for autoguiding and asteroid occultations applications

Cliquer ici pour une version française

The camera tested is the model Watec 120N (CCIR version) equipped with a Sony CCD detector ICX-418ALL (1/2"). The specificity of the camera is the internal long exposure function. The camera allows a maximum integration time of 256 frames, i.e. about 10 seconds. The camera was tested the 18 Nov. 2005 in the author's suburban observatory (Castanet, near Toulouse, South of France) at the period of full moon (naked eye limit magnitude was around 3). The outside temperature was around 4°C. The camera was connected to a computer a GrabBeeX USB2 grabber interface. Iris software was used to digitize (frame by frame mode, not AVI acquisition) and processing.

From left to right, the WAT-120N, the WAT-902H2, the WAT-902H.

The field observed is a well know photometric area of the open cluster M67 (the "dipper asterism"). See
- M. Joner & B. Taylor, PASP 102, 1004 (1990)
- C. Chevalier and S.A. Ilovaisky, A&AS, 90, 225, (1991)

No spectral filter used, therefore the results do not have a rigorous photometric value. They are only representative of the limits magnitude for application like asteroid occultation or autoguiding on faint objets.

The open cluster Messier 67. The "dipper asterism" is near the center of the field. The north is up.
The image was obtained with a Takahashi FSQ-106 refractor (10.6 cm aperture) and a Watec-120N video camera.
Addition of 10 frames exposed each 10.24 s (256 internal stack mode) - mid-gain adjustment.

Detail of the M67 "dipper asterism". The V magnitude of some stars is indicated.
Image taken with a Celestron 11 telescope (28 cm aperture) at F/10 and a Watec-120N camera.
Stack of 3 x 5.12 seconds exposure (128 internal stack).

The following images show the M67 field according the integration time taken with a Celestron 11 at F/10. The gain is adjusted at HI position. The thermal dark signal is subtracted (the master dark is the median of 20 individuals frames taken in the darkness with the given exposure time). The seeing is very bad during the observation, near 5 to 6 arcsec. Only the better resolved images are presented.

Single shot exposed 0.02 second (internal stacking off)

Single shot exposed 0.08 second (2 internal stack)

Single shot exposed 0.16 second (4 internal stack)

Single shot exposed 0.34 second (8 internal stack)

Single shot exposed 0.64 second (16 internal stacking mode)

Single shot exposed 1.28 second (32 internal stacking mode)

Single shot exposed 2.56 second (64 internal stack) 

Single shot exposed 5.12 seconds (128 internal stack)

Single shot exposed 10.24 second (256 internal stack)

The next table summarize the magnitude limit found. The reasonable magnitude is the detectivity limit magnitude minus 1.5, required for autoguiding and for observe with a minimal of precision the occultation of a star by an asteroid.

WAT-120N at the focus of a C11 telescope (D=28 cm)

Video rate (or exposure time)

Limit V magnitude

Reasonable V magnitude

0.02 sec



0.08 sec



0.16 sec



0.32 sec



0.64 sec



1.28 sec



2.56 sec



5.12 sec



10.24 sec



Same table, but for a FSQ-106 refractor:

WAT-120N at the focus of a FSQ-106 refractor (D=10.6 cm)

Video rate (or exposure time)

Limit V magnitude

Reasonable V magnitude

0.08 sec



0.16 sec



0.32 sec



1.28 sec



2.56 sec



Comparison of the Watec 120N and the Watec 902H / Watec 902H2

The WAT-902H and WAT-902H2 (supreme) are sensitive "standard" video. Only the TV vidéo rate is available.

The M67 field is observed with the Watec 902H and the Watec 902H2 at the focus of a Celestron 11:

Watec 902H - One frame (dark offset subtracted). The AGC if off.
The limiting magnitude is estimated to V=12.5

Watec 902H2 - One frame (dark offset subtracted). The AGC if on.
The limiting magnitude is estimated to V=11.4

First conclusion: A Hi AGC is not a good idea for stellar imagery because the increase of the background noise (for identical settings the 902H and 902H2 performances are very similar).
Second conclusion: The detectivity of the WAT-902H and the WAT-120N are nearly identical if the latter is adjusted to the exposure time of 0.08 second, see below:

Watec 902H

Watec 120N @ 0.08 sec

About Deep-Sky observations

The thermal signal correction

The median stack of 10 individual frames exposed each 10.24 s in the darkness (256 internal stack  mode - mid-gain).
Note a fixed pattern and the presence of the hot pixels. The image is named the
master dark frame.

