Welcome to

SPECTRAL OPTICS

Web Page

The FAQ and Tech Notes Page

Frequently Asked Questions

User Comments

The Four Big Myths

Our Tech Support Page

Hot Item Menu

Not all of the Frequently Asked Questions are covered in this Hot Item Menu. This menu only covers the most often asked questions. If you don't see you question here you will need to scan all of the Q/A below.

Q: What do I Focus on - The Star, the Spectrum or the Slit?

A: There are several answers to this question. They depend on the type of telescope you are using (Reflector or Refractor) and if you are doing Visual Observing or taking Pictures. This first answer will deal with focusing for Visual Observations, the next two Questions/Answers will deal with focusing for taking Pictures with Reflector type telescopes and then Refracting telescopes.

There are generally two methods used for Visual Observations of spectrums. 1. Use With-Out a cylindrical lens, and 2. Use With a cylindrical lens.

1. Use With-Out a Cylindrical lens:

Without a cylindrical lens spectrum focus is very simple. Focus on the star's image NOT the spectrum. Its that simple. If this doesn't spread the spectrum enough then defocus the spectrum's image even more. This may generate some criticism by others in the field but don't worry this method works just fine. Look at the out-of-focus picture of Vega where the H-a and H-b spectral lines just jump out at you from the spectrum's unfocused image.

Here, the spectrum was captured using a Refractor and focus was on the Red end of the spectrum (to the right side of the picture furthest end from the star's image). This caused the Blue end of the spectrum to go out of focus, as is the image of the star itself. In the defocused image the H-a line is 30x as tall as the Red end (where it was focused) and the H-b line is 40x as tall as the Red end. This is quite a large amount of defocusing and yet the spectral lines were preserved and plainly visible. The spectral lines are curved here because the spectral lines are arcs of a circle with the center of the circle located at the Red end where it was in focus.

The more you defocus the spectrum the dimmer it gets. You should stop defocusing the image when the spectrum becomes too dim to see well. This method is really effective for spectrums of point objects like stars and doesn't work as well for extended objects like planets.

If you are using a Reflector then that is all there is to it. If you are using a Refractor then as you change your focus the shape of the spectrum will change. In this case you can rack the focus in and out to shift the focus across the image of the spectrum while looking for spectral lines. Generally, you would first focus on the Red end of the star's spectrum (the end furthest from the star's image) and look for lines in the Green and Blue portions of the spectrum and then you would change the focus to the Blue end (the end of the spectrum that is closest to the star's image) and look for lines in the Red portion of the spectrum.

Unlike cylindrical lens this method can easily and quickly compensate for differences in stellar magnitudes and telescope sizes without requiring one to change a lens or lose a spectrum that's too dim and it doesn't cost anything to use.

2. Use With a Cylindrical lens - For Visual Use:

Focusing a spectrum for use with a cylindrical lens is a bit more complicated. First you need to focus on the spectrum itself (instead of the star's image). You need to get the image of the spectrum as small and thin as you can. Then put the cylindrical lens over the eyepiece.

If you are using a Reflector that is all there is to it. If you have a Refractor then it gets a bit more complicated. You will need to pick a point on the spectrum's image and focus it to as small of a line segment piece as you can (you can't focus the entire spectrum down to a long thin line with a refractor as you can with a reflector). After you have chosen your focal point then put on your cylindrical lens. You will usually only be able to see spectral lines around the area where you focused (this is just the reverse of what you have if you are Not using a cylindrical lens).

You will need to select a cylindrical lens that makes a fairly bright image of the star's spectrum you are observing. If you select the wrong size or focal length cylindrical lens then the spectral image may be too dim for you to see any spectral lines (it may be too dim to see the spectrum at all), or the spectrum may be too thin to reveal any spectral lines at all.

Q: Exactly where is the Spectrum's Focus located?

A: The Spectrum's focus is ALWAYS INSIDE the normal image focus point.

The spectrum's focal point is located just inside where the normal image is in sharp focus. What you do first is focus the object itself (its zero order image) as sharply as you can. Use the shortest CCD exposure that you can while still seeing the object well on the monitor (don't worry about its spectrum at this point or even if you can see it). Then after you have the object focused as sharply as possible increase the CCD exposure until you can see its spectrum really good.

Then very slowly turn the telescopes focus knob so that the focal point moves in past the CCD chip (the starting point would be when the object is in sharp focus then the focal point is right on the surface of the CCD chip). You want to move the focal point inward (toward the objective) as the spectrum will focus inside where the object is normally in sharp focus. This would be between where the zero order of the image of the object is in sharp focus, and the objective of the telescope. Remember that each time you turn the focus know you will have to wait for the exposure period before you will be able to see, on the monitor, what difference turning the know made. Ony turn the knob a little bit at a time (generally no more than a 1/4 turn and for some arrangements this may be too much).

If you have a refractor telescope then turn the knob to move the camera in toward the objective. This will move the zero order focus point out past the CCD chip. For a SCT this is a bit more complicated as some knobs turn clockwise to move the focus point out and some move counter-clockwise to move it outward. If you have a SCT, then put in a low power eyepiece and look at a distant object and get the image in sharp focus then loosten the eyepiece set screw and pull the eyepiece out about 1/4" and then tighten the set screw. The turn the focus know until the object is in focus again. Be sure to note the direction you had to turn the knob to get the new focus. This is the direction you will need to turn your SCT to move the focus out past the CCD chip.

The amount you need to defocus the image is a function of your setup. For example if you are using a Meade LPI camera then the central portion of the spectrum (590 nm - the yellow red transition region) will focus 0.40 mm past the CCD chip. If you have Meade DSI camera the central portion of the spectrum will be 0.34 mm past the CCD chip. If you have different spacings then the focal difference will be different.

The B1-S DG grating, designed for Meade's LPI camera, is designed to mount the grating 8.9 mm from the CCD chip. The B1-L DG grating, designed for Meade's DSI I camera, is designed to mount the grating 7.6 mm from the CCD chip. If you know the distance your grating is from the CCD chip, in mm, then multiply this distance by 0.0446. This is for the 500 l/mm compact DG Filters. The greater the separation distance the greater the spectrum focus difference.

In addition it should be mentioned that the blue end of the spectrum will come into focus before the central portion of the spectrum, and conversely the red end of the spectrum will come into focus after the central portion of the spectrum is in focus. On our B1 series units this difference is quite small so you only need to focus the central portion of the spectrum in order to photograph the entire spectrum from the blue end to the red end with little to no loss of detail. This is not necessarily the case when using the 1000 l/mm gratings or the D1 series DG Filters as they can be mounted much further from the CCD chip.

To see an exaggerated example of this look at Mike's first picture. He had a very expensive SBIG camera with a large CCD chip and was using a D1 DG Filter mounted in the SBIG filter wheel. This placed the D1 Filter well over an inch from the SBIG CCD chip. He focused on the spectrum's red end (this is the end pointing away from the zero order image - to the right in this picture). He was using an APO refractor (refractors even APO' have increased blue illumination from CA) and his camera had a heightened blue sensitivity.

For a more generic answer. If you know the separation distance, in mm, between the surface of your CCD chip and the DG Grating surface then you can calculate the center spectral focus offset distance, in mm. If you are using our 500 l/mm grating the grating's focus is: D = X * 0.045. Where X is the total separation distance in mm between the surface of the CCD chip and the grating surface, and D is the distance inside the zero order focal point location, in mm. If you are using our 1000 l/mm DG Grating then the offset is: D = X * 0.193.

For example, if you are using our 500 l/mm grating and the separation distance between the surface of the CCD chip and the grating is 9 mm, then the focal location of the central part of the spectrum will lie 0.40 mm inside the zero order focus location. If you were using a 1000 l/mm grating with a separation of 9 mm the central spectrum's focal location would be 1.74 mm inside the zero order's image location.

The actual equation used to determine the inside focus amount is: x = d * (1 - COS(a) ). Where d is the separation distance (8.9 mm for a LPI camera using a B1 DG Filter), (a) is the spectrum angle to a point on the spectrum (17.16 degrees using a 500 l/mm grating to the 590 nm point), and x is the in-focus amount in mm (the distance you need to move the focus in-toward the objective to focus the spectrum on the CCD chip). In the above calculations the 590 nm point was used as it is near the central point of the spectrum and easily seen on a color photograph. This angle is 17.16 degrees, for a 500 l/mm grating to the 590 nm point. To calculate the angle (a) for a different point on the spectrum use the following equation: a = ASin(G * L / 1000000); where G is the grating line pairs (ie: 500 l/mm), and L is the point on the spectrum in nm (ie: 590 nm).

If you don't know the separation between the CCD chip and the grating you can calculate it from a spectral picture (the picture doesn't need to be in good focus). You need to take a color picture and measure the distance from the center of the zero order image to the yellow/red transition region (this is the 590 nm point). Then measure the full width of the entire picture (from one black edge to the other). The take the ratio of these two measurements and multiply it by the width of the CCD chip used in your camera. This will give you the base side of a right triangle. Then divide this length in mm by 0.309 (if you are using our 500 l/mm grating - and 0.732 if you are using our 1000 l/mm grating). This will give you the separation distance in mm.

