Astrodevice logo

Front Tilting Mechanism

Front Tilting Mechanism is a unique system used in certain Astrodevice adapters. It allows to compensate the tilt of the camera by adjusting the tilt of the whole imaging unit, that is, the camera connected to the adapter and not the camera itself. In presenting this solution, I would like to describe the situation that led to its development. It’s a story showing how necessity becomes the mother of invention and ideas turn on their heads can produce truly amazing results.

Since the information contained in the article refers to the Celestron RASA 8 telescope, I would first encourage readers to read the article on tilt and backfocus. The information contained there will help to better understand the scale of the issue described.

Introduction

Let’s start with the RASA 8 telescope, the Canon Ra camera and the adapter that was to connect them. I developed such an adapter in early 2022. It worked wonderfully and provided essentially perfect 28.70 mm backfocus. With its help, I took great pictures, such as the wide-field image of the constellation Cassiopeia. It was enough to attach the adapter to the camera, shoot the whole thing on the telescope and one could enjoy a really beautiful image.

Cassiopeia
Stacked image of the Cassiopeia region taken with the first version of the CFR R8 adapter
The first version of the CFR R8 adapter mounted on the camera

After taking many similar pictures, I found the adapter to be completely sufficient and the Canon Ra combined with the RASA 8 to give very pleasing results. This turned out to be true, but only until then. Unexpectedly, I experienced a complete surprise.

The problem

After a long break with the Canon, I decided to shoot with it again in the second half of June 2022, a few months after the original adapter was developed. I hooked up the camera, pointed the telescope at the Lion Nebula and…. the image was terrible! The stars had awful shapes, and the farther away from the center of the full-frame sensor, the worse the situation looked. I began to look for the cause, but no ad hoc action produced results: neither stronger or weaker tightening of the connections, nor replacement of filters, nor focus adjustment – it was always bad!

A frame revealing the likely tilt of the sensor
Close-up of the three regions of the frame: mid-left, mid-center and mid-right

For better recognition of the problem, I also inverted the images.

I soon remembered that a few weeks earlier, during a routine check of hand luggage at the airport, my camera fell off the counter. At the time, it seemed that nothing had happened to it: it carried no injuries, nothing was broken, and in daylight it took pictures as before. At the time, I considered myself very lucky and forgot about the whole situation. Now, however, suspicion fell again on the camera: perhaps the fall caused a minimal shift of the sensor?

Investigation

To see what I was dealing with, I decided to investigate again – this time very carefully – the camera’s performance in daylight conditions. I took several pictures looking at different parts of the image. For the first test I used a wide-angle, bright Sigma Art 135 lens.

I wanted to see how the camera handles light coming from a wide range of angles in natural lighting conditions. I also chose this lens because it handles astrophotography very well. Stars have an even, round shape even in the corners when using a full-frame camera. Thus, in the case of the test I conducted, any possible aberrations could not be attributed to a defect in the lens, but precisely to a problem with the camera itself.

Even a rather careful analysis of the image in my opinion does not indicate the presence of any problem. The photo is sharp, the geometry is not crooked, and without specialized equipment , using only subjective judgment, I could not determine the presence of sensor tilt from this photo.

So I had to conduct further research.

Photo of the garden taken with a Sigma Art 135 lens

In the next step, I connected a William Optics SpaceCat 51 telescope to the camera, a magnificent Petzval design with a focal length of 250 mm. This time I positioned the camera against a fence wall with a fine, geometrically uniform texture. I figured that this would allow me to better identify any distortions – they would be more recognizable because the texture across the surface is self-similar.

Uniform wall structure
Considered parts of the image (white)
Considered parts enlarged

However, again, in neither case did I find anything suspicious. If sensor skewing really leads to this situation – its effect may be completely unnoticeable in normal photos.

The next test was quite drastic. Using a special caliper, I very carefully measured the distance of each corner of the sensor from the camera flange.

Warning: do not do a similar measurement without proper preparation, equipment and experience. Touching the sensor with any object, especially one as hard as metal, can lead to scratching the surface of the sensor, cracking it and damaging it irreparably.

I examined each of the four corners, repeating the measurement cycle four times. Finally, I took the average of each measurement and calculated the standard deviations. I found that corner number 4 is the most offset from the rest, and without any doubt this deviation is far beyond the measurement error. 

It turned out that the problematic corner is clearly drooping and the difference in measured distances is almost 0.5 mm! The method of measurement may not be very sophisticated, but it was completely sufficient to determine the existence of tilt in general. The deviation was really quite large! 

In this way, I was able to demonstrate that the problem I am facing is indeed caused by sensor tilt. The tests also showed that under normal daytime conditions such a tilt can go completely unnoticed. Astrophotography, however, sets the bar very high.

Solution

Once I had identified the cause, it remained to find a solution. What was needed was a tilt adjustment mechanism that could compensate for the tilt of the sensor. Dedicated, cooled astrophotography cameras with large sensors are equipped with plates that can be tightened with screws. They are provided with a tilt adjustment mechanism similar to that used, for example, in Astrodevice filter drawers. However, in the case of classic photographic cameras, there is no place to mount a similar mechanism. Especially in the case of the RASA 8 telescope, leaving minimal space due to its extremely small backfocus. The adapter, like a regular lens, fastens rigidly and there is no way to adjust the tilt.

