We have received the following question on our Narrowband (and f/2 Highspeed) filters:
I understand that narrow-band filters suffer shifts of the passband for light rays that come in at an angle. Is that why they are good to a certain f-ratio and why you offer different filters for fast optics (how do they work? are they actually de-tuned to have the correct passpand at a different average angle?)? I would like to ask whether this additionally depends on the field-of-view of the optics? I would expect that since the angle should be larger for wide field imaging at the same f-ratio? So are all of the filter ment to be used with telescopes or can they also be used with shorter focal length camera lenses, for example?
Since this question is of particular interest for many other people, we would like to answer publicly. Please see our answer below:
In general what you assume is very true - but - "the devil is in the detail".
For an aperture ratio of f/2, the filter design-CWL would require to being given some shift in order to compensate for the large angle of the incoming wavefront. The devil in this game is - even if one were to do the coating layer design for a huge incoming angle, at the same time also image information is coming in at zero-degree-angle - respectively at all angles in between. Which essentially would mean that no shift ought to be allowed. A filter made to cover that wide an angle must cope with these contradicting requirements. Hence no one else has succeeded in making f/2-filters that would not suffer from severe image contrast fall off and vignetting effects, either in the corners of the field - or in the middle. And doing these coatings in the appropriate way is about the hardest task in filter coating technology today.
The production of these filters is anything but trivial. It took us around three years and numerous unsuccessful coating runs to finally be able to produce the CMOS-optimized f/2 Highspeed filters with consistently high quality. In essence we are moving to a much more elaborate and critical filter design with very steep slopes on both sides of the spectral window – hence with even better contrast and smaller stars.
We think that our competitors also must go through all of this vicious circle still – and we could not be more thankful that so many customers had the patience to wait until we finally can produce these filters continously in the necessary quality. Now you can finally start imaging monochrome with your prime focus optical equipment
Both families of our new f/2 Highspeed filters with 6.5nm and Ultra-Highspeed Filters with 3.5 / 4nm are entirely designed and exclusively manufactured to work for a focal ratio range of f/3.4 to f/1.8. They are NOT designed to work above that, say from at f/3.5 to f/10 and this may become very apparent to see in the imaging results, if someone would try this out – for these focal ratios, "classic" narrowband filters work better.
Those who will try out our new f/2 Highspeed filters though will be in for experiences like this image, done by Andreas Bringmann (Astrobin: Equinoxx) using the latest prototypes of our new generation of f/2-filters – in conjunction with an ordinary Hyperstar-Celestron 11 Edge-HD OTA).
About Production of Baader filters
Producing filters which will be used consequently and only for a defined and rather tight range of light ray angles (from aperture ratio and wide angle optics) really does make good sense. Most people would not notice that the difference in signal (emission line) throughput will be caused by their improper filters. Everyone will note that the exposure time varies widely, depending on the telescope he uses. In most cases longer delivery time for obtaining similar s/n ratio is attributed to a larger or smaller telescope aperture, but not to the difference in aperture ratio. So the decline in filter-transmission goes unnoticed. Fact remains that a filter designed to work at f/10 will gradually fall off in transmission while the focal ratio of the optics in use becomes shorter. At f/4 the transmission may only be 60% compared to 95% at f/10. To improve on this as good as technology allows, we do have extremely elaborate coating recipees and – most of all – very expensive rare earth coating materials applied – with a large amount of layers accumulating on both sides of the filter-substrate.
The thick coating stacks applied onto our coating substrates make it very necessary to protect the coating edges from peeling and aging. And this extra amount of care against aging we are putting into our production process - as we do explain since years. Contrary to many competitors, our filters are coated as individual sized substrates for each filter size that we offer. We do NOT follow the road of economic optimisation to coat large plates and cut the filters from these plates on demand, in the diameters or sizes required. We could not risk to do this as the thick coating stacks would separate themselves from the substrate when the coating is just brutally cut all through together with the substrate. All these years we have checked if we are right in going this route. When looking at the layer structure of a filter cut out of a larger plate it becomes obvious that the coatings will age from the side while our coatings remain all time sealed all around the filter edges (see image to the right). We will soon emphasize this by announcing a lifetime warranty for all Baader Narrowband-filters. We have emphasized this now clearly by giving each new CMOS-optimized Baader Filter a Limited Lifetime Warranty: Life-Coat™
Many far east companies at this time begin to offer filters at wonderful prices. But once you start to go into details you will learn that their coating designs suffer from severe flaws (e.g. see the test review on uvir.eu) - mostly concerning totally inappropriate off band blocking leading to meager color separation (see also an Article on Cloudy Nights) as well as a huge transmission fall off when the filter has to work at a different than the design f-ratio - not to speak of different temperatures. Regularly filter designs are calculated under the assumption that the application of these filters - as well as the testing of filter specs - happens in the lab. However - astronomical narrowband-filters quite rarely are actually used under lab conditions. So when someone will measure our filters in the lab he might be mistaken to think they do not reach the transmission values they may claim for their own lab designed filters. We have taken pains to come up with coating designs that will work from -30°C to +30°c with an absolute minimum of shift. The secret really lies in the coating materials - the more expensive these are and the more these layers are accumulating, the better the temperature shift as well as the angular shift can be brought under control. So we most carefully aim to achieve a consistence performance across a wide range of working temperatures and f-ratios.
To ensure this standards, we quality inspect each individual filter before it leaves our buildings here. Among else, we utilize a professional Echelle-spectrograph with about 100-times better resolution and ThAr-calibration than the simple educational spectral imagers that are being offered at economy pricing and that recently are being put into use by some companies to offer their own payable "filter inspection service" (found on teleskop-austria.at)
The wealth of filter design knowledge that we gained over the past 20 years is almost "too much information". We could do much better in pricing with choosing cheaper materials for vacuum deposition, use lesser grade coating machines with less exact layer monitoring equipment during the coating process – and let a much worse quality controlled product go out the door - if we were not to absolutely know all the many possible flaws this would cause. The f/2 Highspeed filters weren't available for a much too long time, and we suffered from it as well as our faithful customers. However – the situation did only arise because we NEVER let anything go out this door which is not fully meeting our specs. And still – we are only human. If something like this does happen for once within 20 years – then we call it back. This happened once – and we don't want to let it happen again.
And concluding your question:
As long as you put our regular Narrowbands (6.5 nm) or the new family of Ultra-Narrowband filters (3.5 / 4 nm) in front of your (non-wide angle) DSLR-camera lens, you do not at all need to care for using f/2 filters. Only if you were to put our filters into the back of a fast camera lens – that is between the camera body and the optics – then indeed you would need to choose from the f/2-filter family.
But note: in contrast to the usage with telescopes the light ray angles with camera optics have a much stronger variation, this is due to the short backfocus and mainly back lens and iris size limitations. With full format sensors these light angle differences can exceed the on-axis f/2 light angle spectrum by far! So take care when using strong wide angle lenses! With telephoto there will be less effects.