For example, if you are using a 500 l/mm grating and the distance measured between the zero order image and the yellow/red transition point is 11.1 cm (measured on your computer monitor's screen) and the full width of your picture is 19.8 cm. The ratio between these two lengths is 0.561. Multiply this by the width of the active pixel portion of the CCD chip you are using. For a small chip like that used in the LPI camera this is about 4.9 mm, and 4.9 x 0.561 = 2.75 mm. Then to find the separation distance divide this answer by 0.309, which gives 8.9 mm.

Q: I have a Reflecting Telescope, what do I focus on when I take a spectrum picture?

A: You ALWAYS, ALWAYS Focus on the spectrum's image NOT the star's image.

The grating sits back from the CCD chip from 0.8 cm up to 2.6 cm for some SBIG cameras. Once you have the star in good focus then you need to move the focus point INWARD until the spectrum gets as small as possible. Remember the image MUST focus on the surface of the grating not on the CCD chip to take great spectrum pictures. If you have a refractor you move the focus tube out, the same distance as the separation between the CCD chip and the grating surface, until the spectrum's image is in sharp focus.

This will cause the star to go out of focus. The greater the separation distance between the grating and the CCD chip the more out of focus the star will be. On a SBIG CFW-8 the distance is 25.9 mm, when connected to one of their ST-x cameras, and when the spectrum is in good focus the star will show diffraction rings around it (unless its over exposed). This is fine for these rings will make it much easier to find the center of the star and make wavelength measurements easier.

A Starlight Express CCD camera using the 1.25" nose piece sets the grating of a Type B1/2-L DG Filter at 13.2 mm from the CCD chip. While this is not as much as with the SBIG set up it is still very important to focus on the spectrum. In an LPI using a Type A1 Filter the separation is only 8 mm and the focus difference is noticeable.

Q: I have a Refractor do I need to do anything special when I take spectral photographs?

A: If you have a Refractor then it gets a bit more complicated, even if its a good APO, then when using a Reflector or SCT. Spectral images taken with refractors tend to flare at the ends. This is from secondary color associated with the objective. Refractors whose F/D ratio is smaller than 2.88 * D (where D is the objective diameter in inches) will exhibit some secondary color and this color will shown up as flaring in the Blue and Red ends of spectral images to varying degrees.

The above picture is the spectral image of a F/D 10 - 90 mm Achromatic Refractor. It is right on the edge of the 2.88 * D limit. Yet it still exhibits Blue end flaring of its spectral images. A good work around is to focus on the spectrum 20 % of the way between the Blue end and Red end. This would be around 490 nm. If the image is in B/W then remember the blue end Always points towards the image of the star (the zero order image). This will give the sharpest overall spectral image.

If you don't mind limiting your spectral images to just the visual region then you can screw a Baader Planetarium - Fringe Killer filter into our DG Filter then your spectral images will approach that of reflectors and SCT telescopes. The Fringe Killer filter will cut out most of the secondary color associated with refractors but it will also cut out all of the UV and IR.

If you are planning to use VSpec to process your spectral images you will need to produce straight line spectrums as VSpec won't know how to handle a flared, thick, tall, or curved images and consequently its synthesis may not be valid. If you plan to use VSpec you will need to get your spectrum images as close to a sharp straight line as you can.

Q: How do I Focus if I am using a Slit?

A: This may seem to be a bit more difficult as you can't see the slit to focus on, while you can see the star's image and the image of its spectrum but it isn't.

Start with the star's image in sharp focus with no slit present, then install the slit. The pattern to watch for, as you defocus the star's image, is: 1. Sharp star image and out of focus spectrum image, then 2. Star image going out of focus while the spectrum's image is getting sharper (thinner straight line), 3. Then the star's image getting More out of focus and the spectrum's image going out of focus. and 4. Continue to defocus until the star's double image merges into a single image. 5. Stop when the two dim slit images merge into one.

That's all there is to focusing with a slit. For a more detailed discussion of this along with pictures refer to the Variable Width Slit User Manual.

When using a slit turn OFF all Auto Contrast and Auto Brightness switches/controls and all Sharpness/Enhancement software filters. Set the image brightness range from 0 to maximum (256 or 65535). The slit image is much dimmer than the non-slit image and the Auto control features generally try to over compensate for this and the end result is a very grainy and noisy picture.

While the Variable Width Slit will work with most CCD cameras it is more efficient with cameras that have CCD chips that are 7 mm or more in width.

Q: Why do I get 10x the Resolution with your unit then is predicted by the Resolution Equation?

A: That's because the Resolution Equations predict what the minimum separation is to just resolve two spectral lines of equal magnitude. You aren't doing that. Look at a raw (untouched except for the comments) photograph taken of Vega by Mike using a ST-10XME camera (the ST-10XME has an extended blue sensitivity) on his 106 mm apo refractor. The image of Vega was focused on the grating that's why Vega's image (middle left of center) looks so out of focus and Mike focused on the Red end of Vega's spectrum.

The flaring of the blue end of the spectrum (the left end of the spectrum is the blue end) is extremely evident in this photograph (some of the blue flaring came from the apo's UV scattered light) but we don't lose any resolution because of it, as you can see. Two spectral lines were quickly identified in this image (using the ratio method). The first was the Ha (Hydrogen Alpha) line and the second was the Hb (Hydrogen Beta) line. There are many other lines visible in this photograph but only two lines are needed to answer this question.

The image of Vega when focused on the DG grating was 50u in diameter (at the grating surface). Mike was using one of out 500 l/mm gratings which meant that the disk of Vega, at the grating's surface, only covered 25 of the line pairs. From the Resolution Equation:

R = m N = Lambda / Delta(Lambda)

Here, m = 1 (the order number), N = 25 (50u wide of the 500 l/mm grating), and hence R = 25. Then Delta(Lambda) = Lambda / 25. And we find that the spectral resolution for the small image of Vega would be 200 Angstroms (where Lambda is 500 nm).

Now, take a closer look at the picture. The distance between the two lines is 1702 Angstroms. Now look at the thickness of the Ha line and the Hb line. Using a ruler and simple math we find the thickness of each of these two spectral lines is just under 20 Angstroms. The Resolution Equation Predicted that we couldn't resolve equal peaks closer than 200 Angstroms. But it is very easy to see that in practice we are able to see spectral line widths that are Ten Times (10x) Smaller then the 200 Angstrom Resolution Predicted by the Resolution Equation.

Remember that this picture has not been processed at all, other than to mark the two spectral line locations. With processing the wavelength of these two lines can be determined to within an Angstrom. That would be 10x finer then the thickness of these spectral lines and a Hundred Times (100x) smaller than the Resolution Equation predicts. If calibration pictures had been taken (which they weren't) the position of these lines could be determined to better than 0.1 Angstrom.

Measuring the position of spectral lines is not related to the spectral resolution of the lines (from the Resolution Equation), as it is a process of accurately measuring the positions of the lines after they have been photographed. With good image processing software and calibration pictures 0.01 Angstrom accuracy is not outside the realm of possibility.

In this question I did not attempt to explain why there is a 10x difference between what you are seeing and what the Resolution Equation predicts. Instead I endeavored to show you what you can actually see instead of trying to tell you its useless to look for the Resoltion Equation says you can't see anything.

Q: Why don't I see any Color?

A: Rule #1: If you want to see color you need to use a color camera. Most, not all but most, CCD cameras used for astrophotography are Black and White cameras. This includes the cameras made by SBIG, Starlight Express, Meade (except their LPI unit), and the list goes on and on and on...

There are two categories of color CCD cameras; 1. Those that integrate the CCD light, and 2. Those that don't (these are typically Web Cams). Do Not Confuse Integrating with Stacking - the two are not even somewhat related. Stacking just reduces the background noise and some will argue that it also brings out detail but any detail that Stacking brings out was already there in the first place you just couldn't see it because of all of the noise. Integration on the other hand actually adds more information to the image. Stacking an Integrated image will give you a really great image, while Stacking an un-integrated image will never be really great.

Leading the list of inexpensive color cameras, is Meade's LPI camera. This is a GREAT little camera for the money. It allows Integration as well as Automatic Image Stacking. You will not find a better camera anywhere that will produce such colorful and vivid spectrum images at this price. Meade's DSI camera while advertised as a color camera does not do well with spectrum colors. It is designed for large scale color and consequently turns all of a spectrum's blues and reds into a single violet color no matter what you do (for exposures over 0.5 seconds), you can't even correct this with image processing software.

To get colorful spectrum pictures with a B/W camera you will need to take three pictures of the spectrum using Red, Green and Blue filters - the colored filters are each screwed into the DG Filter (one at a time and a picture is taken then the color filter is replaced with the second colored filter and a picture is taken and so on). Then the three B/W images, just photographed through the three colored filters, are then aligned and a composite image is produced using image processing software.

If you want to see colorful spectrums of dim stars on your monitor, and don't want to spend a lot of money, your only option is to buy one of meade's LPI CCD cameras.

Q: What are those White Streaks I see?

A: That is what a spectrum looks like when you use a Black and White camera. Read the previous question.

Q: What do you have that sets your products apart from everyone else?

