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

Tony Hallas on using a dslr

This topic has been archived.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

#1 shawnhar

shawnhar

    Cosmos

  • *****
  • topic starter
  • Posts: 9,452
  • Joined: 25 Jun 2010

Posted 14 November 2014 - 02:38 PM

https://www.youtube....ZoCJBLAYEs#t=39

 

Awwww man! I only have CS2 and can't do the camera raw thing in photoshop. Interesting talk.



#2 TxStars

TxStars

    Vanguard

  • *****
  • Posts: 2,459
  • Joined: 01 Oct 2005

Posted 14 November 2014 - 05:30 PM

Thank you for posting this.

Tony gave some great information that can help improve DSLR images.



#3 proteus5

proteus5

    Surveyor 1

  • *****
  • Posts: 1,722
  • Joined: 24 Jun 2010

Posted 14 November 2014 - 05:38 PM

Nice find, Great info. Thanks for posting Shawn.



#4 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 7,932
  • Joined: 19 Feb 2007

Posted 14 November 2014 - 06:47 PM

Tony gave a talk recently that I attended where he made a point about one shot color that's proven to be very helpful. His point was to dither images by at least as much as the FWHM of the stars that you're seeing on a particular night. I did that with my OSC CCD camera and it really made a difference in the smoothness of the color. If you're software supports dithering give it a try.

Rgrds-Ross



#5 TimN

TimN

    Soyuz

  • *****
  • Moderators
  • Posts: 3,546
  • Joined: 20 Apr 2008

Posted 15 November 2014 - 11:01 AM

This is very interesting. Some Comments:

 

I currently use "Aggressive Dithering" in SGP without darks. My initial subs were not as bad as Tony showed but it takes me about 30 subs to get the noise down to a level I find satifactory. Tony was getting spectacular results using dithering after 9 subs. Do you think its because he uses Registar as opposed to DSS or PI? Does anyone use Registar that can comment.

 

Has anyone else tried his method of only using subs without bias or flats and fixing those problems in Adobe Raw, then registering /stacking and then feeding the result to your processing program?

 

Certainly seems worth trying after looking at his results. 



#6 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 7,932
  • Joined: 19 Feb 2007

Posted 15 November 2014 - 11:06 AM

What about the camera? I think that part of his point is that the current crop of DSLR's are much better when it comes to noise but I have to suspect that he's using a full frame high end camera to make his images and most DSLR users are using cameras that aren't of the same quality. It would be interesting to know what the specs are on the different cameras. In another thread I noticed that (and confirmed the fact with a bit of internet searching) Nikon's at any price point have measurably lower noise than the Canons that most use for AP. I've had a full framer on my wish list for a while to get wider angle shots than I can now without buying yet another OTA!

Rgrds-Ross



#7 TimN

TimN

    Soyuz

  • *****
  • Moderators
  • Posts: 3,546
  • Joined: 20 Apr 2008

Posted 15 November 2014 - 11:17 AM

Interesting point Ross. I have a Nikon D5100 which has very low noise. I guess I'm wondering if he fed his results from Adobe RAW into another register/stacking program would he get the same noise results? I was just wondering how much Registar has to do with it. I will probably just try it and see.



#8 GJJim

GJJim

    Skylab

  • *****
  • Posts: 4,024
  • Joined: 09 Sep 2006

Posted 15 November 2014 - 12:04 PM

Interesting point Ross. I have a Nikon D5100 which has very low noise. I guess I'm wondering if he fed his results from Adobe RAW into another register/stacking program would he get the same noise results? I was just wondering how much Registar has to do with it. I will probably just try it and see.

I saw an earlier presentation and IIRC the noise Hallas refers to is what he calls "color mottle". It results from the Bayer matrix convolved with the usual sensor noise sources. Dithering of subs has to be done at a scale larger than the mosaic of color "blocks" on the Bayer matrix to smooth this form of noise. 



