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

Samyang 135mm First Light with Nikon D5300

  • Please log in to reply
32 replies to this topic

#1 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 21 February 2017 - 10:43 PM

Hey everyone,

 

I Tested my Samyang 135mm last week, over all I am impressed with the lens and the stars are great shooting wide open.  I have been using my Nikon for about two years and fairly familiar with it but using this lens definitely showed some strange calibration issues while shooting at 200 ISO.  I was able to resolve it by using higher ISO flats with bias to calibrate.  By using the standard calibrations, there were very large rings appearing in the calibrated files.

 

I suppose next time I would want to shoot at 400 ISO at least but at f/2 and without a light pollution filter it is going to be very short subs.  That or I can stop the lens down further but that would introduce diffraction rings.

 

I took apart my current setup to hack together a solution to mount the lens and my guider with the existing brackets and adapters I had, luckily I was able to make it work.  I was able to fit the lens within two ADM rings so I could rotate the camera.  There is just enough room to clamp it down.  I used some painters tape around the lens to give some more texture for ring screws to grip.

 

The lens held its focus very well.  You have to back off from the hard stop infinity just a bit as expected.  It's very smooth and won't move by itself.

 

I was at a semi dark site but Orion is not very high now and without the LP filter it seemed to be picking up huge amounts of light pollution.  It was only after using the background extraction was I able to get the data out of the image, which was unusual for me as I was so used to my other setups performance / results.

 

rosette_135_i_crop_CN1600.jpg

http://www.astrobin.com/full/284605/0/

 

If you pixel peep the full version, I was fairly impressed with the resolution of the Rosette and other areas.

 

I'm still trying to get my head around how this F2 lens is suppose to work and calibrate with it's strong vignette at F2 as this is totally different than my f/6 480mm scope setup.  I don't think I want to invest in a clip in filter right now for this lens.  I will have to try some more tests at different ISO and stopped down to see how it performs.

 

Thanks for viewing!  Comments appreciated.  


  • mikewayne3, moxican, calypsob and 4 others like this

#2 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 21 February 2017 - 10:46 PM

Witch Head.  I cropped this one a bit.

 

Samyang 135mm F/2

ISO 200

50x140" 

25xflats, 60xbias, 14darks.

 

witch3_crop_rotated_CN1600.jpg

 


  • starbob1, Nile, moxican and 4 others like this

#3 petert913

petert913

    Gemini

  • *****
  • Posts: 3344
  • Joined: 27 May 2013
  • Loc: Portland, OR

Posted 21 February 2017 - 11:34 PM

For an unmodified ( I assume) camera,  the colors are great !  That lens is nice. 


  • entilza likes this

#4 leveye

leveye

    Aurora

  • ***--
  • Posts: 4812
  • Joined: 06 Feb 2013
  • Loc: Central Oregon Coast

Posted 22 February 2017 - 12:05 AM

Looks good. My vintage gold band Nikkor 180 ED is a bit sharper in the corners wide open and the colors are amazing so that's what I went with. Give one a try.


Edited by leveye, 22 February 2017 - 12:06 AM.

  • entilza likes this

#5 james7ca

james7ca

    Fly Me to the Moon

  • *****
  • Posts: 7403
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 22 February 2017 - 12:24 AM

Those look good for an f/2 lens. Can you post a 1:1 crop (full size) of Rigel (from the Witch Head shot)?



#6 calypsob

calypsob

    Aurora

  • *****
  • Posts: 4690
  • Joined: 20 Apr 2013

Posted 22 February 2017 - 12:54 AM

Nice first light, looks like you got some good data! Im really impressed with how clean the reflection and dark nebula is off to the right in the rosette image.  The witch came out great as well.   

