since 20050614.                                                                                                                     Copyright    I. Yrjölä  1998...2014

Image Intensified Meteor Video Camera

This is now part of expensive and frustrating history - intensifier was replaced with an off-the-shelf  CCD camera!

Intensified camera pack components (op. 1998 - Oct. 2003)


VCR field-pack components (used 1998 - 2001)

Video processing components ( pre Oct. 2003)

  • A 233 MHz Pentium II or faster PC with DOS
  • A Matrox Meteor II video grabber (or the presenly available grabber type supported by MetRec)
  • Automatic meteor detection/radiant & meteor position software MetRec by S. Molau
  • OH5IY's batch file generator softwares to run MetRec in easier way



    Feb. 1998

    To cut the costs: with on-line analysis, you only need the camera/intensifier, power supply and the PC with proper video grabber.
    Video inserter, GPS, VCR pack, are not needed since you can use PC's internal clock and there is no need to record the session on VHS tape.


    Image Intensifier to boost stellar Lm to mag. +6 ... mag. +9 range,  CCDs - maybe later?

    The most vital and expensive part in such project is the night vision image intensifier. There are night vision equipment using three different generations of image intensifiers available + their evolution types, of which the 2nd generation intensifier being the most suitable and globally available. There are several quality grade units on offer, with the lowest price being the lowest quality with most imperfections and cosmetic faults on the tube screen.

    From this technical note you find 4 pages of text explaining the principles and differences in more detail of different type of intensifiers. Another manufacturer's page on intensifiers. You find out the expensive 3rd generation intensifiers have high IR range gain and lack of gain on the shorter light wavelengths and US made units are not available outside USA. Image intensifiers are quite expensive. There are off-the-shelf night-sights (with lens and IR illuminator) for hunting, etc. purposes. The D-grade 2nd generation consumer quality units costs 700 € or more, but normally from 1800 € up to about 3000 € depending on the quality and manufacturer. The Russian 3rd gen night sights are sold at around 4500 € & up, but of course if you are not satisfied with consumer equipment, the price tag has no real upper limit. It is hard to pick a unit without seeing it and you never know beforehand what you get, or how long it will work.

    Time Coding of taped Video signal

    Note: time coding is not needed for on-line single station system. The PC's internal clock is used then as a timebase. Time coding is necessary only if the session is taken on video tape.

    The cheap DFC-77 clock time-sync receivers do not work in Finland since the German transmitter is practically out of range, so forget it.

    To get UTC & date on the screen, I found the GPT-50 Video Titler with GPS connectivity that seemed to do the job, but was also commercially priced. In case the NMEA sentences from the GPS would not be available, GPT-50 has internal quartz clock for time & date. The GPS needed a TTL/RS-232 converter, which I made by my self. The RMC messages work well providing date, time and position data on screen, but I had to buy the SK8's programming kit. The GPT-50 & SK-8 GPS combination has worked reliably during field testing. The Lassen SK-8 GPS receiver locks-in on satellites in less than two minutes from power-up - about same time that it takes to connect the four connectors and install the camera on the tripod. The small GPS antenna works fine through the rear window of the car. Since I now run the camera only from my home and use PC's own timebase, I need none of these...

    Heater, Power Consumption & misc.

    The heater rings I have made dissipate some 2..3 watts and are made of four 18 ohm 1/2 watt resistors in series. RCA connectors were used in the heater supply cable to allow change of the heater ring with the lens.

    The field pack draws 2.7 amperes of current with the monitor on and 1.6 A with monitor off from the 12 V sealed lead acid battery. Optionally it can also be powered from the car's cigarette lighter socket, or from AC/DC power supply with the appropriate adapter power cables. Coaxial cable with black sheath and white 2*0.75 mm² cable are tied together as camera cable. This color combination can be better seen in darkness on the ground and also on white snow to avoid stumbling on the cable. The VCR, monitor, GPS and video titler are firmly attached to a plywood board, which front edge can be raised and locked to upper position for easier access and visibility to VCR front panel and monitor screen. The board is not attached to the case, so it can be easily removed and units detached, if work has to be done with any of them. The system has been tested in dew summer nights and in freezing cold in the winter (-20 °C), as the project progressed. No dew, or frost has been observed on the lens and the VCR has worked faultlessly. After the session the camera and VCR pack is brought indoors unpowered to warm up and dry overnight. However, later the camera setup was fixed and analyzing the data in real time, made the whole VCR pack unnecessary.

