Astrodevice logo
Filter drawers and adapters - results and experiments

Filter drawers and adapters – results and experiments

In this article, I will present the results I got by testing Astrodevice filter drawers and adapters. As you will be able to further notice, some results subjectively are pretty good and on some you will see more or less optical distortion. I would like to show you real results and not some marketing idealized ones. As I have already discussed in the essay on tilt and bacfocus, in the case of optics with RASA 8, we have so many variables that it is difficult or even impossible to get identical images every time and give some single representative results. Furthermore, copies of telescopes vary, they have different histories, collimation, mirror micro-defects, etc. Hence, the results I have presented here will not necessarily be exactly the ones you will get. If I were to point out one conclusion that I would like you to remember after reading this article, it would be this: I want it to be absolutely clear that although I try to create my designs to work as accurately as possible, the results you get will probably never be totally free of some optical defects. These defects may be small, they may be slightly larger, but they will probably always be noticeable to some extent. My accessories are not intended to correct optical defects in telescopes, fix collimation, or imperfections in cameras. I cannot control ones print quality, precision of assembly, type of the materials used, clamping forces, plasticity of components, screw adjustments, etc. However, I am sure that with proper care you can achieve great results and if you are able to accept possible optical imperfections in exchange for the convenience of use, I think that my designs may give you a lot of joy.  

Stacked images

Here are some examples of stacked images, taken using filter drawers and adapters made by Astrodevice:

Great Orion Nebula (M42)

Telescope: Celestron RASA 8
Camera: ZWO ASI 2600 MM Pro
Filter drawer: FD 900 M42
Filters: Baader Highspeed SII,Ha,OIII
Total time: 3h 20′

Flaming Star Nebula (IC 405)

Telescope: Celestron RASA 8
Camera: ZWO ASI 2600 MM Pro
Filter drawer: FD 900 M87
Filter: Baader Highspeed Ha
Total time: 2h 36′

Rosette Nebula (SH2-275)

Telescope: Celestron RASA 8
Camera: Canon Ra
Adapter: CRF R8
Filter: Triad Ultra Quad-band
Total time: 2h 22′

Cone Nebula (NGC 2264)

Telescope: Celestron RASA 8
Camera: Canon Ra
Adapter: CRF R8
Filter: No filter
Total time: 2h 00′


Telescope: Celestron RASA 8
Camera: Canon Ra
Adapter: CRF R8
Filter: Idas NBZ
Total time: 3h 30′

Single frames

Any optical distortion is most apparent on single frames. When an image is stacked, the irregular star shapes often average. The final result when the image is stacked is therefore usually much better than the result for each individual frame. However, it is worthwhile to look at such individual shots to get a better idea of the effects that may be present in images taken with Astrodevice accessories. The images shown here were analyzed using CCD Inspector, which allows you to measure the curvature of the field from the resulting image.

The above images were taken with a Celestron RASA 8 telescope, ZWO ASI 2600 MM Pro camera, Baader Highspeed OIII filter and FD 900 M87 filter drawer.
The exposure time for each frame was 64 seconds.

Repeatability of optical distortions

When discussing tilt and backfocus, I wrote that large sensor cameras combined with a RASA 8 telescope produce a system that is extremely sensitive to precise configuration. It was mentioned that one factor that is difficult to control is the strength of thread tightening. Depending on how tight you tighten the camera, you will have different image geometry. It is enough to turn it a few degrees more or less, and the image will be completely different. Below you’ll find an example of single frames taken with ASI 2600 MM Pro camera with FD 900 M42 filter drawer screwed on the original Celestron metal adapter. In this setup, besides the drawer calibration itself, we have three variables that we need to control every time we put the camera on the telescope: the thread tightening force on the drawer, the adapter tightening force on the telescope and centering the camera on the telescope’s optical axis. In the experiment shown below I took the camera with the drawer off and put it on the telescope again, each time tightening the threads with a different force. Each time the centricity of the setting was also slightly different.

… and another set:

This experiment demonstrates two important points. First, you should be able to accept that the images you obtain may be slightly different each time you put the camera on the telescope anew. Secondly, it is worth for you to compare both the images and their analysis with each other. These are rather extreme cases of distortion, but it is not impossible that you will encounter them. See to what extent the differences between them are acceptable to you and to what extent they may be an issue you could not reconcile.

Of course, a number of problems can be limited by reduction in the number of degrees of freedom in which conditions of use can change. If you choose the RASA 8 dedicated FD 900 M87 drawer, this is what will happen. There is only one large M87 thread and the layout is completely centric. So we have two potential problems less than in the case of a universal drawer with M42 threads: we do not have one intermediate M42 thread and problems with camera centering. Here are the results obtained with the FD 900 M87 drawer when alternately pulling and attaching the camera.

Using different filters

In the already mentioned article I addressed the problem of backfocus. I pointed out that the 28.73 mm recommended for RASA applies mainly to narrowband filters. I posted a link to a discussion at Cloudy Nights, in which it is stated that for non-filter configurations the backfocus can “back off” by more than a millimeter. In fixed thickness drawers, this can be a problem. That is why it is possible to screw in an additional UV/IR filter in the M87 versions and thus achieve extended backfocus when using broadband filters. I have described this feature here
Regardless of ability to take advantage of this unique feature, it’s worth taking a look at a test that shows how images are obtained “just like that”, when using different filters in the drawers and even using the drawer with no filter at all (with empty filter drawer).
Below you will find the result of the experiment. I did not adjust the drawer to any extent while taking the pictures. I only changed the filters in the drawer. The drawer itself, after screwed onto the telescope, was only regulated in such a way, that I changed the main thread pressure a few times to get subjectively the best image. However, I didn’t spend much time on this activity and after a few minutes I moved on to taking proper pictures.
Here is a description of the 2″ filters used:
  • UV/IR Cut – made by Astronomik
  • Red – “Deep-Sky” series – by Astronomik
  • NBZ – by IDAS
  • S-II – “Ultra-Narrowband” series by Baader
The images are arranged from zero optical density (no filter) to the glass with relatively highest density (S-II). This way you can see how the image systematically changes with the change of the effective backfocus. Please compare images to see what differences will be noticeable when using different single filters with Astrodevice drawers and adapters. 


Measurements made with the CCD inspector allow to objectively measure the curvature of the imaging field. However, each image, in the end, is judged subjectively by the viewer. Some like it, others don’t. For one, the shape of the stars is perfectly fine, for another it may be difficult to accept. Here I have tried to show you the different effects that you can experience when using Astrodevice accessories. You have to decide for yourself if a similar image quality suits you and if this quality, combined with the convenience of my accessories, deserves to be used in your astrophotography kit.

Scroll to Top