 

This lens is surely a different animal to take on, at F2 depth of field is unbelievably small and breathing funny while focusing will literally put it out just a tad.  You definitely need a fine focus mechanism to get it nailed.  I think focus in your rosette image is just a tad out.    I battled with this lens for almost a year not being able to calibrate flats before finally downloading pixinsight and trying out its different functions.  I finally have concluded that I need at least 35 flat frames for every 120 light frames.  This totally cleans up my vignetting.  I am unaware of any other lens with this level of performance at F2.  The zeiss is close but has more coma, on aps-c its a non issue.  I am waiting to see how the new sigma 135 f1.8 fairs, it might be a contender.  

 

How are you currently mounting an LP filter?  



#7 Samir Kharusi

Samir Kharusi

    Mercury-Atlas

  • *****
  • Posts: 2579
  • Joined: 14 Jun 2005
  • Loc: Oman

Posted 22 February 2017 - 03:32 AM

Looks like an excellent lens to me, wide open. For similar lenses (is it an autofocus lens?) on my Canon cameras I just use one-click autofocus on the brightest star available or on a planet. Why would you want to close it down? It's sharp enough wide open, and at f2 subs longer or equal to 30 sec each ought to stack very well. We chase fast f-ratios to be able to use short subs and minimize Integration Time... At f4 to get sharper stars, you have to shoot 4 hours to achieve the stacked SNR you can with one hour at f2. By the way, I found "Star Halo Reduction" in ImagesPlus a very nice tool to sharpen up stars from just about any lens or OTA.



#8 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 22 February 2017 - 09:08 AM

For an unmodified ( I assume) camera,  the colors are great !  That lens is nice. 

 

Thanks Peter! Yes an unmodified.



#9 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 22 February 2017 - 09:12 AM

Nice first light, looks like you got some good data! Im really impressed with how clean the reflection and dark nebula is off to the right in the rosette image.  The witch came out great as well.   

 

This lens is surely a different animal to take on, at F2 depth of field is unbelievably small and breathing funny while focusing will literally put it out just a tad.  You definitely need a fine focus mechanism to get it nailed.  I think focus in your rosette image is just a tad out.    I battled with this lens for almost a year not being able to calibrate flats before finally downloading pixinsight and trying out its different functions.  I finally have concluded that I need at least 35 flat frames for every 120 light frames.  This totally cleans up my vignetting.  I am unaware of any other lens with this level of performance at F2.  The zeiss is close but has more coma, on aps-c its a non issue.  I am waiting to see how the new sigma 135 f1.8 fairs, it might be a contender.  

 

How are you currently mounting an LP filter?  

 

Thanks Wes, your images inspired me to try this lens!  Yes I did not 100% nail focus most likely I did not even use a mask,I was just using the numbers from Sequence generator to find the lowest.  It's a little hard to turn it once in that fine zone spot to really fine focus it as the lens is smooth but a little tight on those final micro-mm movements.  I can try a few more flats.

 

No LP filter at the moment.

 

I'll post some frames shortly.


  • calypsob likes this

#10 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 22 February 2017 - 09:16 AM

Looks like an excellent lens to me, wide open. For similar lenses (is it an autofocus lens?) on my Canon cameras I just use one-click autofocus on the brightest star available or on a planet. Why would you want to close it down? It's sharp enough wide open, and at f2 subs longer or equal to 30 sec each ought to stack very well. We chase fast f-ratios to be able to use short subs and minimize Integration Time... At f4 to get sharper stars, you have to shoot 4 hours to achieve the stacked SNR you can with one hour at f2. By the way, I found "Star Halo Reduction" in ImagesPlus a very nice tool to sharpen up stars from just about any lens or OTA.

Thanks Samir!  It's a manual focus.  I don't want to stop it down for sure, I am just having to deal with this vignette not being calibrarted correctly with 200 ISO at least.  

 

Here is a full calibration done with 200 ISO on the witch.

 

Integrated image stretch:  (Everything seems to be buried in vignette or LP very strange for me)

 

witch_int_pre_background.jpg

 

Background extraction degree 1:

 

witch_int_post_background.jpg

 

Definitely issues here.

 

I believe it's the ISO 200 causing this as it's visible in the flats by dividing the channels together (Thanks Sharksmelly).  I did some tests at higher ISO the flats are ok or appear to.


