p.8 #1 · Star Eater Algorithm and getting in contact with Sony
Schlotkins wrote:
I really hope Sony fixes this... I have a couple A7r's and I'd love to upgrade to the A7rII. But I shot a lot of astro and so will hold off until then.
I can't speak to astrophotography, but I do a lot of low-light, long-exposure nighttime photography and to my surprise, when I compared the A7R and A7RII back-to-back, I found the A7R was actually a better performer overall for this type of photography than the A7RII. And that was with the A7RII's original firmware, too!
I really, really wanted to upgrade to the A7RII, but for my purposes, the "upgrade" was, in reality, a downgrade in disguise. <sigh>
p.8 #2 · Star Eater Algorithm and getting in contact with Sony
Schlotkins wrote:
I really hope Sony fixes this... I have a couple A7r's and I'd love to upgrade to the A7rII. But I shot a lot of astro and so will hold off until then.
Sounds like a good plan at the moment (until Sony fix it with new firmware). Basicly you can get far better Astro images with A7R at 4sec-30sec exposures (not BULB) than with A7RII. Sad but true. I have A7RII, but also Pentax K-1 with same 36mp sony sensor than A7R. Pentax wins hands down. Pentax hasn't anything like star eater NR algorithm, not even in BULB mode...
Hope the “Star Eater” has been fixed in the new version, instead of worsening the situation - “Star Eater” kicks in when exposure time is longer than 2 seconds…
Hope the “Star Eater” has been fixed in the new version, instead of worsening the situation - “Star Eater” kicks in when exposure time is longer than 2 seconds…
Yeah! I already installed (without problems), and I'm very happy to tell: star eater is fixed! At least 4sec-30sec exposures (at least A7RII)... I'm not sure about bulb yet, I have to try it more carefully ...
p.8 #5 · Star Eater Algorithm and getting in contact with Sony
Someone posted on the SAR comments:
---
According to Sony Japan, A7RII will get "長秒時NR設定がOFF時の画質の向上", which means a long exposure IQ improvement with NR set to off. I wonder if they fixed the 12bit RAW problem with bulb
p.8 #7 · Star Eater Algorithm and getting in contact with Sony
Yes, I just tried it very quickly - its one night of a full moon here, but the 3.2 and 4 second exposures I just took of the Milky War are identical apart from a very slight difference in star trailing. A big thanks to everyone who has contributed to raising enough of a collective din to get Sony to finally act...
p.8 #8 · Star Eater Algorithm and getting in contact with Sony
Cheers! It took ~two years to partially fix it But at the same time they admitted that the way it was handled was not good (and thus the filtering bad) as no option was given to enable/disable it. It seems hard for any Japanese company to admit something went wrong. This goes for any big camera company there excluding Fuji perhaps.
p.8 #10 · Star Eater Algorithm and getting in contact with Sony
Jesus man... What do you expect? A written apology?
They went ahead and fixed the issues. I have nothing to complain from Sony and their feedback. It just goes to show they are trying to fix what we ask as per uncompressed RAW option they did few months ago.
IMO Sony is not behind Fuji in terms of FW updates. Remember that Fuji had to add features to be on feature parity with Sony bodies.
p.8 #11 · Star Eater Algorithm and getting in contact with Sony
aikaarska wrote:
Yeah! I already installed (without problems), and I'm very happy to tell: star eater is fixed! At least 4sec-30sec exposures (at least A7RII)... I'm not sure about bulb yet, I have to try it more carefully ...
Anyway, this was a great news ;-)
good news...gotta do that tonight before I head out for the weekend!
p.8 #15 · Star Eater Algorithm and getting in contact with Sony
I just did four quick comparison shots all at ISO 6400 with a lens cap on, all ten seconds long, two with Bulb and two with M, and each pair comparing with LENR turned on and off. After ramping up the exposure by five stops in LR and enlarging to 3:1 it appears that both the bulb and M shots taken without LENR show a similar incidence of random 'bright' pixels, and that the two taken with LENR turned on have obvious suppression of these. I'd like to test at night in the real world, but my impression is that spatial filtering may have also been turned off in Bulb mode as well.
p.8 #16 · Star Eater Algorithm and getting in contact with Sony
Hmmm I'm seeing something interesting.
I created a matrix of single white pixels on a black backdrop. I then focused on these with the 21 loxia at f/5.6 and shot at all the offending settings.
You can see the dim patches in every exposure over 3.2s. Is Spatial filtering really fixed?
p.8 #18 · Star Eater Algorithm and getting in contact with Sony
GMPhotography wrote:
Would that not be lens on remapping the sensor to kill noise
Hi Guy,
Yes, with Long Exposure Noise Reduction on, we expect sensor noise to be mapped out with a darkframe (won't affect genuine bright pixels in the image). However with it off, I am still seeing spatial filtering mistaking bright pixels for hot pixels and trying to punch them out.
Going to do a hard reset (throwing away all my painstaking efforts to avoid the Sony menus through custom buttons) and re-run the test.
p.8 #20 · Star Eater Algorithm and getting in contact with Sony
Last year Pentax introduced spatial filtering to reduce white dot problem in long exposures (some K-1 units had uncalibrated sensor like D810) but it was an option which only affected JPG files. So was this long exposure FW fix some JPG enhancement after all? It says nowhere that the fix was related specially to spatial filtering or "star eater". What is really dubious is that there is still no switch to turn it on/off like it should be.
