Strange behavior of Sigma 40mm F/1.4 Art lens.

[Vroobel]

27 minutes ago, Elp said:

I would note the ZWO Ef adaptor likely isn't mounted square so will cause an element of tilt, I get it on all my SY135s, the weight of the lens doesn't help. A lot of people have to shim to get it right.

 

Oh, yes, the lens is 1.3kg heavy, so I bought a dedicated ring holding it while the Canon or the ASI camera (both around 700g) is just attached to the lens. I wouldn't be worried if I saw marks of the tilt: pinpoint stars along one edge and comas on the opposite edge, but that's not the case. 

Edited April 24 by Vroobel

[Vroobel]

7 minutes ago, Elp said:

Make sure you're measuring with a good quality Vernier caliper too either a manual one or digital, measure each component separately, then again when assembled.

 

There is not too much to measure, only the ZWO filter drawer. I base on the Sigma 44mm flange distance and the ASI 17.5mm back focus. The ZWO must be the remaining 26.5mm long and it is indeed.

I use a high-class vernier caliper which is older than me.  I love to use it, even if I have two digital ones.

[Elp]

Have you got any thin spacer rings to put onto the back of the lens adaptor prior to the camera, that's what I'd try next.

But before that, just because you don't need anything else to test it, try stopping the lens down at each f ratio and see if it improves.

Also I assume you've already checked your 2600 tilt adjustment plate is square to the body and not tilted in any way?

Edited April 24 by Elp

[Vroobel]

Yes, I have a lot of thin plastic spacers starting at 0.2mm and also three paper ones, around 0.1mm each, I described experiments with them earlier in this thread. I reached a compromise between the coma and radially elongated stars at 1.2mm spacer in total, but I used filter then, I think it was the L-eXtreme. I have to check it again without filters, but not tonight, the setup is at home now due to clouds. I'm happy to see that the Sigma lens is OK, I considered returning it under warranty... 

 

There is a very interesting method of setting the iris permanently:

Based on a following CN thread I attached the 6D body, set a BULB mode, pressed the shutter release button and... detached the body - the iris remains in the F/1.8 position. That's a great trick! 

https://www.cloudynights.com/topic/837147-sigma-art-40mm-14-canon-mount-no-aperture-ring/

Edited April 25 by Vroobel

[Vroobel]

34 minutes ago, Elp said:

Also I assume you've already checked your 2600 tilt adjustment plate is square to the body and not tilted in any way?

 

I've never had any bigger problems with tilt on this camera. The only tilt-like effects are related to a bent and it's always visible on the edge aimed down during the session. Here is a masterLight as example with an Askar FMA230: 

 

 

The stars aren't perfect, but take please into account that my Askar FMA230 has its back focus equal to 52.5mm instead of 55mm, so I stopped selecting spacers at an acceptable point which is fixable by a BlurXterminator.

 

Here is the same camera attached to an Altair 102 EDT Triplet APO with a dedicated x0.8 reducer, distortion is neglectable (at least for me). 

 

Edited April 24 by Vroobel

[Elp]

Have you got another astro camera you can try with the lens adaptor? I know the size won't necessarily match the 2600 but you'll likely start to see the distortion if it's a decent sized sensor, and if it reproduces the issue. Process of elimination.

Edited April 25 by Elp

[Vroobel]

Apart from ASI224MC and ASI678MC with their total sensors surface smaller than a half of the APS-C ( 😁 ) I have also a 2600MM Pro which is permanently attached to a filter wheel and used with bigger scopes. I think I have no choice and have to (reluctantly) disassemble it if I want to perform a proper test. 

[Elp]

It never occured for me to ask but does that AF lens have manual focus control when not connected to a camera body?

[Vroobel]

🙂 Yes, it has. It's always switched to the manual mode, even if used with the Canon body. I used a Bahtinov mask earlier, but now I rely on an EAF.

 

 

Edited April 26 by Vroobel

[Elp]

I'd do a test without filters, then try a few different ones. I'm thinking a bodied camera has a few pieces of glass mounted onto the sensor, with the 2600 you may have less though this doesn't explain if you have MF capability why its still causing coma at the edges. I find with my SY135s, if I keep the backspacing consistent, different filters will move the focal point away from the infinity point on the lens significantly.

