p.4 #1 · Sony's star-eater algorithm not fixed by f/w 4.0
JimKasson wrote:
Got it now. I don't think the way to evaluate images for this problem is to look at JPEGs, where the DCT may be doing some of the things the hot-pixel suppression algorithm does. Better to look at raw planes in TIFF form, IMO.
Jim
Yes, I agree with your opinion. I don't have an A7r II (just an A7 II) to replicate the tests of Ari... I need to evaluate if the A7 II is also affected by this, although for the moment I'm still in firmware 3.20. I need to find some time after work.
p.4 #5 · Sony's star-eater algorithm not fixed by f/w 4.0
My A7R II has FW Ver 3.1 Should I never upgrade that to avoid star eater? Or are there more useful items that make it worth upgrading despite star eater?
Like 100-400 GM profiles or 12-24 anti decentering (I wish) setting?
p.4 #11 · Sony's star-eater algorithm not fixed by f/w 4.0
It would be good to have a definitive answer as to when star eater started. The OP said 3.3. But if it started at 2, then I should go ahead and update.
p.4 #12 · Sony's star-eater algorithm not fixed by f/w 4.0
billsnature wrote:
It would be good to have a definitive answer as to when star eater started. The OP said 3.3. But if it started at 2, then I should go ahead and update.
Anyone know with certainty and accuracy?
If you go to Jim Kasson's site (here) you will find a conclusive demonstration that what had only been applied on bulb exposures up until 3.2 become unavoidable on all exposures longer than 3.2 seconds with v. 3.3 of the firmware, and has been ever since, including in V. 4.0.
p.4 #13 · Sony's star-eater algorithm not fixed by f/w 4.0
billsnature wrote:
It would be good to have a definitive answer as to when star eater started. The OP said 3.3. But if it started at 2, then I should go ahead and update.
Anyone know with certainty and accuracy?
My own testing indicates that the "star eater" spatial filtering that kicks in when the exposure time is 4 seconds or longer is present in firmware revision 2.0:
navmannz wrote:
If you go to Jim Kasson's site (here) you will find a conclusive demonstration that what had only been applied on bulb exposures up until 3.2 become unavoidable on all exposures longer than 3.2 seconds with v. 3.3 of the firmware, and has been ever since, including in V. 4.0.
Hmmm... I definitely seem to be getting the effect and I'm relatively sure I'm still on FW 2.0 (unless I updated the camera and forgot about it, which seems unlikely given how much of an ordeal FW updates on Sonys are). I'll double check the FW tonight, and send Jim some dark frames to analyse!
It's possible that they cranked the NR up even further in 3.3...
setting?