    Dedal 41 MCP + CCD-camera's image distortion graph

    The permanent mounting setup needed automatization to turn the camera on and off. This is not so simple task since the MCP is susceptible to damage from "bright" lights. There has to be a timer switch to power-up the camera for the dark hours of the night, but also there was need for a fail-safe in the form of light sensor and need for lower voltage for the shutter motor after it was opened, not to heat up the motor unnecessarily. The shutter is a 36 V DC motor (ex. ink ribbon reel motor) with a gearbox, spring loaded to return-to-shut-position. The light switch uses a photodiode (ex. VCR tape sense) and a relay driver circuit that kills the camera power. The photodiode is mounted by the intensifier lens, outside the shutter. It kills the power to the camera in a fraction of a second, if it detects enough light, but turns it back on immediately when it is dark. The shutter's hold-circuit has some 1 second delay under which period it first keeps the full 12 V to torque the shutter to open position and some 4 volts after that to keep it open. Once voltage goes off, the spring closes the shutter to protect the MCP's photo cathode from bright lights and ambient light exposure degradation during daytime, when the camera is unpowered by the main timer anyway.

    The second phase of the controller upgrade included an far-infra-red (FIR) thermo-switch to turn off the camera, when there is some cloud cover. This FIR-Switch turns power off from everything, but the heater and the FIR detector itself and cuts off the video line. The switch includes two timers, one about 4 minutes and an other some 5 minutes. When T-sky raises over the pre-set level (temperature, say - 20° C), the video and 12 V DC power feed relays open. Later, when the sky clears and T-sky goes again cold, the other timer keeps the camera unpowered for 4 further minutes and then powers it up, which opens the shutter. A minute later video feed is restored by the second timer. These delays are used to prevent rapid unnecessary "off-on" switching.

    Grabbing Video and Finding Meteors

    The meteor detection software MetRec v3.4 worked well, with few problems or ambiguities, but has no mouse support, needs 4 different executable files to be run from DOS prompt, all with long string of parameters. All entries when using the software has to go right step by step, as there are little chances to re-do anything after a wrong entry. Since the software is complex, there are vast number of parameters that can be changed and several of them have to be manually set, though there now is some parameter information exchange between the different executables that are used before actual observing runs with MetRec. The normal procedure requires grabbing a reference image, marking the reference stars (>50 pcs) by use of keyboard cursor keys, editing the MetRec's configuration file for the observation. After this is done, the MetRec itself is executed and after the observation, there is the need to run Post Process of the meteors and delete non-meteors from the log file.

    Here is a short summary of observing results from September 2000. The full data transferred to IMO archives was 26 megabytes.
    day (Sept.) T (hours) n SPO mean mag SPO SPI DAU AAU STA NTA moon
    1 5 5 1.5 2
    4 4.75 18 1.8 1 1
    6 5 19 1.2 2 2 1/4+1 day
    7 5.25 29 1.6 2
    10 4.8 29 0.8
    12 6 13 1.9 full-1 day
    14 6 9 -0.1 2 1 full+1 day
    15 6 15 1.2
    16 6 15 0.7 1 2
    18 6.75 16 0.1 2
    19 7 21 1.6 2
    20 7 18 1.1 1 1
    21 7 14 1.2 1 1/4
    22 6 14 1.1 4 3
    24 6 23 1.0 1 3
    25 7 16 0.8 1
    26 7 38 0.9 2 1 new-1 day
    28 4 16 0.3 2 2 new+1 day
    30 8 21 0.6 1 2 2
    Sept. 2000 total: 114.6 h 349 pcs mean mag 1.1 23 pcs 20 pcs 2 pcs 2 pcs

    Below: A summary table of YRJIL's 2002/2003 and 2003/2004 (as of so far) video observing results from Kuusankoski, Finland. Images, meteor begin/end points, angular velocity and brightness data were measured by MetRec from every one of the observed 1443 meteors. This data was archived by IMO.
                       2002/3                          2003/4 (so far)
    Teff total:  793 h  24 min  57 s    107 h 30 min 31 s
    Nobs:  112   20
    QTY of all meteors:  1443   379  First obs: 18. Aug. 2002, last obs: 11. May 2003
             2002/3 2003/4
      N SPO:   1212    302

      N SIA:      0
      N KCY:      0      1
      N CAP:      0      3
      N NIA:     1      1
      N NDA:      0      4
      N PER:     1     29
      N SDA:     1
      N SPI:     13     18 
      N AUR:      0      4
      N DAU:     13     17
      N ORI:     38
      N STA:     17
      N EGE:     6
      N NTA:     15
      N LEO:      0
      N AMO:      0
      N GIA:      0
      N XOR:     9
      N PHO:      0
      N PUP:      0

      N MON:     2
      N HYD:     4
      M GEM:     14
      N COM:     15
      N URS:     18
      N QUA:     43
      N DCA:     10
      N ACE:      0
      N DLE:     2
      N GNO:      0
      N VIR:     11
      N CSO:      0
      N LYR:     7
      N PPU:      0
      N ETA:      0
      N SAG:      0
      N JBO:      0
      N JPE:      0
      N PHE:      0
      N PAU:      0



    This kind of equipment package is not an everyday investment. We talk about thousands of Euros (€). I do not think you can get the necessary components any cheaper, but if you get state-of-the-art digital video recorder and the best possible MCP with many lenses and so on, you can end up to a budget exceeding 10 000 €. On the other hand, visual observing and plotting meteors on paper maps in the winter is not so easy and the results are not comparable to the accuracy and reproducibility of video system, so this method is gaining popularity within amateur meteor observing. Other applications are taking Aurora videos and satellite observing. Actually, the price tags I gave, do not count the voluntary work by S. Molau on his free software to amateur use. Professional user license for MetRec is 250 €.