A: Much Better Resolution (5 to 10 times the resolution of the competition, except for SBIG) and a Red Shift Spectral Wavelength calibration tool that No One else has or even attempts to offer.

Our unique filters fit down in the barrel of a CCD camera bringing the grating much closer to the CCD chip than any of our competitors can do using the filter thread arrangement they do. Our filters generally lie within 5 mm of the CCD chip while the competitors filters reside about two inches from the CCD chip.

As the spectrum spreads out from the grating the farther it is from the chip the wider the chip has to be to see the entire spectrum. To compensate for this, with the small size of the CCD chip, the competition uses a very low resolution grating (around 200 l/mm) so it won't spread the spectrum as much over the long travel distance to the chip.

Our units on the other hand lie very close to the chip so we can use a much higher resolution grating and still get the entire spectrum on the chip. Most of the less expensive CCD chips are only 5 mm wide and the spectrum needs to completely illuminate the full CCD chip's surface while not falling off either side of the chip. As our filters are much closer to the chip they can produce a spectrum that has five to ten times the resolution of our competitors and do it for less money.

As our units are designed to cover the entire chip with a spectrum and our competitors units can't, this greatly increases our unit's effective resolution as seen by the CCD chip over what the competitior's units can produce. The reason for this is that the competition's products were initially designed to be visual use only units and they didn't see any need to redesign their products to accomodate the new CCD chips that have since become available to the amateur market.

While a few amateurs have adapted the competition's units for their use and produced remarkable results, if they had spent the same amount of time and effort on one of our units they would be able to produce spectrums with significant resolution gains better than anything they had ever done.

Q: Why is your tool so special?

A: Because without knowing what the wavelength of the spectral lines you photograph are you won't be able to determine the wavelength of the line. Without that you won't be able to calculate Red Shifts or know what elements/compounds the spectral lines represent. This is the main goal of astro-spectroscopy. Without this all you have is a colored line.

Q: Why do you have this tool and no one else is offering it?

A: Because we built this for our own use as no one else had anything useable on the market. We designed a special Spectrum Filter to take advantage of the new phase of CCD cameras now available and then we developed a special tool to calibrate the spectrums it produced. After we finished we saw how new and unique our filter was so we filed a patent application on it and decided to market it at a very affordable price to attract more amateurs into this exciting field and we filed a copyright on our software tool and bundeled it with out unique filter.

Q: Why is your calibration so good?

A: Because we derived the actual equation that the spectrum produced by our products follows. It is: y = A * x / Square-root(x * x + 1); where y is the actual wavelength of the line located at the point x within a spectrum, and A is a grating constant. Every manufacturer's spectral device has its own calibration equation and the more complicated it is the more the manufacturer leaves you on your own. The Grism device is a good example.

Q: Other manufacturers offer wavelength calibrations, like SBIG on their SGS Spectroscope?

A: What all the other manufacturers don't tell you is that the spectrums their equipment produces is non-linear with respect to wavelength versus length of the spectrum. This is a big secret as most of them don't know an easy way to actually do non-linear calibration.

None of the other manufacturers of these devices offer any help at all to calibrate the spectrums their devices produce (except Spectral Optics and SBIG) and leave their customers to their own devices to figure this one out on their own, or not. The cheaper the product the more you are on your own, except for Spectral Optics.

To answer your question - NO Manufacturers, other than Spectral Optics and SBIG, offer any type of wavelength calibration directly useable with the spectrums produced by their products. We have a Specially Designed Proprietary Tool that will allow direct calibration of the spectrums All of our devices produce. SBIG offers a Freeware Application for their two spectral devices to assist in calibrating their spectrums.

Q: What is a Grism device?

A: This is a combination of a prism with an attached transmission diffraction grating. The grating is generally glued to the prism and light to be analyzed passes through both items before it comes out as a modified spectrum. They were discussed decades ago in J.B. Sidgwick's 1971 book "Amateur Astronome's Handbook", in his chapter on spectroscopes. Since then a number of others have came up with variations of these designs to solve various short comings of cheap spectrum devices. But none of these variations ever over came the real problem of wavelength calibration. Without that the spectrum produced is little more than star party entertainment.

Q: Do you have anything that I can use with my eyepieces to produce a spectrum as I don't have a CCD camera?

A: Yes, we offer our Types D1 and D3 - DG Spectrum Filters that will screw into any eyepiece that has standard 1.25" filter threads on it. The D1 unit has more resolution (then our D3 unit) and will allow you to see spectral line in the brighter stars (this will require that you purchase a cylindrical lens and mount it on the top of your eyepiece - these are available from Anchor Optical Surplus and Edmund Optics).

Our D3 unit is designed to provide a very colorful visual treat if you would like to see color where most people think there is only Black & White. The D3 unit produces spectrums in four planes so its quite bright and colorful while the D1 unit only produces spectrums in one plane and isn't as bright. With the D1 unit its possible to see spectral lines (if you have a cylindrical lens) but with the D3 unit it isn't.

TIP: The D1 unit is a bit picky about which eyepieces it will work with while the D3 unit isn't. However, you can use the D1 with Any eyepiece if you lay the D1 Filter on top of the eyepiece instead of screwing it in the bottom of the eyepiece. This reduces the width of the spectrum so that it can be easily seen. This also produces a smaller but brighter spectrum.

Q: Which eyepieces work best for viewing spectrums?

A: About any 1.25" eyepiece will work with the D3 unit. The longer focal lengths provide wider fields so you can see more stars and consequently this is what the D3 unit was designed for. Under a rich field of bright stars with a D3 unit one can see vivid colorful spectrums all over the place. The view with a Type D3 - DG Spectrum Filter is Absolutely Stunning.

With the D1 unit its not all that simple. However, the D1 unit has 2.5 x the resolution of the D3 unit which means the D1 unit will resolve the spectral lines within a spectrum (if you have a cylindrical lens). As the D1 unit has much wider dispersion then the D3 unit the D1 unit will only work on eyepieces that satisfy certain criteria (there has to be enough FOV to be able to see both the star and its spectrum).

This following is what is required just to see the spectrum with a D1 Filter which is necessary before you can start to look for spectral lines using a cylindrical lens. (The following paragraphs for this question can be ignored if you are using a D3 Spectrum Filter)

The things you need to know about your eyepiece is: 1 - its focal length in mm, 2 - its Field of View (FOV) Angle which should be available from the manufacturer's spec sheets, and 3 - the length of the eyepiece body (from the bottom threaded end to the top of the top most eyepiece lens not to the top of the eyepiece itself - as Meade's 4000 series eyepieces have the top most lens recessed 10 mm below the top of the eyepiece barrel).

Then take this information and plug your numbers into the following equation:

        X = (FL / 25) * L * Tan (FOV / 2)

Where: FL is the Focal Length of the eyepiece in mm
       L is the body length in mm
       FOV is the Field of View Angle for the eyepiece in degrees
       Tan is the tangent function

A list of a few different 1.25" eyepieces follows:

Name      FL     FOV      L      X 

MA        12      40     56     10
MA        40      37     60     32
4000      26      52     62     31
Epic       3.7    55    114      9
Ultima    12.5    51     48     11
Ultima    18      52     56     20
Ultima    30      54     76     46
SWide     15      65     53     20
Owl New    6      66     71     11
Owl New    9      66     69     16
Owl New   15      66     47     18
Owl New   20      66     58     30

Where: MA = Meade's MA eyepieces
       4000 = Meade's 4000 series eyepieces
       Epic = Orions Epic series eyepieces
       SWide = OPT brand of super wide angle eyepieces
       Owl New = Knight Owls brand of new super wide eyepieces

With the D1 units and most eyepieces, but not all, you will need to put the star's image near one side of the eyepiece field, focus the star and then look in the opposite side of the eyepiece for the star's spectral image (this would be in the dark space just outside of the bright central focus area of the eyepiece). The spectrum is easier to see if you don't get your eye too close to the eyepiece. This assumes that you have aligned the yellow dot on the D1 Filter so that the star's spectrum is in the same direction as the edge you move the star. If the DG Filter is laid on the top of the eyepiece then you will see the star's spectrum in the central bright area of the eyepiece.

What do I do with this number X? It is the effective spectral degree width for that eyepiece. If it is greater than the spread of the spectrum of the grating you are using you will be able to see all of the spectrum, if it isn't then you won't be able to see all of the spectrum. For our 500 l/mm grating the visible spectrum is spread over 11.4 Degrees to 20.4 Degrees. The Red end is at 20.4 Degrees and the Blue end is at 11.4 Degrees.

If: 20.4 > X > 11.4 = then you will be able to see the star and the blue end of its spectrum but not all of the red end of its spectrum.
If: X > 20.4 = then you will be able to see all of a star's spectrum along with the image of the star.
If: X < 11.4 = then you will not be able to see the image of the star and its spectrum at the same time.

TIP: You can use the D1 with Any eyepiece if you lay the D1 Filter on top of the eyepiece instead of screwing it in the bottom of the eyepiece. This reduces the width of the spectrum so that it can be easily seen. This also produces a smaller but brighter spectrum.

Q: Do you have anything that I can use with my film camera so I can photograph spectrums of objects?

A: Yes, go to our Accessories page. We offer standard camera filters from 28 mm up to 72 mm as Type D Filters. These are special order and takes from two to four weeks for shipment.