#9 TimN

TimN

    Soyuz

  • *****
  • Moderators
  • Posts: 3,546
  • Joined: 20 Apr 2008

Posted 15 November 2014 - 12:10 PM

Jim, now that you mention it, he did mention that the dithering had to be a couple of star widths. I don't know if the commercial software allows that much dithering. I use SGP at maximum dither. I'll have to check into this. Thanks!  



#10 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 7,932
  • Joined: 19 Feb 2007

Posted 15 November 2014 - 01:00 PM

I've never personally gone about 5 but Maxim DL appears to allow for at least as much as 20 pixels of dithering either via the mount of via the guider. I use CCDAP and I was able to adjust the guiding parameters so that imaging only starts when you get below one pixel AND to allow for 8 exposures to be made before assuming a guide star problem.

Rgrds-Ross



#11 ZachK

ZachK

    Viking 1

  • -----
  • Posts: 840
  • Joined: 20 Aug 2005

Posted 15 November 2014 - 11:32 PM

Is there an easy way to move the camera around to get that dither? 



#12 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 7,932
  • Joined: 19 Feb 2007

Posted 15 November 2014 - 11:36 PM

If you know the image scale you could theoretically just stop guiding, nudge the mount (say) 3 pixels and start guiding again. Sound like a nuisance - you need software to support I think.

Rgrds-Ross



#13 HJustin

HJustin

    Vostok 1

  • *****
  • Posts: 170
  • Joined: 20 Jun 2005

Posted 16 November 2014 - 05:13 PM

Thanks for posting this. Dithering is on my list of things to try if the skies ever cooperate. 
 

Were there any other videos posted from OPT's? Imaging Symposium at SCAE? 



#14 gdd

gdd

    Mercury-Atlas

  • -----
  • Posts: 2,552
  • Joined: 23 Nov 2005

Posted 16 November 2014 - 05:39 PM

The Tony Hallas video says you need at least 9 dithered images to stack to remove the color mottling. If you are taking a large number of shorter subexposures I would think dividing them up into 9 sets of lights and manually dithering between sets would provide the desired results. DitherMaster would be more convenient, but the manual method should work if you don't want to haul out a laptop.

 

Gale



#15 shawnhar

shawnhar

    Cosmos

  • *****
  • topic starter
  • Posts: 9,452
  • Joined: 25 Jun 2010

Posted 16 November 2014 - 07:07 PM

Thanks for posting this. Dithering is on my list of things to try if the skies ever cooperate. 
 

Were there any other videos posted from OPT's? Imaging Symposium at SCAE? 

 

None that I saw, Paul Stewart posted this vid on Googleplus and he is in my circles, only reason I knew about it.



#16 diurnal

diurnal

    Vostok 1

  • -----
  • Posts: 101
  • Joined: 24 Jul 2013

Posted 16 November 2014 - 10:18 PM

good to see and hear the frustrations of astrophotography

 

thanks for posting



#17 hytham

hytham

    Surveyor 1

  • *****
  • Posts: 1,996
  • Joined: 25 Dec 2012

Posted 17 November 2014 - 08:14 AM

He processes the individual frames BEFORE stacking ...

That seems so bloody non-sensical to me and could be destroying data that is not recoverable. OTOH images that are as wide field as Milky Way imaging generally do not target small scale structures and provide a macro point of view and is not as crucial to retain.

I also have other doubts, but they are not scientifically based.

EDIT: auto correct working its magic

Edited by hytham, 17 November 2014 - 08:15 AM.


#18 terry59

terry59

    Voyager 1

  • *****
  • Posts: 10,189
  • Joined: 18 Jul 2011

Posted 17 November 2014 - 08:16 AM

He processes the individual frames BEFORE stacking ...

That seems so bloody non-sensical to me and could be dedestroying data that is not recoverable. OTOH images that are as wide field as Milky Way imaging generally do not target small scale structures and provide a macro point of view and is not as crucial to retain.

I also have other doubts, but they are not scientifically based.

