Introduction

This section describes the camera system unit and the image acquisition techniques. For information about the telescope, its optical parameters and a description of the data processing procedures follow the links:

Camera System

The experience gained during several years of Hα observations showed that mastering of not always perfect observing and seeing conditions might be even more important than maximum resolution. Key features for the selection of the camera model were

• presence of an electronic shutter, as mechanical shutters had problems with the huge number of exposures on high cadence time series
• high frame rate for use of frame selection
• software controllable exposure times to adapt for changing opacity of the atmosphere
• reasonable high resolution (spatial and intensity)
• digital read-out and progressive scanning which simplifies the data acquisition

The selected JAI Pulnix TM--4100CL features a Kodak interline CCD chip with microlenses and a built-in electronic shutter, the maximum of the spectral sensitivity is close to the observed band. The edge length of 7.4 μm of the 2kx2k square pixels corresponds to 1.04 arcsec/pix which is almost equal to the diffraction limit for the telescope of 1.06 arcsec at 546 nm. According to Nyquist's sampling theorem the image is undersampled but it is good compromise between mostly prevailing seeing conditions and present camera technology. The shutter can be controlled by the length of an external pulse, typical exposure/integration times are in the range of 5 ms and the maximum frame rate is about 10 frames/s. The output is digitized to 10 bit, the lower level as well as the gain can be set according to the incident light level to exploit the full dynamic range but avoid non-linearity due to saturation. The dual tap CCD has the advantage of higher frame rates but the drawback of the need of a careful tuning of the individual gains and lower levels of the A/D-converter for the two taps to achieve equal dense half-images. Details of the camera setup procedures and parameters can be found in the » KPDC - Data Processing Procedures section.

A CameraLink interface transfers the data to a Silicon Software ME3 frame grabber in an Intel based 3 GHz industrial 19" PC and allows also the control of the camera by the image acquisition software.

Figure 1: Spectral sensitivity of the Kodak KAI4021 CCD, both taken from the manufacturer's product description. The selected band is close to the maximum of the sensitivity.	Figure 2: Output voltage vs. illumination. The max. light level for which the output is linear depends on the fixed, factory-set voltage Vsub, therefore the gain of the A/D-converter had to be adjusted to 12 dB to cover the full 10 bit range within the linear part and a noise floor slightly above zero, i. e. of some units per pixel.

Although integration times of less than 1 ms are possible and would be favourable due to the seeing it turned out that smearing which is inherent to interline transfer CCDs does not allow exposure times much shorter than 2...3 ms. It origins from spurious illumination of the interline columns (which are not perfectly masked) during read-out and shows bright (vertical) columns on parts where single bright pixels reside. For attenuation of the incident light level to achieve longer exposure times we put a neutral filter with a transmission of 10% in front of the camera window.

Figure 3: The hump parallel to the pixel columns (y-axis) outside the solar disk is due to smearing of the interline CCD without (left) and with the extra neutral filter T = 10%. The effect is clearly damped on exposure times longer than 2...3 ms which are typical with the filter and clear sky. However there is a higher level of spurious light outside the solar disk - probably straylight from the filter. The intensity of the disk center is about 850 units, but the scale is spread and cut at 200 to make the effect better visible.

The image acquisition application is written in C++ and makes use of the Common Vision Blox library. It is running under Windows XP and grabs continuously frames from the camera. Each frame is evaluated in a user defined rectangle (area-of-interest AOI) with regard to mean pixel value and standard deviation. The mean is used to control the exposure time and keeping the brightness level of the images fairly constant, the standard deviation is a measure for the blurring which is the main factor of the seeing at exposure times of some milliseconds which freeze the image motion component. The image with the best seeing of a consecutive number of frames is then written onto harddisk, the standard format is FITS, JPEG copies are optional. The whole procedure can be repeated after a user defined interval for automatic acquisition of time series. A block diagram and further details of the software which also used for the Hα observations at Kanzelhöhe can be found in Otruba, W.: 2005, Hvar Observatory Bulletin 29, 279.

Figure 4: A screen-shot of the image acqusition application. In the middle of the window the live solar disk image from the cam, in the center the white box indicating the AOI which is used for the calculation of the image parameters. They are displayed in real-time in the right part of the main window. Left: the control buttons for taking dark frames, snapping images right on user requests or for activating the recording of time series. Time intervals, number of frames used for frame selection and AOI-coordinates or a fixed exposure time can be specified via pull-down menues and dialog boxes.