Baz I don't think you will achive your stated goal, IMO the 5DII simply isn't sensitive enough to produce short exposures of the night sky without a significant tradeoff between noise and trailing.
....... long tracked exposures will produce the results you're after, ......... you may be better off investing in some lightweight and portable tracking equipment.
Tracking is no good for me.
My goal is to do time-lapses of much wider views of the milky way WITH various exotic horizons. No good doing longer tracking there or you get blurred foregrounds. 30 seconds is nominal with my current 18mm F3.5, but I could capture much more light in the same or even a shorter exposure, WITHOUT star tails or motion blur on the horizon.
As I said before, there is a middle ground somewhere and I HATE the fisheye effect.
My goal is to do time-lapses of much wider views of the milky way WITH various exotic horizons
I have seen images presented in this manner by an american photographer. I seem to recall reading somewhere that his images were achieved using a composite technique, the terrestrial scene is first imaged, then the starry sky is layered from a tracked image.
I have seen images presented in this manner by an american photographer. I seem to recall reading somewhere that his images were achieved using a composite technique, the terrestrial scene is first imaged, then the starry sky is layered from a tracked image.
If you used the same focal length lens and kept the compass directions the same for both shots (ie the tracked star shot and the terrestrial shot), why would stacking them be any different to stacking 3 RGB shots to produce one image?
If you used the same focal length lens and kept the compass directions the same for both shots (ie the tracked star shot and the terrestrial shot), why would stacking them be any different to stacking 3 RGB shots to produce one image?
Ross.
Hey Ross,
Do you think you could explain this a bit more simple step by step? I am interested in how this is done.
As I said before, there is a middle ground somewhere and I HATE the fisheye effect.
Stand closer to the image. It only looks fisheye because you are at the wrong distance from the picture.
Or you can use a stitching program like AutoPano to correct the field curvature. I'll see if I can find my mosaic of the Subiaco Oval grandstand and the images it was made from to demonstrate.
Baz, just to add to the confusion and to partially dispell the fisheye myth here are some images untouched from the 15mm Canon fisheye. you'll note that only straight lines close to the lens are curved, at distance if the horizon is centred you would be hard pressed to tell the difference. The fourth image is noticeably fishy because I deliberately framed it that way.
Last edited by acropolite; 23-06-2010 at 08:16 PM.
Thanks for the additional info and advice there Andrew. Good to know.
Phil, those 15mm shots are a bit of an eye opener. I can still see the effect and don't prefer it, but it's not as bad as I would have imagined. I've seen a lot worse. Thanks for that.
I use a 17-55mm f2.8 unguided for reasonable and relatively flat wide field astro, but the results have no comparison to a guided prime lens.
For APSC cameras and terrestrial photography the 17-55 makes a dam fine compromise lens for daily use. Note some users have had dust problems, I have not. It gets great reviews from professionals.
My 17-55 was bought 3months old with warranty for AUD850. So keep your eye out for people upgrading from ASPSC to full frame as this lens is APSC only.
Me again!
On the ff 5MK II, 28mm is about as wide as you can go without experiencing some curvature. The 24's give a wider field of course, but one starts to notice distortion in the corners. Its not too bad and can easily be dealt with by sorftware though.
A zoom lens will not give anywhere near as good result as a prime lens and shooting wide open is not usually a good idea with faster primes. In both cases you will notice "blobby" stars though you will decrease the chance of star trails (if shooting un-guided) if shooting at maximum aperture.
Hope this helps.
Again, as I asked in my other thread, different lens. Any filter recommendation for the 16-35? I have never used filters on my current gear, but I see a lot of others do.
Me again!
On the ff 5MK II, 28mm is about as wide as you can go without experiencing some curvature. The 24's give a wider field of course, but one starts to notice distortion in the corners. Its not too bad and can easily be dealt with by sorftware though.
A zoom lens will not give anywhere near as good result as a prime lens and shooting wide open is not usually a good idea with faster primes. In both cases you will notice "blobby" stars though you will decrease the chance of star trails (if shooting un-guided) if shooting at maximum aperture.