Q: Rainbow Optical and Baader Planetarium offer spectrum devices that come with a cylindrical lens, does your Type D unit come with a cylindrical lens for viewing star spectrums?

A: Cylindrical lenses are used to defocus/spread the spectrum perpendicular to its axis to make the spectral lines in the spectrum are more easily visible. These lenses are expensive and very generally quite inefficient and you need a lot of different ones to cover different combinations of stellar magnitudes, eyepieces (yes, you need to match it to your eyepiece) and telescope sizes (see the Next FAQ Question for more details). There are other more cost efficient ways to achieve the same effect without using a cylindrical lens, for this refer to our How to Focus FAQ Question.

The Type D1/D3 Filters can be used with just about any eyepiece, that accepts standard 1.25" filters, for strictly visual observations. The Type D Filter does not come with a cylindrical lens, as is supplied with similar products from competitors. The reason we don't offer a cylindrical lens option is because no visual spectrum filter made by any manufacturer, no matter what they might say, will produce spectrums of dim stars. When a spectrum filter is combined with a cylindrical lens the spectral image is so dim that 3 rd magnitude stars become the limiting magnitude, when using an 8" to 10" telescope.

Even with a cylindrical lens its very difficult to actually see any of the spectral lines present much less identify any of them. Without a cylindrical lens its extremely difficult to see any spectral line information except the very wide ones, but without using a cylindrical lens you can see the spectrums of stars several magnitudes dimmer then is possible when using a cylindrical lens.

This is one of the reasons that we don't offer cylindrical lens with our units. Another is that we would need to offer a lot of different cylindrical lens as they need to be matched to the specific eyepiece and telescope you are using. This is one of the things that other manufacturers don't tell you as they only supply one cylindrical lens. The older GOTO Star Spectroscope supplied three different cylindrical lens with their unit and had instructions on which one to use with different optical systems. To supply a single cylindrical lens that would only work well for 10% of your customers and leave the other 90% to their own devices is not a good way to do business. For more detail on cylindrical lens see our other Cylindrical Lens question below.

Additionally, the software on our CD is specifically for enhancing spectrum images. If a Type D Filter is purchased for exclusive visual use then nothing at all on the CD will be of any use, except the users manual for the Type D Filter. This is not a short coming of our Type D Filter, it is a short coming of All spectrum filters that mount on eyepieces no matter who makes them or how much they cost. That's why none of our competitors offer anything but a simple user manual with their units, as for visual use there is very little that one can actually do or see.

Using a Type D Filter for visual use is fine at a star party where people would like a colorful visual treat or would like to see what a spectrum looks like. But the real power comes when you combine this filter with a camera and produce the image of an object's spectrum. Then you can start to do something with the information available in the spectrum. Then you will need the software on our CD. That's why none of our competitors offer you anything with their spectrum filters as their units were not designed to be used with a camera. A few amateurs have adapted their devices for camera use and have had varying degrees of success.

Q: Which cylindrical lens do I need to buy to see spectral lines?

A: First, you will need a lens that will cover most of the exit lens diameter at the top of the eyepiece that you are going to use it on. Second, is the focal length of the cylindrical lens: Focal lengths from 150 mm to 250 mm is good for smaller diameter telescopes (< 4" in Diameter) and bright stars, 90 mm to 130 mm is good for larger telescopes and 1 to 2 magnitude stars, and 50 mm to 80 mm is good for dim stars (3rd+ magnitude) on larger telescopes.

Anchor Optical Surplus and Edmund Optics sells cylindrical lenses. Edmund has a better selection but they are much more expensive than Anchor Optics. Once you have your cylindrical lens you will need to build a small hat to mount it on. The hat can be made of a cardboard disk with a short cardboard skirt around it. The skirt needs to be able to fit over the eyepiece that you want to use it with.

To use your cylindrical lens, after you have it mounted so it fits over the top of your eyepiece. First, without the cylindrical lens over the eyepiece, find a spectrum that you would like to see and the focus it until the spectrum is a thin straight line or the best that you can. Then put the cylindrical lens over the top of your eyepiece. You will need to rotate the cylindrical lens until its axis is lines up with the axis of the spectrum's image. Then you will see the spectrum still stretched out but you will also see that the spectrum now has height to it.

With the proper cylindrical lens adjusted correctly the thin line spectrum will now be a rectangle with the same length as the thin spectrum had. If the cylindrical lens is too powerful for the telescope or the star is too dim then the stretched spectrum will be too dim to see well. If the cylindrical lens is not strong enough then the spectrum won't be stretched very much from the initial long thin line.

Cylindrical lens need to be matched with the size of the telescope, the eyepiece, and the magnitude of star to be observed. As these parameters change you will need to go to a different focal length cylindrical lens. So when you buy your first cylindrical lens you should buy a range of different focal length ones. The manufacturer of the famous GOTO Star Spectroscope supplied three different cylindrical lens with their unit.

Q: Have you tried out your competition's products?

A: We own a GOTO prism star spectroscope and a Rainbow Optics star spectrum filter. Both of these units are strictly visual observation devices. Sky and Telescope magazine ran a comparison between the Rainbow Optics product against the GOTO spectroscope a few years ago and found that there was little difference between them. If you are only interested in looking at very bright stars to see a faint thin, sometimes colored, line then you won't be disappointed.

If you are going this route then you might want to consider one of our Type D3 units as they sell for about one half the price and produce superior results.

Q: What kinds of things can I determine from a spectrum?

A: There are two things: 1. By using VSpec and Spectrum (both of these packages are included on our CD) you can determine what a lot of the elements corresponding to the spectral lines, in your spectrum, are. 2. You can determine if the object is moving toward or away from you and how fast its moving if your spectrum is also calibrated. We are the only company that has a calibration tool and its UCC Copyrighted so no one else can offer it, not even those companies located outside of the United States.

Q: I don't understand the difference between the two (set out in the last question above)?

A: By using VSpec you can see the spectral lines in your spectrum but you won't know what elements these line are from. By using Spectrum and matching up the pattern of spectral lines you can determine the elements that are producing a lot of the spectral lines in your spectrum. Once this is know then you can determine the rest (rest - meaning zero velocity) wavelength of each of the identified lines.

Without the ability to accurately calibrate the spectrum you photographed in wavelength (remember spectrums produced by Any means are always non-linear by nature) you will have no idea if your observed lines have been Red Shifted or not. This is a really big thing in astronomy - trying to locate black holes by observing the speed of the material in its accretion disk and the stars orbiting the black hole. They have found there are black holes in the center of all of the large galaxies.

If you can calibrate your spectrum, then after you have identified your lines using Spectrum all you need to do is determine your photographed wavelength of one of these lines and compare the difference between what you found in your photograph and the rest wavelength. Once this is know then the Red Shift of the objects spectrum can be calculated with reasonable precision. Once this is known the objects speed can be readily determined.

Q: Are you going to offer anything for other CCD cameras?

A: The problem here is that we will have to buy a camera to make a DG Filter to fit it. This is very expensive as there are hundreds and hundreds of makes and models being sold around the world. What we are doing is developing something for the cameras that we get the most request for.

Q: Why do you talk about using Photoshop for image processing when you have a CD with image processing software on it?

A: The image processing software included on our CD is IRIS. It is an application that will allow you to rotate an image (very important for VSpec) and to convert an image into a FIT file that VSpec will accept.

While IRIS has a few functions that can be used to process a spectral image it is very limited in this respect, as is most astronomical image processing software. Photoshop on the other hand is used to process a more general variety of images. This lends itself well to processing spectral images.

We have developed a copyrighted process using Photoshop 7 to process spectral images. If you use Photoshop 7 instead of IRIS, to process your spectral images, you will be able to analyze spectrums one to two magnitudes fainter than is possible using the available tools in IRIS.

Photoshop 7 was the last version of Photoshop that didn't require activation. The newer versions of Photoshop have new features that are not used to enhance spectral images so purchasing the most recent version is not necessary. Versions as far back as Photoshop 5 have all of the tools necessary to process spectral images.

One thing you need to avoid is using any type of image processing software that will alter the content of the spectral information. This generally leaves out most of the techniques used to enhance planetary and deep sky images, stacking, etc. You need to preserve the information in the spectrum. All our image processing does is to bring out the detail that is too dim to be adequately seen in the image but was there all the time.

If you don't have Photoshop you can still use our included software to process your spectral images, you just won't be able to photograph as dim a magnitude object as would be possible if you were using Photoshop to process your images. Photoshop 7 and older versions are sold on ebay every day.

Q: How good is VSpec' calibration?

A: VSpec doesn't have any automatic calibration. It does have several calibrate selections but all of them require that the user have determined at some wavelength points PRIOR to using VSpec. As the spectrum is non-linear with respect to distance along the spectrum a minimum of three points is required for non-linear calibration.

Our spectral tool will assist you in calculating these points prior to using VSpec. VSpec doesn't actually use the wavelength in any of its calculations so you don't have to do this to use VSpec. Calibration (but not necessarily VSpec calibration) is required if you intend to measure Red Shifts. With the tools included on our CD you can identify elements corresponding to the spectral lines found in your spectrums with out worrying about doing a non-linear calibration of VSpec.

