95 replies to this topic

[t_image]

I'd love to see people with the same analytical zeal and energy focus on studying how to effectively lobby a manufacturer about an issue that isn't the most popular issue for complaints.

For example, how many members of this:

https://esupport.son...info/1523/US/EN

(for down under):

https://prosupportpr...sonyanz.com/faq

are there out there also on the sony support community forum or DPR or vlogging about this issue?

I've seen some popular camera vloggers complain that they weren't invited to the a9 press release.

Interesting though Sony does say members (above) has its privileges:

"Dedicated members only phone number
 Priority support with digital imaging specialists"

"Invitation to special events
Evaluation Loaners (try before you buy)"

 

For example, maybe someone could find historical fingerprints of the discovery, complaints and then the final change that happened with the Nikon star-eater issue. Was there something immediately detectable like a groundswell of internet complaints?

 

How many popular vloggers/ twitter/ instagram/ facebook/ [social media] leaders in photography or the Sony fanboy world are mentioning about the complaints of star-eater????

How about the same analysis of how it came about Sony introduced a touch screen to the a6500 or more customize-able buttons to the a7x m2 line?

 

How many people at trade shows are winning the ear of the engineers (not the evangelists) that are often there from Sony?

 

From a cursory search and glance about who is talking about the "star-eater" or spatial filtering algorithm,

[and who isn't],

I'm not too surprised Sony isn't immediately responding to this issue (adding a menu on/off feature) since there are many other issue requests that are often in the conversation of the populous, and the ones that seem to more likely have Sony's ear.....

the ones with lots of social media followers and not just a wordpress website blog......

It's frustrating to me to watch youtube videos of individuals that have gotten access to Sony's people in more than just a support email exchange

 

I mean its great there have been people that spent quite a bit of energy fleshing out this issue analytically,

but now where is the same skilled lobbying approach?

 

I get the CN lobbying may help inform the uninformed of some limitations to AP,

but do you think a corporation that supplies sensors to dedicated AP cameras is really going to try and satisfy a small niche community by making a consumer camera line that could completely rival or replace such after a few lonely voices calling in a crowded room of other conversations?

 

I'd love to see Exxon build a gasoline powered automobile that gets 100 miles to the gallon. And don't tell me the tech isn't there.

Edited by t_image, 30 April 2017 - 10:43 PM.

[sharkmelley]

Let me put on my dispassionate non-astrophotography hat for a while. Ironically, I think the reason star eater is so much worse on the Mk2 versions of the A7R and A7S is because Sony did respond to user complaints by introducing the infamous firmware fix.

 

The complaints against the Mk2 versions of the A7x cameras were that there was too much colour noise in images with a few seconds of exposure.  This might be because those cameras have in-camera stabilization (IBIS) which means it is far more difficult to conduct heat way from the sensor.  Although the Sony cameras all have internal hot pixel mapping to remove hot pixels from every image, the warm pixels (which are unmapped) remain in the image and those warm pixels were more obvious than on the Mk1 versions.  So in response to complaints about this perceived increase in colour noise, Sony introduced the infamous firmware 3.30 fix that introduced spatial filtering to deal with these warm pixels in all exposures of 4 seconds and above instead of just in bulb mode.

 

There is no way that Sony will ever remove this spatial filtering because it serves an obvious useful purpose.  If instead they introduce a menu setting to switch off spatial filtering  they would get complaints from less knowledgeable customers who accidentally switch it off.  They would want to know why Sony has a introduced a menu setting that gives them colour noise.

 

There is only one long-term solution that I can see:  Sony should do similar to Nikon.  Nikon upgraded their star eating hot pixel suppression (HPS) algorithm from a very basic spatial filtering algorithm (the same one Sony is using) to a more sophisticated version that removed single hot/warm pixels but left groups of hot/warm pixels untouched.  So whereas any star fully contained within a 2x2 pixel area would previously be deleted (just like Sony is presently doing), the improved HPS algorithm now leaves them more or less untouched.

 

For those technically interested, here is a discussion on the more sophisticated Nikon HPS algorithm:

https://www.dpreview...s/post/37071846

 

Now the question is, how to persuade Sony that implementing a more advanced HPS algorithm makes good business sense.  How to effectively lobby Sony to do this, I've no idea.

