newbie ccd astrophotography

[ollypenrice]

I neither refocus between filters nor shoot flats for each colour. A lot depends on your optics. If they are very well colour corrected and are not at an exotic F ratio (and you are not working at extremely fine pixel scales) then this should work for you (based on my experience.) Why is oversampling bad? Because it's a waste of time, quite literally, and time is the one darned commodity you can't buy in this game. Well, you can buy a multiple rig, of course... 

In practice I have found OSC slower. I ran both OSC and mono Atik 4000s for a few years. On bright targets I found the OSC not very different from the mono and said so in an Astronomy Now shootout article using M42 as a test case. However, the fainter the the target the greater the mono speed advantage became clear. If you are shooting a faint Ha target, for instance, your OSC is stuck at about 25% efficiency compared with a mono which can catch faint Ha on all its pixels.* It does depend on the target but there are no targets on which OSC is faster. Take a difficult case like the Squid Nebula. I seriously doubt that you would catch it all all in an OSC, which is why nobody did. Catch it in a mono with O111 filter, however, and you can see it it a single long sub. As you can see Simeis 142 in a single Ha sub. It is way faster to blend NB data into (L)RGB than it is to try to haul NB data out of broadband filters as fitted to OSC or used with mono.

Olly

*In fact this is being incredibly kind to red filters (whether in  OSC or mono) bacause the higher contrasts of the Ha filter can be processed into the final image quite easily.

[DaveS]

If you're plagued by LP as I am then NB gives you a way to cut through it that OSC can't, even with a LP filter. Can also image in moderate moonlight which somehow tends to coincide with clear nights.

[Thalestris24]

"UK stop-start-stop-scream imaging conditions" - yep, that just about sums it up! Plus add "damned city lp"! I've even been using dslrs lately rather than my cooled osc ccd simply for the convenience. Mind you, I hope to get a cooled mono dslr in the autumn

Louise

[kalasinman]

FYI on the mono-NB IS faster side, see this thread where Olly beats 90 minutes of exposure on NaN of a new Nikon D810a with 41 seconds of total integration with a CCD and NB.

http://stargazerslounge.com/topic/246119-first-light-d810a-dead-of-the-ccd/page-2

[Geordie mc]

Interested to hear that you don't take flats for each filter nor refocus between filters. I have always done this as I read that you should. I will try without and see what happens. I would have thought that temperature change during a long session would effect focus though, even if the filters were genuinely parfocal.

[swag72]

I have to say that I have never found my filters to be parfocal. So I always refocus between a filter change and also at 1 degree temperature changes. An auto focuser has revolutionised my imaging for sure.

[Physopto]

Astro Engineering a Telrad dew shield from Astroboot , £5 plus p&p. Bargain and it works. Make sure that you flatten it out for a pattern.

Floppy A4 files are ideal,

Nick.

I have to say that I have never found my filters to be parfocal. So I always refocus between a filter change and also at 1 degree temperature changes. An auto focuser has revolutionised my imaging for sure.

Sara

What F ratio are you imaging at, also are all the filters from the same manufacturer and revision/batch?

Derek

[kalasinman]

FYI on the mono-NB IS faster side, see this thread where Olly beats 90 minutes of exposure on NaN of a new Nikon D810a with 41 seconds of total integration with a CCD and NB.

http://stargazerslounge.com/topic/246119-first-light-d810a-dead-of-the-ccd/page-2

Sorry, taken in by a joke.

[Stub Mandrel]

So who said it was bad?

The FLO calculator sends  a disparaging message about unsuitability if your sampling rate is too high, and my earlier message got the reply "an OSC version of the camera would be stuck at an unproductive 0.4 arcsecs per pixel or a still unsatisfactory 0.6 with reducer."

I realise that oversampling is a 'waste' of data, but most people can't afford a different camera to suit each different optical setup. I just wanted to know if there was a practical objectioon to oversampling such as increased noise, artefacts or significantly reduced efficiency (I have worked out that smaller pixels have smaller quantum wells, but I think LP will get me long before I collect too much light from DSOs!)

[Geordie mc]

Saving my pennies for an autofocus unit as we speak. Any recommendations much appreciated.

[swag72]

Sara

What F ratio are you imaging at, also are all the filters from the same manufacturer and revision/batch?

Derek

I use Astrodon 3nm's for my narrowband and Baader for RGB's - I've used them in an F10, F5, and F3.9 system and never once found them to be parfocal (obviously the Astrodons with each other and the Baaders with each other) - I've also used them in different optics, an SCT, RC and a refractor.

It doesn't bother me as I have all of the offsets loaded in SGP and so on filter change everything is done for me.

I base my opinions on parfocality with this experience of the use of different scopes and f ratio's.