Q: What do I need to know about using your DG Filters with a SBIG CCD camera?

A: Our Type D - DG Filters will mount in SBIG's CFW-8 Filter Wheel which accepts standard 1.25" filters. Additionally SBIG offers an adapter where the T-thread can be replace with a T-thread to Ring Adapter which also accepts standard 1.25" filters. As such our Type D1/D2 - DG Filters will fit in them as well. However, the various combinations of devices and distances available produces widely differing results. The following table shows generally what to expect with different combinations of chip sizes, DG Filters and placement locations for our Type D - DG Filters when used with a SBIG CCD camera:

                Red Spectrum Limit (nm)

SBIG CCD        CFW-8     |  T-thread Adapter
             D1      D2*  |     D1     D2*   (DG Filter Type)
ST-7/402    410     560   |    750    710
ST-8        770     715   |   1260    885
ST-9        590     645   |   1025    805
ST-10       825     735   |   1320    900

DG Filters transmission starts to drop off at 1000 nm.

The Type D2 - DG Filters pass band is computed from 390 nm on one side/edge of the chip/monitor to the column indicated Red Limit cut off wavelength in nm.

Q: What do I need to know about using your DG Filters with a Starlight Express CCD camera?

A: Starlight Express offers a standard 1.25" threaded nosepiece that accepts standard 1.25" filters. As such our Type B1/B2 - DG Filters will screw into the threads at the front of the nosepiece. But the various combinations of devices and distances available produces widely differing results. The following table shows generally what to expect with different combinations of chip sizes, DG Filters and placement locations:

                Red Spectrum Limit (nm)

Starlight CCD Camera   +   1.25" Nosepiece Adapter 
                               B1         B2*  (DG Filter Type)
MX-7C/716                     880       1210
MX-5/516/5C/HX-516            695       1050
MX-916                       1100       1405
SXV-H9C/M9                   1125       1425

DG Filters transmission starts to drop off at 1000 nm as does the response of the MX and HX cameras.

The Type B2x DG Filters pass band is computed from 390 nm on one side/edge of the chip/monitor to the column indicated Red Limit cut off wavelength in nm.

Q: What are the dimension of your Type B - DG Filters, so I can measure my camera and see if it will fit?

A: There are four versions of the Type B unit. They are: B1-S, B1-L, B2-S and B2-L. The barrel of all four of them is the same diameter, which is 17/32" in Outside Diameter. The S suffix units are 'shorter' in length then are the L suffix units - L = 'longer'.

The S suffix units extend 18 mm down from the base of the DG Filter cell (this is measured from the top of the filter threads on the filter cell). The filter threads extend down from the filter cell 2 mm.

The L suffix units extend 21 mm down from the base of the DG Filter cell (this is measured from the top of the filter threads on the filter cell). The filter threads extend down from the filter cell 2 mm.

The DG Filter cell extends 8.03 mm above its threads. This means that when screwed into a threaded barrel (that accepts standard 1.25" filters) the top of the filter cell will stand up 8.03 mm above the flange on the barrel its threaded into.

You can't directly select a S or L suffix Type B DG Filter. You can only select a Type B1 or a Type B2. Then based on which camera option you select when you place an order the type shown on your order will have the suffix attached to it. This was done to simplify the selection process.

For those that want to force a selection for a specific unit you can select a LPI camera to get the S suffix unit and specify a DSI camera to get the L suffix part. Then in the comment section of your order tell us which camera you are actually going to use it on.

Q: What are the dimension of your Type A - DG Filters?

A: There are two versions of the Type A unit. They are: A1 and A2. The barrel of both of them is the same diameter, which is 9/16" in Outside Diameter.

Both Type A DG Filters have the same length of 7/8". It was designed to be inserted into an empty lens well/housing that has in Inside Diameter of 9/16", like Meade's LPI camera. It is held in place by the friction between the filter and the inside wall of the lens well.

Q: Do your DG Spectrum Filters have/use a slit?

A: You generally only need a slit if you are making spectral pictures of extended objects. Stars are not extended objects. Planets and nebula are extended objects. Our units do not incorporate a slit in their design.

Adding a slit would cut down on the light and increase the manufacturing complexity and cost. To answer your question No our DG Spectrum Filters do not incorporate a slit from the factory.

However, we do offer a Variable Width Slit that can be used on extended objects with our DG Filters. For information about it refer to our Accessories Page.

Q: How do I increase the Resolution of your DG Filters?

A: Are you are asking about trying to resolve finer line positions/wavelengths, or more spectral lines? If its - How do I resolve finer line positions/wavelengths? Then please read our FAQ Question How can your DG Filters measure wavelengths to 0.01 Angstroms for Red Shift measurements when the CCD cameras, with the Grating filter, only have a resolution of 10 or more Angstroms per pixel? and our FAQ Question Can I do Red Shift measurements with your DG Filters?

If its - How do I see/resolve more spectral lines? Then you will first need to know more about how spectrometers work and what limitations resolution. The current line of DG Filters were designed to be used by amateur astronomers for use with the new generation of inexpensive CCD cameras on the market today. These new cameras and their software opened up a whole new era of astrophotography that is now within the reach and temperament of most amateur astronomers.

The number of spectral lines that can be resolved/seen is a function of:
1) the number of line pairs available on the grating,
2) the spacing between the grating and the CCD chip and
3) the percent of illumination of the grating by an object.

The present system was designed to be used with the least amount of auxiliary equipment, time and training. Otherwise the average amateur astronomer would not be interested in using it. In short the DG Filter line had to be inexpensive, require little to no training or learning curve, require no expensive attachments, and to work with the equipment that the amateur astronomer happen to have - his/her telescope and CCD camera.

These requirements produced a small/compact unit that works very close to the CCD chip so as to fill the entire CCD chip with spectral information without having to resort to using expensive and bulky lens systems. This is what the DG Filter line set out to accomplish and it did. Today using our DG Filters more amateur astronomers are able to casually observe astronomical spectrums with only a few minutes of work.

The trade off for simplicity and cost was in line resolution, not line position. Line position is a function of how well you can measure the location of a line that has been photographed. Line resolution is how close together you can see different spectral lines before you take the picture. We will examine each of the three requirements set out above in detail.

1) Number of Line Pairs. This is simply the number of ruled lines the grating has. Usually it is specified as number of lines per mm or inch. A grating of 500 l/mm is about optimum for amateur astronomy use as it provides good resolution along with extended IR coverage. There were no trade-offs with the DG Filter line here.

2) Spacing between the grating and the CCD chip. The DG system was designed to use the new line of inexpensive CCD cameras on the market. All of these cameras had a very small chip (around 4.8 mm wide - 640 pixels). In order to cover this small area with both the zero and first order spectrums the DG Filter needed to be very close to the CCD chip (generally within a 1/2"). This is rather close spacing for a spectrometer. Because of this the spectrum produced does not have a lot of spread before the spectral lines overlap. When spectral lines overlap all the CCD chip sees is a single thick line. Most commercial spectrometers have separations of at least a foot and some up to three feet. As this would require a CCD chip several inches in width it easy to see why this would be a costly problem for the amateur and just not practical for the mass market.

3) The percent of the grating that is illuminated by the object. As the DG system was designed to be quick and easy to use and setup, it does not employ any external optics. Instead it uses the optics of the telescope alone. The telescope image is used to focus the image onto the grating's surface and then to the CCD chip. Typically the image of a 1st magnitude star through a telescope with a focal length of 1500 mm will illuminate a 50u circle of the grating. This would only illuminate 25 lines of the grating. From the grating equation this would correspond to a resolution of 200Angstroms. But in practice we see line resolutions of closer to 20 Angstroms. For more about this difference read our FAQ Question Why do I get 10x the Resolution with your unit then is predicted by the Resolution Equation?

In the DG Filter arrangement the spectral image is imaged onto the CCD chip without using any auxiliary optics (which would increase the spectral spread so that the zero and first order images no longer fit on these narrow CCD chips). By having both of these images on the chip at the same time greatly simplifies the line identification process which is appealing to the average amateur. For those amateurs who want more resolution the DG line also incorporated a 1000 l/mm grating that produces a much wider spread as it allows the first order image to completely cover the entire width of these CCD chips.

Using a standard Type 1 grating (500 l/mm) DG Filter on a Meade DSI or LPI camera the thickness of a star's spectral line will vary between 20 Angstroms (A) and 50 A depending on the brightness of the star and the telescope used. This means that spectral lines that are the very same intensity will just be resolved when they are 1.5x the line thickness (30A to 75A) apart. For spectral lines of much different intensities this spacing increases up to around 2.5x (50A to 125A). With the short spacing between the grating and the chip coupled with the small chip size this is about the best that can be had. A Type 2 grating (1000 l/mm) will decrease the line thickness at the expense of only being able to see the first order image, which makes it much harder to identify the spectral lines. As a point of reference each pixel in these two camera's covers approximately 13 A.

