Historical Measurements

This document summarises the filters available for use with the ING telescopes, see also ING Technical Note no. 96. Table 2 lists the filters available for CCD imaging. All of these are 50 mm square, except where otherwise noted, and can be used at all the imaging instruments: JKT Cass, INT Prime, WHT Prime and AUX port. Apart from the filter characteristics we also give some telescope specific information. Table 2 lists the larger, circular, filters available for imaging and for Fabry-Perot spectroscopy with TAURUS-2 on the WHT. The information in this document comes from various sources: the La Palma Technical Notes 45, 73, 75 and 90, the INT Prime Focus manual, and a List of Optical Components compiled by Chris Benn.

Imaging filters
Non-imaging filters

Imaging filters

Our collection consists of broad band and narrow band filters. For broad band imaging we have several sets of BVRI filters, which are not meant to leave their telescope. Our set of narrow band filters at the moment is still rather limited, except for (redshifted) H_alpha filters, of which we have more than 15. All filters are square, and have a length and width of more or less 50mm. Some of the larger filters do not fit into every filterwheel. There are however filterwheels to accomodate each filter.

Here are 2 tables with a summary of the filter characteristics:

All filters except redshifted H_alpha

Redshifted H_alpha filters

Broad band filters
The Kitt Peak Interference Filters (BVRI)

These were purchased from Kitt Peak, as originally specified by J Mould. They have flat-topped profiles and are intended for use with CCD detectors (unlike earlier filter systems). Figure 1 shows details of these filters. Three sets exist; one for the INT, one for the JKT and a spare set at Herstmonceux.

The transmission curves are given here:

Johnson U Kitt Peak B Kitt Peak V Kitt Peak R Kitt Peak I

and some ascii tables with the filters scans are given here:

Johnson U Kitt Peak B Kitt Peak V Kitt Peak R Kitt Peak I

The original specifications are given below:

All filters to be 5cm x 5cm, 3mm thick, blocked to 1.1 micron, better than 75% peak transmission (except B 70%), MgF2 coated, no pinholes, optical quality glass.

R: Filter 7200Å = short-pass hard-coated on 3mm OG590, blocked to 1.1 micron, 75% or better peak transmission, MgF2 coated on glass side.

I: Filter 9000Å = Short-pass hard-coated on 3mm RG-M9, blocked 1.1 micron, 75% or better peak transmission, MgF2 coated on glass side.

V: Filter 6000Å = Short-pass hard-coated on 3mm GG 495 blocked to 1.1 micron, 75% or better peak transmission, MgF2 coated on glass side. (Sent back did not come up to specification, 1/6/81)

B: Filter 4900Å = Short-pass, hard-coated on the glass substrate indicated below, blocked to 1.1 micron, 65% or better peak transmission, MgF2 Coated on glass side. Glass substrate to be cemented combination of Schott GG 385 and either Hoya GM-500 or Hoya C-500 (whichever will adequately suppress the red leak). Total thickness to be 3mm as with the other items.

Broad band filters
The RGO glass broad band filters (UBVRIZ)

These filters have been made at RGO from various cemented combinations of Schott glass. R W Argyle specified the combinations which are similar to those in use at the AAT, SAAO and elsewhere. Table 1 shows details of these filters. Three sets exist; one for the INT, one for the JKT and one at Herstmonceux.

The transmission curves are given here:

RGO U RGO B RGO V RGO R RGO I

Gunn Z

and some ascii tables with the filters scans are given here:

RGO U RGO B RGO V RGO R RGO I

Gunn Z

Broadband filters
The Harris set

To replace a deteriorating set of Kitt Peak filters, 3 sets of large BVR glass colour filters were purchased in 1990 for CCD imaging.

The filters consist of cemented stacks of Schott glass, following the following recipe (see also NOAO newsletter 11,12 and 14:
B: 1mm BG12 + 2mm BG39 + 1mm GG385
V: 2mm BG39 + 2mm GG495
R: 2mm OG570 + 2mm KG3
I: 4mm RG9 (not an original 'Harris' filter)

It is thought that

When the filter response curves are convolved with a typical CCD response, the glass filters provide a closer match to the standard Johnson B and V and the Kron-Cousins R bandpass.
The transmission of the glass filters is comparable to that of the interference filters, and does not deteriorate with time.
Glass filters are significantly cheaper than interference filters of the same size.

Note that the transmission in B is significantly (about 30%) lower than in Kitt Peak B.

Here are the transmission curves:

Harris B Harris V Harris R

and the ascii tables with the transmission curves:

Harris B Harris V Harris R

Narrow band filters

At present, we have 31 narrow band filters, mostly of 50Å bandwidth, comprising a set of emission line filters and a set of redshifted H_alpha redshifted filters. Only one set exists and therefore the filters are shared between the INT and JKT. These filters were specified for a focal ratio of 4.5 which explains the difference in wavelength between the actual peak wavelength and the required peak wavelength. The original specification for the emission line filters is given below:

Unless otherwise specified the effects of an f/4.5 converging beam of light has been taken into account and the central wavelength duly reddened. A correction of +2 Angstroms has also been made to correct the central wavelength for operation at +10 degrees Celsius on the assumption that the filters will be made, tested and specified at +20 degrees Celsius.