The M82 galaxy. WAT-120N at the focus of a Takahashi FSQ-106 refractor (10.6 cm aperture). Direct addition of 10 individual raw frames exposed each 10.24 s.
La galaxie M82. WAT-120N disposée au foyer d'une lunette FSQ 106 (ouverture de 10,6 cm). Addition directe de 10 images brutes exposées chaque 10,24 s.

 The same set of images after subtraction of the dark master and registration. The hot pixels disappeared well (except for a saturated zone on the left). On the other hand, the periodic structure of the master dark frame is finds in the final image - see below.

The origin of the vertical parasitic structures is probably the frame grabber because the level of the phenomenon is independent of the Watec camera gain. The quality of grabber is essential for the final result. Some devices can have qualities higher than others. The type of interface can also have an importance: here an USB interface, but it is possible to find PCMCIA cards (IMPERX grabber for example). Some camcorder can be used via the s-video input/output.

The grabber model used for the tests.

For the same model of grabber, the result can be different from a serial number to an other... see below the very high contrasted images (typical RMS noise of only 0.4 ADU) of the bias signal for the same WAT-120N, digitized with three distincts GrabeeX devices (median sum of 20 successives frames):

Grabber #1

Grabber #2

Grabber #3

Important note

Put always the GAMMA switch at position OFF. It is required for image processing. In the contrary case, the correction of the thermal dark current is not effective.

Gamma switch position: LO. Hot pixels are overcorrected.

Gamma switch position: HI. Hot pixels are also overcorrected, because for the two situations, the intensities scale is not linear.

The best solution for attenuate the fixed pattern if it is present, is to shift the pointing direction of the telescope between each image (diphering technique). By simple average effect, the pattern is erased during the stacking (and after registration of course).

Stack of 30 x 10.24 s diphered images (256 internal stack mode). Gain at mid-position.
FSQ-106 refractor. Suburban conditions.

The drizzle algorithm is applied to the 30 frames for increase the final resolution, here by a factor 1.3, very significant (it is possible because the very sharp image produced by the refractor is undersampled by the camera pixel's). The Watec-120N image shows a partial dark circle around the stars (fisheye artifact). This can be result of an internal sharpening algorithm. The exact origin is not clear for the moment (the camera, the grabber ?), some authors report no fisheye effect with their camera, some authors report a problem (click here for a link).

If the fixed pattern is definitively present in the processed image, frequency filtering in the Fourier domain can be the last help solution...

Messier 51 image. Stack of 30 x 10.24 s images (256 frames mode). The cumulative integration time is of 317 secondes (FSQ-106 refractor). The image is affected by the vertical pattern.

The Fourier image (after a FFT). The frequency peaks of the pattern are clearly visible (noted by yellow tics).

After suppression of the pattern peak in the Fourier space (Iris FFILL procedure), the inverse Fourier transform give a correct image. Remember, the very severe light-polluted sky condition.

The Watec 120N control box.

Some applications

Faint star autoguider for spectrograph

The WAT-120N used for center and guiding star on the entrance slit of a LHIRES spectrograph. A camera WAT-902H + a telephoto lens are used as an electronic finder.

Star autoguiding with the WAT-120N. The entrance slit is the faint horizontal dark line.

Scan the sun surface
Spectroheliographic mode of LHIRES

The WAT-120N is used for monochromatic scan of the Sun (Halpha, He, Calcium, ...). A Watec 902K help to position the sun image on the entrance slit of the spectrograph..

Flare on the Sun - H
a image..
Takahashi FS128 refractor, stopped to 50 mm.
LHIRES spectrograph. WAT-120N. Nov 18, 2005
The big interest for the application of the 120N model is the inactive automatic gain (AGC OFF mode).