To increase the resolution using these CCD cameras the third requirement, 3), can be changed. This will require two lenses be added. One positive lens will be placed between the telescope's focal point and the DG Filter (a D1 DG Filter would work best here and a B1 would not have the room to work using this method). Then a second positive achromat would be placed between the DG Filter and the CCD chip. A holder arrangement would need to be fabricated to rigidly hold both lenses and the DG Filter in place and one that would both allow attachment of the CCD camera to the holder and the holder to the telescope.

The first lens would be placed at its focal length from the focal point of the telescope (a slit may be used here but is not required). This will change the telescope's image into a parallel beam the diameter of the lens (the lens would need to be no smaller than 10 mm in diameter to work with a D1 Filter). The second lens would be mounted at its focal distance from the CCD chip and placed in back of the D1 Filter (in line with the parallel beam from the first lens).

Using this arrangement the telescope would focus an object to a point where the first lens would change the point image to a very wide parallel beam that then goes through the D1 Filter at which point the second lens picks up this exit beam and focuses it on the CCD chip. To complicate matters the second lens would need to be offset by 17 degrees from the telescope's optical axis so it would focus the spectrum (which is bent at a 17 degree angle) onto the CCD chip. Additionally the second lens would need to have a focal length of less than 15 mm (otherwise the spectral image would be larger than the CCD chip and be lost). The second lens also needs to be the same diameter as the first lens.

In the above example the number of lines illuminated would increase from 25 lines to 5000 lines. The increase in resolution would be from 200 A using the standard DG Filter with no lenses, to 1 A with these two 100 mm diameter lenses - a 200x increase. This increase is significant but fabricating the housing would be costly and the two lenses needed would be extremely expensive (several hundred dollars each if you can find them on the surplus market).

To increase the resolution per the second requirement, 2), can also be done. This would negate using any of the inexpensive CCD camera's on the market as it will require an extremely large CCD chip size. The chip size would need to be 40% of the separation distance (in order to show both the zero and first order images at the same time as the DG system presently does). For a separation of only 6" (which would give you a resolution increase of 16x) you would need a CCD chip 61 mm wide. If you only wanted the first order image this could be had using a CCD chip only 31 mm wide. CCD Chips of this size are beyond the reach of most amateurs today. Besides, this isn't the way the professionals do it. They use a scanning photo-multiplier system (also extremely expensive).

As far as increasing the resolution per the first requirement, 1). You couldn't do any better than using either a 500 l/mm (or 600 l/mm) or a 1000 l/mm grating. Both of which are already offered in the DG Filter line.

In Conclusion: The most cost effective method and the one that yields the greatest resolution gains is modifying requirement 3) above.

CAUTION: There are three things that you need to consider when trying to increase your line resolution. They are:

1. Your photographs won't show line widths any narrower then the base pixel width of the CCD camera you are using. For example, using a DSI (or LPI) camera and a 500 l/mm DG Filter, each pixel will cover 13 Angstroms. If you get the line resolution of the grating down to, say, 1 Angstrom then you will cut down your sensitivity by a factor of ten and your photographed line widths will still be 13 A wide but ten times dimmer.

2. To do Red Shift measurements you need a grating resolution, at the CCD chip, of about twice the pixel resolution. For a 500 l/mm DG Filter this would be 20 A to 25 A. It would be half this using a 1000 l/mm DG Filter but don't try to do Red Shifts using a 1000 l/mm grating right away as its extremely difficult to do red shift measurements using a 1000 l/mm grating - more on this under the NOTE at the bottom of the Red Shift FAQ Question.

3. If your line resolution is smaller than your base pixel width of your CCD camera you will Not be able to do any Red Shift Measurements for objects that have Red Shifts smaller than the base pixel resolution of your CCD camera. As most stars have Red Shifts of less than 1 Angstrom this about rules out doing any Red Shift measurements with CCD cameras that have pixel resolutions greater than 1 A.

Q: How can I add/build a slit to use with my DG Filter?

A: A few users have improvised a slit into our devices. They take aluminum foil and paint it flat black (on both sides of the foil) and carefully glue a small piece of the foil with a thin slit cut in it, using an exacto knife or razor blade, to one edge of the grating of the DG Filter (or at a location on the DG Filter's grating' outer surface just above the edge of their CCD chip if the chip is smaller than the DG Filter's open diameter). This way if they move the object to one side they will have its full unimpeded spectrum and if the move it to the other side the object encounters the slit and the spectrum produced is only from the light from the object that passed through the slit.

With one of our Compact units both the image of the object through the slit and its spectrum are visible on the monitor which makes object and spectrum placement reasonably easy.

The downside here is the accuracy of the tracking required when using a slit. Using a slit greatly increases the need for an extremely well guide scope and generally short exposures. Additionally, using a slit greatly cuts down on the light getting to the camera which requires longer exposures which requires better tracking, and on and on.

Additionally, for a slit to be effective it needs to be separated a distance from the grating material. If you build a slit on the grating material it will not perform as well as a slit that is separated from the grating material.

Q: Is the DG Spectrum Filter's grating blazed?

A: Blazed gratings have five inherent short-comings when it comes to using them for CCD astrophotography. They are:

1. They allow more intensity to pass in the center of the visual region at the expense of having decreased efficiency over the rest of the visual region.

2. They generally only have a useable band width from 420 nm to 670 nm, when blazed for 550 nm (or 500 nm either). By their nature the UV and the IR regions are cut off in order for a slight increase in efficiency where our eyes are most sensitive.

3. Blazed gratings are designed for a parallel light beam passing directly perpendicular to the normal grating surface. If the light reaches the grating at angles other than 90 degrees then the blaze angle changes and its band width warps. This is generally not a problem for telescopes with a F/D > 20 (a F/D 10 SCT with a 2X Barlow). However, for short scopes or scopes with focal reducers there is a problem. For F/D <6 the incoming light rays hit the grating blazing at two widely differing angles. This skews the blaze point wavelength and narrows the band width.

4. In addition to the skewing of the blaze point and narrowing of the band width there is another problem. The light coming in from the half of the objective that is perpendicular to the blaze surface is refracted into a bright spectrum and the light from the other half of the objective hits the blazing on its edge. This edge illumination is mostly lost because of the steep angle of attack at the blazing surface. The bottom line is that 40% of the objective's light is lost from the spectrum for all short F/D ratio telescopes using a blazed grating.

5. For CCD work using a camera that has the IR blocking filter removed, using a blazed grating will drastically cut down the width of the spectrum available. If a blazed grating is used on a CCD camera that has the IR Blocking filter in place there will be little gained out side of the visual region.

To answer your question. No, our gratings are not blazed. This way they produce a spectrum that is more uniform throughout its band width, of 380 nm to 940 nm. They were designed for CCD photography using a camera that does not have an IR Blocking filter in place.

Also read Myth 1 on our Big Myths Page.

Q: Why do manufacturers blaze their gratings?

A: Blazed gratings are specifically made to respond to a small specific region of light. For amateur astronomy use this would be for observing star spectral in the visible region (as your eyes can't see outside of the visible region). The blazing is generally for 550 nm which is the region our eyes are most sensitive to.

The draw back here is that a blazed grating only passes a very small part of an objects spectrum. If all you are interested in doing is visual observation of spectrums then a blazed grating will give you a little more colored light but not very much more. Additionally, blazed gratings that are designed for visual work have gratings of around 200 l/mm. This type of grating produces a very short spectrum compared to much longer spectrums produced by more dense gratings, like the ones we manufacture.

A typical blazed grating has an efficiency of around 60% at 550 nm (depending on the manufacturer), with a spectral width of 420 nm to 670 nm where the efficiency falls off rapidly as the wavelength departs from 550 nm. A non-blazed grating has a fairly flat efficiency curve of around 50% over the spectral bandwidth of 370 nm to 970 nm. Do not confuse Efficiency with Transmission as they are Not the same thing.

For wide bandwidth CCD work the non-blazed grating is far superior to any blazed grating. For visual work the blazed grating gives a bit more light at 550 nm then our Compact grating does. However, blazed gratings generally cost several times what a non-blazed grating cost to manufacturer.

Q: Explain the pro/cons of blazed gratings in simple terms?

A: Blazed gratings are used by manufacturing companies to increase the gain of an optical sensor over a very small (generally less than 10 nm) portion of the spectrum. Blazed grating are extremely expensive. The initial mold could cost tens of thousands of dollars. Becuaes of this Amateurs and most companies that build equipment for Amateurs select a blazed grating that is already in use and they modify their intended use, sometimes to the detriment of their end user, to produce a product for the Amateur market.

A blazed grating concentrates both wings of a spectrum into a single order. The theoretical gain is two (2) but in practice the gain is closer to 80% at the blaze wavelength. The blaze wavelength is typically a narrow bandwidth of around 100 Angstroms wide (sometimes its only a few Angstroms wide). In this narrow band the gain in one of the first order spectrum wings approaches 180%. The gain outside of this blaze bandwidth drops off rapidly. Most blazed gratings have their bandwidth in the IR because the industrial demand is there rather than in the visual region.

Except for the zero order spectrum (which is the full color image of the spectrum target) each spectral order has two wings, one stretching out to each side of the zero order image. As the spectral intensity for each order is divided between two wings each wing (you can generally only see one wing at a time) only has 50% of the light intensity for that order. The lower the order (like the 1st order) the brighter it is and the shorter its length. The blue end of the spectrum always points to the zero order image and the other wing for that order (which is on the other side of the zero order image).

