UHC filter on an unmodded DSLR

[MikeWilson]

Do-able or not?

I really want to get more defined views of nebulosity from my rather heavily light polluted garden, but I don't particularly wish to sacrifice my Canon EOS 1000D DSLR to do it.

I had bought the DSLR in order to sacrifice it to the gods of Astrophotography, but I've really taken to it for daytime photography and although I know it can be temporarily unmodified with a clip-on filter, these such filters won't work with the daytime EF-S lenses.

I don't think I'm ready to consider CCD yet, but I want to get more detail. I have a 2" focusser. What should I do? UHC? OIII or none of the above on an unmodded DSLR?

[MikeWilson]

Shameless bump - anyone know the answer to this?

[SteveL]

Although you`ll get a lot of OIII signal through a UHC filter, I think an unmoddified Canon will only allow ~20-25% of Hydrogen Alpha signal through to the chip due to its own over zealous IR blocking filter.

Here is quite a telling image from 12dstring's website (350d mod page)

[Ben Ritchie]

I think UHC is H-beta rather than H-alpha, so will work with an unmodified DSLR - both wavelengths are in the blue/green. The standard filter will also block much of the IR-leak from a visual filter too, although there may be some overlap in the far red.

It won't match a modified DSLR as the H-beta signal is always much weaker than H-alpha, but I guess it should cut through light pollution pretty well. The image will be a rather sickly green, but Photoshop can deal with that.

[SteveL]

I did just have to go check

UHC-E (visual I think)

UHC clip for EOS

[Ben Ritchie]

I stand corrected

The net effect should be the same though, you'll get some line emission ([O III] and H-beta) but unless you modify the DSLR you're missing most of H-alpha, which is the strongest signal by far.

[MikeWilson]

Ok,

I can pick up another 1000D DSLR, refurbished from eBay for around £220 and remove the filter myself. Combined with a UHC filter, would I still be able to extract colour data given that I'm simultaneously recording multiple narrowband frequencies?

Ideally I'd like seperate Ha, Hb, O-III filters, but that's going to get expensive for the 2" flavours.

If I get just one filter for my modified (or unmodified) DSLR now, which one should I get?

edit: I'm choosing to improve the capture system rather than adding a guidescope & autoguider. I can guide quite accurately for 60 seconds, slightly less so at 120 seconds to a maximum of 145 seconds. Presumably I would benefit more from a more responsive camera than I would from being able to take longer subs...... (either route would cost about the same).

[Ben Ritchie]

As far as I can see there's no sense in buying a H-beta filter for a modified DSLR, in practice it's just the same as a very low transmission H-alpha filter - you're seeing the same gas, and the emissivity of H-beta is considerably lower.

[MikeWilson]

Am I right in saying that (please can you check my learned assumptions );

1. A UHC filter on an unmodified DSLR will cut the light pollution more or less completely with the Hb, O-III coming through as strong as they possibly can, but the more common Ha signal (as strong as O-III) will hardly register at all (around 10%) due to the inbuilt IR cut filter.

2. If a DSLR has the UV and IR cut filter removed and one attempts to image with a UHC filter, you will get halos around bright stars caused by IR. A separate UV/IR cut filter (a less zealous/more accurate one!) is required in addition to the UHC filter.

2b. A separate UV/IR filter is NOT required when one uses a Ha, Hb, O-III or S-II filter as these filters are optimised for those specific frequencies and will already cut the unwanted IR.

3. A UHC filter is unable to capture composite colour image of objects as the camera sensor will register Hb and O-III as greenish and the Ha as redish with a particularly weak blue channel.

Am I right?

And if so, would I be best going the modified DSLR route + 2" UHC filter + 2" UV/IR filter or should I instead concentrate on getting a guiding system up and running to correct star trails and allow me to take longer exposures?

Or, should I start looking for a cheap entry level CCD setup?

Initially I expect the DSLR imaging route to suit me just fine for the months/years(?) ahead, but I don't want to acquire kit that I later offload because I have 2" filters that don't work in a 1.25" filter wheel, or whatever... I'm not in a rush, but would like to know what my options (as far as colour DSLR imaging) are from here.

[MikeWilson]

This, chaps, looks awesome:

I use the Skywatcher Light Pollution Filter which cuts a little of the near infrared (700nm->800nm, but then let's it all in). It's also surprisingly high on the mercury lamp spectrum, too. So presumably the astronomik cls-ccd filter doubles up as a suitable IR cut filter. (nothing past 700nm gets through).

So one of those, combined with any UHC filter ought to allow me to access the photons from emission nebulae. But not in colour. I think.

[SteveL]

The CLS is really a light pollution filter, and to be used by itself. It allows in all the light but blocks the usual LP wavelengths.

A UHC or narrowband filter is the other way round... it blocks everything, but allows in only specific wavelengths.

I had (or maybe still have, will have to check) a CLS filter and also a Hutech IDAS filter, In my specific location the CLS was too harsh, giving a colour tinge to the images. The IDAS works better for me, and its the one I use for all my colour imaging.

[BlueAstra]

One thing I've often wondered about is that although you can replace the internal filter with one that lets through more Ha, the 1000d still has a Bayer RGB matrix in place over the pixels. So 2 downsides, there's only 1 R pixel for every group of 4 pixels (GGBR), and what is the spectral transmission graph of the RGB matrix?

