Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Star eater in Action - sony A7rII

This topic has been archived. This means that you cannot reply to this topic.
95 replies to this topic

#76 nampramos

nampramos

    Lift Off

  • -----
  • Posts: 19
  • Joined: 17 Apr 2017

Posted 15 May 2017 - 01:14 PM

On that link it shows up 3.20 and not the 3.10 firmware version.



#77 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 13,401
  • Joined: 13 Jan 2012

Posted 15 May 2017 - 01:27 PM

Use 3.20. Mark is dealing with issues may not concern you. Regards



#78 nampramos

nampramos

    Lift Off

  • -----
  • Posts: 19
  • Joined: 17 Apr 2017

Posted 15 May 2017 - 01:31 PM

What do you mean? Can you confirm that the spatial filter that kicks in at 4s isnĀ“t on the 3.20 FW for the A7S Mark I?



#79 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,773
  • Joined: 14 Jan 2008

Posted 15 May 2017 - 01:49 PM

 

One question: from which firmware did Sony introduce this so called "star eater monster" laugh.gif

 

I read the entire thread and could not find an answer to that.

 

I have an A7S Mark I with FW 1.10 and would like to know which FW I can load on it.

 

Thanks!

I'm running the A7S with firmware version 3.10 and the star eater is only in bulb mode.   Version 3.20 is available but there is no way I'm going to risk updating it, just in case.

 

Mark

 

I updated my firmware about 2 months ago and surely that was 3.20 as its been out for a while.

 

The star eater is not an issue in M mode, only bulb. You can see the comparison images I posted earlier. 



#80 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,773
  • Joined: 14 Jan 2008

Posted 15 May 2017 - 11:30 PM

I wanted to see how the A7S star eater algorithm works on the EdgeHD 8" at 2032mm.

 

On the left is a 10 minute shot at ISO 100, cropped.

 

On the right is a 30 second shot at ISO 3200 (when adjusted for ISO stops, the total brightness is the same as about 16 minutes at ISO 100)

 

The ISO 100 shot seems to be very usable with little evidence of the star eater algorithm.

 

In fact, due to the lower noise at ISO 100, a couple of stars are actually more visible than the ISO 3200 even though the exposure has less brightness. 

 

So it seems like shooting at very low ISO values for the A7S can result in good results as long as the focal length is high. 

 

ISO100vs3200.png



#81 bwallan

bwallan

    Viking 1

  • -----
  • Posts: 563
  • Joined: 08 Jul 2010

Posted 16 May 2017 - 12:12 AM

I wanted to see how the A7S star eater algorithm works on the EdgeHD 8" at 2032mm.

 

On the left is a 10 minute shot at ISO 100, cropped.

 

On the right is a 30 second shot at ISO 3200 (when adjusted for ISO stops, the total brightness is the same as about 16 minutes at ISO 100)

 

The ISO 100 shot seems to be very usable with little evidence of the star eater algorithm.

 

In fact, due to the lower noise at ISO 100, a couple of stars are actually more visible than the ISO 3200 even though the exposure has less brightness. 

 

So it seems like shooting at very low ISO values for the A7S can result in good results as long as the focal length is high. 

 

ISO100vs3200.png

Better star color in the ISO 100 image.  Other than that both look reasonable.

 

bwa



#82 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,773
  • Joined: 14 Jan 2008

Posted 16 May 2017 - 01:14 AM

the faint stars along the middle and left seem to have better SNR??



#83 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,773
  • Joined: 14 Jan 2008

Posted 16 May 2017 - 01:26 AM

 

I wanted to see how the A7S star eater algorithm works on the EdgeHD 8" at 2032mm.

 

On the left is a 10 minute shot at ISO 100, cropped.

 

On the right is a 30 second shot at ISO 3200 (when adjusted for ISO stops, the total brightness is the same as about 16 minutes at ISO 100)

 

The ISO 100 shot seems to be very usable with little evidence of the star eater algorithm.

 

In fact, due to the lower noise at ISO 100, a couple of stars are actually more visible than the ISO 3200 even though the exposure has less brightness. 

 

So it seems like shooting at very low ISO values for the A7S can result in good results as long as the focal length is high. 

 

ISO100vs3200.png

Better star color in the ISO 100 image.  Other than that both look reasonable.

 

bwa

 

 

Here is another example, I used an arrow to point to the star that is more visible in ISO 100 shot. 

Capture_edited.jpg



#84 sharkmelley

sharkmelley

    Fly Me to the Moon

  • *****
  • Posts: 5,614
  • Joined: 19 Feb 2013

Posted 16 May 2017 - 01:50 AM

You collected 20x as many photons in the ISO100 image so, yes, you would definitely expect faint stars to be more visible.

Mark

#85 nampramos

nampramos

    Lift Off

  • -----
  • Posts: 19
  • Joined: 17 Apr 2017

Posted 16 May 2017 - 01:51 AM

 

 

One question: from which firmware did Sony introduce this so called "star eater monster" laugh.gif

 

I read the entire thread and could not find an answer to that.

 

I have an A7S Mark I with FW 1.10 and would like to know which FW I can load on it.

 

Thanks!