The first image below shows the zero order spectrum (this picture is the spectrum of the star Rigel) and on each side of the zero order image is seen its first order spectrum. Both wings of the first order spectrum are shown.

The next picture is Rigel moved over to one side of the picture so all of one wing of its first order spectrum can be seen in relation to the zero order image.

These two pictures were taken with a 5" Maksutov-Cassegrain using a Meade DSI I camera with a B1 spectrum filter.

Blazed gratings skew the detector response outside of the blaze bandwidth. The response of a non-blazed grating is much more linear than the response from a blazed grating which has a bell shaped response curve on either side of the blaze bandwidth. As blazed gratings usually have their bandpass in the IR region their response in the visual region is generally very non-linear falling off sharply in the green and blue portions of the spectrum.

With a blazed grating the response in the visible region is difference for the two wings (for any spectral order). This means that when you use a blazed grating the spectrum it produces is different depending on which wing you are using. In a non-blazed grating this is not the case and both wings have the same response. This means that with a non-blazed grating you can photograph either wing of a spectral order equally well. This is generally important for some optical orientations and when using slits.

The maximum possible gain within the bandwidth of a blazed grating is 2x which is 0.75 magnitude (remember a magnitude is a difference of 2.512x in light intensity). The actual gain achieved is generally around 1.8x which equals 0.65 magnitudes within the blaze bandwidth. The blaze bandpass is rarely specified in the Amateur market as the manufacturers don't want to tell their customers that they may not be selling them anything useful, just something expensive.

Q: I am confused by what Efficiency and Transmission are?

A: Diffraction Efficiency is what most manufacturers specify for their gratings. This is a good thing to know if you are designing a system but if you are using a system in the field it is of little use. Efficiency percentages numbers tend to be rather high compared to Absolute Transmission percentage numbers, which looks good to the casual consumer.

Diffraction Efficiency compares the amount of light that is diffracted by a grating, at a specific wavelength(s), to the amount of light that is diffracted by a spectral mirror for the same wavelength(s), for the same diffraction order and the same surface coating. Diffraction Efficiency doesn't have much practical meaning for end users, unless they want to tell everyone that their grating stands up well against an equivalent spectral mirror...

Absolute Transmission on the other hand measures the actual amount of light that gets diffracted through the grating for a specific wavelength(s). This is the amount of light that is available for your camera to use. Manufacturers either specify Relative Transmission or Absolute Transmission. Relative Transmission usually gives a larger percentage so its used a lot of the time to make low transmission gratings look good.

Absolute Transmission is easy to use. If a grating has an Absolute Transmission of 40% at 650 nm then that means that the light available from the grating (for use by a camera or any other light sensor) will be one (1) magnitude dimmer than the total light coming from the object (16% would be 2 magnitudes, 6.4% would be 3 magnitudes, 2.6% would be 4 magnitudes, etc). Remember that a difference of one magnitude is 2.5:1 and 1/2.5 = 40%. So if you want to photograph the Hydrogen Alpha line in Denebola's spectrum you will need a camera/telescope system that will be able to capture 3.1 magnitude objects (Denebola's magnitude is 2.1).

The Transmission Graphs for our Diffraction Gratings are Here if you would like to see what one looks like.

Q: Can I do Red Shift measurements with your DG Filters?

A: Yes. Because of the accuracy needed it is a lot more tedious to calibrate the images you take but we have a detailed tutorial that will guide you through all of the steps. We also have a Radial Velocity list of several thousand stars down to a 6th magnitude that you can use to practice on.

You should realize that in order to do Red Shift measurements you will need to perfect a technique where you can measure image distances to one part in a million; 1:1,000,000. If you limit your Red Shift measurements to radial velocities over 50 km/sec and stay only in the green portion of your spectrums you can reduce this accuracy down to 1:50,000. With care you can measure to one part in 5,000, 1:5,000, but going beyond this becomes very tedious and time consuming. While this can be done with our DG Filters it is not quick, easy or trivial.

NOTE: There are two things you should remember when you set out to do Red Shift measurements. They are:

1. Start with stars that have large Red Shifts (Rad Vel >90 km/sec) in the Radial Velocity Table located on the CD. Don't start out trying to see if you can measure a 6 km/sec star's shift. Working with star's whose Radial Velocity is > 90 km/sec will only require you to measure the line accuracy to a few Angstroms which is a lot easier than trying to measure down to a hundredth of an Angstrom.

2. Do Not start out using a 1000 l/mm DG Filter to do Red Shift measurements. To do Red Shift measurements using a 1000 l/mm grating you will need to capture the spectrum of two stars on the same photograph. The second star needs to be one that has a Red Shift at least ten times less than the star you are measuring. This is because you will have to use the lines in the less shifted star as a reference indicator so you can measure the shift in the lines of the star of interest. You do not have to do this using the 500 l/mm DG Filter as the zero order image is there and it is used as the reference point. With the 500 l/mm DG Filter Red Shift measurements are a lot easier.

Q: How can your DG Filters measure wavelengths to 0.01 Angstroms for Red Shift measurements when the CCD cameras, with the Grating filter, only have a resolution of 10 or more Angstroms per pixel?

A: We are trying to compare Apples to Oranges here. The 10, or so, Angstroms of resolution is resolution as predicted by the Grating Resolution Equation and or by the Pixel spacing. This is the Static Resolution of the camera/grating package that is required to separate two spectral lines of the same intensity - Apples.

The grating equation only predicts what the minimum spectral line resolution is. This means how close together spectral lines can be before they merge with others near them. These grating equations Do Not predict what the minimum limit of wavelength resolution is.

When you take an integrated picture of a spectrum with a camera/grating package that has a low static resolution, like 10 Angstroms, and you analyze the spacings on the picture using optical measurement techniques to determine the location of an unknown line relative to a reference point you are no longer talking about Static Resolution for you already have a spectral image. This optical measurement can be done with a great deal of precision which can be many times greater than the static resolution of the system - Oranges. This fine measurement of spectral line location is the Dynamic Resolution and it is different than the static resolution of the system.

When you focus a spectrum on a CCD chip the image of an individual line can spread across several different pixels. The software used to process the image uses special algorithms to determine where the details of the line' image is located based on the different percentages of illumination on the various adjacent pixels, and then determines the line location accordingly even if this location is in-between pixels.

Wavelength resolution in a spectrum's picture is a statistical function that is not related to the standard grating equations. What we are talking about here is how accurately we can determine the center of the spectral lines that we photograph, not how many lines that can be photographed or how close they can be together.

VSpec has developed a very special algorithm just to determine spectral line locations like this, and SBIG has a special software tool that is used to determine wavelength measurements to 0.01 A with their 600 l/mm grating. We have also developed a very special set of software tools along with a copywritten process that allows you to determine wavelength locations to fractions of an Angstrom with our 500 l/mm and 1000 l/mm grating units.

Most image processing software with sharpening functions can help do this to varying degrees with reasonable success. What you lose with CCD chips that have 20 A/pixel versus 1 A/pixel is total number of lines. The 20 A/pixel arrangement will exhibit fewer spectral lines than will the 1 A/pixel arrangement (assuming the same sensitivity), but the accuracy of the lines recorded by the 20 A/pixel chip can be as accurate, with respect to wavelength determination, as those recorded by the 1 A/pixel chip, and the measurements can be as accurate as 0.01 A with either of these arrangements.

With the 20 A/pixel arrangement you will onlt see the more pronounced spectral lines, while with a 1 A/pixel arrangement you will see those as well as fainter lines. What you lose with a 20 A/pixel arrangement is the wealth of dimmer spectral lines. But the wavelength of the lines you can see can be determined to sub Angstrom limits if the unit is calibrated sufficiently well.

SBIG has a really great top of the line spectrograph, the SGS that sells for $4900. It only works with their ST-7 and ST-8 cameras. Their unit is spected from 2.4 Angstrom/pixel (A/pixel) to 38 A/pixel depending on the configuration chosen. Yet, they make Red Shift measurements that require accuracies of better than 0.01 A/pixel "Measurement of Stellar proper motion to +/- 6 km/sec accuracy".

Using the SGS package that only has 2.4 Angstrom Static Resolution (their CCD camera and 600 l/mm grating) they are able to measure Red Shifts as low as 6 km/sec. This requires the spectral line measurements to be made to 0.01 Angstrom Dynamic Resolution. This is 240 x times finer Resolution than the Static Resolution of the camera and grating package they are using.

How do they do it? They don't lie, its all done with the software. Its really just Apples versus Oranges (Static versus Dynamic Resolution). Static and Dynamic Resolution is covered in more detail in Myth 4 on our Big Myths of Spectroscopy Page.

FYI: 10 A (Angstrom) = 1 nm (nanometer) and visible light goes from 400 nm to 700 nm.

Q: Explain to me how a camera with 10A/pixel and a grating resolution of 100A can see 1A differences?

