Bahtinov frustration.

[frugal]

After my comments yesterday about trying to rush into things I decided to take a step back and approach things in a more deliberate manner. My main problem so far appears to be focus, so I spent this afternoon making up a variety of Bahtinov and Y masks. For the Bahtinov masks I used someones clever suggestion of printing them onto OHP transparancies rather than trying to cut out lots of flimsy bars. For the Y-masks I cannibalised an empty plastic box of Heros from Christmas.

In order to make sure that I had a full selection of masks to play with when it cleared up I printed off 3 bahtinov masks using the Astrojargon generator, one with a Bahtinov Factor of 150 (the standard), one at 50 (recommended for smaller apertures) and one at 25 just to see if going any smaller made any difference. For the Y-masks I made them at about 5mm, 8mm, 13mm and 18mm.

This evening there was a clear spell, so I took the lot outside and tried each of the different masks to see what effect they had. The problem was that I didn't really see diffraction spikes on any of them at any exposure length that the Bahtinov software in BYE was able to really latch onto.

All of the imaging was done with a Canon 60D with a 18-135mm f3.5-5.6 EFS lens. The Snapshots in Frame and Focus were all taken at ISO12800.

For the Bahtinov masks, the two angled spikes seemed to be swamped by secondary spikes next to the main one. For the Y masks, the spikes were not long enough to avoid getting swallowed up in the glare from the star. The image below shows a snapshot from each exposure as it was shown in the BYE Frame and Focus zoom window.

The star I was focussing on was the bright star in this image I took at the same time (to get a real idea of the focussing). The star is the middle star from the W of Cassiopeia: 

Do DSLR lenses just not get enough light in to make use of one of these masks? Am I trying to use a star that is too bright? Am I using the wrong ISO? Should I just go back to using FWHM?

[baggywrinkle]

I made a mask but used the 3 equalateral triangles (Scheiner mask) holes. This is less complicated and works well as you bring together 3 images to one point of focus.

[swag72]

A couple of thoughts from me - I use a Bahtinov mask for focus with a CCD camera, it's much more sensitive than a DSLR. I use the brightest star in the sky , currently Capella in Auriga and then use 5 second exposures in order to combat seeing and let the image settle down a little.

Did you experiment with the ISO level? Could you perhaps try a lower ISO for the moment and see if you can get the correct pattern using different exposure times. Maybe also try a brighter star?

The other alternative of course is that he mask just isn't the right size at all. I had a small mask for a 66mm telescope, the slits were tiny. I dread to think how small they should be for your lens - Perhaps they are not thin enough in order to get the normal pattern. Can you post a picture of the mask you used?

[wxsatuser]

Looks like the Y mask was near.

I use exposures of 10secs at ISO 400 with my 80ED.

[ronin]

so I spent this afternoon making up a variety of Bahtinov and Y masks. For the Bahtinov masks I used someones clever suggestion of printing them onto OHP transparancies rather than trying to cut out lots of flimsy bars.

First problem I can see is that there is just one Bahtinov mask to a scope. The bar size and spacing is determined by the scope diameter and focal length. May find that the number of bars present also depends on the scope parameters. You need to locate a mask generator and feed the scope parameters into one for it to determine the mask that you need. You indicate 3 Bahtinov masks made so at least 2 are incorrect and unless the scope details match the third is very likely to be as well.

Check any printer as even if you had the correct mask on the screen I think that often a printer will resize things (slightly) to fit the size of the medium that it is printing onto. So it could start off correct on the PC screen and incorrect on the transparancy.

[frugal]

I was using ɣ-Cassiopeia which has a magnitude of 2.15 so it should have been bright enough.

These are the masks that I was using:

I used the correct values for aperture and focal length for the lens that I was using (135mm focal length, 95mm diameter (to go over the lens hood)). I have confirmed that changing the aperture value does not affect the thickness of the generated slits or bars, just the size of the circle. The only thing I changed between masks was the Bahtinov Factor. The standard is 150 (top right), but the literature suggests that you can drop it to 50 for small apertures in order to get the third order diffraction spikes (middle mask). The last mask was done with the same aperture but a Bahtinov factor of 25 as I had space on the sheet of acetate for one more.

