I'm about to start playing around with long astro exposures with my 7D. Although I have a tracking mount, I also have light pollution, so I'm unlikely to go beyond 4 minutes or so.

Planning to use Backyard EOS ($30) to run the camera and automate the exposure sequence, including the taking of dark frames that will later be subtracted. The multiple short exposures will be stacked together using Deep Sky Stacker (free), and I think between the shorter exposures, the stacking, and the dark frame subtraction, I won't have any of the white noise.

Mark

n0b0 wrote:
Now let's talk over VERY long exposure noise.

26 minutes exposure test, ISO100 with zero Luminance and Chrominance NR settings.

As you can see, lots of noise, though for some reason, the noise I see in DPP looks more red and blue than the white dot noise after I converted the RAW to JPG.

Is it normal to get this much noise?

In advance: Please pardon the potential stupidity of this question... I have never done astrophotography.....

Is there any chance whatsoever that a good number of those dots evident in the 100% crop could be stars? Only reason I ask is that the vast majority of stars are not visible to the naked eye. A 26 minute exposure may be enough to catch them.... I have seen pictures from NASA that look a lot like this one. I doubt their sensors are that noisy.

I have found those white dots are not affected even by 100% luminance and color NR.

Hello

I had decided not to contribute to this thread, but since the dreaded 'sparkles' have been mentioned, This at ISO640 +.75 in ACR 6.7. No sharpening or NR in ACR. NR in Neatimage, with 75% noise removal and one dose of 'Sharpen more' in CS5
http://www.dpreview.com/galleries/4402957911/photos/2180359/sparkles2_mg_2146
Click on original to see the full size image

Mike Engles

pKai wrote:
Pixel Perfect wrote:
pKai wrote:

When home (hardly ever), I do have a 24" HP 1920x1200 external monitor, but I'm just as happy doing post on the laptop's 15.6" 1920x1200 screen alone.... The extra real estate of the 24" is nice, though...


What extra real estate? Exactly the same


Adding a second monitor of the same resolution DOUBLES the amount of available real estate, doesn't it?

Sorry I thought you meant your 24" monitor had more real estate than your laptop monitor at the same resolution.

Yes double the real estate.

I know so many of you are just devoted fans of ACR and LR, but have you tried a mix of raw converters, including DPP? I know of several people have seen strange image artifacts with 3rd party processors, but when they go to DPP, the images are cleaner. Just a thought?

Here's my contribution for this very interesting thread. To me, if one is shooting above ISO1600, you're in light that is beyond the capabilities of the 7D IMHO. Frankly anything above ISO800 in bad light is horrible. But for still subjects, ISO400 in really really bad light (like in the samples to follow) is still quite acceptable with a little post processing. I just used Noise Ninja on the BG leaving the subject masked out. Resized, WB adjustment, clutter removal and sharpened a touch.

1. Unedited from RAW file

Did you use LR to initially process this image? That 100% crop looks like ISO 3200 to me.
Be careful with LR; if you set the detail to more than about 15 at ISO 400 it creates a a graininess in areas of low texture, which makes the image look noisier than it really is. No ther RAW converter I've used does this and it took me a long time to realise this was why my 7D shots looked so grainy, even at modest ISO.

For my 7D my LR settings at say ISO 400 for sharpening and NR would be something like

Amount: 45-50
Radius: 0.7-1.0
Detail: 15
Mask: 35

LR: 12
CR: 20

As ISO increases I increase mask and reduce detail further. Even at ISO 100 I don't use more than 25 on detail, unless there are no important regions of low texture, in which case I may go to 35.

I used Bridge to open the image, Adobe Camera RAW 6.7 and then Photoshop to edit.

Hello

I bought a 7D in May of this year and within just one weeks use it became obvious to me that the 7D is basically a fairweather camera and Conrads's opinions echo mine precisely.

Mike Engles

mikeengles wrote:
Hello

I bought a 7D in May of this year and within just one weeks use it became obvious to me that the 7D is basically a fairweather camera and Conrads's opinions echo mine precisely.

Mike Engles




+1 Big time!

If the weather/light is bad, go spend time with your wife/girlfriend!

Pixel Perfect wrote:
Did you use LR to initially process this image? That 100% crop looks like ISO 3200 to me.
Be careful with LR; if you set the detail to more than about 15 at ISO 400 it creates a a graininess in areas of low texture, which makes the image look noisier than it really is. No ther RAW converter I've used does this and it took me a long time to realise this was why my 7D shots looked so grainy, even at modest ISO.

For my 7D my LR settings at say ISO 400 for sharpening and NR would be something like

Amount: 45-50
Radius: 0.7-1.0
Detail: 15
Mask: 35

LR: 12
CR: 20

As ISO increases I increase mask and reduce detail further. Even at ISO 100 I don't use more than 25 on detail, unless there are no important regions of low texture, in which case I may go to 35.

I agree completely. To me, the 100% crop looks more like what I'd expect at ISO 2500-3200 with the 7D. I can't figure out what has gone wrong with this one, but it certainly is not what I'd expect at ISO 400 or 800 or even somewhat higher. How much of an exposure adjustment was done before processing? My first guess is that it looks to me like the result of some extreme adjustments done prior to conversion.

