Shutter speed

[Sniper]

Hi all!

Is there by any chans a way of getting a longer shutter speed than 8 seconds in a G1???

I read that somebody had taken a photo with an aperture value of 9.5. My camera stops at 8.0. Do I have an old firmware or is there a trick that I have missed!? :9

As you can guess I´m quit new to this, and I´m happy for answers.

Thank you

[oops]

Mine stops at 8 also. The only option to this that I can imagine is a different iso setting; ie: you can increase ccd sensitivity times 2 with 100. This should result in twice the seconds; 16?

Every time I think I'm right, I'm wrong. Oops!

[Balthus]

Since I'm not an expert I may be wrong, but maybe some expert could say if it's correct or not.
Increasing the sensibility will introduce noise in your picture (hot pixels). It will look grainy with some strange spots.

What I was thinking to take long shot is to take 3 or 4 short shots and use a program that would make an "addition" of these pictures. But I don't know if it's feasible and if so, how to proceed?

I've heard about another technic which is called "Dark field substraction"? But I have no idea what it does and how it works...

Can some expert give an opinion?

[gandini]

balthus wrote:
Since I'm not an expert I may be wrong, but maybe some expert could say if it's correct or not.
Increasing the sensibility will introduce noise in your picture (hot pixels). It will look grainy with some strange spots.

This is true, and the G1, like all other digitals, will show more noise at higher ISO ratings. Try to keep as low an ISO as practical.

What I was thinking to take long shot is to take 3 or 4 short shots and use a program that would make an "addition" of these pictures. But I don't know if it's feasible and if so, how to proceed?

This doesn't make any sense to me. Say the scene has a luminosity range of 0-10 (10 is brightest value). If you take a short exposure, you will only capture the brightest, say, 25% of the scene--those areas with luminosity values 7-10, and the range of values in your final captured image will be 0-3. All scene values of 0-6 will record as 0 on the sensor as insufficient exposure was given for them to active the sensors. You can keep doing this, say 3 or 4 times, as you suggest, but all you'll have is 4 copies of a very dark image.
If you try to "add" them using photo imaging software, you'll never be able to successfully re-create all the original scene values of 0-6 since they were never recorded! This won't work. To record a scene value of 1, you must expose sufficiently for this value to energize a sensor to that value (10% above dormant in my example.)

I've heard about another technic which is called "Dark field substraction"? But I have no idea what it does and how it works...

Can some expert give an opinion?

Dark field subtraction, as I know it, refers to a technique to eliminate hot pixels (it won't eliminate dead or stuck pixels) that occur with long exposures (endemic to CCDs) by taking a zero-light exposure, creating a mask in say PhotoShop and using subtractive blend to rid the offending pixels from your long exposure image. Do a search for hot pixel removal, or the like on dpreview for more, and better, explanation.

cheers.

[Balthus]

Hey, I've found an interresting experience about what I previously said...
Have a look there :
http://www.virtue.nu/rozee/composite/composite.htm

The guy took 28 pictures in a nearly total darkness and combine them together...

[Sniper]

This got really interesting. Can you or can you not make an "addition" of pictures in order to create a great picture of the stars, for example?

Unfortunatly my G1 is shipped away to got some dead pixels fixed, so I´m not able to test it myself. But soon I will.

But this is still very interesting.....

[Balthus]

I don't know how it has been done...

But as Gandini said, it seems that there were sufficent light for the ccd to capture the image but with lot of noise...
I tried with the first image in the dark to auto adjust the level in PS5. The image was more bright but with a lot of noise.

I don't know how this person combined them together, if it's just by layering the pictures or if he applied some transformations between layers to make the noise disappeared??

Gandini, do you have an idea of how it could have been processed?

[gandini]

balthus wrote:
I don't know how it has been done...

deletions...

Gandini, do you have an idea of how it could have been processed?


This *is* getting very interesting! I (we) should perform some experiments, but in the meantime, I've analysed the website you gave us.

The original image seems to have a luminosity range of about 0-3, as I predicted. The author then auto levels the image, which expands the input to a full range (0-10) output, which appears quite blocky and very noisy. Still, it's impressive to "see" how much information the CCD is recording!
The next image (#3) is the full 28 image product. The key seems to be the processing of this image. The 28 originals are blended using the screen mode. From PS help files:

"Screen Looks at each channel's color information and multiplies the inverse of the blend and base colors. The result color is always a lighter color. Screening with black leaves the color unchanged. Screening with white produces white. The effect is similar to projecting multiple photographic slides on top of each other. "

Auto levels makes sure the output is full range, desaturation will reduce any noise in the color channels.
The final image, when compared to the lit color image below it still shows more contrast, and less information, but pretty good given the limited input information.

I still contend that the process can't process information that's not there in the first place, but I must modify my comments to reflect what appears to be the presence of more information than I would ever have guessed.

cheers,