Obviously you get a big improvement replacing the filter, judging by the results on the forum, but I'm also reluctant to do it since I use it a lot for normal photography. I think my route is either to get a second 1000d to modify, or a dedicated mono CCD. However, a mono CCD with the same field as the 1000d is £££££, so perhaps a modded second 1000d is favourite. Or perhaps get a 550D for normal photgraphy and mod the 1000d...hmmm.

[MikeWilson]

The CLS is really a light pollution filter, and to be used by itself. It allows in all the light but blocks the usual LP wavelengths.

A UHC or narrowband filter is the other way round... it blocks everything, but allows in only specific wavelengths.

I had (or maybe still have, will have to check) a CLS filter and also a Hutech IDAS filter, In my specific location the CLS was too harsh, giving a colour tinge to the images. The IDAS works better for me, and its the one I use for all my colour imaging.

That's interesting, I haven't heard of the IDAS filter before.

In any case - using a broad spectrum and a narrow band filter together would work though, wouldn't it? Sure you'd lose maybe 2-4% of the total potential transmission, but that's not a lot to ask given the benefits of reduced noise, etc.

Here's my thoughts:

Get an unmodified second hand 1000D. Modify it manually by binning the filter. It shall become the astro-DSLR. Purchase an Astronomik CLS clip filter which will live in the camera semi-permanently. This is the new IR cut filter with an added benefit of cutting sodium/mercury also. I will then get a UHC filter and attempt DSO imaging through that.

Later, it is possible to isolate the various spectrums with individual filters, but my budget won't allow that at the moment. Not with 2" filters, anyway.

Or would I be best just going in with a modified DSLR plus a 2" Ha filter? (My thoughts are - UHC as a first step - separate filters later).

But I'm sold on the Astronomik CLS. I have the Skywatcher LPF which although is brill on the orange lights, it's not so good on the mercury ones. Let me dig up a chart...

[MikeWilson]

Skywatcher LPF.

[MikeWilson]

I just had a thought - if I get a modded DSLR, a UHC filter and a CLS filter - take a sub of an emission nebula and then debayer the raw sub file and extract the red channel....

Will I have just taken a Ha image?

Furthermore - if I then extract the green from the debayered image (two greens?) will I then have O-III + Hb?

Then I can merge the two and somehow amplify the blue to get an RGB image?

edit: This appears to have been discussed before (I'm reading this thread now: http://stargazerslounge.com/imaging-discussion/105032-dslr-bayer-matrix-question.html)

[3ggman]

I know nothing about any of the filters mentioned here or their usage, however I recently modded my 1000D and due to the non stop cloud cover I find myself using it more during the day, including for work and have no problems with it at all.

I will always do a custom white balance for any given situation, (the factory white balance settings seem to be the only thing really effected by the mod, as they are useless afterward) autofocus still works perfectly too.

[MikeWilson]

I know nothing about any of the filters mentioned here or their usage, however I recently modded my 1000D and due to the non stop cloud cover I find myself using it more during the day, including for work and have no problems with it at all.

I will always do a custom white balance for any given situation, (the factory white balance settings seem to be the only thing really effected by the mod, as they are useless afterward) autofocus still works perfectly too.

That's very interesting... could work out cheaper than getting a second hand 1000D body and modding it. Still, I'm very reluctant to potentially sacrifice my 1000D.

[3ggman]

Now I could be wrong about this but when I done the mod, the camera not working during the day was my biggest fear. The autofocus not working I could live with.

The filter you remove on the 1000D is 0.8mm thick which is negligible when it comes to shift in the focal point, in the 350D or/and 400D (not sure which one or if both) the filter is a lot thicker I believe so when its removed the focal point can shift by 2 or 3mm which would render the autofocus useless.

The color change in images (all the extra red) is very simple to fix by using a custom white balance.

Just read up about it and see if more people get the same result as this is just me experience and could be a fluke.

[MikeWilson]

Very interesting. I'm still not sure I want to be playing around with the white balance continually while using the camera for non-astro use, so I think a second (refurbished/second hand) 1000D is in order.

How long did it take you to do the mod? Unpicking all those ribbon cables seems scary... It seems too easy to forget reattaching one!

[3ggman]

Yea, refurbished ones are not that expensive. I would say it took max 2 hours to do. its quite an easy mod to do if you take your time and all that.

[themos]

I did mine end of August and ruined my camera. It's worked once since then but I suspect the ribbon cables have been damaged somewhat.

It's not a question of forgetting to connect them, it's that there's no procedure to check WHICH one hasn't "sat" properly.

[MikeWilson]

@3ggman & thermos:

Very interesting reports!

Thermos, when you say you 'ruined' your camera, in what way? If it's just cables, surely you can re-open your camera and reseat the cables, or was something else more permanently damaged?

I'm thinking of cleaning out my desk, running air filters for 12 hours beforehand and then getting down to work with a screwdriver and a toothpick

[themos]

I've done the dis/connect business quite a few times already. After one of those, it worked. But then it didn't. It needs professional help, don't know at what cost yet.

[3ggman]

I did mine end of August and ruined my camera. It's worked once since then but I suspect the ribbon cables have been damaged somewhat.

It's not a question of forgetting to connect them, it's that there's no procedure to check WHICH one hasn't "sat" properly.

If you mark the cables with a felt tip pen before you take them out (just a put a line across the back of them) you will know exactly how far they must go back in.

[MikeWilson]

If you mark the cables with a felt tip pen before you take them out (just a put a line across the back of them) you will know exactly how far they must go back in.

+1 - what a brilliant suggestion. Thanks for that