Astrophotography Light Pollution Filters Slam-Down for Broadband Imaging

Here we will be comparing various filters for broad-band imaging in the big city.. Typical borad-band images would include Galaxies, Globular Clusters, etc.

Page Links

 


Introduction

The primary purpose of imaging with a filter in the city is to litigate light pollution and help suppress the noise so that the signal of the target object becomes easier to identify.  This is rather easily achieved for nebula type objects that emit at very specific wavelengths but much more challenging to accomplish for full-spectrum sources such as galaxies and globular clusters.  Cutting down on city glare is becoming even more challenging as city lighting transitions from Mercury type lighting that emits at defined wavelengths to LED lighting that generally emits broad spectrum lighting.

 

The purpose of this study is to determine what the best filter(if any) is for imaging broad-band objects (Galaxies, Globular Clusters, Open Clusters ect.) in the big city.

 

Downloads

Link

Format

Size

Comment

PDF

N/A

Filter Study PDF download

FIT

1.2Gb

Raw files captured at telescope

XISF

1.5Gb

NGC-5907 Master Files (PixInsight format)

FIT

1.5GB

NGC-5907 Master Files (FIT)

FIT

343 Mb

Batch Stretched Images

FIT

380 Mb

Custom Stretch Images

 

Imaging Session Details

 

The Target

 


The Imaging Process

Take 3 images, 5 minute exposures for a total of 15min integration time with each of the following configuration:

 

| C-11 | Corrector | {Identified Filter Below} | Camera |

 

Filter

Imaging

Time Range

Imaging

Altitude Range

None

01:58 – 02:14

55.0° - 57.0°

Baader Moon

02:39 – 02:54

59.9° - 61.5°

OPT Triad

03:19 – 03:35

63.8° - 65.0°

Astronomik CLS-CCD

04:20 – 04:36

67.0° - 67.1°

Astronomik CLS-UHD

05:18 – 05:33

65.7° -   64.8°

 


The Filters

 

Manufacture

Filter

List Price

Description

Specs

 

Chart

Wavelength Chart

None

N/A

Free

No filter usage, this is our baseline

None

 

 

 

Baader

Moon and Skyglow Filter

$132

 

Pollution Filter

 • Filters out specific wavelengths, especially those caused by streetlight and most importantly their scattered light which lightens up the night sky

• Selective blocking retains natural colours intact but with RGB significantly enhanced, differences in colour and brightness persist.

• 95% transmission in the selected spectral range.

OPT

TRIAD

$775

Tri-band Narrowband filter

H alpha/656.3/3nm

Oiii/H beta/ Center 493/18nm

Transmission over 90%

Astronomik

CLS-CCD

$200

Popular narrow-band filter

• 95% transmission at 486nm (H-beta)

• 95% transmission at 496nm (O III )

• 95% transmission at 501nm (O III )

• 97% transmission at 656nm (H alpha)

• 1st Passband 450nm to 520nm

• 2nd passband 640nm to 690nm

Astronomik

UHC

$200

UHC Filter

97% Transmission for the following wavelenghts

  • 486nm (H-beta)

  • 496nm (O III)

  • 501nm (O II)

  • 656nm (H alpha)

Optimal for F/3.5 to F10 systems

 

 


The Hardware

Hardware

Manufacture and Model

Specifications

Telescope

Celestron C-11

Note: This model C-11 is no longer sold, link is to latest model of similar specs

  • Type: Schmidt-Cassegrain

  • Aperture: 279.4 mm (11")

  • Focal Length: 2,800 mm (110")

  • Focal Ratio: f/10

  • Secondary Mirror Obs: 95mm = 12%

  • Weight: 27.5 lbs

Mount

iOptron CEM60EC

  • Type: Center Balanced Equatorial Mount (CEM)

  • Payload: 60lbs

  • Controller: Go2Nova 8407+

  • Mount Weight: 27 lbs

  • Resolution: 0.06 arc seconds

  • GPS: Internal 32-channel GPS

  • Autoguide Port: ST-4

Focal Reducer/Corrector

Starizona SCT Corrector (LF)

  • Focal Length: 2,000mm

  • Focal Ratio: f/7.2

  • Image Circle: 42mm

  • Mechanical Backfocus: 132mm from top of mounting threads

Camera

QHYCCD QHY128c

  • Read Noise: 1e- to 4e-

  • Pixel Size: 5.97 um

  • Well Capacity: 74ke-

  • Bit Depth: 14 bit

  • Dynamic Range: 14 stops

  • Sensor: Sony  Exmor IMX1238

  • Dark Current: 0.0006e-/p/s @ -15C

 


Analysis

The purpose of utilizing filters when imaging is to eliminate unwanted signal from the image resulting in an increased signal.  As a result the Signal to Noise ratio should be improved.  As a result when multiple images are stacked together the target object is more prominent than what it would appear without the filter.  

Software used for Stacking and image analysis is PixInsight.