In all cases the thickness of the filter including blocking filters should not exceed 9mm and should be the same for all filters wherever possible. The effective refractive index should be near 2.1. Blocking in all cases should be 3000-12000 Angstroms and the tolerance on the central wavelength should be better than 3 Angstroms. The tolerance on the dimensions of each filter should be smaller than 0.5mm. Filters should be free from pinholes greater than 0.03mm diameter in the central 25mm diameter circle and free from pinholes greater than 0.1mm diameter elsewhere.

Here the transmission curves for the filters:

1. The emission line set (except H_alpha)

[OII] 3727 [SII] 4075 HeII 4691 H beta [OIII] 5007 He II 5881 [OI] 6306 [OI] 6306 large scan H alpha [NII] 6590 [SII] 6730 [OII] 7331 [SIII] 9069 [SIII] 9539 HeII 10837

2. The old redshifted H_alpha filters:

H_alpha 6556 H_alpha 6563 H_alpha 6607 H_alpha 6652 H_alpha 6700 H_alpha 6742 H_alpha 6789 H_alpha 6835 H_alpha 6877 H_alpha 6925 H_alpha 6970

3. The new redshifted H_alpha filters:

H_alpha 6565/45 H_alpha 6565/60 H_alpha 6595/45 H_alpha 6625/45 H_alpha 6655/45 H_alpha 6685/45 H_alpha 6715/45

and here we present the tables for the scans:

1. The emission line set (except H_alpha)

[OII] 3727 [SII] 4075 HeII 4691 H beta [OIII] 5007 He II 5881 [OI] 6306 [OI] 6306 large scan H alpha [NII] 6590 [SII] 6730 [OII] 7331 [SIII] 9069 [SIII] 9539 HeII 10837

2. The old redshifted H_alpha filters:

H_alpha 6563 H_alpha 6607 H_alpha 6652 H_alpha 6700 H_alpha 6742 H_alpha 6789 H_alpha 6835 H_alpha 6877 H_alpha 6925 H_alpha 6970

3. The new redshifted H_alpha filters:

H_alpha 6565/45 H_alpha 6565/60 H_alpha 6595/45 H_alpha 6625/45 H_alpha 6655/45 H_alpha 6685/45 H_alpha 6715/45

General remarks

The filters marked with an asterisc in Table 2 have been scanned in October, 1994. No lightleaks were found, except for the [OI] 6306 filter We plan to scan the rest of the interference filters in the first half of 1995. Many filters are already old, which means that especially their edges are not as smooth as they could possibly be. To prevent stray light coming in through the edges, especially on INT-PF, be sure to tape off the edges before observing. Also, when calculating filter response curves: be sure that you also take into account the relative transmission of the CCD.

Non-imaging filters

This is a list of interference filters for use at the WHT, either for narrow-band imaging, or as order-sorting filters for instruments such as TAURUS or UES.

The filters have been scanned using the Guildford Spectrometer at RGO. The measurements covered a region of about 5mm in diameter close to the center of the filters. The measurements were taken with the filters at room temperature. Note that temperature changes will affect the transmission; the central wavelength of an interference filter shifts to the red with increasing temperature, with a temperature coefficient of about 0.2 A/degree C.

The transmission of an interference filter also depends on the angle of incidence of the light on it. This has the advantage that it is possible to fine-tune the central wavelength by tilting the filter; the central wavelength shifts to the blue with increasing tilt angle. For small tilt angles (smaller than 10 degrees) the effective wavelength is given by:

lambda_eff = lambda * ( 1 - (theta)**2/28954 )

It should however be noted that tilting an interference filter also affects the bandpass and the peak transmission; the bandpass increases and the peak transmission decreases with increasing tilt angle. Here is a summary of the filters.

Set 1

Set 2

Here are the transmission curves:

3721/28 3737/50 3742/40 3760/34 3769/30 3770/50 4550/100 4550/300 4770/320 4868/15 4880/100 4883/15 4898/15 4912/15 4962/15 4974/15 5000/350 5009/15 5010/100 5012/20 5021/15 5032/20 5033/15 5045/15 5052/20 5065/15 5072/20 5092/20 5108/16 5124/16 5145/15 5164/16 5175/15 5194/16 5216/16 5220/380 5232/16 5247/17 5267/15 5340/15 5700/450 5905/10 5960/490 6240/540 6303/15 6345/18 6565/15 6568/17 6577/15 6589/15 6589/16 6590/130 6601/15 6610/17 6613/15 6631/17 6631/17 6652/17 6667/15 6673/17 6676/15 6689/15 6692/18 6702/17 6714/18 6726/18 6730/50 6741/19 6754/19 6765/18 6770/50 6778/18 6792/18 6803/19 6816/18 6826/17 6838/17 6851/19 6859/16 6872/16 6884/18

And here are the tables with the filters transmissions:

Filters for use with the ING Telescopes – Historical Measurements