 

Mark

Edited by sharkmelley, 01 May 2017 - 01:15 AM.

[mmalik]

Ironically, I think the reason star eater is so much worse on the Mk2 versions of the A7R and A7S is because Sony did respond to user complaints by introducing the infamous firmware fix.

Agree; sometimes I think we shouldn't have complained too much about it (star eating I mean) in first place. a7S was the best and last thing that happened to astronomy and I feel we ruined it for ourselves in two ways, a) Sony doing a half hearted fix for the problem, b) Sony introducing IBIS, a feature not needed in astronomy or low light imaging [I even had to come up with an IBIS version of the 'Cold finger Inside...', an unnecessary burden].

 

 

Mark knows, but for everyone's sake I'll reiterate once again. Sony a7S with its large pixels was never intended for high resolution wide angle photography; folks who did it were doing it wrong in my opinion. There were R versions for that. a7S shines with long focal length narrow angle photography if folks will once for all understand it. It is misappropriation of resources in my opinion between S and R versions that has contributed to the confused state of mind with the users and the Sony alike. Regards

Edited by mmalik, 01 May 2017 - 02:05 PM.

[sharkmelley]

 

Ironically, I think the reason star eater is so much worse on the Mk2 versions of the A7R and A7S is because Sony did respond to user complaints by introducing the infamous firmware fix.

Agree; sometimes I think we shouldn't have complained too much about it (star eating I mean) in first place. a7S was the best and last thing that happened to astronomy and I feel we ruined it for ourselves in two ways, a) Sony doing a half hearted fix for the problem,

 

 

I'm not sure I agree.  If Sony had listened to us concerning the star-eating issue then maybe they wouldn't have made a half-hearted fix to the colour noise problem of the IBIS (the Mk2) cameras i.e. by imposing crude spatial filtering on all exposures 4sec and greater instead of bulb mode only.

 

 

Mark knows, but for everyone's sake I'll reiterate once again. Sony a7S with its large pixels was never intended for high resolution wide angle photography; folks who did it were doing it wrong in my opinion. There were R versions for that.

I'm not sure I understand your point - the R version suffers from the same star eating issue as the S version. It just becomes more obvious at a slightly different focal length. Both versions will suffer from spatial filtering when used with top quality lenses/scopes. 

 

It's also crazy to suggest that Sony is selling a camera that they never intended to be used with standard wide angle lenses. 

 

 

a7S shines with long focal length narrow angle photography if folks will once for all understand it.

Only because long focal length narrow angle photography is less affected by the star eater issue since the stars are well sampled and do not fit inside the 2x2 box that is deleted by Sony spatial filtering.  It is an happy accident that the A7S still works OK with long focal lengths.

 

 

It is misappropriation of resources in my opinion between S and R versions that has contributed to the confused state of mind with the users and the Sony alike.

I completely disagree.  Both versions suffer from star eater - there is no real difference between the two versions in that regard.  As I said above, neither version is suited for high resolution astrophotography with top quality lenses/scopes.  Sony has now made this far worse with the firmware "fix" for the Mark 2 (i.e. IBIS) versions.

 

I don't have a particular grudge against Sony.  I'm realistic enough to realise that using a consumer camera (Sony, Nikon or Canon) for a very specialist niche such as stacked long exposure astrophotography will inevitably hit the compromises that the manufacturer has built in for the consumer market.  It's just that Sony has 2 crucial compromises that Canon and Nikon do not have:

• the "star eater" spatial filtering (bulb mode for Mk1 models and 4sec for Mk2 models)
• digital scaling leading to potential coloured concentric banding issues

If I had known about these issues before buying the A7S as an astro-camera I would never have bought it.  I spent hard-earned cash on a camera that unknown to me had serious failings for astrophotography.  However, it's a risk I took because this information was not available 2 years ago when I bought my camera. Therefore I have tried to let potential buyers know in advance so they can make an informed decision.

 

I feel really sorry for those who bought the Mk 2 versions having understood the limitations but decided they can live with 30sec exposures (to work around the star eater issue) and high ISOs (to work around the concentric banding issue).  They now find that a Sony firmware update has made it impossible in any practical sense to avoid star-eating behaviour.

 

Firmware updates are an 2-edged sword.  They can give extra functionality or they can remove functionality.

