Skip to the end of navigation

Welcome to the Community!

Alpha - SLT & DSLR Cameras

Showing results for 
Search instead for 
Do you mean 
Reply
ADF Apprentice
Apprentice
Posts: 8
Registered: ‎04-18-2017
Message 71 of 111 (762 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

So this completely killing sharpness in all kinds of photos rather than just Astro... hmmmm

Specialist
Posts: 22
Registered: ‎01-19-2017
Message 72 of 111 (757 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

Yes, anything shot with exposures longer than 3.2 seconds - I don't think a lot of photographers realise that, although astrophotography is probably most affected.

 

-John

Specialist
Posts: 16
Registered: ‎04-17-2017
Message 73 of 111 (703 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

the effect on tree foliage is the most obvious. disappearance of reflections, loss of resolution + typical pattern of median filter ("worms") like here

 

i will see if i can conduct the same kind of tests in daylight

 

ebb9b3c11dacc03caff198b692113f9c.jpg

Enthusiast
Posts: 2
Registered: ‎05-02-2017
Message 74 of 111 (700 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

thanks for the examples guys. With the ND-Filter a part from the loss of detail could also be the result of the ND-Filter maybe, no?Question is, will it get even worse for a--lets say 3 minute exposure at daytime (e.g. for fine art B&W photography).

Specialist
Posts: 16
Registered: ‎04-17-2017
Message 75 of 111 (698 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

[ Edited ]

no, both images are taken with the same ND filter successively, so there should be not difference between those. any difference is from the fiddling of the raw for exposure >3.2s

 

there should be no difference between 4s, 10s, 3minutes, etc... the only difference is between  =<3.2s and >3.2s

 

 

 

 

Specialist
Posts: 22
Registered: ‎01-19-2017
Message 76 of 111 (672 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

[ Edited ]

Spoutnicko is absolutely correct on both counts.

 

First, the neutral density filter was used for both images, which were shot a few seconds apart on a still, cloud-free morning so the difference that you see here is solely a result of Sony's median filtering.  

 

Second, there is basically a threshold at 4 seconds at which the filtering kicks in with v. 3.3 of the firmware. All exposures of this length or longer have the median filtering applied, and that will affect all images, regardless of their length.

 

The dishonest aspect of this for me is that Sony, while calling these 'raw' files, are giving us files in which they apply in-camera modification to remove noise between the data coming off the sensor and it being written to the card.

 

I have no problem with that being applied to a jpeg, or possibly even a video file, but the bottom line for me is:

 

                                  Raw files should be raw!

Apprentice
Posts: 3
Registered: ‎04-18-2017
Message 77 of 111 (564 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

Here is another blog post write up on this issue. Keep this issue spreading around because this MUST be fixed. Email Sony here: SonyListens@am.sony.com 

 

http://www.lonelyspeck.com/why-i-no-longer-recommend-sony-cameras-for-astrophotography-an-open-lette...

 

 

 

Apprentice
Posts: 12
Registered: ‎02-27-2017
Message 78 of 111 (530 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

petapixel has also now run a story on this - it's a major issue, especially given the cost of these cameras... and it's completely self-imposed and fixable. Sony just needs to decide how long they want to keep hurting themselves but not addressing this:

https://petapixel.com/2017/05/04/star-eater-issue-no-longer-recommend-sony-cameras-astrophotography/
Apprentice
Posts: 8
Registered: ‎05-04-2017
Message 79 of 111 (452 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

This greatly affects fine detail in all photographs, not just astrophotography.

Review sites that compare the A7RII to other cameras should re-run the tests of the cameras at 4s in order for it to be an honest representation of what consumers can expect.

Sony should adjust their marketing which mentions the lack of AA filter to get finest details of those finest details are erased by their in camera algorithm. Quite dishonest in my opinion and grounds for return of the camera (at least in Australia).

Either that or they can just release a fixed firmware.
Enthusiast
Posts: 1
Registered: ‎05-04-2017
Message 80 of 111 (437 Views)

Re: star-eater algorithm in Bulb-mode shots - please offer alternative

Wow, Sony. First the whole lossy RAW debacle (I'm still waiting for lossless compression, which *every* other major camera maker offers) and now this. What's so hard to understand about the idea that RAW means the absolute minimum of in-camera processing; nothing more than — maybe — just a little massaging of the data to meet the published specs of your RAW format?

 

I can't even imagine the failure of both management and engineering that lead to the decision to apply aggressive noise reduction to RAW files, especially 3.2 second shots which should not suffer from significant sensor noise to start with. The examples shown here would be unacceptable even in bulb mode, for 30+ second exposures, but the idea to apply an aggressive median filter to sub-4-second RAW shots should have been shot down immediately and at every level, from the engineer who proposed it to the management chain that apparently decided it was a good idea. (Or perhaps the problem is that someone high enough in management decided they knew better and insisted on this "feature"? I don't expect we'll ever really know.)

 

I've recently invested quite the sum in an A7RII and lenses (including so far two GM series lenses — and I know you know how much those cost), switching from Nikon as I very much appreciate the compactness of the Alpha series and I thought Sony had at least mostly figured out how to make serious, pro-level camera gear. (Not just the sensors at their core.) I had fully intended to buy an A7RIII/A9R (depending on specs and what actually ends up being a real product) as a second body, but that's on hold now. Clearly my switch was premature, and I can only hope that Sony addresses this promptly with a firmware update, allowing experienced photographers to use the same post-processing techniques they always have to remove hot pixels without significantly impacting detail. (Maybe we could get a smart built-in dark-frame noise reduction option instead of this nonsense? Any maybe, pretty please, we can also finally have some lossless RAW compression to complement it???)

 

The only alternative to a prompt response from Sony would mean that I've wasted my money and will need to switch to another system from a manufacturer who actually respects their customers (*cough* Fuji? *cough*). Needless to say, I would not be looking back any time soon.

Get Social

Share your ideas Watch YouTube Support Videos follow us on Twitter Visit us on Facebook