Flat Frame Aquisition Question

[ammcdavid]

Hello all,
I spent several hours today learning a little more of the preprocessing needed before one can actually do any useful processing.....creating the foundation as it were.  Lots more to know in the digital world compared to what I did in film.  I did as was suggested by a member here and took 50 bias frames....that was kinda cool and easy using the Canon software and my computer to control the camera....it even dumps the resulting files to the folder of my choosing with the naming convention I created.  In PI I saved all files as fts so I could use the II tool to integrate them.  After tediously saving each individual file as a fts I learned that there is a Batch Conversion tool in the Scripts section......hey, I am learning!  So, I have darks, bias, but no flats.

Anyway, forgive the somewhat non-PI question but I have seen a few references to the "best" way to gather flats.  I searched on flats here and came up with some interesting processing threads on flats (imagine that on this forum  ) but nothing on actually aquiring the raw files.  I know some prefer a flat twilight sky and others use artificial techniques.  Personally, I would prefer the artificial method since I can control that to a large degree.  Can anyone point me in the right direction for gathering flats with a Canon DSLR?  How many of these should I aquire to create my Master Flat???

As always, your patience and guidance is appreciated.

Andy

[Carlos Milovic]

Hi Andy

I do DSLR, with common lenses. The way I take flats is using a rather opaque white plastic (can't remember the name) in front of the lenses, very close, and I take several pics, changing the camera orientation and position over the plastic (pointing to the sky, through it, plain daylight). It may not be a perfect system, but any large gradient may be corrected later with background modelation on the image.

[ammcdavid]

Hi Andy

I do DSLR, with common lenses. The way I take flats is using a rather opaque white plastic (can't remember the name) in front of the lenses, very close, and I take several pics, changing the camera orientation and position over the plastic (pointing to the sky, through it, plain daylight). It may not be a perfect system, but any large gradient may be corrected later with background modelation on the image.

Thank you Carlos for the quick reply.  You alluded to a opaque white plastic and, yes, I remember reading about this specific plastic but I too cannot remember the name of it.....something like ESD I think.  Anyway, is this a like a thick plastic pane (like Lexan) that you use to basically make a light box or is it more like a thin sheet that you simply drape over the camera while pointing the lens toward a uniform light source?  Sorry for the additional questions......I am also looking for some pointers on the proper technique in addition to the right materials for quality flat aquisition.  Thanks again.

Andy

[Carlos Milovic]

Yeah, the one used for the light boxes. I borrow one from Daniel Verschatse, each time I'm at his observatory. When I'm on my own, I do the same with paper or a white t-shirt.

[Niall Saunders]

Andy,

If you just want to 'get your feet wet' in the field of Flats - then just try photographing through some sheets of plain white paper. If you want to get a little bit more adventurous, build a very simple 'light-box' out of whatever you can find lying around (preferably 'opaque' ), and put in multiple 'layers' of white paper, spaced aboout 1/2" apart (-ish). These layers will act as diffusers, and if you keep the 'side-walls' white as well there will be multiple random reflection paths off the walls as well as back and forth between each layer.

Another method I have used is to make the diffuser out of 5mm thick white artist's foam-board. Yep, sounds crazy 'cos that stuff should be 'totally opaque'. But, it isn't!! With a bright enought light behind it (I use an EL panel about 5" away, and just leave it at 'full power') you will find that you need exposures of a few seconds to get 50% ADU readings (camera dependent, obviously).

Light box design is a great use of rainy-night time

[ammcdavid]

Thanks Carlos & Niall!  Sounds like I get to put my mechanical engineering degree to work  .  I will probably make the Rube Goldberg of light boxes  .  So much for instant gratification in this hobby...  .  I am sure I can come up with somthing useful.

Andy

[Carlos Milovic]

Lots more to know in the digital world compared to what I did in film.

I wanted to comment this. It is not that film were simpler or easier per se. It was a combination of two factors that kept all this preprocessing steps in a much more basic level:

- Image processing, aimed to the amateur astronomer, was just beginning. At that time, very few people did state of the art processing, or applied complex processes. Mainly, because PS was the only resourse, and computers were much slower than today. A deconvolution may take hours with old computers. I remember when I used to go for a walk after applying the noise reduction routine to my scanned images, with the first PI versions and my old Pentium of 133Mhz .