Considering the problem, however, I thought that it would be possible to mount the mechanism on the other side – on the side of the telescope. Instead of tilting the plane of contact between the adapter and the camera, the plane of contact between the adapter and the RASA could be tilted. Normally, the adapter should sit flat against the flange of the telescope, but if I added adjustment screws there, I would get a mechanism that could compensate for the tilt.

But there has been another difficulty: the screws protruding from the adapter could easily scratch the telescope’s metal flange. It is made of blackened steel and if a protruding adjustment screw sticks to it, a mark will certainly be left on it.

So I thought of an additional washer. If an additional safety washer were given between the telescope flange and the protruding screw, it could take the pressure on itself and prevent scratching the telescope flange. However, an additional washer increases the distance, which moves us away from perfect backfocus. Does such a solution make sense in this case?

CRF R8 adapter with front tilt adjustment screws
Safety washer with special flange for easy installation on the telescope
Safety washer after installation on RASA 8

Test

To test this, I printed a suitable washer of 0.6 mm thickness. I also redesigned the camera adapter, equipping it with front tilt adjustment screws.

The first step in the test assembly was to put the camera back on with the adapter, without the screws extended, flat, without the washer. After taking a test shot, I put the washer on, thereby moving the sensor 0.6 mm away. I re-sharpened the image using the Celestron electric focuser and took the picture again.

Test frame without a washer
Test frame without a washer (mid-left, mid-center, mid-right closeup)
Test frame with a 0.6 mm washer
Test frame with a 0.6 mm washer (mid-left, mid-center, mid-right closeup)

It turned out that the images were slightly different, but at first glance not so much as to abandon the idea of being able to move back with the backfocus by 0.6 mm. Moreover, it seems to me that adding this pad even improved the result a bit. The reason probably is that during the fall the camera suffered not only a skewing of the sensor, but also a forward shift along the optical axis. The attached washer compensated to some extent for this problem as well.  By the way, knowing the thickness of the pad, I found out the necessary number of steps of the focuser, which he had to overcome in order to sharpen the image again. With this, I calculated that for every 0.1 mm of backfocus, the focuser takes about 100 steps.

Again I put on the adapter without the washer, again without the adjustment screws extended. And again I focused the center and noted the current position of the focuser. I then changed the focuser position, sharpening the image in the corner where the stars had the worst shape. Again, I noted the focuser position and calculated that the difference in focus corresponded to roughly 0.5 mm of offset. If the calculations were correct, that’s roughly how much I had to point offset the plane of adhesion of the adapter to the camera. “Roughly” because the titlt adjustment screws are not exactly in the corners of the sensor, but a little further away. The actual extension of the adjustment screw therefore had to be slightly larger.

I removed the adapter, put on a protective washer, extended one of the screws by 0.7 mm and put the camera on the telescope. I focused and… the image was almost perfect! Once again, I took off the adapter, made a second minor adjustment and…. done!

Test frame with a 0.6 mm washer and a quick tilt correction
Test frame with a 0.6 mm washer and a quick tilt correction (details)

Once I understood what the problem was, the whole tilt adjustment took me less than 5 minutes. I got a very good image, according to my subjective feeling, of course. Most importantly – it turned out evenly across the entire plane of the full-frame sensor. Encouraged by the result, using the adjusted adapter I took a 1.5 hour series of images of the Lion nebula, the result of which seems to me very satisfactory.

Full frame image of Lion Nebula (Sh2-132) taken with Canon Ra, RASA 8, Idas NBZ filter and CRF R8 adapter with tilt correction and 0.6mm protective washer

Front Tilting Mechanism

I used the experience gained from the tests to create a mechanism I called Front Tilting Mechanism. Without delay, I equipped with it both adapters I designed for mirrorless cameras of the Sony E-mount and Canon EOS-R systems. All of these adapters sold starting in July 2022 are already equipped with the described tilt adjustment mechanism. In the future, the FTM will also be installed in other designs where it is impossible to adjust the tilt from the camera side.

By the way, it turned out that destructive for the image in RASA is not so much the wrong backfocus, but mainly tilt. This is because tilt causes problems, which further skyrocket with improper backfucus. However, if there is no tilt – just moving the boackfocus even by 0.6 mm doesn’t seem to make a very big difference – at least one that I subjectively felt couldn’t be easily accepted.

Final notes

It should be borne in mind that sensor tilt is not anything normal. A properly seated sensor should be perfectly parallel to the plane of the telescope mount. Any tilt adjustment mechanism should be regarded as an assistive tool in a temporary, abnormal situation. If your camera has a tilted sensor, you should return it to an authorized service center for repair. Solutions such as FTM are intended to help you get through the time of the defect, help diagnose it and provide ad hoc assistance. Don’t expect that a tilted sensor compensated for by an artificially tilted mounting plate and additional washers will give the same image as a sensor mounted the right way. Even if you get a very good image, be aware that your camera should go to a good service center as soon as possible.

Scroll to Top