Edited by entilza, 22 February 2017 - 09:29 AM.


#11 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 22 February 2017 - 09:19 AM

Those look good for an f/2 lens. Can you post a 1:1 crop (full size) of Rigel (from the Witch Head shot)?

Thanks James Here are a few files I've shared:
 

https://drive.google...V0lCcFVPOXVqOEE



#12 calypsob

calypsob

    Aurora

  • *****
  • Posts: 4690
  • Joined: 20 Apr 2013

Posted 22 February 2017 - 10:04 AM

 

Nice first light, looks like you got some good data! Im really impressed with how clean the reflection and dark nebula is off to the right in the rosette image.  The witch came out great as well.   

 

This lens is surely a different animal to take on, at F2 depth of field is unbelievably small and breathing funny while focusing will literally put it out just a tad.  You definitely need a fine focus mechanism to get it nailed.  I think focus in your rosette image is just a tad out.    I battled with this lens for almost a year not being able to calibrate flats before finally downloading pixinsight and trying out its different functions.  I finally have concluded that I need at least 35 flat frames for every 120 light frames.  This totally cleans up my vignetting.  I am unaware of any other lens with this level of performance at F2.  The zeiss is close but has more coma, on aps-c its a non issue.  I am waiting to see how the new sigma 135 f1.8 fairs, it might be a contender.  

 

How are you currently mounting an LP filter?  

 

Thanks Wes, your images inspired me to try this lens!  Yes I did not 100% nail focus most likely I did not even use a mask,I was just using the numbers from Sequence generator to find the lowest.  It's a little hard to turn it once in that fine zone spot to really fine focus it as the lens is smooth but a little tight on those final micro-mm movements.  I can try a few more flats.

 

No LP filter at the moment.

 

I'll post some frames shortly.

 

Im glad you decided to try this thing out, it really is great to be able to shoot F2, you can get massive integrations in a short period.  You will nail focus in time, for now these are some great first light images for sure! 



#13 james7ca

james7ca

    Fly Me to the Moon

  • *****
  • Posts: 7403
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 22 February 2017 - 11:29 AM

 

Those look good for an f/2 lens. Can you post a 1:1 crop (full size) of Rigel (from the Witch Head shot)?

Thanks James Here are a few files I've shared:
 

https://drive.google...V0lCcFVPOXVqOEE

 

Thanks.



#14 dciobota

dciobota

    Surveyor 1

  • *****
  • Posts: 1646
  • Joined: 07 Aug 2007
  • Loc: Omaha, Nebrrrraska

Posted 26 February 2017 - 01:36 PM

Excellent pics.  I bought the Rokinon version (same lens) and found out the same thing, the vignetting is making flats very difficult.  I did manage to take a decent set of flats (I always use 50) under gray skies that worked.  As far as focus, I find I can get consistent good focus using a bahtinov mask and liveview at 10x on polaris.  Alternatively, if the temps change a lot, I usually do the same on a synced star in the neihgborhood of my target.  I don't know if I've just been lucky but it seems the focus stayed consistent.



#15 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 26 February 2017 - 02:21 PM

Thank You! May I ask which camera do you have? And what ISO?

I am still unable to calibrate at f2 with 200 ISO without those massive rings. I tried some room light wall flats same issue. Even tried long 10 second flats. Nothing.

Higher ISO seem to work. I am sort of disappointed at the moment. Calibrating with higher ISO flats just was a bandaid and introduced a lot of noise in the calibrations for some reason.

Not even sure what to do perhaps try 400 or 800 iso but will have low exposures maybe 30" without a LP filter.

Or I can just stop it down until I get to very dark skies..

Anyone have any other ideas? Thanks.
  • calypsob likes this

#16 dciobota

dciobota

    Surveyor 1

  • *****
  • Posts: 1646
  • Joined: 07 Aug 2007
  • Loc: Omaha, Nebrrrraska

Posted 27 February 2017 - 12:51 PM

Hi Martin,

 

 I haven't shot at iso200, just 100 and 400.  But the flats I took at iso100.  I use DSS for stacking, and even though it warns me about flats not matching iso if I shoot at say 400, it works fine.