- Contrary to what a first glance may tell, preprocessing film data is by far harder than CCD data. It is non linear. Something as simple (today) as flat fielding was a real complex task. Yeah, you may end with a flat background, but stars and other objects literally were messed up.

So, yes, today there is much more to learn. But it is worth. Despite romantic memories, the results that are achieved today are superior to those with film. Even for starters, getting to a point were we get decent images has a shorter period. Everything is more dynamic today.

[Nocturnal]

I wonder if anyone ever flattened film images? I wouldn't know how to do it but then again dark room techniques border on the voodoo sometimes. I should ask my dad, he used to have a dark room in his (our) house.

[Carlos Milovic]

I had a dark room I did apply synthetic flats to scanned film images, with good results, but only as an approximation (now you understand where the Divide process, in PixelMath, came from? and now is in the Obsolete category...).

[ammcdavid]

I wonder if anyone ever flattened film images? I wouldn't know how to do it but then again dark room techniques border on the voodoo sometimes. I should ask my dad, he used to have a dark room in his (our) house.

Well, that is just it.....my experiences with film were just one exposure (no stacking) and only processing (stretching and unsharp masking for the most part).  I used slide film and actually got some decent photos....some manually guided up to 40 minutes in below freezing temps.  Now, I plan on getting to a point where I can stay inside and warm when gathering lights but on the processing and certainly pre-processing fronts, I have a very limited and/or narrow experience base.

Andy

[Nocturnal]

Oh I'm sure it's possible to scan film images and film flats and flatten on the computer. I was talking about flattening purely optically, without computer assistance.

[Carlos Milovic]

Well... if you shooted a gray card, to the dynamic range's center (neutral gray), and then positive the negative, I think that you may be able to use it as a mask that evens the illumination, to some degree Anyway, I think it would be a p*** i* t** a** XD

[ammcdavid]

Okay, as I ponder how to go about effectively creating some flat fields, I considered what I could use as a source of light that wouldn't be a point source that would create unwanted gradients.  Well, I have a thought that I think might actually work pretty well but then again...... .  My computer screen is a 32" flat screen and I figured out how to make the screen white and it is pretty uniformly bright and white all across.  I am planning on getting some white opaque plastic from a hobby store and putting that maybe even two sheets in between the camera and the computer screen & then take some flats.  I have pretty much convinced myself this should work but what do you all think???  At the very least, it is worth a try.

Andy

[Niall Saunders]

Hi Andy,

Yes, that approach (using a PC Monitor) has been used in the past, and is effectively not really any different from using an EL Panel.

The key is to ensure that any 'local luminance variation' has been well enough 'diffused' before it gets to the OTA.

Personally, I have found that ALL of my 'considered approaches' have worked for me. I have shone torches on expanded polystyrene foam (from several feet away) and pointed the OTA at that foam (on a side wall of the observatory). I have stretched a tee-shirt over an embroidery frame, and hung that over the end of the dewshield, and illuminated THAT with an LED torch, handheld (which guarantees 'random wobbly illumination ), from the other end of the observatory. I have a single 12V torch bulb, still in its original 5" parabolic reflector, shining through five layers of 'photo-quality' inkjet card stock (which has a pure white 'chalky' surface on both sides), each separated from the next by an inch or so, and this is what I currently, and very successfully, use on my Moonfish ED80. I have an A4 EL panel shining through a single layer of 5mm white artist's foam-board, at a distance of about 5", and this is what I currently use on my 8" LX90 SCT. I eventually plan to change the ED80 lightbox over to an A5 EL panel - but will probably wait until the torch bulb 'blows'

Basically, if your design, when carfeully considered, 'feels right', then it will probably 'work right' too. Granted, there are a whole bunch of SkyFlat 'snobs' out there who will tell you that ANYTHING other than properly acquired skyflats, under a clear New Mexico sky, are just not worth the effort. But I can also think of a whole load of folk who will just say 'make something, test it, and if it seems to to the job it is probably good enough'.

And, in any case, if it looks as if it was made out of old fish boxes - but it still does the job - it makes your imaging results all the more impressive