This page aims at presenting images recorded with a T250F3.5 telescope and a thinned backside illuminated Marconi47-10 CCD and an Halpha filter.
Most of the nebulae and deep sky objects are emitting Halpha
light that having exactly a wavelength of 656.3 nm (Red light). The advantages
of using an Alpha filter are the following :
1. It weaken the sky background : in 60s I get 65 ADU/pixel/min with the Halpha
filter, and 1750 ADU/pixel/min with a R filter You can even make the images
during full moon..
2. Since the nebula emits mainly Halpha light, the filter does not absorb
the nebula itself
3. Advantages 1+2 = better signal to noise for the nebula !!!
To place the filter in a converging beam is not really good, because in that case, the beam is not perpendicular to the filter and the transmission peak shifts towards the blue. This is nevertheless what I did : The filter was located at 45 mm from the CCD, dealing with a F3.5 incoming beam. The next plot shows the transmission peak shifting versus the incoming F/D, the filter has a gradient index of 2. The red curve is the wavelength shift of the filter for the incoming beam at its maximum angle (TetaMax) at a given F/D, and the blue curve is a photometric mean of the shift for all the angle between zero and TetaMax. The more the angle is open the larger is its shifting contribution thru a larger annular ring.

For instance at F3.5, I've got a TetaMax of 8.13°, leading to a maximum shift of 16.5A (Angstrom) and the mean photometric shift is 8.2A So this is nting dramatic when using a 100A FWHM filter.

I have used the 50x50mm H-Alpha filter from Edmund Scientific, which is a 155Euros filter.
Here are the main features, available in three shape 11.8 and 24.2 mm diameter and 50x50mm.
Blocking <=0.1%, X-ray to 1.2µ
Full Width-Half Maximum (FWHM) 10 nm or 100A
FILTER INT 656NM 11.80MM DIA 656 H emission line NT43-086 $42.40
FILTER INT 656NM 24.15MM DIA 656 H emission line NT43-138 $79.50
FILTER INT 656NM 50.8MM SQ 656 H emission line NT43-190 $153.70
Here is the transmittance plot of this filter (blue curve) (credits to C.Buil at http://astrosurf.com/buil/us/filter/filter.htm) :
Here is a spectra of the sky background affected by light pollution (credits to C.Buil at http://astrosurf.com/buil/) :

Fortunately, the filter is open at 6500A... avoiding the light pollution of the Sodium street lights.

This is the south (left) and south west (right) telescope horizon, the two
most disturbing street light located at 7 and 15m from the telescope.

Center part of the Veil Nebula (20h48m30.370s +31°38'37.68''),
11x60 sec exposure, moon was up and 75% illuminated.

Ngc6888, 11x60 sec exposure, moon was up and 75% illuminated. There
is gas from a large nebula spread out in the whole image, this is NOT a flat
fielding problem.

Pelican nebula (20h52m47.900s +44°09'49.98''), 11x60 sec exposure,
moon was up and 75% illuminated.

M27, negative and positive image, 11x60 sec of exposure, moon was up and
75% illuminated. Note the extensions of the nebula rarely seen.

Cocoon nebula, left no filter, right with filter : 10 x 120s

Bubble nebula (high contrast view): 10 x 120s
Conclusion
Observations using Halpha filter is really promising in skies that are largely jeopardized by light pollution. Attention must be provided to the readout noise of the CCD, the dark current level (especially the hot pixels) . Once those steps are in control, very nice results can be achieved, despite the sky background. The telescope can be used full time during bright time of the Moon. The Halpha filter helps to reveal nebula features that are rarely seen because they are hidden by stars and/or other features of the nebula at different wavelengths.