Scheiner and Hartmann Masks
Bringing a telescope to sharp focus at high magnification can be a real trial. An experienced visual observer will be constantly manipulating the focus. This is not an option with CCD imaging since you must set the system at optimum focus and then collect your frames. A focus mask is one way around this problem. As the name implies, we are talking about a mask that blocks most of the light that would enter your telescope through its aperture. The mask does, however, have two or more openings cut into it. If the mask has two cut-outs it is known as a Scheiner Mask, while three or more makes it a Hartmann Mask. Shown below is a Hartmann Mask with three triangular openings:
Now, a few basic "design issues" with respect to Scheiner/Hartmann (S/H) Masks:
Diameter of Mask - the minimum mask diameter is equal to the clear aperture of your telescope. The one above is 127mm to match my Orion 127mm MCT. If you are going to make the mask into a lens-cap-like arrangement, the mask may be somewhat larger.
Color of the Mask - most masks are cut from posterboard or other relatively thin cardboard stock. The mask should be flat black on the surface that faces into the telescope, but the color of the other side is of no consequence. I use two sided poster board - matte black on one side and white on the other. I use the white side for all my plotting of the holes!
Shape of Openings - The simplest S/H Masks have circular openings. The reasons for my use of triangles will be evident shortly.
Number of Openings - the mask must have at least two openings but there is little benefit to having more than four. Three is very common.
Arrangement and Orientation of Openings - In theory, the openings could be anywhere on the mask, but a symmetrical arrangement results in a pattern that is easier to use. A mask with two openings would thus have them spaced at 180o, a three opening mask (above) would space them at 120o, and a four opening mast would use a 90o spacing.
Size of Openings - The larger the openings, the easier the mask is to use, subject to the following constraints:
The outer edge of the openings must not extend beyond the limits of your clear aperture. I always offset inward by a few millimeters (as shown above), just to be sure.
The openings should not impinge on the central obstruction if your telescope has one.
The openings should not impinge on each other.
The mask will let you do a very precise job setting the focus, but the object you are viewing must be a point-source. That means, for all practical purposes, a single bright star. Other comparatively faint stars in the field should not be a problem. However, avoid bright double stars, DSOs, planets, or the moon. OK, lets pick our a nice bright star - in this case, Spica, a very respectable first magnitude star! Get the star onto the imaging chip and do a rough focus. Now drop the mask (black-side down) over the aperture and here is what you will see:
If the star is considerably out of focus you will be something like [A] - three faint triangles. The further out-of-focus the optics may be, the fainter the images and the further apart they will be. Begin to adjust focus - if the triangles get further apart and fainter, you are going in the wrong direction. As you start to approach focus, you would see something like [B] - the triangles will start to converge and they will get brighter as they do so. You have hit the precise optical focus when all three sub-images converge, as in [C]. Here is where the business of using triangular openings comes in! A conventional S/H Mask has round apertures and you are watching for the overlap of the thee circular spots of light. Under moderate to poor seeing, the three circles of light will writhe around, twisting and moving, making it very difficult to say exactly when they overlap. With the triangles, a marvelous thing happens as the come together - the development of six very distinct diffraction spikes! At the precise point of focus:
- The spikes are arranged in precise symmetry
- The spikes reach maximum elongation
- The merged images peak in brightness
- The image achieves almost a crystalline appearance
Under poor-seeing, the central body may be unstable, but the diffraction spikes will be as obvious as ever! In short, its pretty easy to know when you hit focus. It you run past the focal point, the image starts to dim, the diffraction spikes disappear, and you will see the triangles start to diverge. That's the time to back up until you hit the very distinctive image in [C
The following directions do not assume the use of a GoTo mount, although a good one can make life a bit simpler. I do assume you have a targeting eyepiece of some sort to assist you in aligning the scope to put the target on the imaging chip. The ideal eyepiece will have a field of view significantly wider than the camera but which will still allow you to place the target with precision. A flip-mirror is a useful accessory, but ONLY if it is properly aligned and the flip-mirror, targeting lens, and camera hold their mechanical alignment. If not, you are better off using a simple diagonal, swapping the targeting lens and camera as required. I am also assuming that you are using a Barlow or PowerMate. If so, everything can be done with the accessory lens in place if you are careful and methodical.
- If using a GoTo mount, use your target as the alignment object. Otherwise, manually slew and align to the target. Either way, you should end up with the target precisely framed in the FOV of the targeting eyepiece.
- Now, with the greatest accuracy you can muster, adjust the crosshairs on your finder so they are absolutely centered on the target. For any of this to work, you must have a real finderscope, a red-dot unit will not cut it. I have a nice right-angle 9x50 from Orion that is well-suited to the task.
- Now, either GoTo or manually slew to your focus star and carefully center the star in the FOV of the targeting eyepiece.
- Now swap the lens for the camera and center the star on the camera chip and do a rough focus. Put the mask over the aperture and focus as indicated earlier. FROM THIS POINT ON, DO NOT ALTER THE FOCUS ADJUSTMENT! If your telescope lets you lock the focus, do so. Now remove the mask!
- Using the GoTo function (or manually), slew back to the target. Carefully adjust the alignment to place the target directly under the crosshairs and it should be somewhere on the camera chip. If not, swap the target eyepiece for the camera, center up the target, swap the camera back in and capture your image sequence(s).
As you gain experience, you will find that you start saving a lot of time:
- The imaging Barlow or PowerMate stays on the scope - no need to swap it in and out
- You will do less swapping of the target lens and camera.
