Canon Cameras - Filter replacement or removal ?

[RobertM]

Silly question ... time to rethink perhaps ?

I've been mulling over replacing the filter on my Canon but this article on the Astronomiser web site has made me rethink - http://www.astronomiser.co.uk/eosfilter.htm

It seems that on cameras with a separate dust removal glass (sensor cleaning) a filter removal may in fact be a better option.  The reason is that it's been found out that the sensor cleaning cover glass is in fact an extremely effective UV/IR cut filter as well as acting as a dust seal.  A disadvantage of the removal rather than replacement are that there would be a minor change to the autofocus position.  This is due to the difference is light path length and may not be noticable under normal use.

Filter removal also means two fewer optical surfaces in the light path, which has got to be a good thing.

Note this only applies to camers with 'Sensor Cleaning' i.e. models introduced from the 400D.  For earlier models such as the 300D/350D a replacement is still recommended to protect the CMOS sensor.

Robert

[mickw]

I had the filter replaced with my 400 rather than take a chance on just the removal not being quite as efficient.

Nice of the guy to be so honest though.

[MarkS]

Robert,

Thanks for that - it's very interesting information which I wasn't aware of.

Since I'll be using it astro-work only, the decision is a no-brainer - just remove the rear filter.  I might give it a go with my 400D this weekend.

Also, removing glass means that rays now come to focus [Edit: See later post] *behind* the CCD.  This is fortuitous because it means when using a standard lens, objects at infinity can still be brought to focus (though it will be necessary to focus manually rather than using autofocus).
I'm beginning to think that the 400D might be the DSLR of choice (subject to confirmation by post-mod comparison with 350D).  See also, my thread where I "re-evaluate" the 400D:
http://forum.orpington-astronomy.org.uk/index.php?topic=6268.0

Mark

[MarkS]

Whilst cycling home tonight I realised I got this the wrong way round.  Removing the glass filter means that rays now come to focus further in *front* of the CCD.  This is maybe unfortunate because it means when using a standard lens, it may be impossible to bring objects at infinity to focus - it depends how far you can adjust the focuser beyond the infinity marking - most lenses allow this to some extent but will it be enough?

If it is critical to use the modded camera with standard lenses, it may be necessary to use the Baader replacement filter (or clear glass).

Mark

[MarkS]

Just modded my 400D by removing the IR filter (leaving the anti-alias/dust removal filter in place).

The only description I've found is here:
https://xythos.lsu.edu/users/jcham21/400d/

The IR filter is very thin (0.55mm according to Ian's micrometer) and cannot be removed without the glass cracking.

The whole job took me 1 hour 40 minutes.

Mark

[MarkS]

I've just done a quick comparison of the 400D with the 350D using an Ha filter.  Slightly disappointing. 

It appears that at the Ha wavelength, the 2 sensors have almost identical Quantum Efficiency (per unit area of CCD).  But since the 400D pixels are 80% of the area of the 350D pixels this means that for a given F-ratio and exposure time, the a 400D pixel will only collect 80% of the photons of a 350D pixel.

I'll perform a more extensive sensor sensitivity test of the 300D, 350D and 400D at the weekend.

Mark