 

Mark

Edited by sharkmelley, 02 May 2017 - 12:01 AM.

[bwallan]

Gees, after a 4-1/2 hiatus we're back to complaining about the so called "star-eater" problem.  Can't we find anything better to complain about?

 

Just use the cameras with lots of subs and get on with imaging!

[t_image]

So I just ran across this video:

https://www.youtube....h?v=6KnWD6yK2Yk

(Sony S3CA)---an a7sII camera in the form factor of something that might fit on hyperstar.

 

a7sx users may find this conversation from an interviewer and a Sony rep interesting.

Of course the interviewer is all about video not stills, but notice what the Sony guy says.....

 

Spoiler: doesn't sound like Sony even thinks about AP. But it does think video with the a7sx.....

[bwallan]

So I just ran across this video:

https://www.youtube....h?v=6KnWD6yK2Yk

(Sony S3CA)---an a7sII camera in the form factor of something that might fit on hyperstar.

...

Thanks for posting this link.  I'm not really interested in the camera BUT would really like access to the software used to run it if it was applicable to the standard A7 line of cameras.

 

bwa

[sharkmelley]

On Brian Smith's site I have seen the first indication that Sony is working on some kind of fix:

 

http://briansmith.co...re-update-3-30/

 

In the comments section at the bottom:

May 4, 2017

"This is an example of an update that fixed a problem for one group had the undesired side-effect on another group. Photographers who long-exposure daylight exposure of several minutes in duration requested built-in long exposure noise reduction. The update solved that but ate a few stars in the process. Sony engineers are aware of the issue and working hard to find a solution that makes both groups of users happy."

 

Mark

[StuartJPP]

On Brian Smith's site I have seen the first indication that Sony is working on some kind of fix:

 

http://briansmith.co...re-update-3-30/

 

In the comments section at the bottom:

May 4, 2017

"This is an example of an update that fixed a problem for one group had the undesired side-effect on another group. Photographers who long-exposure daylight exposure of several minutes in duration requested built-in long exposure noise reduction. The update solved that but ate a few stars in the process. Sony engineers are aware of the issue and working hard to find a solution that makes both groups of users happy."

 

Mark

To make both groups of users happy, Sony should put a menu option in and let the user choose...simple. The hardest part would be coming up with a name for it and updating the documentations, not actually doing it.

[t_image]

The fact that Ian Norman now has also contacted Sony about the "star eater" issue is a positive thing since it seems he may have been a registered "Pro User" by the like of feedback he received from Sony.

http://www.lonelyspe...letter-to-sony/

 

Although it has taken a while, getting enough chatter and momentum from the astro community about this issue may get the Sony engineers working on a creative solution....

[sharkmelley]

 

On Brian Smith's site I have seen the first indication that Sony is working on some kind of fix:

 

http://briansmith.co...re-update-3-30/

 

In the comments section at the bottom:

May 4, 2017

"This is an example of an update that fixed a problem for one group had the undesired side-effect on another group. Photographers who long-exposure daylight exposure of several minutes in duration requested built-in long exposure noise reduction. The update solved that but ate a few stars in the process. Sony engineers are aware of the issue and working hard to find a solution that makes both groups of users happy."

 

Mark

To make both groups of users happy, Sony should put a menu option in and let the user choose...simple. The hardest part would be coming up with a name for it and updating the documentations, not actually doing it.

 

Or they could take the Nikon approach - use a more sophisticated Hot Pixel Suppression algorithm which removes single hot pixels but leaves stars intact.

 

Mark

[navmannz]

I rang the Sony help line here in NZ about it last week and sent them the link to my blog post that documents the problem (https://johnleathwic...trophotography/) - two days later I was asked to send the raw files that I show in the blog, which I did, and since then have have had a couple of queries asking me to confirm details about the lens, serial numbers, accessories, etc.. I certainly got the impression that they are taking it seriously. 

 

-John

[sharkmelley]

I rang the Sony help line here in NZ about it last week and sent them the link to my blog post that documents the problem (https://johnleathwic...trophotography/) - two days later I was asked to send the raw files that I show in the blog, which I did, and since then have have had a couple of queries asking me to confirm details about the lens, serial numbers, accessories, etc.. I certainly got the impression that they are taking it seriously. 

