M42 in Near IR, [SII] and Halpha
- this is a project I've been meaning to start for some time. It is a widefield using Near IR longpass, [SII] and Halpha
the bright stars in the Trapezium really severely limited my exposure time in Near IR, otherwise the sensor would bloom terribly. I ended up taking 15 second exposures. I had to take 100 of those subexposures to get 25 minutes total exposure.
the Halpha and [SII] data were 25 exposures of 60 seconds each.
each exposure with this camera, IMG6303, is 12.3MB. So I had a lot of raw data to crunch: 150 exposures of 12.3MB each.
Even so, the data is a bit grainy so I may shoot more data another night soon. Who'd have thought that 150 exposures would be too few?
it takes about one minute to cycle through a 15 second exposure roughly, just as a data point.
[Non-text portions of this message have been removed]
- --- In firstname.lastname@example.org, "Rick Wiggins"
> There are no wrongs here (other than I have not won the Lottery
> yet!). I do think my proposed theory is correct. Idon't think you
> gaining much after 10 to 20 exposures to average down the noise. AsI think the signal does build up, slowly.
> far as building up the signal, I don't think that happens. Signals
> that are so weak as to vary sow up on one frame and not another are
> probably interpreted as noise. Signals that do show up in all
i forgot the equation but I believe the signal adds in an rms
fashion. and the noise adds too but there is a net win in SNR as you
take more data from what i remember last time stan went through this
i need to look at my references a bit and get back.
no question that the 100 frame data is better than 20 of them. I
already looked at that.
> but vary, will be recognized as signal and will be improved. Trydecided
> using the technique that I suggested if you have time. That's the
> same as some use for LRBG of M42. They shoot 10 secs, 5 mins,etc.
> They use the sweet spot out of each set with layers and masks. This
> should work for you as well. I just don't think that super high
> numbers of frames buys you much in this situation. I hope that Stan
> or one of the SNR Gurus chimes in on this one.
> Tonight I will be continuing on my light box project.
> Thanks, Rick
> --- In email@example.com, "rdcrisp" <rdcrisp@s...> wrote:
> > here is why i did what i did, right or wrong
> > i wanted to image the m42 object without blooming the trapezium
> > and using my nir longpass filter.
> > i could expose for about 15 seconds without blooming too badly.
> > is it. so my maxiumum exposure time was limited to 15 seconds.
> > with 15 seconds my ability to capture the faint nebulosity is
> > challenged to say the least. so to get as much as I could i
> > to take a lot of exposures in hopes that the faint signal wouldsignal
> > accumulate.
> > this is a situation where low read noise is nice because in 15
> > seconds there is a part of the nebula where the accumulated
> > just above the noise floor set by the read noise. so the lower
> > read noise, the better I capture the faint nebulosity.again
> > So while the maximum exposure time was limited by avoiding
> > the faint nebulosity detection/capture was limited by the noise
> > of the camera and the read noise is the biggest part of that as I
> > understand.
> > That was the "why" that I did what I did, right or wrong.
> > still the image is noiser than I'd like. I may fiddle with it
> > tonight.for
> > rdc
> > --- In firstname.lastname@example.org, "Rick Wiggins"
> > <rickwiggins@e...> wrote:
> > >
> > > Richard,
> > > The shot is very interesting and quite well done. I am curious
> > about
> > > the "100" frames. I have seen this done by several others as
> > I
> > > must be missing something. ROn and Stan please chime in if you
> > > agree/disagree/or can shed some light on my thoery here. The
> > > exposure of 15 secs captures all the signal for the bright
> > > of the frame (probably near 50% saturation levels in the
> > > sections). The series of exposures does not increase the signal
> > > the bright areas, but serves to randomize the noise allowing
> > > reduction of noise in an averaged (or similarly processed)you
> > > image. The combination of the captured signal and
> > > noise gives higher SNR and a higher quality image. So, after
> > havethe
> > > captured enough images to drive down noise (probably 10-20),
> > > subsequent frames don't buy you much; they do reduce noise, butthe
> > only
> > > marginally so. Wouldn't it be better to take 10 more frames at
> > longer
> > > exposures; say 10 at 1 minute, and 10 at 10 minutes, and then
> > > composite these combined frames to use each combination's best
> > > areas of the frame? If the blooming would kill you, then I
> > understand
> > > why you wouldn't do this with this camera; but if that's the
> > > what about using a ABG camera to take the long exposures for
> > > dimmer sections and then combine. It just seems to me thatafter
> > > frames or so, there is marginal (and probably invisible to our
> > > improvement. Is there something else going on when stacking
> > I
> > > mean other then noise averaging?
> > > Thanks for listening to my thoeries.
> > > Rick
> > > --- In email@example.com, "Richard Crisp \(SBC\)"
> > > <rdcrisp@s...> wrote:
> > > > this is a project I've been meaning to start for some time.
> > a
> > > widefield using Near IR longpass, [SII] and Halpha
> > > >
> > > > the bright stars in the Trapezium really severely limited my
> > > exposure time in Near IR, otherwise the sensor would bloom
> > terribly.
> > > I ended up taking 15 second exposures. I had to take 100 of
> > > subexposures to get 25 minutes total exposure.each.
> > > >
> > > > the Halpha and [SII] data were 25 exposures of 60 seconds
> > > >a
> > > > each exposure with this camera, IMG6303, is 12.3MB. So I had
> > lothttp://www.narrowbandimaging.com/m42_ap180_6303_NirS2Ha_page.htm
> > > of raw data to crunch: 150 exposures of 12.3MB each.
> > > >
> > > > Even so, the data is a bit grainy so I may shoot more data
> > another
> > > night soon. Who'd have thought that 150 exposures would be too
> > > >
> > > >
> > > >exposure
> > > > it takes about one minute to cycle through a 15 second
> > > roughly, just as a data point.
> > > >
> > > >
> > > >
> > > > [Non-text portions of this message have been removed]