 

Raw File Statistics

Here we took the each of the tree raw files generated for each configuration and performed analysis on the files utilizing the SubframeSelector tool in PixInsight.  As expected the filters appear to have helped raising the SNRWeight value in most instances.   While I’m not certain of the equation governing the Signal to Noise Ratio Weight (SNRWeight) in Pixinsight, I think we can determine the relative signal change since the Noise value is readily available.

 

Index

Filter

SNRWeight

(average)

FWHM

(avg)

Noise

(avg)

*Signal

Comments

1-3

None

1.06

7.28

312.8

331.6

 

4-6

Baader Moon

1.38

6.11

131.3

181.2

Seems to be quite good

7-9

OPT Triad

1.12

5.81

26.6

29.8

 

10-12

Astronomik CLS-CCD

1.05

7.49

78.1

82.0

 

13-15

Astronomik CLS-UHD

1.45

0.34

31.0

45.0

Seems to be quite good.  Not sure what to make of the FWHM value.

 

 

 

 

 

 

 

 

Stacked Files Statistics

Utilizing PixInsight each of the three raw files were processed using the process flow indicated below.  In addition, we also processed the group of filtered files together.  All sets of images were processed identically.

Image Press flow:

| Image Calibration | Cosmetic Correction | Debayer | Star Alignment | Image Integration |

 

 

Index

Filter

SNRWeight

(average)

FWHM

(avg)

Noise

(avg)

*Signal

Comments

1

None

155.5

7.19

17.72

2,755

 

2

Baader Moon

81.0

5.77

9.67

783

 

3

OPT Triad

5.0

5.83

2.62

13.0

This is a narrow band filter, so I expect low noise here.  But it's cutting out a lot of light also.

4

Astronomik CLS-CCD

27.1

7.39

5.13

138

 

5

Astronomik CLS-UHD

14.8

6.09

3.73

55.0

 

6

All filters Stacked

(excluding None)

128.2

6.49

2.06

264

This just for fun, not in consideration

 

 

 

 

 

 


 

Processing Photos in PixInsight

Each of the images were processed identically in PixInsight for the following steps (In order of operation):

| Dynamic Crop | Dynamic Background Extraction | Automatic Background Extraction |

| Background Neutralization | Photometric Color Calibration |

| Multiscale Linear Transform-Luminance | Multiscale Linear Transform-Chrominance |

 

Images were then duplicated and broken up into two identical sets to have final stretch performed in two differing processes:

1. Batch Stretch - Performed stretch on the “All-Filters Combined” image and then this stretch was applied to the rest of the images so the complete set all received identical stretch.

2. Custom Stretch – Stretch each image based on background noise where I tried to drop the background noise to just under detectable amounts for each image.  This is a subjective process but may be more representative of real world processing.

 

 


Batch Stretched Images

Click on the image to view full scale image in new window

01 - No Filter

02 - Baader Moon

 

03 - OPT Triad

04 - Astronmik CLS-CCD

 

05 - Astronmik CLS-UHD

06 - All Filters Combined

 

 


Custom Stretched Images

Click on the image to view full scale image in new window

01 - No Filter

02 - Baader Moon

 

03 - OPT Triad

04 - Astronmik CLS-CCD

 

05 - Astronmik CLS-UHD

06 - All Filters Combined

 

 


 

Conclusion

Statistical analysis of the Raw files leads us to conclude that the Baader Moon and the Astronomik CLS-UHD filters provide the best Signal to Noise Ratio.  Yet when images have been stacked, the None and Baader Moon come in as showing the best SNR.  This data seems to be in conflict since process used stack the photos was identical for all stack photos.

 

Inspection of the final stretch images is difficult and subjective.  For the images where identically stretch is performed on all images the Astronmik CLS-CCD filter seems to be able to provide the best contract between the galaxy and the background.  Images with Individual stretch are by nature of the stretching process much more difficult to judge, but to me it seems that the Astromik CLS-CCD and Astronmik CLS-UHD filter images appear to be the better images I am not overly confident in this).

 

We are trying to minimalize noise while keeping as much signal as possible.  My final conclusion is that the Astronmik CLS-CCD or UHD filters may be the best filter for imaging broad spectrum objects such as galaxies and globular clusters in the city measuring in with a Bortle Scale 7 while the Baader Moon filter should be considered in Darker sites that measure at Bortle scale 4.  Real world examples seem to bare this out.

 

 


Real World Examples

 

 

M042: The Great Orion Nebula

Whale and Hockey Stick Galaxies  

Galaxy Cluster in Canes Venatici

C-11 | LF Reducer | Baader Skyglow | QHY128c | 155min

Massacre Grounds Trailhead, AZ

M042: The Great Orion Nebula

M106 Galaxy Group  

Galaxy Cluster in Canes Venatici

C-11 | LF Reducer | Astonomik UHC filter | QHY128c | 85min

Chandler, AZ

M042: The Great Orion Nebula

NGC5907: The Splinter Galaxy

Spiral Galaxy in Draco

C-11 | LF Reducer | Astonomik UHC filter | QHY128c | 150min

Chandler, AZ

 

 

 

Astronomy

Home