I have experimented with different ISO levels, but that did not get me a more pronounced spike, all it did was clear up some of the background noise. Here is a crop of the same star from a sample shot of 120s at ISO800:

I am certainly not getting anything like 

I just seem to be getting the three horizontal spikes in the middle. Is it just that I am not getting enough light into the lens?

[wxsatuser]

Thinking aloud.........these lenses have many elements.......does that cause a problem???

[frugal]

Thinking aloud.........these lenses have many elements.......does that cause a problem???

I have no idea. I have seen a number of people swear by Bahtinov masks, even on DLSR lenses, and I have never seen anyone claim that they do not work on DSLRs, so I have to assume it is something I am doing.

[wxsatuser]

Thinking aloud again.

You printed on transparency........is it that???

[swag72]

The masks that you buy are made of thick black plastic material. These need to be totally black and unable to let any light in apart from through the slits. Is this the case?

Also, the bought ones sit directly against the lens hood or dew shield with no ripples etc. this may be important?

[m37]

Try a hartmann mask, way easier for getting you started. I find it easier to see three triangles converging than to judge by eye whether diffraction spikes are in the exact centre. Horses for courses, of course, but give it a whirl. Mine cost about 70p to make and it is just awesome.

There are generators online that do the calculations for you although I just cut three triangles with a ruler, compasses and craft knife and it worked an absolute treat.

I use it in backyard eos and when the triangles are converged I lower the exposure so it is a bit dimmer to better define the individual triangles then fine tune it with the FWHM analyser in BYEOS.

Also cut it from black plastic or card, transparencies are not really solid enough to be accurate.

Cheers And good luck

Chris

[davew]

Some years ago I did a fair amount of experimenting with camera lenses and masks ( Some would call it messing around instead of getting anything done  ) and I still use the results with what I regard as success. Y masks do work with both DSLRs and cooled cameras through narrowband filters. I can also say I use camera lenses almost exclusively and only use my refractor when I get a rush of blood to the head !

I use a Y mask with a width of around 2 to 3mm. This works with all focal lengths from 50mm to 630mm that I know of. There is a slight problem when you get below 100mm though because you'll have to zoom in so far that the bars become indistinct on the DSLR rear screen. I always use a netbook now to determine focus because it shows up so much better but still a struggle below 100mm.

What Sara says is correct re a bright star. No way would I use ɣ-Cassiopeia. I know people say you should chose a star in the area of where you're going to shoot but if you think about it, after a few hours the stars will have risen or fallen a long way. Only if you need to re focus at regular intervals do you need to focus in the local area.

I think that only one mask can be used per scope is a myth. Only the spikes change not the focus point. ( No offence meant to anyone )

The bars get more distinct the closer you are to focus and using a zoom lens may flair the star very slightly making them more difficult to see.

Use the best ISO to be able to make a judgment. Depends on your conditions.

I use around 4 to 5 seconds with a DSLR and around 8 seconds through a NB filter.

I attached two images showing the best and worst of Y mask focusing. The worst was still usable and so much better than spending half the night chasing FWHM, which is what seems to happen with very wide lenses. { Edit. It won't show the bitmap so click on and open rather than save }

Just spotted the Hartmann mask post. Nice idea but I couldn't get it to work due to the fact the mask had to be so tiny. Do try it though.

Dave.

Focus 105.bmp

[martin_h]

I bet its because you printed them on to transparent plastic...not the best optical material i would have thought...........try a clear unprinted bit over the lens and look at a star I bet that wont focus and will be blured.

[ollypenrice]

I agree with Martin. Shooting through a transparency will do two things, scalter the light like crazy and (ouch!!) change the focal length of the system. Now that really would be a waste of time!

Have you tried measuring the FWHM in a camera control programme? I used to do this when using camera lenses.

If using bright stars for focusing they should be at the same elevation as your target. The atmosphere affects focus.

Olly

[frugal]

I agree with Martin. Shooting through a transparency will do two things, scalter the light like crazy and (ouch!!) change the focal length of the system. Now that really would be a waste of time!

Have you tried measuring the FWHM in a camera control programme? I used to do this when using camera lenses.

If using bright stars for focusing they should be at the same elevation as your target. The atmosphere affects focus.

Olly

I used the FWHM method first to ensure that the shots were sufficiently focused that I would get a reasonable diffraction spike. I got the number down to 3.2 which is about as good as I have ever got it.

