1 Video Camera for Photometry: It can be done.. ….but… IOTA July 12, 2014 John Menke x x x .

1

Video Camera for Photometry:It can be done..

….but…

IOTA July 12, 2014

John Menkewww.menkescientific.com

2

If you are measuring the time, you are doing occultation/timing

If you are measuring the intensity, you are doing photometry

3

What is the biggest problem using video for photometry?

xSeductively easy: intensity vs time!

4

Now that I have your attention:

Software issues: software not really suited to photometry

Hardware issues: video cameras not really suited to photometry

and

Occultation skills <> Photometry skills

5

Problems:

Photometry not as easy as it looks

Photometry with video can be done.. But all the details must be right (and it’s tougher than with CCD)

More sources of error: photometry is Quantitative observing

6

So, the best way to do video photometry

is

Don’tx

(Use a CCD camera..)

7

Why not…

First, the software…

The observing SYSTEM is the softwareplus the camera plus the person

8

Actually, there are two programs

Both excellent for timing

Tangra somewhat better for photometry

LiMovie-older, less sophisticated

Tangra-newer, Hristo Pavlov,more advanced

9

Video timing software not designed for photometry

Hard/impossible to measure individual pixels (and you have many more to check…)

Hard to handle long series (hours of data)

(Designed for timing, not intensity)

Hard to use reference or comparison stars

Photometric tools, calibrations, corrections not integrated

Next, the camera..Software not realtime

10

Video cams complex..May deliberately be non-linear (gamma<>1)

AutoExp algorithms may depend on image, exposure often hard to control or know

Readout /ADC may not handle stars properly

Designed for non-point, mid-level brightness image

High readout noise (100,000 video images are worse than a dozen 5min CCD images)

Limited dynamic range (8 vs 16 bit in CCD)

11

Video cams not tightly specified

Actual saturation – 8bit=255 limit

For bright targets, Saturation and Roll-Off issues

Behavior approaching saturation (Roll-Off)

Saturated Pixels:

12

Typical frame near occultation

Io

Saturation

13

JEE events challenges..

Satellite images bright 5-6mag

Occulting image is changing shape and size with time (from two to one PSF)

Occulting image changing with time

Getting brighter, not dimmer!

These are all unusual—we’re not used to them

14

But

Sometimes video is the only way

Fast timing required (NEOs, flares, etc)

No other instrument available

So, how to avoid errors..

15

Ya’ gotta’…

Use appropriate tools (linear camera, photometric software such as MaximDL)

Learn photometric methods

Apply rigorously

Be skeptical of results (lots of possible errors)

Know/measure all the relevant characteristics of the sensors

Gotta’ practice!

16

Many sources of error..

Analysis-improper comparison, inappropriate software

Sensor issues-non-linearity. saturation, exp varies, color, unstable gain, noise

Astro/Sky-sky transparency, altitude/ (extinction and color), scintillation, color mismatch, dew

17

Basic photometric methods

Assure sensor always within linear response range (or results will be junk)

Must use reference intensity source (comparison object)

Must use software allowing easy individual pixel measurement, analysis

Evaluate need for transformations (color, atmospheric extinction)

Document, document, document!

18

And/or Photometric software..(MaximDL or CCDSoft—used for photometry)

Can use for video—two steps, though not real time (actually, neither is Li, etc)

Here’s how..

Use Occultation software..Must compensate for its limitations

or both..

19

Convert video to JPGs

Use VirtualDub to select ~100 frames

Export/ImageSequence/..JPG

Paste into Maxim or other photometry prog, examine pixels, take averages, etc.

Examine pixels, average, etc.

20

Whichever you do…

Must assure that the camera AND software is/are linear in the regime of interest

Must assure that you have reference/comp objects

Now, about linearity…

Must account for all potential errors

21

How to measure cam linearityDifficult to do right (great exercise)

eg., Video a star cluster with known values of same color (or use filter), all in same frame

Convert video to JPGs

Measure JPG star inten. in Maxim

Plot

..but have to use same conditions

22

Use Comp (reference) object..

Comparison object use can correct for many, though not all astro/sky errors

Beware variable comps, different colors, etc

Due to short exp, comps often not available, then find nearby (move scope during run)

(Differential Photometry)

23

Minimum To do List..

Read how-to photometry books

Practice: DO several “easy” variable stars (eg., WW Ursa Major)

Be skeptical of your results- seek anomalies

Peruse AAVSO, MPO sites

Study http://www.hristopavlov.net/Articles/VideoPhotometry.html for a ten page paper evaluating Tangra, and on doing video photometry

24

Again, strongly consider..

Use video where needed—primarily for fast response, accept limited dynamic range

But, use CCD whenever possible

Video used for slow objects have huge files, higher noise, hard to analyze

Wider dynamic range (x255), cam and software designed for photometry

25

26

Ever wonder how an 8 bit camera can yield higher precision results?

Averaging depends on noise to gain precision 1

2

3

255

*

*

*******

******

******

************************************************

bins

oo

o

eg, three points

noise morenoise

27

Video Cameras

Sensor Readout ADCUSB

Sensor Readout

ADCRCA

PC

28