A: Its really not all that hard to see. Lets look at a worse case. That would be using Meade's LPI camera. (Please do not take this worse case analysis to mean that Meade's LPI is not a good CCD camera - as it one of my Favorite CCD cameras bar none!) This camera has 640 pixels across the CCD chip's width. Using a B1 DG Filter this camera will produce spectral lines that are around 25 A thick (the actual width depends somewhat on the brightness of the star, the atmosphere and the focal length of the telescope). You might want to refer to another FAQ Question Q: Why do I get 10x the Resolution with your unit then is predicted by the Resolution Equation? if you would like to see how to measure spectral line thickness. Here, Mike's 1096 pixel camera was cut down to a chip close to the one used in the LPI camera using a 2x2 binning setting.

Now, if you measure Mikes photograph width and divide it by the distance between the lines we can come up with the effective width of the CCD chip. I used a mm ruler and just placed it on my monitor screen and found the photograph was 190 mm wide and the two Hydrogen lines were 32 mm apart. Dividing 190 by 32 = 5.9, then multiplying this by 1702 A (the wavelength difference between these two lines). This give us 10,042 A as the effective width of the CCD chip. Then we simply divide this last number by 640 to get the number of Angstroms spread across each pixel. This turns out to be 15.7 A/pixel (for the LPI and a B1 DG Filter the number is closer to 13 A but we are doing a worse case here).

Now, we have found that the grating won't produce spectral lines thinner than 25 A and the camera can't see any close together than 15.7 A. When I tell you this is a great match for doing Red Shifting much of the audience will walk out. However, they should have stayed for here is the answer, and its so simple.

As long as your spectral line thickness (from the grating) is 1.5x to 3x the pixel resolution you are fine. If your line resolution from your grating is much less than your pixel resolution you can't do Red Shift measurements at all. For those of you who think I have lost it please read the CAUTION in another FAQ Question, before you go on, Q: How do I increase the Resolution of your DG Filters?

The Meade LPI has 256 level of gray scale. More expensive CCD camera have 65,536 level of gray scale which is better. But as I stated at the beginning we will do a worse case analysis. As the spectral line from an LPI is seen to be larger than the pixel width of the CCD chip in Angstroms the spectral line will cover at least two and maybe three pixels at the same time. If the spectral line width were less than that of a pixel then it would only illuminate a single pixel. In this case it could wander all over the single pixel and there would be no way to measure its position any closer than the width of a single pixel which would negate doing any Red Shift measurements.

But in the real world the spectral line covers several pixels at the very same time. By making the assumption that there are no other spectral lines resolved that are very close to the spectral line of interest, the location of the center of that spectral line would be a simple computation of the image density of the two or three pixels that the line fell across. As the line would illuminate only one of the pixels fully the other illuminated pixel(s) would have a different gray scale value then the pixel that was fully illuminated. By using this technique it is possible to measure the center of the spectral line very precisely.

How precisely? Well again doing a worse case analysis we find the LPI's pixels are each 15.7 A wide. But it can produce gray scale levels that are 1/256 different. So the center of the spectral line can be measured to an accuracy of 15.7/256 = 0.06 A. In the Red Shift Radial Velocity that is on the CD that comes with a DG Filter a lot of stars have Red Shifts that are over 60 km/sec. This means that you only need 1 A measurement resolution to measure their Red Shift. This is 16x the positional resolution the LPI is capable of. The DSI with 65,536 gray levels can produce even finer resolutions (finer than you will be able to measure using readily available optical techniques).

Granted a spectral line width of 25 A will preclude capturing spectral line much closer than 70 A (of unequal intensities). But the ones you can see will be solid lines. A wide spectral line width should in no way should inhibit anyone from making Red Shift measurements.

Q: How do I make a good spectral line calibration source?

A: Go to Ace hardware and buy a 1/4" steel ball bearing ($1.50) and a can of flat black spray paint ($4.00). Then go to Home Depot and buy a small 6" battery operated fluorescent light and white 6" fluorescent tube ($15.00) to fit the light body (if it doesn't come with a tube) and four C cell alkaline batteries ($5.00) - the batteries are for the light. Find a clear location that you can see clearly from your telescope site.

Mount the steel ball with a piece of tape or modeling clay to something that is dark in color (black would be best - like an old cardboard packing box you painted with flat black spray paint) a few feet off of the ground (it might require two boxes sitting on top on each other to get to the height of the chair seat. Sit the fluorescent light on a chair seat no closer than 2' to the steel ball. Put a piece of cardboard in back of the lamp and paint both sides of the cardboard flat black so you don't see the glow from the tube in the scope and you won’t get any reflection from around the fluorescent tube.

The steel ball needs to be at least 10 times the focal length of the telescope in distance from the telescope site. Then wait for a dark night. Turn on the fluorescent light and illuminate the steel ball and focus the telescope on the steel ball. Hook up your camera and focus on the spectrum (not the ball) then take a few short pictures. Do not use any filters on your software other than to remove the Dark Frames. Make a new set of Dark Frames before you do this.

What you are doing is building an artificial star and illuminating it with a spectral line source instead of a conventional light. As long as the image of the light source, from the steel ball, is below the resolution of your scope this will work fine (the dimensions of the parts and set up I gave you above are good for scopes from 1" to 20" - Reflector or Refractor). If you need more information on building/using artificial stars read the book by Harold Richard Suiter - Star Testing Astronomical Telescopes, published by Willmann Bell.

This will only work if you are using a Type B or D DG Filter and have adjusted it so that it fits tight in its threads when the spectral image is horizontal. If you have done all of this you will have a good reference spectrum that you can use over and over again. If you are using a Type A DG Filter you will have to take a calibration picture each time you need a really accurate calibration.

The four main lines you will see are:

 Hg Line          Intensity
Center (A)          Units  

 4046.56             1800
 4358.33             4000
 5460.74             1100
 6149.50             1000

These images can be used to calibrate all of your spectral images.

You might want to check out the article about building a semi-permanent Calibration Source for easy use with your telescope.

Q: Why don't D1/D2 DG Filters work with a Toucam camera?

A: The reason the D3, visual DG Filter, is recommended for use with a Toucam is because the distance between the DG Filter, when screwed onto the 1.25" filter holder for the camera, is too far from the CCD chip to produce the image of the star AND its spectrum at the same time.

Using the D3 DG Filter with a Toucam will produce spectrums that has spectral lines. However, using a D1 DG Filter on a Toucam would be the same as using a B2 DG Filter on a DSI or LPI. With a B2 DG Filter you can't see BOTH the image of the star and its spectrum at the same time. This makes it extremely difficult to use with dim stars.

A stars spectrum is two magnitudes dimmer than the star itself. So if your Toucam camera can just barely photograph a 3rd magnitude star then the dimmest star that you can just get a spectrum of is a 1st magnitude star (+2 - +1 = +3).

As the Toucam camera can't be integrated its ability to capture dim spectrums is extremely limited. Using something like a D1 DG Filter on a Toucam will just multiply the problems as now the spectrum is so dispersed from the star's image that you can't see anything on the monitor but a dim spectrum (that you might not be able to actually see until AFTER the image has been processed).

The problem is that the D1 and D2 DG Filters have been designed to work with cameras that have a separation of around one centimeter between the DG Filter and the surface of the CCD chip. In the Toucam the distance is more like three centimeters. So to overcome this excessive separation you use a D3, instead of a D1, which will produce both images at the same time and make it easier to photograph the spectrum even if you can't see the spectrum on the monitor. If you can see the star's image and you move it over to the edge of the monitor its spectrum will be on the CCD chip (assuming that you have gone through the horizontal orientation procedure set out on the Tech Support Page.

While the D3 was specifically designed to be used with eyepieces it works well with the Toucam because both have a separation of several centimeters. The big draw back with the Toucam is that it can't be integrated and therefore won't capture very dim objects.

By the way you will need the 1.25" filter adapter for your Toucam in order to use any DG Filters with it.

Q: Can I take the same Great pictures with my Toucam as I have seen on your web site?

A: If you like the pictures on our web site, most were taken with a Meade LPI camera, and expect the same with your Toucam you will be disappointed. The Toucam won't integrate images but it will just allow stacking of images (which is not the same thing). As such the Toucam won't be able to photograph objects as dim as a LPI camera will. However, you can photograph bright stars with your Toucam. The spectrum of a star is 2 magnitudes dimmer than the star itself. So if you can just barely photograph a 5rd magnitude star then the dimmest spectrum you will be able to photograph will be from a 3nd magnitude star (+5 - +2 = +3).

A LPI camera will capture the spectrum of a 7th magnitude star with an 8" telescope. It can do this because Meade’s LPI software and hardware together can integrate the image where the Toucam's hardware board isn't set up for integrating the images it sees.

If you have the 1.25" filter adapter for your Toucam then you can use our D3 DG Filter with your camera. This DG Filter was designed for strictly visual use with eyepieces so it produces a really vivid set of spectrums for any object. The D1 DG Filter has been used with a Toucam before but its a lot harder to use as the distance between the Toucam's CCD chip and the DG Filter is so great. To be effective the separation distance needs to be less than a centimeter and on the Toucam its around three centimeters. If the separation is too long then the image of the star and its spectrum won’t fit on the CCD chip at the same time.

I have seen used LPI cameras on Astromart and ebay for under $100 USD if you would like to upgrade to a camera that will allow you to take much dimmer pictures than your Toucam is capable of doing.