Sub exposures - how 'bright' should they be?

[mg (Mitchell)]

Hi everyone.

I have a question regarding subs and how "exposed" they should be.
I usually just make sure the histogram is away from the left edge, but would like to know how others do it. I use a DSLR.

I've read a bit on the net but it's hard to find real examples, and a picture paints a thousand words!
Would be great if someone could post a single sub / histogram that shows what they look like straight out of the camera. That would help me with getting my exposures correct.

Thanks!

[nebulosity. (Jo)]

Good question, I'm interested in this as well. Was reading a great bit of info the other night (can't remember the website sorry) about dslr imaging and it was recommended that the left side of the histogram should have around about a 15 to 20% gap from the lefthand edge. The idea is to get the faint details lifted out of the noise level. Exposing any more than this would have no advantage as it would cut down the dynamic range.

Not sure if that answers your question, I'll have a look and see if I can find the website again as it has some great info.

Cheers
Jo

[2stroke (Jay)]

As much as you can without clipping and getting blooming if your camera doesn't have anti bloom. Never really had any issues apart from blooming with a few extremely bright stars. You just don't want your brightest pixels exceeding the graphs range other wise you will your loosing data. Take a few images of the moon and see for your self, do 1 at .5 secs and another at 5mins and you will see. Maxim has an anti-blooming feature which helps for bright stars to btw. Thats a 4min sub as you can see i could go way more, it will clip with the brightest star first and bloom but who cares i could fix that. Just experiment for your self, lol trying to find the bloom i had with 1 star a while back now to show you the graph from that.

[mg (Mitchell)]

Originally Posted by nebulosity.

Good question, I'm interested in this as well. Was reading a great bit of info the other night (can't remember the website sorry) about dslr imaging and it was recommended that the left side of the histogram should have around about a 15 to 20% gap from the lefthand edge. The idea is to get the faint details lifted out of the noise level. Exposing any more than this would have no advantage as it would cut down the dynamic range.

Not sure if that answers your question, I'll have a look and see if I can find the website again as it has some great info.

Cheers
Jo

Sounds like the site I was reading too, and thats the basic principle I have been working on.

Originally Posted by 2stroke

As much as you can without clipping and getting blooming if your camera doesn't have anti bloom. Never really had any issues apart from blooming with a few extremely bright stars. You just don't want your brightest pixels exceeding the graphs range other wise you will your loosing data. Take a few images of the moon and see for your self, do 1 at .5 secs and another at 5mins and you will see. Maxim has an anti-blooming feature which helps for bright stars to btw. Thats a 4min sub as you can see i could go way more, it will clip with the brightest star first and bloom but who cares i could fix that. Just experiment for your self, lol trying to find the bloom i had with 1 star a while back now to show you the graph from that.

Thanks for the example. My histograms are usually about the same as this. I might try pushing it further and see how I go. I can always hit delete.

[cometcatcher (Kevin)]

I just make sure the RGB channels clear the left edge. Too bright just increases noise over signal.