NGC 5823 - What ISO 200 can do ?

[jsmoraes]

GSO 305 mm - Canon T3 - ISO 200 - 5 x 4 min (20 min) - OAG - Skyglow filter - Coma Corrector

Canon resolution with full resolution mode in http://www.astrobin.com/113069/0/

For some DSOs the low ISO shows better star shape and color. (with my set of equipments, my sky, my Canon and DSS) ;)

[Camelopardalis]

Forgive me Jorge, but it looks a little out of focus to me :confused2:

[jsmoraes]

Dunk, this image hasn't a star with spikes, therefore it is hard to say about focus. Althought if you analyse the center of stars, they don't show problem with focus.
I didn't strong sharp. I always prefer some blur around stars, it seems to me more natural than a well delimited spot.

Really, I can not say if there is some problem with focus. At Astrobin, in full resolution mode, the image is with native resolution of CR2 RAW file from Canon T3: around 0.7 arcsec ( I prefer to think around 1.4 arcmin because of the oscilation of guinding). With my set of equipment and Canon when I have problem with focus the stars loose the round shape. They strech vertical or horizontal.

note:
I have only one identification of star in Astrobin.

If I consider the resolution of 0.7"/px, the star HD 133063 has 17 pixels, or 11.9 arcsec of diameter.

If I consider the resolution of 1.4"/px, it has 23.8 arcsec

If we find the apparent angular diameter of this star... we can know if the image is with good focus. I didn't find this information, yet.

[Camelopardalis]

My apologies Jorge - it just looked like the effect I get see with my camera when the stars are not focused, they bloat a little. However, when that happens I don't get to see the faint companion stars like you have shown here.

[jsmoraes]

Good observation, Dunk. Despite I have no spikes to analyse, the presence of faint companion with good shape and as a dot can be a good pattern.

I did the focus with BackyardEOS and lowest FWHM with a star not in binary system, but some times I am betrayed by refraction.

At Astrobin, with option of full resolution mode, you can see the photo with my Canon T3 resolution. It seems to me that was a good capture, At least for a Canon camera.

[Camelopardalis]

Yes, I downloaded for a look :) however, the brighter stars are about the same size as those I see at ISO1600.

What's the advantage with using ISO200 with this camera? I'm curious now :D

[jsmoraes]

Good question.Let me try explain it.

1 - Canon has a tendence to saturate the bright stars.

2 - Canon has a tendence to make all stars white because of the saturation in RGB values.

3 - My Canon (both 350D and T3) has much noise in Red channel.

With low ISO, for some DSOs like clusters, these problems are less intense.

I give attention to the histogram. If 120 seconds with ISO 800 is the same than 240 seconds with ISO 400, I use the ISO 400.
Analysing the first image, if the stars are saturated or all are white ...I change ISO to 200.

With nebula and galaxy it is more complicated. They are or have areas very faint. But I try to follow the same procedure.

There is an amount of photos (from Canon) that is the ideal to stack in DSS. Use more than it can saturate the RGB values of stars. But you need at least 10 photos to get good performance with DSS, for noise reduction.
With low ISO you can take largest number of photos without the RGB saturation issues.

Four photos of 60 sec has the "same" signal than one photo of 240 sec. But the four photos of 60 sec will have more noise than one photo of 240 sec. With low ISO I can use largest time of individual exposition.

[Camelopardalis]

This is brilliant, thanks for your tips Jorge :)

I had recently taken some exposures of globs and found them over saturated, and largely colourless as you describe...sadly, I got distracted by other fainter objects and didn't get a change to go back to them. However, I will try this the next time I get a chance :thumbsup:

Another question: what is the "ideal" number for photos from a Canon - I'm using a 1100D, which I believe is the same as the T3.

[jsmoraes]

Very good question !

For DSS, ten photos is a good number. It suggest you to use Kappa Sigma. I read from DSS team that more than 30 photos will not improve the quality of the result. At least to noise issue.

The number of photos with Canon 1100D (or T3) depends from the level of signal.
The problem is saturation of RGB values when you stack.

With CCD cameras you can do hours of photo. My experience says that it is not true for Canon.

Each case is a particular case. I did captures with 1 to 2 hours. 20 to 40 photos. And, in some cases, the result was best with less than the total number of photos. I lost time, because I didn't use all the photos.

With my set of equipments and my Canon, my experience says that to cluster, the total time between 20 to 40 minutes is a good time. The ISO and individual time of exposition depends from histogram, as I said.

For galaxy and faint nebulas, 1 to 1.5 hours. Bright nebulas 40 to 60 minutes.

Normally I try to make the distance from the left mark of the histogram and the beginning of the curve with 1/8 or 1/4 of the curve area.

The level (height) of red, green and blue curves with the most similarity. This change with different ISO.

I don't consider if exist or not width difference between the red, green and blue curves. This width depends from the target; nebula and cluster have different shape.

For nebula and galaxy some times I use more than one set of exposure. For example: 20 or 40 photos with ISO 400 and 4 minutes and 2 to 5 photos with ISO 800 (or 1600) and 2 or 4 minutes. I stack them together, as group, in DSS. Or I stack them separated, and mix them with aid of layer mask in Photoshop.

The problem with nebula and galaxy is to catch the faint information and do not get saturation of the bright core.

As you can see there isn't an absolute answer to the question. The quality of the sky, light pollution and seeing have much influence in the result.

note:

When you work with L,R,G and B filter and mono camera, you can balance the amount of information during the graphic process. But with color camera, mainly Canon, you can't do it.

Mono camera with filter has all pixels to catch specific color information.

Color camera have more Green pixel than Red and Blue, and has a pattern. For our Canon (1100D or T3) it is RGBG.

It means that the Red and Blue informations is separated by 3 pixels. We can see like Red - noting - nothing - nothing - Red. In other words there are three black holes between the red (and blue) informations. You can think the red (and blue) pixel having 4 times the pixel size.

You can't think as 4272 pixels divided by 3. For me, it works like 3 different cameras assembled as one. For example: the red camera has 1068 pixels with size of 20.72 microns, without the real sensibility of a pixel with size of 20.72. This giant virtual pixel works as a 5.18 micron pixel.

That's seems to me the great problem of color camera in saturation of RGB information. What the internal electronic amplifier (ISO) make with these informations... I don't know.

Maybe I am talking nonsense... but I know that it is easy to saturate the RGB information with Canon... this, I know !

[jsmoraes]

I forgot to say:
Because of the Canon bayer pattern, our stars are chubby ! :lol:

[Camelopardalis]

Great stuff Jorge thank you - this is great teaching for me :)

So it's the camera giving us the chubby stars and not the scope? That's good to know...I noticed on the pics I posted earlier that my stars didn't look any smaller in my wide field scope or my C8 :confused2:

For the price of the camera, I'm really happy with the results...it was almost certain to be my gateway imager ;)

[jsmoraes]

With the same set of equipments, mono camera, with similar number of pixels, will show better quality of stars than color camera. The Airy disk wil be an Airy disk. Not as an extension of the nucleous. Therefore, not chubby !

[Camelopardalis]

:)