[Vroobel]

40 minutes ago, Elp said:

I find with my SY135s, if I keep the backspacing consistent, different filters will move the focal point away from the infinity point on the lens significantly.

 

It happens to the Sigma as well, but I don't remember how much it moves. I'll try to record the values bot for the Canon and astro-cameras. 

Edited April 26 by Vroobel

[Vroobel]

After several approaches to the Sigma 40mm F/1.4 Art and the ASI 2600MC Pro cam with different apertures, I stopped my experiments at F/2.0 which still isn't perfect, but supported by the power of BlurXterminator, I can obtain nearly perfect stars.

I compared results of 50 x 15s subframes stacked with bias, dark and flat frames. The lens was joined with the camera with no spacers used and with 1.4 mm spacer in total which at F/2.0 seems to be a compromise between the aperture and star shape.  Each master file is presented as it is, with BlurX using 'Correction Only' and with BlurX using the following settings: 'Sharpen Stars = 0.15' and 'Adjust Star Halos = -50'.

 

Sigma 40mm @ F/2.0 - no spacer - IDAS LPS-P2:

 

Sigma 40mm @ F/2.0 - 1.4mm spacer - IDAS LPS-P2:

 

Sigma 40mm @ F/2.0 - no spacer - Optolong L-eXtreme:

 

Sigma 40mm @ F/2.0 - 1.4mm spacer - Optolong L-eXtreme:

 

As you can see, it's acceptable. 

Just at the end, another test of the lens joined directly to the Canon 6D (full frame) with no filter and with an FF OPTOLONG L-Pro filter looks as follows.

 

Sigma 40mm @ F/1.4 or F/1.8 - no filter:

 

Sigma 40mm @ F/1.4 or F/1.8 - OPTOLONG L-Pro:

 

Well, it looks like the lens works perfectly with the full-frame sensor, while use of filter brings distortions. The lens joined with the APS-C sensor using a dedicated 26.5mm thick ZWO filter drawer behaves strangely and becomes manageable af F/2.0 only thanks to sophisticated software.

I was advised to try it with another camera. I will do the test with my ASI 2600MM Pro camera ASAP which means when I finish imaging the Lion Nebula in SHO. Of course, I'll update the conclusion.

CS, 

Edited May 16 by Vroobel

[michael8554]

You might have to an consider obscure problem,  such as the sensor attachments in the astrocam being too tight.

Michael

[Vroobel]

I'm not sure if I understood. Could you please explain it? 

[TiffsAndAstro]

Do you have the lens wide open at f1.4? From reviews I've seen lots of fast lenses have issues wide open. 

Have you tried stopping down and seeing if problems are still visible?

Edited May 17 by TiffsAndAstro

[Vroobel]

10 minutes ago, TiffsAndAstro said:

Do you have the lens wide open at f1.4? 

Yes - with Canon 6D I can do it easily. 

No - with 2600MC not. 

 

Every lens loses its distortion and aberration at a higher F ratio. But I didn't pay £629 for the top class cine lens to work with it at F/5.6 or higher. 

[TiffsAndAstro]

20 minutes ago, Vroobel said:

Yes - with Canon 6D I can do it easily. 

No - with 2600MC not. 

 

Every lens loses its distortion and aberration at a higher F ratio. But I didn't pay £629 for the top class cine lens to work with it at F/5.6 or higher. 

I totally agree, but I've watched a lot of (I think?) Christopher frosts tests on YouTube and a lot of lenses struggle with sharpness in the corners wide open, admittedly not astrophotography tests though.

[Elp]

From my experience testing a variety of lenses this much is true, but their Canon works fine wide open apparently so something else is going on.

[TiffsAndAstro]

44 minutes ago, Vroobel said:

Yes - with Canon 6D I can do it easily. 

No - with 2600MC not. 

 

Every lens loses its distortion and aberration at a higher F ratio. But I didn't pay £629 for the top class cine lens to work with it at F/5.6 or higher. 

Someone makes  a canon EF adapter for zwo cameras that can control canon lenses. I wonder if the adapter might magically fix this? Maybe the lens needs a electric signal to keep iris in position or something?