 

 The camera is an EOS M3, same sensor size (and same sensor I think, maybe) as the T6.  It is not modified though.

 

 The way I shoot flats is in the daytime.  I have a cylinder with white typing paper over the opening that fits over the sunshade of the lens.  I usually try to shoot at noon, away from the sun, and overcast days are best because they diffuse the light pretty well.  My first set was on a clear day, and I think that messed up the flats as the sky brightness varied from zenith to horizon.  This lens seems very sensitive to that.

 

Hope this helps.



#17 calypsob

calypsob

    Aurora

  • *****
  • Posts: 4690
  • Joined: 20 Apr 2013

Posted 01 March 2017 - 11:07 PM

Thank You! May I ask which camera do you have? And what ISO?

I am still unable to calibrate at f2 with 200 ISO without those massive rings. I tried some room light wall flats same issue. Even tried long 10 second flats. Nothing.

Higher ISO seem to work. I am sort of disappointed at the moment. Calibrating with higher ISO flats just was a bandaid and introduced a lot of noise in the calibrations for some reason.

Not even sure what to do perhaps try 400 or 800 iso but will have low exposures maybe 30" without a LP filter.

Or I can just stop it down until I get to very dark skies..

Anyone have any other ideas? Thanks.

Martin I struggled with this lens for almost a year in DSS before finally jumping over to Pi and immediately resolving my issues.  I think you and Jon Rista were the first ones to start giving me advice to calibrate everything actually.  I will try and tell you what I do because I can always correct the vignetting edge to edge on my 550d 100%, and 96% on my 60d.

 

I am not familiar with Nikons, but I always match iso with Canon.  If I shoot iso 1600 lights, then bias, darks, and flats are also iso 1600.  

 

I shoot flats using an LCD computer monitor with a white tshirt over it.  I put the hood on the lens, turn the lights off in the room and stack some books to rest the camera on and push the lens hood flush to the monitor.  I use manual exposure to center the histogram, I am usually shooting 1/250 of a second at F2 iso 1600. Without moving the camera I will shoot 35-50 flats.  This completely eliminates vignetting.  After processing a pass in ABE is mandatory at this Fov and then you re ready to start processing.  In the field I have used my ipad and a white tshirt to take flats with notepad open.  Again this is very effective but you need to take 30-50 flats, I assume that you are building a signal by taking multiple flats which allows you to correctly illuminate the dark corners.  The more the better.

 

The flip up mirror creates a gradient at F2 which is almost impossible to remove, but it takes up a small sliver of the bottom of the image.  It is almost gone with 50 flats, maybe still 4% remaining

I removed the mirror from the 550d so there is no shadow gradient at all.

 

Let me know if this helps you at all, I am glad to share some raw data with you if you want to see how bad vignetting i on my individual subs vs final calibrated output stack/pre abe.  

 

 

I am not sure if this is of huge help but here is a contour plot applied to the master flat for my mono 60d and the Samyang 135mm F2 at F2. 

 

 

33035643581_148cd36ed0.jpgSamyang 135mm Contour Plot with debayered 60d at F2 by Wes Schwarz, on Flickr


Edited by calypsob, 01 March 2017 - 11:11 PM.


#18 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 03 March 2017 - 09:41 AM

Hey Wes, thanks for responding!  I can show you the issue with just analyzing a single flat sub.  If you can provide a raw flat file at 200 ISO I would appreciate to see if that issue is present on your camera as well at low ISO.  It would appear at higher ISO at least with the D5300 it gets resolved.  Also stopping the lens down also corrects this.

 

Take your raw flat sub extract the channels 

divide the green channel by the red channel

auto stretch (boosted helps show it)

 

 

Attached Thumbnails

  • sammy_d5300_issue1.jpg

Edited by entilza, 03 March 2017 - 09:41 AM.


