flying birds coming at you at high speed at 10, 20 or even 30fps but the human eye is really not that challenging AF-wise. Out of curiosity, have you encountered any issues where the face is quite a bit larger, more in the sense of portraits or at least environmental portraits? Again, I am not questioning your photography and I am genuinely curious and I hope you don’t mind…
Thank you,
Joshua
Thanks, Joshua. You are of course right in what you're saying. I'm not expecting absolute 100% hit rate either. The only reason I mention this in this thread is that my impression is, that on 1.30 eyeAF was more resilient to such distractions. As I mentioned, I tested this scenario on 1.30 as I wanted to see what:
* Improves the accuracy of the Eye AF feature
actually means in practice and noticed an improvement. Then on my first casual shots after 1.31 update, I noticed eyeAF seemed to regress to its previous state but I didn't think much of it at that time. I don't know if there's a specifc size of eye which makes eyeAF work more reliably, but I think it's mostly about the contrasty objects near the eyeAF frame. Anyways, if I could directly test 1.30 against 1.31, I would .
Edit: An example where the eye occupies more of the frame:
And let's be clear, I'm not usually hunting for this but this time I played aroud some obstacles (a spoon in this case ) looking to trigger this effect.Normally I take a lot of slice-of-life kinda snaps of my kids so I just notice if eyeAF feels off. Like you guys do with BIF ...
Donzo98 wrote: j4nu wrote:
I might not be in the BIF crowd, but I just don't think this is how eyeAF is supposed to work on a modern Sony camera:
I actually did some test shots on 1.30, where I noticed an improvement. After upgrading to 1.31, unintentionally this time I noticed some regression pretty quickly...
How do you unintentionally upgrade your FW??
It’s not that simple…
Unfortunately the upgrade part was intentional . I didn't intend to do any test shots regarding eyeAF accuracy, I just noticed it when taking casual pictures after the upgrade...
---------------------------------------------
AGeoJO wrote: j4nu wrote:
I might not be in the BIF crowd, but I just don't think this is how eyeAF is supposed to work on a modern Sony camera:
I actually did some test shots on 1.30, where I noticed an improvement. After upgrading to 1.31, unintentionally this time I noticed some regression pretty quickly...
Jan, no question that both crops show front focusing. I have seen plenty of your images and I am not questioning your photography. But the face, let alone the eye is tiny relative to the entire frame on both. The AF can be tricked by something contrasty in the front of the target. Plus, the green square is just an overlay. While I could see that the camera may have a tough time trying to focus on the eye of flying birds coming at you at high speed at 10, 20 or even 30fps but the human eye is really not that challenging AF-wise. Out of curiosity, have you encountered any issues where the face is quite a bit larger, more in the sense of portraits or at least environmental portraits? Again, I am not questioning your photography and I am genuinely curious and I hope you don’t mind…
Thank you,
Joshua
Thanks, Joshua. You are of course right in what you're saying. I'm not expecting absolute 100% hit rate either. The only reason I mention this in this thread is that my impression is, that on 1.30 eyeAF was more resilient to such distractions. As I mentioned, I tested this scenario on 1.30 as I wanted to see what:
* Improves the accuracy of the Eye AF feature
actually means in practice and noticed an improvement. Then on my first casual shots after 1.31 update, I noticed eyeAF seemed to regress to its previous state but I didn't think much of it at that time. I don't know if there's a specifc size of eye which makes eyeAF work more reliably, but I think it's mostly about the contrasty objects near the eyeAF frame. Anyways, if I could directly test 1.30 against 1.31, I would .
Mar 22, 2023 at 12:36 AM
Previous versions of j4nu's message #16199498 « A1 to get firmware update in February? »
. I didn't intend to do any test shots regarding eyeAF accuracy, I just noticed it when taking casual pictures after the upgrade...
...