 

Did you cheat and read Ivo's response in the DSLR forum? Kidding...I know better



#19 hytham

hytham

    Surveyor 1

  • *****
  • Posts: 1,996
  • Joined: 25 Dec 2012

Posted 17 November 2014 - 09:01 AM

Hah! I was completely blown away by that step and immediately thought the purpose of stacking was to eliminate this supposed non-random/fixed noise. Looking at my dslr images I don't have any colour mottle at all, but what I do see is the typical walking noise that is present when dithering is not used. I own a T3i which is a pile in comparison to a 6d.

#20 Footbag

Footbag

    Cosmos

  • *****
  • Posts: 9,115
  • Joined: 13 Apr 2009

Posted 17 November 2014 - 09:24 AM

I think this can be summed up by looking at one of the things he said.  This is the lazy way to do it.  He's not trying for the best results.  It's more of a good enough is good enough strategy.  

 

On the other hand, we jump through hoops to preserve data integrity.  It would be nice if he pointed out that better results could be obtained by doing it differently.



#21 jerry10137

jerry10137

    Viking 1

  • *****
  • Vendors
  • Posts: 783
  • Joined: 26 Jan 2013

Posted 17 November 2014 - 10:18 AM

I think this can be summed up by looking at one of the things he said.  This is the lazy way to do it.  He's not trying for the best results.  It's more of a good enough is good enough strategy.  

 

On the other hand, we jump through hoops to preserve data integrity.  It would be nice if he pointed out that better results could be obtained by doing it differently.

^^ very much agree with this



#22 hytham

hytham

    Surveyor 1

  • *****
  • Posts: 1,996
  • Joined: 25 Dec 2012

Posted 17 November 2014 - 12:03 PM

Is it honestly lazy though?

 

All he has done is moved when some of the steps occur and created more work. I cannot see why he didn't do them afterwatds and I readily admit that if he did explain the reasoning in the video, I was a wee bit distracted at the time and missed it.

 

Also the statement made by the gent over at reddit asserts that the colour mottle occurs only when imported into adobe raw and if this is true, that is more unnecessary work created.

 

Doesn't seem very lazy :)



#23 G. Hatfield

G. Hatfield

    Surveyor 1

  • *****
  • Posts: 1,572
  • Joined: 31 Jan 2009

Posted 18 November 2014 - 12:48 PM

I love this guy!  I bought my modified 6D last year and have been using his techniques since then.  No darks, no flats and the use of Camera Raw up front.... before Registar.  Works like a charm at least for me.  I also use PixInsight on the combined image if I need to get rid of a pesky gradient or increase color, etc.  I mentioned on this forum that I was doing the noise reduction before registration sometime back and someone stated that this would interfere with the registation process.  Not true with Registar.... I asked the company.  As for new cameras, Tony's video on this topic (at least the one I have) used images from and old Canon 5D and the results were equally as good (considering the source).

 

The only thing I have not tried is the dithering.  Next on my list.

 

George



#24 gdd

gdd

    Mercury-Atlas

  • -----
  • Posts: 2,552
  • Joined: 23 Nov 2005

Posted 18 November 2014 - 01:46 PM

A side by side comparison of the same lights processed conventionally and by the Tony Hallas method would be interesting. And also conventionally using the same lights but with the other frames Tony does not use.

 

Gale



#25 Synon

Synon

    Mariner 2

  • *****
  • Posts: 261
  • Joined: 02 Jun 2012

Posted 18 November 2014 - 04:26 PM

Is it honestly lazy though?

 

All he has done is moved when some of the steps occur and created more work. I cannot see why he didn't do them afterwatds and I readily admit that if he did explain the reasoning in the video, I was a wee bit distracted at the time and missed it.

 

Also the statement made by the gent over at reddit asserts that the colour mottle occurs only when imported into adobe raw and if this is true, that is more unnecessary work created.

 

Doesn't seem very lazy :)

 

Hopefully I'm not asking a stupid question, but is it possible to convert a TIFF back to a RAW image? My only guess is he is using functions that are only available for RAW images, if he stacked and created a TIFF file he could not use this same features if it can't be converted back. Just a guess (I'm not a PC with photoshop so I can't test this to be sure).




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