#19 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 03 March 2017 - 09:43 AM

Here is another example showing a single flat at ISO 200, 400 and 800 note the improvement on increase ISO.

 

Again this is by dividing the green channel by Red  and blue by red.

 

Me by choosing 200 for my initial test was as it appears probably the worst choice given the results here.  At 400 at least it looks as if it would be more manageable.

Attached Thumbnails

  • sammy_d5300_issue2.jpg

Edited by entilza, 03 March 2017 - 09:44 AM.


#20 nathang123

nathang123

    Mariner 2

  • -----
  • Posts: 248
  • Joined: 02 Feb 2015

Posted 04 March 2017 - 02:02 AM

I have to say, that is a very weird artifact you've discovered. Looks to be an awesome lens, though. 



#21 james7ca

james7ca

    Fly Me to the Moon

  • *****
  • Posts: 7403
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 04 March 2017 - 09:48 AM

Martin, I appreciate your work with and examples from this lens.

 

IMO, it may not be too long before many of us are using short-focus, fast lenses with the latest crop of CMOS sensors to do wide-field imaging. As the pixels get smaller it actually makes sense to compensate by using shorter focus lenses/scopes.  I suspect this lens would produce some good results with the Sony IMX178.

 

Here is a framing on the Orion nebula with that lens/sensor combination...

Attached Thumbnails

  • Samyang 135mm with Sony IMX178.jpg

Edited by james7ca, 04 March 2017 - 10:38 AM.

  • entilza likes this

#22 Samir Kharusi

Samir Kharusi

    Mercury-Atlas

  • *****
  • Posts: 2579
  • Joined: 14 Jun 2005
  • Loc: Oman

Posted 13 March 2017 - 12:43 AM

Simple physics at f2 leads to very obvious vignetting with most optics. Several years back I found out that my C14 Hyperstar (f1.9) was very sensitive as to how I took flats. Ended up using sky flats at dawn with a T-shirt over the huge dew cap. I would suggest that flats ought to be taken at the same ISO as the Lights. Too weary at this time to verify the maths behind this statement (all that pixel division business cool.gif , maybe not true?). Anyway a quickie way of verifying whether your procedures work is to calibrate any one of the flat frames with your Master Flat. It should be quite obvious whether what you are doing actually flattens that "flat". Luckily you can do all such testing in daytime.


  • Miguelo likes this

#23 sharkmelley

sharkmelley

    Vanguard

  • *****
  • Posts: 2316
  • Joined: 19 Feb 2013

Posted 13 March 2017 - 01:35 AM

If you can make a raw exposure available, I can check if the concentric rings are caused by a similar issue to mine. 

 

Mark



#24 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 13 March 2017 - 09:52 AM

Simple physics at f2 leads to very obvious vignetting with most optics. Several years back I found out that my C14 Hyperstar (f1.9) was very sensitive as to how I took flats. Ended up using sky flats at dawn with a T-shirt over the huge dew cap. I would suggest that flats ought to be taken at the same ISO as the Lights. Too weary at this time to verify the maths behind this statement (all that pixel division business cool.gif , maybe not true?). Anyway a quickie way of verifying whether your procedures work is to calibrate any one of the flat frames with your Master Flat. It should be quite obvious whether what you are doing actually flattens that "flat". Luckily you can do all such testing in daytime.

Thanks Samir!  I have yet to try sky flats but I have tried wall flats and it showed the same issue.  I even tried some crazy long 10 second exposures with no luck.  I will try a sky flat once these outdoor temperatures settle down :)

 

I tried calibrating one of the flat frames with the master flat as you suggested and it seemed to work ok, not sure what that means, other than the flats were taken ok?



#25 entilza

entilza

    Soyuz

  • *****
  • topic starter
  • Posts: 3743
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 13 March 2017 - 09:58 AM

If you can make a raw exposure available, I can check if the concentric rings are caused by a similar issue to mine. 

 

Mark

 

Hey Mark, thanks for your time!  I've linked a few here along with some master flats.  https://drive.google...NlFJTm0wbWhMUFk




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