Hope this helps.
The zeiss 21 and the canon 14 have pretty much nil distortion in my experience.
At last a thread where I can talk about something I know!
You got very good recommendations already.
I'd add the Carl Zeiss Jena Flektogon 20mm F2.8 in M42 mount with an M42 to EOS adapter. That is cheap and quite good.
I wouldn't discard the fisheye, if you use DxO optics to develop the RAWs you can correct the fisheye distortion almost completely for excellent results.
Finally be careful with ISO, it is not true that higher ISOs produce more noise it's the opposite. The reason why high ISO images seem to have more noise is because of the shorter exposure times and not because of the ISO itself.
F4 30'' at ISO100 has more noise than F4 30'' at ISO200. Believe me.
Usually the signal is
S = [(So +/- Npre) * IsoGain ] +/- Npost
Where
So = Signal (photons)
Npre = Noise before ISO amplification (photon fluctuations + read noise + thermal noise)
Npost = Noise after ISO amplification (A/D conversion + rounding)
While Npre is multiplied by ISO gain Npost is not and that's the reason why as you use a higher ISO you get a better Signal to Noise Ratio.
At last a thread where I can talk about something I know!
You got very good recommendations already.
I'd add the Carl Zeiss Jena Flektogon 20mm F2.8 in M42 mount with an M42 to EOS adapter. That is cheap and quite good.
I wouldn't discard the fisheye, if you use DxO optics to develop the RAWs you can correct the fisheye distortion almost completely for excellent results.
Finally be careful with ISO, it is not true that higher ISOs produce more noise it's the opposite. The reason why high ISO images seem to have more noise is because of the shorter exposure times and not because of the ISO itself.
F4 30'' at ISO100 has more noise than F4 30'' at ISO200. Believe me.
Usually the signal is
S = [(So +/- Npre) * IsoGain ] +/- Npost
Where
So = Signal (photons)
Npre = Noise before ISO amplification (photon fluctuations + read noise + thermal noise)
Npost = Noise after ISO amplification (A/D conversion + rounding)
While Npre is multiplied by ISO gain Npost is not and that's the reason why as you use a higher ISO you get a better Signal to Noise Ratio.
Finally be careful with ISO, it is not true that higher ISOs produce more noise it's the opposite. The reason why high ISO images seem to have more noise is because of the shorter exposure times and not because of the ISO itself.
F4 30'' at ISO100 has more noise than F4 30'' at ISO200. Believe me.
I dunno about that.
The math sounds impressive but I have some landscape photos where the exposure length was about 2 minutes at 100 iso on my 1ds mk3. There is darn near zero noise in them.
If I did 2 minutes at 1600 iso I can't imagine it would have near zero noise in it.
The math sounds impressive but I have some landscape photos where the exposure length was about 2 minutes at 100 iso on my 1ds mk3. There is darn near zero noise in them.
If I did 2 minutes at 1600 iso I can't imagine it would have near zero noise in it.
The 100% crops which I can post if you like have pretty much zero noise.
I think you haven't understood what I said.
If you want to minimize noise the key is to expose to the right, in other words get as much light as you can, your first and always preferred way to do this is to increase exposure time.
In all your shots the key to low noise is to just shot at ISO100 with the longest exposure time you can without blowing any channel.
The key is that you shoot at ISO100 not because ISO100 has less noise (it doesn't) but because ISO100 allows you to expose longer!
Now a completely different scenario: Stars. If you don't want trails your exposure time is limited by your focal lenght. Let's say you can't expose more than 20 seconds.
Then you should shoot at the highest ISO you can without blowing any channel for 20 seconds.
The highest the ISO the better the S2N ratio will be.
In short words:
1) Expose as much as you can
2) When you can't expose more if you have room in the histogram increase the ISO
This is a new way of looking at it for me. I think I will conduct some interesting and informative test-shots at different temperatures at night and check out a 100% crop of each.