Flats question

[bigal1]

Hi folks well i've started to take flats although for now only using a laptop screen but it does seem to have made a big difference in processing but what i wanted to know is do my flats have to be taken at the same iso as my lights ? a lot of sources on the net say to use the lowest iso for flats ( iso 100 ) but when i stack in DSS is warns me that the flat iso does not match up with the lights ? they still seemed to work but it got me questioning my methods , any advice is much appreciated thanks .

Regards Alan

[Black Knight]

I use the same ISO for flats as I do for light frames. Laptop screen like you, camera set for Av mode, with a t-shirt secured over the scope using a rubber band:). I aim for 30 to 60% on the histogram.

Sent from my mobile using Tapatalk, so please excuse the speeling and granma!

[rfdesigner]

on a point of scientific priciple I would always use the same ISO.

you want to eliminate all possible variables other than the one thing you're trying to measure (in this case star light), voluntarily introducing another variable is bad practice. I think in practice it won't make much if any difference, but I still wouldn't do it.

Derek

[bigal1]

Thanks for that guys

[Spikey]

I use Av mode with my 1000D Alan but thats with a flats plan on APT. Before that I used to use ISO 100 and just try different exposure lengths until my histogram peak was between 1/3rd & 1/2 of the way across.

[steppenwolf]

I like the idea of not introducing any more variables but in reality a flat is only capturing the distribution of brightness in the light cone so it shouldn't make a difference what ISO you use but it can't do any harm to match the lights on this one parameter can it!

I hate flats but they sure do make a huge difference to your images....

[Jessun]

I hate flats but they sure do make a huge difference to your images....

Hear hear...

[ollypenrice]

Yup, a pain but essential. Actually they are not a pain once you have a system that works for you but the system I use on my own scopes doesn't work on Yves 14 inch and we are still faffing around to find out why.

'We do not do these things because they are easy.'

Olly

[vdb]

lowest iso ...

why? you can introduce noise with flats, and as your flats are not shot with the same exposure time as the lights, the equal iso makes no sense at all ...

For this reason some processing software will even blur/denoise flats as a processing step ... some go even that far to shoot flat darks to remove as much noise as possible ...

Yves

[ollypenrice]

Good point from Yves. Shoot in the lowest noise format possible. Maybe try to do your flats when it's cold as well if you have no cooling.

Olly

[rfdesigner]

In electronics it is terribly easy to time things nearly perfectly. So we can vary time in images at will and account for it, hence using short flats, this will hold for DSLRs as they have shutters, for shutterless CCD cameras I'd be more concerned.

When we try and measure voltage/current/resistance we do very badly. We're pretty chuffed at work to measure 1 ohm in 47k on our temperature sensor, but we can measure to 1 in 1million when it comes to time without any problem what so ever, therefore so long as there is a good shutter I'm happy to use different length shots.

So when it comes to ISO settings for flats what I'm worried about is that the bias settings on the CMOS chip itself will be in some way altered such that the pixel to pixel offsets etc. dont match perfectly between ISO settings. This isn't a problem with CCD imagers, but I'd need to study the schematic on a CMOS sensor before decalaring it free from any variation between ISO settings. Certainly if there is more than one readout ADC in the camera then the ISO for each ADC will be subtly different. i.e. half the frame is ISO 199 / 401 and the other half is 200 /399, under this condition it would be better to fix on just one ISO setting. When doing flats it isn't hard to simply take more frames to get the noise down.

Re smearing Flats. I would never concider this a good theoretical idea, although it may be desirable in practice. When the pixels are formed on the CMOS / CCD sensor the area of each pixel will not be perfectly identical due to imperfections in the lithography. So the sensitivity of each pixel will not be identical, my atik383L+ has some noticable variation. By not smearing we have the opportunity to correct for this variation in pixel area. However we would need to take a good many flats with noise well below the varition we are trying to correct, if we don't do that then we are better off smearing to reduce the noise and just live with the pixel to pixel gain error.

Of course all the above is silicon theory, what we do in practice will shy away from perfection for any number of valid reasons, and boy do some of my shots shy away from perfection

Derek