For comparison purposes, below is an ISO 3200 shot from a couple of days ago and a 100% crop. Note that the shot was taken on an extremely cloudy day, and I normally wouldn't even bother to shoot on a day like that because I don't like the pictures from cloudy days for reasons other than noise. Also note that I don't have nearly the skills as some of the others do at processing, post processing, and dealing with noise. For this shot, I basically just used default settings for dealing with noise in ACR. The converted TIFF just had a slight shadow/highlight adjustment and some sharpening. While the noise is visible in the background of the 100%, I think it disappears nicely in the full image, and, at ISO 3200, I'll take it, which is why the image that Conrad posted at ISO 400 puzzles me. And, again, I know that some of the others posting in this thread could do a lot better with the processing than I did.

With my 7d, I am happy to shoot just about anything at 1250 iso.

Here is a sample at 1250 iso. 500mm lens f6.7 1.4xiii extender and a shutter speed a bit on the low side. 100% crop, processed in adobe raw only sharpen =31, radius=.5 detail=25, masking=51 luminanceNR=31 colorNR=43

Also, Overexposing 1/2 a stop is usually about right for me.

Conrads comment on 400iso or 800iso in bad light does not make sense to my way of thinking. I just use the iso needed for whatever light is there. Boosting brightness in post seems to cause more noise.

JP

pKai wrote:
n0b0 wrote:
Now let's talk over VERY long exposure noise.

26 minutes exposure test, ISO100 with zero Luminance and Chrominance NR settings.

As you can see, lots of noise, though for some reason, the noise I see in DPP looks more red and blue than the white dot noise after I converted the RAW to JPG.

Is it normal to get this much noise?

In advance: Please pardon the potential stupidity of this question... I have never done astrophotography.....

Is there any chance whatsoever that a good number of those dots evident in the 100% crop could be stars? Only reason I ask is that the vast majority of stars are not visible to the naked eye. A 26 minute exposure may be enough to catch them.... I have seen pictures from NASA that look a lot like this one. I doubt their sensors are that noisy.


No they're not stars because as you can see in the photo, all the stars became circular streaks due to earth's rotation. I didn't use any tracker to keep the stars still.

I really don't know what those white spots are, the only thing I can think of is hot pixels.

Conrad's ISO400 100% crop looks like it has a very strong NR applied on it.

uz2work wrote:

For comparison purposes, below is an ISO 3200 shot from a couple of days ago and a 100% crop. Note that the shot was taken on an extremely cloudy day, and I normally wouldn't even bother to shoot on a day like that because I don't like the pictures from cloudy days for reasons other than noise. Also note that I don't have nearly the skills as some of the others do at processing, post processing, and dealing with noise. For this shot, I basically just used default settings for dealing with noise in ACR. The converted TIFF just had a slight shadow/highlight adjustment and some sharpening. While the noise is visible in the background of the 100%, I think it disappears nicely in the full image, and, at ISO 3200, I'll take it, which is why the image that Conrad posted at ISO 400 puzzles me. And, again, I know that some of the others posting in this thread could do a lot better with the processing than I did.

Pixel Perfect wrote:
uz2work wrote:

For comparison purposes, below is an ISO 3200 shot from a couple of days ago and a 100% crop. Note that the shot was taken on an extremely cloudy day, and I normally wouldn't even bother to shoot on a day like that because I don't like the pictures from cloudy days for reasons other than noise. Also note that I don't have nearly the skills as some of the others do at processing, post processing, and dealing with noise. For this shot, I basically just used default settings for dealing with noise in ACR. The converted TIFF just had a slight shadow/highlight adjustment and some sharpening. While the noise is visible in the background of the 100%, I think it disappears nicely in the full image, and, at ISO 3200, I'll take it, which is why the image that Conrad posted at ISO 400 puzzles me. And, again, I know that some of the others posting in this thread could do a lot better with the processing than I did.




Les


Les what was the detail setting in ACR you used? Was it 25 which IIRC is the default. If so try moving it down to 5 as I can see a bit of that graininess pattern LR introduces with the detail slider in the eagles head feathers.

Wayne,

I don't recall, and I don't have access to the computer on which the file is at the moment. In any case, I normally wouldn't shoot under those conditions anyway. I actually took the picture for two reasons. First, after seeing some of the posts in this thread, I thought I'd give ISO 3200 a try. Also, I took the picture because I wanted to get a better look at what was under the eagle's talon than what I could get with the naked eye. I also should have pointed out that the full picture posted is actually cropped to less than 1/2 of the original frame, which further deteriorates image quality, and regardless of the graininess that may be in the bird's head at 100%, to my eye, it does disappear in the full image.

Les

John_T wrote:
I have found those white dots are not affected even by 100% luminance and color NR.

Did you use DPP or C1 I take it?

ACR pretty much gets rid of all the 7D sparkles automatically

I agree that something was processed incorrectly with Conrad's example. Also 1/5000th at ISO 400 isn't poor lighting as described, so I am not sure what happened there?

Here is a ISO 4000 raw with no JPG post processing, no noiseware, etc.



How about a really low light shot at 12800 that was shot to the right 1 full stop? Again no post processing, but the JPG is very easy to work on.



Final result after processing...

n0b0 wrote:
Conrad's ISO400 100% crop looks like it has a very strong NR applied on it.

Looks to me that way too. I'd be curious to see Conrad's unedited 100% crop with more moderate NR/sharpening settings from ACR or DPP.