Eta-Carina (HDI process)

[74tuc]

Here is a pic of the nebula eta-Carina - the star eta-Carina just makes an appearance - RHS.

The image was processed to extract max detail.

The experiments into this type of processing is continuing!

Camera: EOS 350D @ Iso 400

Total exposure - 2048 sec.
Frame (#11) longest exposure 1047 sec.

Jerry.

[Doug]

Nice work Jerry, how many exposure groups were used and what values?
Star co;ours are good and rich as well as the nebulosity.
Doug

[RB]

Great work Jerry, I like the deep colours you've achieved in this shot.

[h0ughy]

ooohhhhh yeeessssss please tell all, what did you do?

[iceman]

WOw that's magnificent! Well done!

[tornado33]

Very nice image there, it does show a big range of brightness there.
Scott

[Striker]

Very nice Jerry.

[Ric]

Great image Jerry, love all the detail and colour.

Cheers

[74tuc]

Hello All,

Thanks for the good words.

The pics Iv'e put on this forum are the result of trying some new techniques. The idea is to produce artistic effects in astrophotography - pics with no scientific merit but are nice looking - technical perfection comes later!!

These techniques come from the world of computer animation, HD displays etc. and have produced amazing results in terrestrial phtography.

To discuss these techniques will require a separate discussion thread but I'll present you with the info for you to have a go and see what develops.

The program I'm using is the trial version of Photomatix Pro Beta version 2.4 and is available from:
http://www.hdrsoft.com/download/download20win.php
There is plenty of info at this site.

In keeping with the idea of exposure blending and because of the huge dynamic range of our images I choose a range of exposures:

1, 2, 4, 8, ..... 1024 secs. This gives an apparent range of 10 extra bits.

The pics with the mount continuously tracking and combined in Photomatix to produce a single HDR image. This image was tone compressed to produce the results displayed here.

For the quieter and more sensitive astronomical cameras I would try the following exposures:

Going up in 1 magnitude steps I would tru these exposures

0.01,0.0251,0.0631,0.159,0.398,1.00 ,2.52,6.31,15.84,39.8,100,251,631,1 585 secs. This is a range of 13 magnitudes and is a good starting point.

Jerry.:)

[sheeny]

Wow!:eyepop:Excellent work Jerry!

Al.

[Doug]

Thanks for the info Jerry. There is also an article on IIS on HDR processing using Photoshop CS2. It will shortly be undergoing a revision....I hope.

cheers,
Doug

[avandonk]

That is very good work Jerry. I am impressed. How do you think the software would handle a linear tiff produced by ImagesPlus. I think I would lose what marbles I have left if I had to go to taking a series of different exposures.
If I send you a suitable file (linear tiff) would you show what can be achieved with it.
Otherwise I would be willing to produce lets say three? different exposures. How many do you suggest. The Canon 5DH has 14.3 stops dynamic range. The A to D converter is 12 bits. Any advice is gratefully accepted.
I do understand that to get star colours a short one second exposure is all that is needed for the brightest stars and so on up. The dynamic range is huge! The other end of course is the very dim nebulosity at an exposure of many hundreds of seconds.

Again very nice work.

Bert

[h0ughy]

Jerry I would be most interested in you starting another thread to discuss and work through the processing of this image (and others). Fantastic work, and I can see now you would have to dedicate a night and object.

[74tuc]

To answer a few questions.

1. All frames were converted from the Canon RAW to 16 bit tiff files so tiff files from any source should be ok. The main thing is that the frame alignment must be very good. Photomatix has a frame alignment option but it did not work well for me - perhaps I didn't try hard enough.

2 The Canon in-camera noise reduction was used - nothing else. I merely took the tiff files and had Photomatix combine them, its as simple as that. I was told that CS2 had some similar procedures but they are not as good as Photomatix. Adobe has modified the algorithms and the new CS3 should be an improvement.

3. My main motivation is to see what other astrophotographers come up with in this area - there are some pretty creative people out there. Playing around with astrophotography you find that we keep doing the same things - I know the results are good - very good but why not try something new?

Bert,

12 bit dynamic range is what we get for most CCD's so we get what we get.
I could have a go at processing a few frames for you - but you should also have a go. OK! take 3 pics with your 5D. Convert the CRW files and convert them to *.jpg and e-mail (send me a message via this forum) them to me lets see what we can do.