[kalasinman]

The FLO calculator sends  a disparaging message about unsuitability if your sampling rate is too high, and my earlier message got the reply "an OSC version of the camera would be stuck at an unproductive 0.4 arcsecs per pixel or a still unsatisfactory 0.6 with reducer."

I realise that oversampling is a 'waste' of data, but most people can't afford a different camera to suit each different optical setup. I just wanted to know if there was a practical objectioon to oversampling such as increased noise, artefacts or significantly reduced efficiency (I have worked out that smaller pixels have smaller quantum wells, but I think LP will get me long before I collect too much light from DSOs!)

The FLO calculator doesn't take images.

It is long focal length that creates tracking and guiding dificulties.

If one has an OAG, as I do ,using a guider with small pixels gives a wider FOV for the guider (so more stars) and makes finer tracking adjustments. Pixel scale still must be within reason. For example I'm imaging at 1.3" and guiding at 1.08". If seeing is poor, I bin 2x2 or even 3x3, but if seeing is that bad, imaging may be a waste of time anyway.

[ollypenrice]

FYI on the mono-NB IS faster side, see this thread where Olly beats 90 minutes of exposure on NaN of a new Nikon D810a with 41 seconds of total integration with a CCD and NB.

http://stargazerslounge.com/topic/246119-first-light-d810a-dead-of-the-ccd/page-2

No, I was joking about the times. I'm sorry but this hasn't been clear to everyone! My apologies for having created confusion here but I picked impossibly short times and then said 'would I lie to you?' (I thought I was very obviously lying!) Not my most successful joke, then!    Sorry. The image had 5.5 hours.

I would say that OSC is slower than mono by about 6 to 4 from theory. From practice I think the mono advantage is greater. (In theory an hour per channel RGB should give the same signal as an hour of luminance. I never find that it does, though.) 

Olly

[Geordie mc]

Just had a quick check and it was Nik Szymanek who said you should take a new flat field for any change in the imaging train and this was usually when a filter change takes place.

To be clear, I do think mono is clearly better and more flexible. Just think OSC can be quicker for certain objects. I have read that colour cameras are less sensitive and understand the Breyer matrix part, but never read they are 4x less sensitive, though I can see how it seems that way. Aren't 2 filters the same colour and if so which ? I suppose that would make it rather less than 4x depending on colour of object which is dominant?

You could choose a OSC with bigger pixels to get round the scale issue as you cannot bin.

There must be some plus side to OSC surely or no one would buy them.

Going back to autofocus, anyone tried the Lakeside units. Look good on the website. Seem to remember a self build unit based around anduino processor cropping up on here too. How hard/expensive is this option?

[Thalestris24]

Hi

Um, you can certainly bin the qhy8l osc (2x2< 4x4). I guess the capability depends on the particular osc camera's sensor chip and readout modes available. You can't bin a dslr cmos chip.

Louise

[kalasinman]

Many Bayer filters use 2 greens and one red and blue.

[spartan2007]

ok guys any ideas on a atik camera or other brand  that would suit a newbie for deep sky objects with a cgem 925 edge hd  i have 2500 up to maybe 3000 dollars australian for my budget ,cheers

[ollypenrice]

Just had a quick check and it was Nik Szymanek who said you should take a new flat field for any change in the imaging train and this was usually when a filter change takes place.

To be clear, I do think mono is clearly better and more flexible. Just think OSC can be quicker for certain objects. I have read that colour cameras are less sensitive and understand the Breyer matrix part, but never read they are 4x less sensitive, though I can see how it seems that way. Aren't 2 filters the same colour and if so which ? I suppose that would make it rather less than 4x depending on colour of object which is dominant?

You could choose a OSC with bigger pixels to get round the scale issue as you cannot bin.

There must be some plus side to OSC surely or no one would buy them.

Going back to autofocus, anyone tried the Lakeside units. Look good on the website. Seem to remember a self build unit based around anduino processor cropping up on here too. How hard/expensive is this option?

Nik's right in principle but I found that I didn't really need multiple flats in practice. If in doubt do multiples.

Here's the sum for OSC (or RGB) versus LRGB. I'll use an hour as the time unit but a minute would be just the same. The Bayer matrix does have double green filters but a green filter still blocks 2/3 of the light (by blocking red and blue so I can't see that this affects the calculation. Whether having two green per blue and red is a good idea for astronomy is another matter. I think it is almost certainly a bad idea, in fact.

OSC is, by definition, colour filtered all the time, so catches 1/3 of the incident light (because colour filters block 2/3) so in 4 hours it gets 4 x 1/3 (One and a third) 'full illumination hours.'

RGB, also filtered all the time, gets 3 x 1/3 = 1 'full illumination hour.' But then we shoot an hour of Luminance which catches all the colours and gives one full illumination hour as well. That makes two full illumination hours hours, as compared with only one and a third for OSC.  