I think the use of the OHP transparency has caused a lot of my problems. I will have to stick with the FWHM approach for now as the Y-mask does not seem to have got me anywhere and my skills with a craft knife are not up to cutting 80 slits in a piece of plastic thick enough not to flex without making a mess of it ;( 

[davew]

I agree with Martin. Shooting through a transparency will do two things, scalter the light like crazy and (ouch!!) change the focal length of the system. Now that really would be a waste of time!

Have you tried measuring the FWHM in a camera control programme? I used to do this when using camera lenses.

If using bright stars for focusing they should be at the same elevation as your target. The atmosphere affects focus.

Olly

I sort of agree about atmosphere but the objects we image don't stay at the same elevation. I suppose it depends on how many times you want to re focus in a night and at what focal ratio you're imaging at.

Dave

Edit - bod speeling

[Dave_D]

 The atmosphere affects focus.

Olly

atmosphere at my location is uniformly garbage  i just focus on capella which is usually overhead and thats it...

[Zakalwe]

I bet its because you printed them on to transparent plastic...not the best optical material i would have thought...........try a clear unprinted bit over the lens and look at a star I bet that wont focus and will be blured.

^This^

Think about it. Put an EP into your scope and then place a piece of transparency over it (especially one thats been coated to allow it to absorb liquid ink, as you have used). Do you think that the image at the EP will affected or not? Will you be even able to focus with a piece of plastic in front of it?

Lots of people use "proper" Bhatinov masks (myself included). And they use them with DSLRs and CCDs (again, me included). The only thing that you re doing differently is using one printed onto a piece of plastic. Remove that variable and then report back.

[steppenwolf]

I would urge you to persevere with the Bahtinov mask and not fiddle with the Hartmann in its various forms. A Hartmann starts off working really well but just as you reach the critical focus point, the waters get muddied - not so the Bahtinov mask which clearly shows the transition from intra to extra focus.

I think there are several factors in play here:-

1. the wrong star for certain  - you need Vega, Deneb, Capella, Arcturus or Betelguese as suitable candidates.

2. the transparency will not give a proper diffraction, you need a sharp cut edge to get a crisp diffraction spike

3. the mask must be flat

I have found that even inappropriately proportioned Bahtinov masks work well. I'd go for the top left mask but cut out of laser printed thick paper covered in transparent cover film (applied before cutting!!).

[Zakalwe]

A small Bhatinov mask only costs around £14 so I personally can't see any advantage to messing about making one from scratch. Just get the nearest one to size for your DSLR lens...as long as its centred (there or thereabouts) on the lens then it will be OK. They don't have to be made exactly for each lens

http://www.firstlightoptics.com/bahtinov-focus-masks/starsharp-bahtinov-focus-masks.html

[m37]

Fair enough, think I will get a Bahtinov for my new scope and try them both in a shoot-out. Be interesting to see if there is any indication of extra precision with the Bahtinov when the Hartmann says I'm in focus.

[dph1nm]

If using bright stars for focusing they should be at the same elevation as your target. The atmosphere affects focus.

The atmosphere affects the seeing but I don't see how it can affect the focus of the telescope, which is a local phenomenon? Unless you have a very heavy lens/mirror which distorts under its own weight - then observing at low elevation can change the focus, although it is nothing to do with the atmosphere.

NigelM

[ollypenrice]

The atmosphere affects the seeing but I don't see how it can affect the focus of the telescope, which is a local phenomenon? Unless you have a very heavy lens/mirror which distorts under its own weight - then observing at low elevation can change the focus, although it is nothing to do with the atmosphere.

NigelM

Yes, I thnk you might be right. Rays coming into the scope will be parallel so really the biggest differential atmospheric effect will be in their apparent position. This certainly does change a little, which is why naively perfect sidereal tracking won't do. Unguided mounts have to have an atmospheric model built in to their motor control. Maybe focus won't be affected, as you say. Time for some tests!

Olly

[frugal]

Ok, mea-culpa. It looks like my abilities to cut straight lines out of a bit of plastic are far worse than I had imagined. I gave in and bought a StarWave Bahtinov mask (as the Starwave ones advertise themselves based on aperture, not focal length). Taking a 30 second shot at 100mm was okay, but even a 10 second shot of Capella at ISO1600 gave me huge spikes. 

Of course now I have it focussed, it has clouded over completely