Houghy,

OK a little discussion thread might be in order - I'll get my thoughts in order and get back to you. I'll have a go at Omega Centauri this evening to see how star clusters turn out. For me its one object per evening otherwise I'll end up not going to work:thumbsup:


Clear skies,

Jerry:)

[avandonk]

Thanks Jerry. Since I did not have a clear sky handy took a series of exposures from a 1/60 (pic1) to 1/8000 (pic2) at 100 ISO at f/11 of an object with a lot of dynamic range ie a light globe.
The third picture is the result of Combining Highlights and shadows Auto of ten jpg's.
I will definitely try it with some astro images. Thanks again for making things even more complicated:)

Bert

[avandonk]

Here is the tiff version HDR Combine and then tone mapping.
I just realized this would be fantastic for Crescent Moon images to show
the Earthlight lit part without blowing out the Sun lit crescent.

Bert

[74tuc]

Very good Bert. Just what I thought the program would do.

See!! I've got a new idea from you - the moon. Didn't even think about it. Thanks.

I will continue to experiment with this code.

I have some even more amazing code that produces pics of Jupiter (single frame) with amazing detail - this is still under test.

Please keep us posted with your moon pics. I'll also have a go.

Jerry.:)

[niko]

Hi guys,

as someone new to the hobby/art/science of astronomy and still in the world of afocal photography this photo and the subsequent discussion has blown me away.

Very inspiring while at the same time completely daunting!

Still, it's one step at a time, beginning with curve adjustments in Photoshop of my moon photos.

Totally loving this website and the forums.

Thanks for all the hints, tricks and time you lot put in

Cheers

Nick

[74tuc]

Hi Nick,

Re: "... with curve adjustments in Photoshop of my moon photos ..."

To get excellent moon pics is surprisingly difficult - alot of things have to come together. So continue along this path and when you get an excellent pic pat your self on the back:thumbsup:

Jerry:)

[JohnH]

Thanks for that shot and the heads up on the potential of HDR - this is news to me and I will give it a shot as soon as possible...these shots just hit you in the eye, amazing.

[fringe_dweller]

Kinda reminds a little me of what Mike (iceman) was doing with that technique of normal daylight shots - with one underexposed and one over exposed shot combined? for extra dynamic range - hhhmmm

great stuff Jerry!

[niko]

Thanks for the encouraging words Jerry - had a pretty good night on Saturday with some pleasing results - so a small pat is perhaps in order

Your photo is truly inspiring

Nick

[74tuc]

Hi Nick,

OK.

Well ..... ? Where's the pics?

Lets have a look.

Jerry.:)

[Astroman]

amazing stuff Jerry, the mind is boggling at what other advances in artistic astrophotography is out there... the deepsky is the limit I guess...

[ving]

great stuff jerry! :)

have you tried "easy hdr"? works in a similar way i believe

[74tuc]

Hello David et al.

Just had a look at the "Easy HDR" site looks good and worth a try.

My current interest is in tone mapping and the algorithms used - please let me try to explain. The processing we are describing as HDR is actually a double barrelled process:

1. Creating a file that is capable of representing (encoding) the scene you are looking at rather than the common (& limited) colour gamut that is represented by the standard SRGB(TIFF format). To this end a number of encoding standards were invented. Eg. The formats XYZE and RGBE each encode in 32 bits per pixel. The problem is that no currently available display can display the colour gamut that may be accommodated by these formats so:

2. To represent these colours in the current displays (printers, CRT's etc) a process called tone mapping is used. The reason for tone mapping is best described in "A review of tone mapping techniques" by Kate Devlin, quote:

"The ultimate aim of realistic graphics is the creation of images that provoke the same response and sensation as a viewer would have to a real scene i.e. the images are physically or perceptually accurate when compared to reality."

A lot of this work is involved in modelling the human visual system and there is much research going on in this area so as some firms develop these algorithms they put them up for sale. As a result of this some code currently available will perform better than others - pity they do not publish their research results so we can really tell how much better one piece of code is better than another.


Jerry.:)

[niko]

Okay Jerry, you asked for them and so here they are!!!

Unprocessed images of the moon taken last Saturday night. Haven't figured out why the blue edge appears and while I tried the moon filter and even covering up the largest aperture of our 8" dob the images remain either too bright or too dull (with filter etc).

Anyway, much happier with these results and gives me encouragement to take another step forward.

Jerry, thanks again for your words of encouragement

[74tuc]

Hi Nick,

I've replied in the "Solar System Area".

Jerry.