I have never been able to see anything wrong with this simple calculation* and saw it repeated, if I'm not mistaken, by R Jay GaBany. And, as I said earlier, this is being kind to OSC because according to the sum above, an hour per channel in R, G and B should give the same signal as an hour of luminance. I have never  found that it does. I extract a synthetic luminance from my RGB to add to my real luminance and weight them according to their signal strenght. The three hours of RGB in fact fall consistently well short of the hour of lum.

Olly

* If anyone reading this can see something wrong with it I promise that I'm all ears.

[kalasinman]

ok guys any ideas on a atik camera or other brand  that would suit a newbie for deep sky objects with a cgem 925 edge hd  i have 2500 up to maybe 3000 dollars australian for my budget ,cheers

QHY9? Change left for a filter set.

[ollypenrice]

Whatever you buy needs to be able to offer pixels getting you at least an arcsecond per pixel. This will almost certainly mean binning. I have had severe difficulty (ie total failure) binning a Starlight Xpress H36 and one of our guests had the same difficulties with binning their 8300 chipped camera. So you need to be sure that the camera you choose will indeed bin successfully.

Personally I'd be thinking about either the Atik 460 or the Atik 8300. The 460 has a cleaner, faster chip but the 8300 has a bigger one and at long focal legths things get a little crowded on small chips...

The calculator you need for pixel scale and FOV is here. http://www.12dstring.me.uk/fov.htm

Olly

[MattJenko]

I also use a single flat on occasion for all the filters. It works for me because most of the visible dust bunnies are on the my sensor, not on the filters, and the filter dust doesn't show up as much anyway, especially on my small chip. This is the same reason I haven't made new flats for a good number of sessions and images now. I do leave camera and fliterwheel attached to the scope when I tear up/tear down though.

I also refocus between each filter change, and I use Baader 1.25" LRGB + NB filters. I have a Lakeside motofocuser, which is a god-send, but it still takes a little while, so I don't rotate between filters round robin style (L,R,G,B,L,R,G,B etc) as this would taker forever and also gives fate more opportunity to make a mess of things. Watching a meridian flip happen during an autofocus routine was a scary sight.

I think a lot of the debate about OSC and mono comes from the (mis)perceptions about sub duration, number of subs and actual data gathered. It can take a while to gather 3 or 4 sets of subs and get enough subs to stack using a decent algorithm in a mono setup and the equivalent sub number is easier to reach in OSC. The underlying data is less, but you have what you need for a stackable image using a clipping method etc in a much shorter time using an OSC setup. It means you can create a nicely stacked image in a quicker timeframe, but the amount of data actually used in the image is less. It is a perception thing, and also a sanity thing, as there is not much more frustrating a scenario where you have some glorious R+G data but the Blue was cut short by cloud and the forecast is gloomy. I'd take the reduced OSC data at that exact moment, even though in the long run I am more than happy with my mono.

Matt

[spartan2007]

Do any of you guys know anything about the gstar cameras either the gstar ex colour or thr gstar ex2 mono heard this may suit a beginer.

[toxic]

i use the atik 383l+mono (8300 kaf) and have no problems with binning and got this camera for the size of the chip because i wanted something as near as possible to the eos size chip and it is not to far away for the same-ish field of viewwith the 200Pds. and i use 1 flat for all which was with a luminance filter.i have noticed that my images have improved a lot since the changeover from my 1100d (not a comparison just a fact) and the Light-pollution  going mono with filters in my opinion is the best way to beat LP.and it is quicker .and isnt mono noise a lot easier to deal with than color noise.

just my 2 pennies worth.

[Woodwind]

Some earlier posts talk about the cost of going down one route and then changing to the other later on (colour v mono etc). For that reason I started with a standard reasonable DSLR (which costs less if you don't buy the lenses). You can learn a huge amount and especially get the hang of the image processing. Then you can sell the DSLR quite easily - or indeed use it as a normal camera if you decide you want to go the dedicated astro-imaging route.

I have had a Canon 450 adapted for astro imaging by having the IR filter removed etc. I know it is only 12megs which is nothing these days - but as I get back not the game of imaging it will give me many years of useful work and only cost a few hundreed to buy and modify.

If this repeats anything said elsewhere - my apologies!

[Scott]

I get that. Olly only uses a flat from one filter for all, and says it works fine. Auto focus and parfocal filters? I dream of a CCD, but a OSC would cost the same and I'd feel like halfway there. No NB, no binning, no false color pallets.

I use a mono ccd and if I had my time again, I'd do the same, However, don't for one minute think that they cost the same. Throw in filters (extra if you wish to do NB) and filterwheel (all but essential) you have probably spent about half again.