 

-John

Looking at the recent comments on http://www.lonelyspe...letter-to-sony/ it does actually appear that Sony is taking the problem seriously.  Certainly on the Mk2 versions, at least.

 

If they also fix bulb mode on the Sony A7S and the other Mk1 versions then that would be really cool!   The Sony A7S could finally end up being the best (reasonably priced) camera for astrophotography.

 

Mark

[bwallan]

Looking at the recent comments on http://www.lonelyspe...letter-to-sony/ it does actually appear that Sony is taking the problem seriously.  Certainly on the Mk2 versions, at least.

 

If they also fix bulb mode on the Sony A7S and the other Mk1 versions then that would be really cool!   The Sony A7S could finally end up being the best (reasonably priced) camera for astrophotography.

 

Mark

I may never have encountered the problem but if they fix before I ever do run into it, I'm a happy puppy!  But I suspect we'll only see a fix for the Mk2 cameras, although I'd love to see them put it into an A7S firmware upgrade...

 

bwa

[whwang]

I remember that Sony had already introduced a few features/corrections via firmware updates to its Mk2 cameras, but had ignored the Mk1 in these updates. I will not hold my breath for the correction of star eater problem on Mk1.  I am ready to sell my modified A7R in a price that I wouldn't believe 2 years ago.

 

At this moment, the best and yet reasonably priced camera for astrophotography is a used D800 for modification. In Japan, it is now possible to buy a very good used D800 with less than $1100 (no sale tax for foreigners). If you have a friend traveling to Japan recently, ask him/her to buy one for you.

[bwallan]

I remember that Sony had already introduced a few features/corrections via firmware updates to its Mk2 cameras, but had ignored the Mk1 in these updates. I will not hold my breath for the correction of star eater problem on Mk1.  I am ready to sell my modified A7R in a price that I wouldn't believe 2 years ago.

 

At this moment, the best and yet reasonably priced camera for astrophotography is a used D800 for modification. In Japan, it is now possible to buy a very good used D800 with less than $1100 (no sale tax for foreigners). If you have a friend traveling to Japan recently, ask him/her to buy one for you.

I think you're throwing away one of the best cameras available for astro-imaging simply because you've somehow convinced yourself there is a serious problem.  It is your choice but I'll definitely continue using my A7S and A7R II!

 

But if you're going the Nikon route I would suggest the D810a instead of the D800.

 

bwa

[whwang]

Take a look at the pictures I took with D800 and D810A.  Then you will agree that the A7R is totally dispensable for me.   For me to consider using Sony again, Sony has to work hard and earn it.

 

Also, what make you think D810A may be a better choice?  I am among the first ones to characterize D810A on this forum and I had tested its H-alpha quantum efficiency and read noise against D800.  I see no evidence that D810A is superior to a modified D800 in these two aspects. D810A may have lower dark noise, but not significantly lower.

Edited by whwang, 08 May 2017 - 11:20 PM.

[sharkmelley]

 

I remember that Sony had already introduced a few features/corrections via firmware updates to its Mk2 cameras, but had ignored the Mk1 in these updates. I will not hold my breath for the correction of star eater problem on Mk1.  I am ready to sell my modified A7R in a price that I wouldn't believe 2 years ago.

 

At this moment, the best and yet reasonably priced camera for astrophotography is a used D800 for modification. In Japan, it is now possible to buy a very good used D800 with less than $1100 (no sale tax for foreigners). If you have a friend traveling to Japan recently, ask him/her to buy one for you.

I think you're throwing away one of the best cameras available for astro-imaging simply because you've somehow convinced yourself there is a serious problem.  It is your choice but I'll definitely continue using my A7S and A7R II!

 

But if you're going the Nikon route I would suggest the D810a instead of the D800.

 

bwa

 

Talking about the Sony A7S specifically, there is more than one serious problem, at least for calibrated frame deep-sky astrophotography:

• Star eater spatial filtering in bulb mode
• Bulb mode drops the camera into 12bit mode
• Coloured horizontal bands on the left hand side of the split sensor
• Digital scaling of data by a scalar close to 1.0 (evidenced by histogram gaps) which causes coloured concentric banding
• Flat frame calibration leaves coloured wide vertical swathes in the data

The first 4 problems have workarounds and I'm experimenting to see if there is a possible workaround to the problem 5.  But why should I have to do all that?