I'm running the A7S with firmware version 3.10 and the star eater is only in bulb mode.   Version 3.20 is available but there is no way I'm going to risk updating it, just in case.

 

Mark

 

I updated my firmware about 2 months ago and surely that was 3.20 as its been out for a while.

 

The star eater is not an issue in M mode, only bulb. You can see the comparison images I posted earlier. 

 

 

Thank you!

So this problem is actually a Mark II firmware problem, that only affects the Mark I cameras in Bulb mode.



#86 charotarguy

charotarguy

    Apollo

  • *****
  • Posts: 1,320
  • Joined: 10 Sep 2014

Posted 24 May 2017 - 01:23 PM

Any update on this? Have not heard anything new on whether Sony is addressing this or not.



#87 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,773
  • Joined: 14 Jan 2008

Posted 24 May 2017 - 02:03 PM

this issue has been around for years...not expecting it to go away all of a sudden



#88 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 13,401
  • Joined: 13 Jan 2012

Posted 24 May 2017 - 04:37 PM

Any update on this? Have not heard anything new on whether Sony is addressing this or not.

If you use longer FL (i.e., bit over sampled stars), this is a NON-issue. Most of this is academic discussion. Regards


Edited by mmalik, 24 May 2017 - 07:02 PM.


#89 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,773
  • Joined: 14 Jan 2008

Posted 24 May 2017 - 05:28 PM

yes at fl = 2032mm

 

almost no stars lost....but the center of stars seem a bit softer



#90 whwang

whwang

    Skylab

  • *****
  • Posts: 4,201
  • Joined: 20 Mar 2013

Posted 24 May 2017 - 07:28 PM

 

Any update on this? Have not heard anything new on whether Sony is addressing this or not.

If you use longer FL (i.e., bit over sampled stars), this is a NON-issue. Most of this is academic discussion. Regards

 

Let's guess whether most people on this forum use short focal length (<800 mm) or long focal length (>1500mm).



#91 charotarguy

charotarguy

    Apollo

  • *****
  • Posts: 1,320
  • Joined: 10 Sep 2014

Posted 08 June 2017 - 01:01 PM

Looks like Sony just came out with new firmware for the A7x ii cameras. 



#92 navmannz

navmannz

    Lift Off

  • -----
  • Posts: 4
  • Joined: 29 Apr 2017

Posted 08 June 2017 - 04:01 PM

I did a quick test on it last night here in NZ, where we had a clear night with a moon one day off full.

Exposures at 3.2 and 4 seconds with LNR turned off showed no difference in detail, apart from the

expected very small difference in star trailing, so it looks to me like the median-filtering is switched

out when LNR is disabled. A big thanks to all those who put shoulders to the wheel on getting Sony

to listen on this, and a particular 'hats-off' to Mark Shelley (sharkmelley) for his persistence and

grace in the face of sometimes curious reactions from the disinterested...

 

-John


Edited by navmannz, 08 June 2017 - 05:42 PM.


#93 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,773
  • Joined: 14 Jan 2008

Posted 08 June 2017 - 08:09 PM

is the bulb mode also fixed?

 

how about for exposures > 30 seconds?

 

and i dont see an update for the A7S (mark 1). will they ever get a patch?



#94 sharkmelley

sharkmelley

    Fly Me to the Moon

  • *****
  • Posts: 5,614
  • Joined: 19 Feb 2013

Posted 09 June 2017 - 02:09 AM

Hi navmannz,

If your test had trailing stars then I respectfully suggest it is not a good test. Star eater never affected trailed stars in the first place.

I haven't yet seen convincing evidence that the latest firmware upgrade has fixed the issue. The test needs to be done on stars that are contained within a 2x2 pixel box in the raw file.

Unfortunately my camera is the A7S which has no firmware upgrade.

Mark

Edited by sharkmelley, 09 June 2017 - 02:55 AM.


#95 joelin

joelin

    Mercury-Atlas

  • *****
  • Posts: 2,773
  • Joined: 14 Jan 2008

Posted 09 June 2017 - 11:41 AM

I did a quick test on it last night here in NZ, where we had a clear night with a moon one day off full.

Exposures at 3.2 and 4 seconds with LNR turned off showed no difference in detail, apart from the

expected very small difference in star trailing, so it looks to me like the median-filtering is switched

out when LNR is disabled. A big thanks to all those who put shoulders to the wheel on getting Sony

to listen on this, and a particular 'hats-off' to Mark Shelley (sharkmelley) for his persistence and

grace in the face of sometimes curious reactions from the disinterested...

 

-John

i would suggest using a short focal length lens...say 50mm or less and expose for 5 seconds or less...this should avoid any trailing 



#96 sharkmelley

sharkmelley

    Fly Me to the Moon

  • *****
  • Posts: 5,614
  • Joined: 19 Feb 2013

Posted 09 June 2017 - 02:15 PM

Most likely Sony is still applying a spatial filtering algorithm but a less destructive one. 

 

I would be interested in seeing some dark frames:

https://www.cloudyni...ware/?p=7930702

 

Mark


Edited by sharkmelley, 09 June 2017 - 02:15 PM.



CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics