Long Exposure with MFD: Show us photos

[Audii-Dudii]

Amazing, i love landscapes long exposure shots more than nightspots, i started photography with nightspots as my favorite, but not so long i changed it to landscapes without regret, but still shooting nightspots.

For many years, I loved landscapes and daylight long-exposure shots as well. Unfortunately, I no longer have the time to photograph landscapes or even during daylight hours, as my so-called "day" job usually keeps me busy late into the evening as well. (I joke about working half-days -- i.e. 12 hours -- but it's not far from the truth. Besides, it's not the long days that bother me so much as the fact that my workweek usually has six or seven of them.)

Anyway, photographing at night has been a revelation for me and I usually manage two or even three outings a week, whereas if I was to try to photograph during the day, I'd only be managing one or two outings a month at best.

 

[jlm]

nice one, Ed, magical.

 

[archivue]

Please, can you post a 100% crop as well with your images... noise being the problem, i'd like to see the texture...

thanks you very much !

Considering white balance in mixed lights situation, i select a fixed color temp... 3200, 3400 or 5600 and process it like that, then i fixe different part of the tiff 16 bits with U points... i don't have any use of a grey card in this kind of situation !

 

[P. Chong]

25s, ISO100, H3D-39 with 28mm

51s, ISO50, H3D-39 with 28mm

 

[ptomsu]

Peter,

this again shows (me - I own exactly the same camera and 28) what you already can achieve with this relatively "old" sensor design, especially in times when one starts thinking (WRT all the recent posts) that you have to have a P40+ or a H4D40 to get such results!

Truth is, newer sensor technology makes life easier for sure, but some creativity and willingness to work and think out of the box makes same results also possible in many areas with older technology. What we should learn of this - do not always run after the latest and greatest, this can make life much more relaxed and easier sometimes

Great shots, great samples of what can be achieved!

Peter

 

[jlm]

here is a 100% crop, at least i think so

 

[BradleyGibson]

The sun rose enough during the exposures to cause a gradient in the sky. There are a couple of other issues as well--I plan to go back to get it right, so I won't bother to fix the light trails at the seams in this trial version.

Original is 206MP.

 

[ondebanks]

Ursa Major/Plough/BigDipper...you choose. Taken in weak moonlight.

Mamiya 645AFD, Kodak DCS 645M, MC Arsat 30mm at f3.5, ISO 400, 12x 30sec

 

[ondebanks]

Ursa Major/Plough/Big Dipper...you choose. Taken in weak moonlight.

Mamiya 645AFD, Kodak Proback DCS645M at ISO 400, MC Arsat 30mm at f3.5, 12x 30sec.

Exact same set of frames, but processed in two different ways:
- locking onto the celestial coordinates and letting the Earth do the turning
- as originally seen by the camera on fixed tripod

(Ignore the colour balance difference; I wasn't paying attention to that).

 

[ondebanks]

Orion (in outer suburban skies)

Mamiya 645AFD, Kodak Proback DCS645M at ISO 400, 80/1.9 at f2.8, 59 sec.

What makes the Kodak one of the best MFDBs for this sort of shot is that its IR-block filter actually transmits the 656 nm emission line of ionised hydrogen. So Orion's nebulae look bright and red, as they should, and not a dim washed-out green, as virtually all stock MFDBs and DSLRs record them.

OTOH, like all MFDBs, the readnoise is terrible in comparison to DSLRs at high ISO. So the signal to noise falls off a cliff as you approach the fainter stars and lower surface-brightness nebulae.

In terms of star detection (limiting S/N), I would rate the Kodak back as equal to fresh ISO 800-1000 film. MFD should be much better than that, given the sensor's much higher q.e., but it's the readnoise and the dark thermal noise which set it back. Much more signal, but also much more noise.

In terms of colour, and especially sharpness, the Kodak back is far better than fast film.

BTW the Mamiya 80/1.9 is stunningly sharp across the whole field at f2.8 - much better corner performance than anything I've seen from a Zeiss Planar at f2.8.

 

[Bob]

30 second multiple exposure
NSFW: http://forum.getdpi.com/gallery/files/1/8/2010-12-15_ambercandace_8350-900.jpg

 

[Woody Campbell]

Inspiring thread - keep the images coming.

 

[ondebanks]

Now for something rather longer...25 minutes. Or rather, 50x 30 second exposures.

Perseus-Cassiopeia region, featuring the Double Cluster & the Double Nebula (aka "Heart and Soul" nebulae), the "E.T." Cluster, and several more deep-sky treasures.

Mamiya 645AFD, Kodak Proback DCS645M at ISO 400, 110/2.8 at f2.8, 30 sec x 50 frames, EQ6 equatorial mount tracking (but no guiding corrections). Taken from my back yard.

No sharpening or filtering.

2 views: Full frame, and central region at 50% zoom.

Again, the colours are realistic and the resolution is spectacular...but you'll just have to take my word for it because the forum software applies some horrible jpeg compression which really smears them. A starfield, with its countless sharp dark-bright transitions, is a red rag to a jpeg-algorithm bull.

The 110/2.8 is another fine M645 lens, wide open: on individual frames, the stars are pretty much Nyquist sampled right across the field. The lack of an AA filter helps, of course. The final stacked version is not quite as sharp because of the interpolation involved in registering the frames. And then the additional forum jpeg compression kicked in on top of that.

The limiting S/N is "not bad", but a FFDSLR would detect much fainter stars in the same amount of exposures. However, the DSLR would need to be modified to pick up the red of the nebulae.

Ray

 

[Woody Campbell]

Now for something rather longer...25 minutes. Or rather, 50x 30 second exposures.

Perseus-Cassiopeia region, featuring the Double Cluster & the Double Nebula (aka "Heart and Soul" nebulae), the "E.T." Cluster, and several more deep-sky treasures.

Mamiya 645AFD, Kodak Proback DCS645M at ISO 400, 110/2.8 at f2.8, 30 sec x 50 frames, EQ6 equatorial mount tracking (but no guiding corrections). Taken from my back yard.

No sharpening or filtering.

2 views: Full frame, and central region at 50% zoom.

Again, the colours are realistic and the resolution is spectacular...but you'll just have to take my word for it because the forum software applies some horrible jpeg compression which really smears them. A starfield, with its countless sharp dark-bright transitions, is a red rag to a jpeg-algorithm bull.

The 110/2.8 is another fine M645 lens, wide open: on individual frames, the stars are pretty much Nyquist sampled right across the field. The lack of an AA filter helps, of course. The final stacked version is not quite as sharp because of the interpolation involved in registering the frames. And then the additional forum jpeg compression kicked in on top of that.

The limiting S/N is "not bad", but a FFDSLR would detect much fainter stars in the same amount of exposures. However, the DSLR would need to be modified to pick up the red of the nebulae.

Ray

Ray - How are you doing the stacking?

 

[ondebanks]

Woody, I use Deep Sky Stacker. It's free and it's astonishingly good.

It detects stars in every frame and uses a sophisticated pattern-matching algorithm to automatically compute the registration between frames, even if there are shifts and rotations between them. If so desired, it will also apply calibration frames (bias, dark, flat-field) to the input frames. Then it stacks them, offering a large choice of statistical algorithms for transient noise rejection and S/N weighting.

I choose to let it operate directly on the RAW input frames (it uses DCRAW as its RAW decoder), and for maximum control and precision I prefer it to output FITS format files (the astronomical standard). I then combine the FITS colour channels the way I want them in another free package, SAOimage ds9. But Deep Sky Stacker will also input tiffs and jpegs and output tiffs.

http://deepskystacker.free.fr/english/index.html
http://hea-www.harvard.edu/RD/ds9/

 

[Woody Campbell]

Woody, I use Deep Sky Stacker. It's free and it's astonishingly good.

It detects stars in every frame and uses a sophisticated pattern-matching algorithm to automatically compute the registration between frames, even if there are shifts and rotations between them. If so desired, it will also apply calibration frames (bias, dark, flat-field) to the input frames. Then it stacks them, offering a large choice of statistical algorithms for transient noise rejection and S/N weighting.

I choose to let it operate directly on the RAW input frames (it uses DCRAW as its RAW decoder), and for maximum control and precision I prefer it to output FITS format files (the astronomical standard). I then combine the FITS colour channels the way I want them in another free package, SAOimage ds9. But Deep Sky Stacker will also input tiffs and jpegs and output tiffs.

http://deepskystacker.free.fr/english/index.html
http://hea-www.harvard.edu/RD/ds9/

Thanks - their website is very informative. I can set a camera up on an equitorial mount this weekend (weather permitting) and give it a try. It might also be interesting to see how the software handles non-astro subjects.

 

[perjorgen]

Here are a few from aout trip to Botswana

Kalahari desert
15*64 sec 28mm f/4 iso 800

Baobab trees
64 sec 28mm f/4 iso 50 (+1.7 EV)

Victoria falls
64 sec 28mm f/4 iso 50 (+2.5 EV)

I have some problems with noise, but I am working on an algorithm to get rit of it (it might take a while though)

---
http://botswana.perjorgen.com
http://www.antarctic.dk

 

[Hauxon]

Some nice pictures here. Anybody using ND or ND-grads for long exposure shots? Filters like the Lee Big Stopper are great for daylight long exposures.

 

[jlm]

another from underground

 

[jlm]

another