    Camera & VCR packIf you feel you wish get involved in video meteor observing and setting up such a system too, I wish you luck. As you see, I did not have too much with it - with the destruction of the intensifiers. I can not encourage anyone to get involved, or not - you have to consider if it is worthwhile. You perhaps know better now what to expect. I believe this page adds a practical real-life view on setting up of video meteor camera compared to more theoretical material available on the net on this kind of project. The wisdom being - you don't need anything else but the intensified camera (now, the CCD camera), videograbber, PC and MetRec and a dark backyard site to run the system.

    One more thing; if you ever get an image intensifier on your hands, it WILL be DAMAGED or DESTROYED from bright lights, so be more than careful! Even the Moon can be too bright!  Also I would like to point out that the Dedal factory on their website says their generation 2 intensifiers life-time is only some 1000...3000 hours, that is only 125 to 375 8-hour nights! Of course the life-time depends on the condions of use - the amount of light (beam current), but still it makes me wonder... Also, never operate or store it in temperatures above normal room temperature (above +30 °C) and never expose the photocathode for extended periods to daylight or even worse, direct sunlight, even if the intensifier is not switched on!

    #1   Dedal 41 intensifier degraded to useless condition after about 300 h of use

    The first Dedal 41 intensifier (model not in production anymore) was mysteriously degraded (frankly, for this purpose: destroyed) and had become useless due to high noise after the summer pause of 2001. This resulted unpredictable and unfortunately high operating costs of the intensifier (replacement costs of 1200 €).

    #2   Dedal 41 intensifier powersupply failed partially after about 1000 h of use!

     The second Dedal 41, MCP manufactured in 1998 week 47, sold in March 2002,  lost a little of its brightness and later the image size began to fluctuate randomly by -15% and the phosphor plate eventually settled to be underilluminated  by that 15%  with clear brightening on the outer rim (see image below). I suspect it is a failure in the high voltage power supply system, but since the tube and PSU are integrated to PVC molding and hard to disassemble without damaging them, any attempt to further troubleshoot or repair (no schematics) is almost futile. So, the MCP would need to be replaced... another 1200$ ? No thanks, enough is enough.

    After intensifier's warranty expires,  it may go bad any night for many reasons without you actually causing it and even an extended home insurance that is said to "cover" all hobby equipment also, will not pay for it, because the "cause of the damage could not be detemined" and "is not related to any known unexpected sudden event" (such as lightning strike, falling, ...)(quoting my previous insurance company's statement...).

    The Watec WAT-902H CCD camera with sensitive EXview HAD CCD sensor has been there for a few years - nobody just believed it would come so close, or actually beat Dedal 41 Gen. 2 intensifier with a price tag of just 318 € (Aseko, Finland), + fast F0.8 lens. The MCP camera gave a circular FOV of about 35 degrees vertically, while the WAT-902H with a 6 mm lens gives 43 degrees vertical FOV, with a  Limiting magnitude for 32 x integrated image only about 0.3...0.4 magnitudes worse. Since the imageof the CCD camera  is rectangular with 4/3 ratio, it gives more (full) detection area with 450 line resolution and much less geometrical edge distortion, which all compensate the sensitivity difference. I do not see any reason to buy fragile 2000 € or even a 1000 € or $ Gen. 2 intensifiers, since the off-the shelf WAT-902H CCD camera does almost as well. Nobody is forbidding you to get more CCD cameras and running two MetRec systems at same time if you think the system now is too cheap. What is the best thing; Watec has 3 year factory warranty for their cameras, while the intensifiers I used, had just one year, and which they just survived. Maybe this page should be labeled "The $ 4 000 page...".  and the CCD page "The $ 400 page..".

    The background image is  of a bright Perseid taken with this intensified Gen. 2  camera. Constellation of Ursa Major and trees 35 m away illuminated by Dedal 41's IR source - digicam photo through the intensifier ocular.


    Updated: 26. Sept. 2006

    Link to CCD Meteor video camera
    Link to intensified (MCP) Meteor video camera
    Link to skylight polarisation during twilight and it's reduction by filters

    Back to OH5IY's main page