- No matter what the seeing, all your image AVIs will be shot with optimal focus
If temperatures are changing as you work, you may want to revisit your focus star for a check now and again. Most of the time I can move back and forth without having to resort to the target eyepiece, which would, of course, be required if you have to slew manually. If you have a GoTo mount, at the very least use the target for a one-star alignment. That way the mount knows the target's location very accurately and any return from your focus start will certainly place the target within the FOV of the target eyepiece. A step up from this is to do a two-star alignment using the target and the focus star. That greatly facilitates moving back and forth between the two objects if you want to check focus over time. If you them move on to another target, you should repeat the alignment so that returns to both the target and focus star are done with the highest possible precision. As long as the finderscope is not moved, you should not have to repeat the step where you center the crosshairs on the target. If your finderscope is taken off the OTA between observing sessions, doing the crosshair calibration will be a must at the start of each operating session!
While bright focus stars are helpful, there is considerable latitude. As noted, I have used Denebola extensively for the past week or so (magnitude 2.1) with no problems. I have practiced with stars down to magnitude 3 and it seems to work. The main factor appears to be available camera gain. If you use stars around the first magnitude, you can often turn down the gain a bit so that any fainter starts in the field will not complicate your focus determination.
Does It Work
Well, I would hardly write it up if it didn't! Here is a Saturn image made using the triangular Hartmann Mask with Denebola as the focus star. I image was captured with a 2.5X PowerMate (f/30) with a 4,000 frame AVI aligned, optimized, stacked, and processed in RegiStax5:
Compared to my earlier efforts there are three significant improvements:
- Titan - I have always had to dig down in the noise or resort to very high gain settings to locate any of the moons - even Titan. This time, Titan could be located (dark but present) in the original image. It required minimal enhancement and looks like a disk!
- Cloud Bands - the cloud bands are very subtle in the case of small-aperture monochrome images, but this example is better than anything I have taken to date.
- Rings - the rings are hard to catch these days as they are gradually approaching edge-on and thus are moderately dark, However, at either end you can see slight darkening, representing the Cassini Division. Paradoxically, I have this on earlier low-powered (f/12) images (easier to focus), but never, until now, at high-power (f/30).
Needless to say, I am interested in using the focus mask both for Lunar imaging as well as Jupiter in a few weeks!
One of the hottest focus masks in play today is the Bahtinov Mask, invented by the Pavel Bahtinov, a Russian astronomer: Take one look, and it is obvious we are dealing with something a bit different from a basic Hartmann or Scheiner Mask:
The purpose of the slits running in three different directions is to generate three distinct diffraction spikes:
Two of the spikes cross at a shallow angle with orientation depending on how the mask is positioned. As you focus, a third diffraction spike moves across the two-spike cross (up and down with this orientation) as focus is adjusted. Optimum focus is obstained when the third spike precisely bisects the other two, as shown above. The pattern is said to be very resistant to poor- seeing and coming to the proper focus is very easy.
Although there are on-line Bahtinov generators, many will prefer to purchase a mask. One source is Spike-A , who provide black-anodized masks ranging from 102mm ($70) to 407mm ($99). They can also produce custom masks for any telecope. While the aluminum masks are very durable, they are also pricey, so you may wish to look at Lumensa, who produces a line of laser-cut stiff plastic masks in sizes ranging from 80mm ($20) to 279mm ($50). They can produce custom masks in this size range at costs comparable to similar stock sizes.
I purchased a 130mm Lumensa mask for my 127mm MCT, which is actualy the one pictured above. Black nylon screws permit the mask to simply be hung over the objective/corector lens/open aperture of your telescope. The first night out with the new mask I opted for Spica as a focus star and here is the pattern I saw:
Where, we might ask, are all the sharp diffraction spikes? At first I was ready to blame really poor seeing, but the pattern basically looks this way all the time! Fortunately, it still works just fine for adjusting focus, but it doesn't look like anyone else's! Some experimentation will be required!
OK, why does my mask, while perfectly functional, produce such a crummy diffraction pattern? Well, it seems there are two reasons:
Note that, despite these pattern issues, the Bahtinov is a quick and accurate focus accessory, its just not pretty!
So - Hartmann or Bahtinov?
Both masks do the job, so which one has the edge? One evening in late June, I was waiting at 4AM to see if Jupiter would move out from behind my planet-eating maple tree before the sun started to wash-out the sky. It was getting obvious that the sun would probably win the race, so I decided to run a little test on Altair under conditions of mediocre seeing.
I kept going back and forth, using one mask to focus and then cross-checking the results with the other. As well as I could tell, it was pretty much a dead heat. The central "body" of each figure was strongly effected by the seeing, but when the third "spike" was centered with the Bahtinov or the diffraction spikes blossomed with the triangular Hartmann, the star was always in perfect focus. Just for fun, never having imaged a star, I took Altair's picture and was rewarded with a tight pin-point of light. Both masks work eaqually well* and I consider them equally convenient. In practical terms, I tend to use the Bahtinov a bit more, since it is plastic and I can leave it out in the dew with no ill-effects, but functionally, I have not seen an advantage for either one.
* This does not include the traditional Scheiner or Hartmann mask with the circular apertures. Under marginal seeing it is impossible to get an unambiguous focus because the merged images will writhe around in a most disturbing way! The Bahtinov will win hands-down when tested against a traditional Hartmann or Scheiner Mask.
Ralph E. Taggart email@example.com