p.4 #5 · Star Eater Algorithm and getting in contact with Sony
I have confirmation from reading a German sony forum with plenty of samples that this issue ocurred as far back as fw 3.00! Specifically spatial filtering at 4.0s
I have official word they are intending to fix the issue. Timeframe unknown but source is a friend at Sony who would know.
p.4 #6 · Star Eater Algorithm and getting in contact with Sony
It was back in October 2015 when some update made liveview noise in dark appear colored chunky mess versus fine grained noise. Also, the screen update was tiny bit slower. I think this was a serious downgrade in performance.
Then I took some milkyway photos in November and immediately noticed that why does the sky look a bit soft compared to D810 output using the same lens, maybe the lens had suffered some impact or adapter issue etc. and left it at that. Now it makes sense.
p.4 #7 · Star Eater Algorithm and getting in contact with Sony
thrice wrote:
I have confirmation from reading a German sony forum with plenty of samples that this issue ocurred as far back as fw 3.00! Specifically spatial filtering at 4.0s.
I think that spatial filtering (at 4.0s) came to the A7RII with firmware v.3.30, before that the limit for filtering was Bulb exposure only, and /but you had the chance to shoot without filtering in normal exposure mode (no Bulb) <30sec...
thrice wrote:
I have official word they are intending to fix the issue. Timeframe unknown but source is a friend at Sony who would know.
...This is very good to hear. I can wait a bit longer if official update is coming.
p.4 #8 · Star Eater Algorithm and getting in contact with Sony
MJKoski wrote:
It was back in October 2015 when some update made liveview noise in dark appear colored chunky mess versus fine grained noise. Also, the screen update was tiny bit slower. I think this was a serious downgrade in performance.
Then I took some milkyway photos in November and immediately noticed that why does the sky look a bit soft compared to D810 output using the same lens, maybe the lens had suffered some impact or adapter issue etc. and left it at that. Now it makes sense.
...And you were shooting in Bulb mode then?
I thought it was fairly well known fact that 4sec star eater filtering came with firmware v.3.30, and before that it was in Bulb mode only.
p.4 #10 · Star Eater Algorithm and getting in contact with Sony
aikaarska wrote:
I thought it was fairly well known fact that 4sec star eater filtering came with firmware v.3.30, and before that it was in Bulb mode only.
I'm back on 3.20 and it definitely occurs beyond 3.2s. Are you reciting what you've read online or what you've tested yourself?
Taken at 17th of February, 2016. A7R2, ISO100, 30 seconds, f/5.6. I always thought something is not right with that pixel level crop because it looks like someone smudged it with a layer of heavy handed noise reduction. This damn thing explains it all. That is downright cellphone quality at pixel level right there. You are looking at a pile of sand which now looks like some alien moon material.
p.4 #14 · Star Eater Algorithm and getting in contact with Sony
But, then Sony comes up with a proper fix with number x.y you better be sure to not use a number higher that that NOW Unless, you wanna modify all future FW fixes.
p.4 #17 · Star Eater Algorithm and getting in contact with Sony
MJKoski wrote:
Apparently fine grains of sand also qualify as stars.
I suspect you are right on this, but we really can't tell anything without seeing one from less than 4 seconds with that image.
I still cant really understand Sony's thinking on this at all... they develop a 42MP camera that has this amazing resolution capability, they develop all these lenses that are supposed to be able to deliver IQ into that camera and beyond presumably, and then they start chipping away at the actual performance of the camera through things like this.
I get that there will be sensor noise to address, and for mere snapshooters, it probably makes sense to apply noise filtering in the body, but to force that on the advanced users as well is really dumbfounding.
I've said in the past that they need to eliminate this from the Bulb setting along with their fix! Make it an electable option in the settings... 'NO, Yes-4 sec, Yes-Bulb' or something similar. For normal shooting, I might choose 'Yes-Bulb' to avoid having to pay special attention to an occasional longer exposure, but for astro, I'd choose 'NO'.
However, MJKoski, I'm not sure about that your crop above. I agree with mjm6, that we can't tell anything for sure without seeing one from less than 4 seconds with that same sample....
p.4 #19 · Star Eater Algorithm and getting in contact with Sony
Let's go shoot the same pile of sand. You live close enough
If it only was this one incident. But we have cRAW also in the history which makes one wonder what were they thinking, if anything at all. This is the very reason I suspect A9 design may contain some nasty shortcuts to achieve what it tries to achieve.
p.4 #20 · Star Eater Algorithm and getting in contact with Sony
BTW - I have also 36mp Pentax K-1 (and A7RII)... So I have the opportunity to make direct comparisons between these cameras, also with same lens. For example 3.2 sec and longers exposure comparisons...