 

Strangely enough, the raw compression on the Sony has never caused me a single problem but it is the single feature you will find most complaints about.

 

How many other A7x models are affected?  Who knows.

 

So, I'm very close to ditching Sony now and buying a Nikon 800/810/810A.

 

Mark

Edited by sharkmelley, 09 May 2017 - 12:34 AM.

[Aryeh95]

I captured this a few days ago with the A7sii and it seems fine to me https://youtu.be/sB8Jocv7A1Q

Sent from my Nexus 6P using Tapatalk

[sharkmelley]

I captured this a few days ago with the A7sii and it seems fine to me https://youtu.be/sB8Jocv7A1Q

Sent from my Nexus 6P using Tapatalk

I think you mean the A7Sii is adequate for creating a timelapse video.  Which firmware version is it running, by the way?

 

Mark

[Aryeh95]

I captured this a few days ago with the A7sii and it seems fine to me https://youtu.be/sB8Jocv7A1Q

Sent from my Nexus 6P using Tapatalk

I think you mean the A7Sii is adequate for creating a timelapse video. Which firmware version is it running, by the way?

Mark

I'm not sure about the firmware version, its coming back from kolarivision tomorrow. But I'm assuming that if the stars survive at 15mm focal length, then at 800mm+ it should be much better.

Sent from my Nexus 6P using Tapatalk

[sharkmelley]

 

 

I captured this a few days ago with the A7sii and it seems fine to me https://youtu.be/sB8Jocv7A1Q

Sent from my Nexus 6P using Tapatalk

I think you mean the A7Sii is adequate for creating a timelapse video. Which firmware version is it running, by the way?

Mark

I'm not sure about the firmware version, its coming back from kolarivision tomorrow. But I'm assuming that if the stars survive at 15mm focal length, then at 800mm+ it should be much better.

Sent from my Nexus 6P using Tapatalk

 

Stars can certainly survive with a 15mm lens: they will survive if they extend beyond a 2x2 pixel "box" because of bad optics, star trailing or focus issues.  Stars that are completely contained within a 2x2 box will be deleted.

 

Mark

[whwang]

No matter what lenses and what focal lengths you use, stars won't be totally eaten.  I believe many stars will survive, especially the brighter ones, although perhaps with strongly altered core profile.  The star eater won't show up clearly in videos.  Only in carefully taken still images plus side-by-side comparisons with/without star eater in action, you start to feel the camera is not up to the $$$ you invested.

 

Mark, if you want to replace your A7S, do you think D750 is better than D8XX?  It has larger pixels, though not as large as A7S.

 

Cheers,

Wei-Hao

[Aryeh95]

No matter what lenses and what focal lengths you use, stars won't be totally eaten. I believe many stars will survive, especially the brighter ones, although perhaps with strongly altered core profile. The star eater won't show up clearly in videos. Only in carefully taken still images plus side-by-side comparisons with/without star eater in action, you start to feel the camera is not up to the $$$ you invested.

Mark, if you want to replace your A7S, do you think D750 is better than D8XX? It has larger pixels, though not as large as A7S.

Cheers,
Wei-Hao

I've used both for landscape astrophotography and I the d810 has a clear disadvantage in terms of low light noise. The D810a might be on par but I haven't used it enough to know for sure. So if you're deciding between the regular d810 and the d750 and you don't need the extra resolution then definitely go with the d750. Its cheaper, lighter, has a flippy screen which is useful for focusers in weird positions and it also at least a 1 stop advantage in terms of high iso snr.

Sent from my Nexus 6P using Tapatalk

[sharkmelley]

 

Mark, if you want to replace your A7S, do you think D750 is better than D8XX?  It has larger pixels, though not as large as A7S.

 

 

Thanks - I'll add the D750 to my shortlist.  I already have sign-off from the "financial controller".  She is understandably frustrated at the number of evenings I spend trying to solve Sony A7S issues.  In some ways I enjoy the intellectual challenge of understanding what is going on but it does divert time away from actual astrophotography.  Unfortunately I have a big backlog of data waiting to be processed.  If I can find a workaround to the latest problem I've identified (the purple and green swathes) then I have a chance of processing it.

 

Mark