Why low ISO and what about long time exposition ?

[jsmoraes]

Well, you can see it. Better than many words to explain !

All photos single RAW converted and reduced to JPG in Photoshop CS3. No process was done.

photo 1 60 sec ISO 400
photo 2 60 sec ISO 800
photo 3 60 sec ISO 1600

photo 4 60 sec ISO 400
photo 5 120 sec ISO 400
photo 6 180 sec ISO 400

[dannat]

you have increased signal but noise has also increased -you would have to trim the blk/wht areas of your image to bring the noisier images back to a blk background

[nebulosity.]

That's interesting Jorge, I do benefit from higher ISO's but that's because I have really dark skies.

Jo

[jsmoraes]

You are right. Some times we need high ISO because the target is very faint. But it is not a LAW.

Maybe the most important is that some areas or details of target would be missing inside the background glow.

M42 isn't a good example, I know. But you can see that you loose all information and details of Trapezium.
Aniltak has nebulosity around. That nebulosity will be lost too.

Many people ask me why I use low ISO, sub with no more than 4 minutes. Well those photos show that: in my special case the ISO 400 and 800 are the best for most targets. Not all targets.

More than 4 minutes... and I will capture only the sky glow !

The camera used, 350D, is more sensible, because it hasn't the loss in IR filter. But it would be very similar to camera 1100D.

Did you see my photo published here from M 42 in another message ? With my Canon 350D I don't believe I can do more than 2 minutes for subs. It will explode the target !

[cometcatcher]

Just looking at your 60sec ISO 400 images compared to mine. Your ISO 400 images are very bright. Much brighter than mine. My Pentax RAW files are probably different in nature to your Canon, but still there is a big difference.

What light pollution colour zone would you live in Jorge? Orange - Red?

[cometcatcher]

Okay just looking at the light pollution map for Saquarema, RJ , Brazil. It's not that bad. Yellow, green and blue zones. And you are a long way away from Rio.

How bright is your observatory? It appears that you are getting light leak in from somewhere.

[dylan_odonnell]

I might be wrong here, but your camera has a "native" ISO usually around 100-150 and setting it to higher ISOs simply increases the entire exposure (noise and all) in the camera instead of later in your post processing.

In this sense, it's convenient to be able to see your target in your subs and then stack for noise reduction etc which is the benefit of higher ISOs. Ultimately you could shoot everything at ISO 100 and amplify everything yourself in post later if you wanted.

I tend to shoot at a "high enough" ISO to see the target but theres really no need to push it higher in camera if you are going to push it later in post anyway.
Please let me know if I'm off track!

d

[dannat]

yes many diff cameras will have a diff most effective base iso for astro tragets -some cam's stop at iso400, others at 800, 1600, 3200 etc -best to test your individual camera to see what its optimal iso is & stick with that

[jsmoraes]

Yes, Kevin. Actually I don't know what colour zone I am. I have light pollution form Saquarema and Araruama city. They are 14 km far from me. And I have some street light very near my observatory. I needed to build a new one because the light incoming inside my OTA !
Look my old observatory here: http://astronomia-e-astrofotos.1069742.n5.nabble.com/file/n283/cf-9097.jpg

And my new observatory here: http://astronomia-e-astrofotos.1069742.n5.nabble.com/file/n1762/IMG_1062A.jpg

Best view of the new here:
http://astronomia-e-astrofotos.1069742.n5.nabble.com/Reforma-do-Observatorio-Vilatur-td1723.html

Now I haven't problem with incoming light.

As I always say my main problem with light pollution is my proximity to the beach and a lake that is source of salt. I have that salt fog atmosphere. I need wash my main mirror each 4-6 month !

I have much reflexion of street light on this salt fog creating a glowing dome around me.



yes, you are right. I don't know the native ISO of 350D, but as its maximum ISO is 1600, probably its native ISO should be around 400. Just the ISO I most like to use. With Canon 1100D maybe the best ISO, for me, should be 800.

The number of subs is my main strategy to get quality and amount of details.

Now I am perceiving an increase of noise, mainly in red channel. Only in red channel. I back to 350D to test if it has something with the IR filter. My 350D hasn't it.

Any Halfa photo with my 1100D presents very too much noise. A granulation white large spot noise. As little stars.
I saw much less noise, and normal noise, with 350D.

I don't know if this increasing in noise has source on this filter or if my atmosphere here, summer, was changing. I have an ozone hole over Atlantic near South America. Maybe I am having much intensity of IR and NIR radiation that is causing this amount of noise in red channel. Actually I don't know.

M 42 isn't a good object for test, but it was in good position after I installed and calibrate focus in Canon and guiding cameras. You can see the quality (despite of overexposition) of Halfa with 350D in http://www.iceinspace.com.au/forum/showthread.php?t=130271

350D has problem because it needs more cable and USB port to remote control. Very old, it has bad contact in connectors. Less resolution than 1100D, its pixels are larger. But if the noise in red channel is less, I will return to use 350D.

I published this message because there are many ask about why I use low ISO and short sub exposure. I do it because of my site. Each case is a case. We can not think as LAW: you must use high ISO and long exposition subs.

The same about the focus issue, here in my site. The same salt fog atmosphere produce an intense refraction. My stars never stop to dance. They change continuouly its shape with much intensity. My set has high resolution and it perceive it. It is very hard to say: well, I am with good focus.

I don't know if it is the same for you, but for me the focus with strong stars is different than with faint stars. I can have appearance of good focus with brightness stars and bad appearance with faint stars. And focus faint stars that never stop to dance... it is problem.

