17 replies to this topic

[Mike I. Jones]

Chris Lord has developed a very practical improvement on the Bahtinov mask that is much easier to make, yet produces the same general through-focus diffraction pattern behavior. I show a ZEMAX simulation here of the "Y-Mask" developed by Chris. A 6" f/8 objective with no central obstruction is shown here. The defocus is approximately equal to the Rayleigh 1/4 defocus shift given by

4(wavelength)(focal ratio^2)

At f/8 and 0.55um the 1/4 wave depth of focus is 0.00554", so the defocus simulated here is less than 1/4 wave.

A central obstruction has no real effect on the shifting of the central spike relative to the X spikes.

This mask is far simpler to make than all the slats a Bahtinov requires, and also gives a brighter image at focus, by about one f-stop.

The Y-Mask is simply the Y-shaped obstruction to the aperture. The blue in the pupil is where light is transmitted.

The Y-Mask will be the first focusing mask I'll make. There is no reason to spend the time cutting out a complicated Bahtinov mask with this innovation.

Mike

Attached Thumbnails

• 

[Mike I. Jones]

Here's a direct comparison between Chris Lord's new Y-Mask and the Bahtinov mask for the same 6" f/8 system at 0.55um and ±0.005" defocus, plotted at the same scale. The Y-Mask not only gives a brighter image, it is also a much cleaner one, as it does not generate all the grating sidelobes the Bahtinov makes (similar to a Ronchi grating). It's definitely a clever improvement over the Bahtinov.
Mike

Attached Thumbnails

• 

[Dennis Sakva]

I would be careful calling it a better alternative (even though it can be) before real world testing. In Case of Chris's mask star brightness may seriously overpower the closest spikes, while further away spikes are too dim to see.

[Luigi]

It's pretty the same thing, just fewer edges, so the diffraction spikes will be proportionally dimmer. I guess it'll be OK if you can still see the spikes on stars in or near the intended FOV. If not, add a few more parallel lines to each leg and arm of your Y to make the spikes brighter.

[Luke S]

This looks similar to what I have used for a long time, I just put a piece of masking tape over the front of the telescope, centered on the aperture. The dew shield or lens cell is high enough to prevent the tape from touching the optics, makes a nice diffraction pattern that is easy to focus with. (It is in focus when there is only one diffraction spike on either side, rather than two parallel diffraction spikes.)

[wh48gs]

I like the Y-mask better too - why complicate it if it can be simple? Spike brightness can be easily controlled by choosing appropriate opening width, and the pattern is more precise, due to the spikes intersecting at the center. I have no doubt which one I'd use.

Vla

[Mike I. Jones]

I will repeat the post I made over in the CCD Forum here. Dennis makes a valid point about the Bahtinov having higher diffraction energy in the sidelobes relative to the central peak. It is easy to see in the ZEMAX analysis that the Bahtinov sidelobes are brighter than those produced by the Y-Mask, at least for the mask geometries I simulated.

However, the Y-Mask blocks only a few percent of light to focus in contrast to the Bahtinov, which blocks at least 50% of the pupil. The Y-Mask sidelobes are fainter, but the overall image is brighter by almost an f-stop. The Y-mask sidelobes are also narrower than the Bahtinov, allowing more precise location of focus to be determined, if they can be imaged well.

This analytical work is a good example of real-world results with real imaging equipment, stars of different spectra and brightness, and degrees of turbulence present not necessarily agreeing with the very precise computer calculations under pristine conditions. It is best validated with field testing of Bahtinov and Y-Masks with several different line width and spacing parameters, using direct empirical photographic and visual comparisons. A particular Bahtinov mask design might be superior to a given Y-Mask design in locating best focus when using a particular star under some range of seeing conditions. The Y-mask might do a better job under different conditions and line widths. The two mask configurations could also be a wash, depending on the mask geometries used, the brightness and spectrum of the star being used, and the degree of turbulence present.

Mike

[daniel_h]

very interestng Mike - do you have a template for us to try? what angle is the Y cut at?
thanks daniel

[Mike I. Jones]

The line widths and angles are all things to experiment with at this point using whatever CCD/DSLR camera you have. I ran the example at plus or minus 20 degrees slope with 0.3" wide opaque strips across the 6" aperture. The Y-Mask sidelobes are not as bright as the Bahtinov's, but as Luigi points out, more strips parallel to each Y-Mask strip can be added to augment the diffraction sidelobe brightness. So in a way the mask kind of migrates back toward being a Bahtinov, but perhaps with not as many slats.

[Dennis Sakva]

Just a clarification,
Original Bahtinov's idea was to cut slots depending on focal length. The longer the focal length is the fewer slots you need. The formula was f/s=150...200, or 3f/s=150...200 where f - focal length, s - grid period.
So in case of 150mm/1200mm scope the "recommended" period would be either 1200/150=8mm (4mm slots) or 3600/200=18 (9mm slots). Which would require 18 cuts in case of 4 mm slots (not very technological) or merely 8 cuts in case of 9mm slots.
Mike, would you please test how mask with 4mm and 9mm cuts will work in case of your scope? If you have time (and desire) for that of course.

Later people decided that the mask works good even if the original formula was often ignored.

[Mike I. Jones]

Sure, glad to look at the 4mm and 9mm slots, but it will be sometime next week before I get some time to. Lots going on at work and with weekend summer company at home.

Mike

[Dennis Sakva]

Thanks a lot, Mike!
Your simulations look VERY realistic!

[amateur]

Very similar to what I did: simplified Bahtinov mask, just place 3 wires in front of the scope.
I found that in practice the Bahtinov mask is superior for visual, since the bright diffraction lobes from the grating are easier to see than the broad lines from a single wire.

Here are some simulated diffraction patterns for different masks.

Attached Thumbnails

• 

[groz]

The Y-Mask not only gives a brighter image, it is also a much cleaner one, as it does not generate all the grating sidelobes the Bahtinov makes (similar to a Ronchi grating).

Have you actually _used_ both masks to get precise focus on a telescope ?

Hint, we use the b-mask here all the time. Those sidelobes you seem to dislike, are the way to get the focus PRECISE. first you rough it in using the main spikes, and, when you get that really close, the crispness of the side lobes 'finishes' the job of focussing.

[Mike I. Jones]

In the months that have passed since I first analyzed the Y-mask and Bahtinov, and after doing some purely subjective experimenting with both the Y and B masks, I think I have come to like the Bahtinov a little better for focusing on fainter stars, where the brighter diffraction sidelobes definitely home in on focus, and the Y for brighter stars. They each have advantages and disadvantages, and I'm pretty much _good_ with either.
Mike

[Mert]

Like you state Mike, for me it has been like going
to the optometrist, the Bahtinov really helped me
getting on the road again! :bow:2Pavel :-))
I'd like to be able to distinguish the difference
each mask offers but for me IMHO it is
useless to try to improve upon what for me has
proven to be the solution

[Janez]

Has anybody tested and compared Bahtinov mask and Y mask for visual use. The eye works more on level of contrast (brightness difference) than on brightness alone. It seems that Bahtinov maks gives better results, since the difference in brightness of star and difractions is smaller, which gives better contrast. Y mask is simpler to make, however the larger difference in brightness of star and difractions makes it probably less usefull for visual use. So what are the practical findings?

[avarakin]

This mask is a very interesting idea!
How about going one step further and making spider vanes of reflector using the same shape?
This would be a built in Bahtinov mask!
Obviously astro pictures with reflectors like this would have a different distinct diffraction pattern from what we normally see, but I think it would look much more interesting than just boring perpendicular spikes we see in regular reflectors.

Alex