Cuiv lazy geek had a video on the adapter.

[Vroobel]

31 minutes ago, TiffsAndAstro said:

Someone makes  a canon EF adapter for zwo cameras that can control canon lenses. I wonder if the adapter might magically fix this? Maybe the lens needs a electric signal to keep iris in position or something?

Cuiv lazy geek had a video on the adapter.

Yes, a Russian company called Astromechanics makes the ASCOM-controllable adaptors. Unfortunately, my assumption is to use this lens (and maybe Sigma 105 Art or 135mm Art in future) with ASIAir, that's why I didn't buy the Russian adaptor and used an EAF instead. I have to admit that it was really tempting me... Moreover, I really need filters under my Bortle 9 sky. The Canon 6D is kept for a darker sky. 

I cannot believe that one of the most experienced company like ZWO made their EF/M42 filter drawer wrong. As @Elp wrote, something else is going on. 

Edited May 17 by Vroobel

[michael8554]

23 hours ago, michael8554 said:

You might have to an consider obscure problem,  such as the sensor attachments in the astrocam being too tight.

Michael

I mean like Pinched Optics - the sensor may be mounted slightly distorted.

Is the lens bad on the 2600MM and on the MC  ?

Michael

[Vroobel]

I tested the lens with the 2600MC camera till now. Yesterday I finished my Lion Nebula by collecting RGB stars, so it's possible to repeat the tests with 2600MM one. I'm going to perform the simple test with filters but without spacers. I'll give up if it fails. 

[Vroobel]

I didn't want to detach the mono camera from my current setup, but finally, I did it. I replaced the MC with MM and ran it with ASIair Plus. During a few hours, I was able to capture literally three pics before I started an autofocus procedure, but I didn't save them, I didn't know that they were the last pictures taken with this mini setup. My ASI 2600MM Pro was continuously refusing cooperation, every time I received a message that the capturing of a frame failed. I wasn't able to perform the autofocus because of the same reason, which means an unknown reason. I realised that usually this problem is related to not enough power delivered to the camera (my MC camera is powered from the ASIair). I connected the camera to a 12V 10A PSU, but nothing changed, it still was unable to properly take even one picture, also after swapping USB 3.0 cables linking the 2600MM and 120MM mini cameras with the ASIair. I was very close to panicking, 2600MM isn't cheap. I gave up and replaced the MM with MC and screwed the MM back to the 4" APO setup - I wanted to perform a starfield test using the Sigma 40mm and the MC camera and also have a plan to start imaging a Crescent Nebula in SHO.

Both cameras work and ASIair works too. The only result of swapping the cameras is dust on the MM sensor, so I have to unscrew it from the filter wheel again, which I don't like. I gave up, I don't want to do another test, I'm happy with the fact that the PixInsight and BlurXterminator can give me a round shape of stars captured at F/2.0:

 

 

It's a result of stacking 137 subs with bias, dark and flat frames. I used the BlurXterminator with its 'Correct Only' feature. Below is proof that the 2600MM works as well - two first hours of the Crescent in SHO:

 

 

The mystery remains unsolved.

Everything above concerns the fight against an APS-C camera and it does not change the fact that this lens is unbeatable if it is to work with a camera 'dedicated' to it, in my case the Canon 6D.

 

Edited Sunday at 15:39 by Vroobel

[Elp]

I know you've probably given up with using it with an astro camera. Thinking about it a little more I'm wondering whether glass distances are contributing to the issue. A camera body has typically 2 or 3 pieces of glass over the sensor, and they're mounted right on top of it. An asi camera, although I don't think the sensor is bare won't have the same, its AR glass (or whichever one is in the 2600) is mounted further away from the sensor, and then if you're using a filter another distance and refraction of light path to consider. Although it doesn't help your issue, I wonder if this is what's causing it.

[Vroobel]

Of course, it's possible. It looks like the 40mm (or less) is very sensitive to anything not typical in the optical path. I'm very curious how the Sigma 135mm (or 105mm) behaves with the APS-C astro-camera, but presently I have no money to find it out.   

Thanks anyway for thinking about it.