My stars always present a more chubby shape because it was the sum of those distortions, causing an appearance of out focus. But if you see the line of spikes, you will see that they are single line. So the image is near the good focus. Perhaps, in good focus.
:thumbsup:

[rmuhlack]

Native ISO or "unity gain" is a flawed concept, as described here by Roger Clark: http://www.clarkvision.com/articles/digital.sensor.performance.summary/#unity_gain

[jsmoraes]

I am reading many texts where mates discuss about noise, noise source, noise theory, noise distribution and noise of noise.
But I think that many people are forgeting the eletronic noise caused by the transistors. No mater if it is integrated circuit or an old single transistor. There is noise with solid state components. More than the old vacuum tube.
And just because this you need high refrigeration of solid state electronic circuits.

I understand as native ISO of a camera the area linear of the amplifier, where these noise is with less value to SN rate. The noise is constant up to 60-70 % of the curve of amplification. After that position you enter in the distorted curve area of the amplifier, where the signal has less amplification than the noise generated. Is the condition near of saturation level of solid state electronic compoments.

People that work only with photography don't consider those issues. But when you use the camera as plain people you see it very well. The same time, same object, same place, same luminosity, same camera, same sensor, same all things... and if you change from ISO 100 to ISO 6400 you perceive the increment of noise in the photo.
Many of noise sources are equals, and generate the same intensity of noise. The unique thing really different is the amplification, made by the amplifier.

All knowledge of sources of noise discussed are very important. And can be used to improve the quality of a capture.

But as our automobile, car, even with the best mechanics, you need gasoline. Your car stop to work. Doesn't want to work again. First question: Is there gasoline in the tank ?

The best sensor, used by NASA, if the amplifier is a trash... the photo will be a trash. The amplifier is the gasoline to car. Bad quality... bad performance.

The other electronic task inside the cameras: convertion analog/digital doesn't cause many noise. It only can limit the range of information: 8 bits, 12 bits, 14 bits or n bits

Conclusion:
quality of the sensor and the electronic amplifier are the most important factors. People only discuss about the sensor and its characteristics.

Richard, the text of the link seems to be very good about the noise. I will copy to my computer and read it again with more attention.
If we can not deal with the quality of the amplifier, we must manage very well the others factors that harm our photos. ;)

:question:
note: if the camera is a trash... no knowledge about noise will help :lol:

Will be it the case of my Canon 1100D with the red channel noise issues ? :confused2:

[Camelopardalis]

I use ISO1600 almost exclusively with my 1100D when at a dark site...after reading a bunch of threads on SGL (and elsewhere) about signal to noise performance, temperature, etc.

Beyond ISO1600 the red speckle with my 1100D is unbearable. At ISO1600 in cooler temps it's almost non-existent, whereas in summer temps, it's rears its head in subs beyond 2 minutes. In winter, somewhere between 5 and 8 minutes I was getting images that were over-saturated.

Maybe a light pollution filter is worth considering?

[John K]

I find with my Canon 400D ISO1600 and 200 seconds is the optimal.

The 350D and 400D are very similar I think.

I use to think that perhaps longer exposures in lower ISO would provide benefits, but if you Google for graphs of noise/ISO/temperature for your camera you will find I think that ISO 1600 will be optimal for your camera - shorter exposures at higher temperatures and longer when it's colder.

[jsmoraes]

warning: 1600 is the maximum ISO for 350D

[cometcatcher]

Problem is those old cameras are pretty noisy at high ISO. Modern cameras are better at high ISO.

[rmuhlack]

My experience with a 400D was very similar, with ISO1600 (the maximum for this camera) and sub exposure times of 180-240s giving the best results: http://www.iceinspace.com.au/forum/showthread.php?t=106775&highlight=snr

[glend]

But those old cameras produce much less heat and thus are less noisy at their optimal speed - read the thread on DSLR cooling box.

[Camelopardalis]

Regardless, my 1100D behaviour more or less correlates with Richard's plots. Around 15C or higher, 2 minutes is about my lot. Below 10C 3 and 4 minutes become practical, below 5C the darks are almost completely clean. Maybe it's because the 1100D is an "old" camera ;)

[jsmoraes]

Wow ! By my side I never have less than 30C. Normally I work with 35 - 45 C. Few times I must work with 45 - 55 C. Umidity never less than 50 %.

Perhaps this is the reason to use ISO 400 and subs with 4 minutes, or ISO 800 and subs 3 minutes. Never more ISO. The time of exposition is because of sky glow.

[Camelopardalis]

Sounds like you need a cooler box Jorge ;)

[ZeroID]

I'm getting 90 secs or better from a 450D at 1600 ISO but ambient at night is about 12*-16*. A cooler box should get me down another 10* or more. Currently the sensor tends to stay about 24-25* after a few frames have been taken. 10* lower is another 75% noise reduction according to most information I've read.
35-45* must be cooking the sensor just about, noise will be a big issue. I'd seriously look at either a Peltier system or even a water\ice\glycol circulating system to get it down.

[jsmoraes]

yes, Dunk. I agree... Although... Canon + home-cooler = best buy ATIK camera with cooler or QHY camera with cooler.

Better camera, less noise, less weight, less problem to focus with OAG (you don't need spacers), less problem with distance of Coma corrector, less difficulty to intall. And perhaps others less ....

My actual problem is the less money to solve the problem ! :sadeyes:

However, it always will be done ! With or without noises ! :thumbsup:

[Camelopardalis]

Have a look at this thread Jorge... http://www.iceinspace.com.au/forum/showthread.php?t=129469

I'm working on mine (slowly) but the individual parts should be relatively inexpensive, and if you have any old PC parts laying around you might be able to save there too.

It's not likely to give the temperature differential of a cooled CCD like an Atik but I figure 10-15C would be a good start.