Hi

I’ve been taking flats and am getting a bit confused by the fits viewer.

When the stats show 16384 (my camera is 14bit so I believe that’s the maximum) for my images surely that means that the pixels are blown out? If it’s that for the averages, minimum surely the whole image is so why is it only a lighter grey? Is it rendering things to the equivalent of 18% grey?
If I click on the red button to show pixels that are saturated I’m not seeing anything change on either my 16384 flats or on star images with some pixels at 16384. Could someone tell me what’s supposed to happen in case I’m missing something please?

Also the histogram doesn’t always look like it’s updating between images and seems inconsistent with what is being reported in the stats. Is there some way of zooming in so I can see whether the exposure is up against the left edge or not and what is the bottom scale in?

Sorry for all the questions just happened to be a clear night (well what passes for night in the uk at this time of year - more like dusk at 1 in the morning:) ) so had ended up actually trying to take some pictures rather than just reading about taking pictures….and then the questions came…



Regards

Nigel

ZWO camera scales output to full 0-65535 range even thought they are 12 or 14 bit. To be sure take flat with like 10 second to be 100% that it pixels are saturated. Then measure maximum. AFAIK that red button is just hardcoded to 65000. I succesfull use automatic flat wizard where I set target ADU 40000 and just let Ekos do everything else.

The sony is definitely 14bit. I’ve had a look at a few of my lights and the centres of the stars hit no more than 16384.
I’ve now tried an led light panel and tried the auto flats function which gave an exposure of approx 0.6secs for an adu of 8000+-1000 and that seemed to give a nice mid range exposure.

If the saturated level is hard coded to 65000 then I think it would be pointless for dslr type cameras and probably a lot of others. Perhaps if it cannot get the actual setting from the driver perhaps a setting for the viewer or a guess based on a maximum showing up equivalent to 12, 14 or 16 bits could be used.

Nigel,

It may be decoded using 14bit but that does not equate to an actual brightness level only to the number of differentiated levels.
The 14bit (16,384 levels) values are returned as 16bit (65,536 levels) values so each level will be equal to 4 when shown.
This means you should think in terms of 65,536 when setting a flats level.

Paul

Hi Paul

The sv305 driver has an option to basically bit shift the 12 bit reading it gets to 16 bits. However I’ve not seen such an option in the sony dslr driver.

If what you say is true then something else appears to be wrong as I’ve yet to see any of my images pass 16384. I’ve looked at them in fits viewer stats and astroimagej.

The 3d plot in astroimagej of my flats shows a nice dome shape due to vignetting. However the ones I took using the t-shirt method showed a flat area in the middle compared with ones taken with a led panel. To me this indicates overexposure and this again occurred at 16384.

The cameras in your sig both appear to be ccd and look to be 16bit. Do you have any images from a cmos 12/14bit camera where they have been multiplied up?

Regards

Nigel

Nigel,

Quite happy to learn something if I am wrong.

There are some cameras out there with larger than 64K and I am pretty sure they still display only 64K levels.

Paul