Replace Trius 694 Mono with ASI 294 MM-Pro?

[DaveS]

Yes, I know it's a bit like apples and oranges, but the plate scale is about the same (And similarly barking) while the FoV is significantly wider.

The origin of my query is from putting my old SX Trius 694 on the ODK 12 (Yep, I said it was barking) to go after the HCGs which would be tiny in the FoV of my G3 16200.

I have run into all kinds of issues, the main ones being having sufficient stars in the field for plate solving to get a meaningful fix, but more crucially the ASA Sequence autofocus is throwing so many issues as to be dysfunctional. I suspect driver conflicts, but don't know enough to sort them out.

On paper the ASI 294 *should* be a better camera, with lower read noise, higher QE and higher DR. With the short BF, and on a ZWO FW I think I will be able to "get away with" the 1.25" Baader filters that are already in said wheel.

BUT...

There's always a "gotcha" , and I was wondering what gotchas I should be aware of, and which might be a deal breaker. I am aware that the ASI 294 has amp-glow which will need Dark Frame calibration. The Trius 694 just needs Bias frames.

[DaveS]

Wow, nobody has any issues at all! Magic!

[StevieDvd]

I think the jury is still on a long lunch break.

 

[Adam J]

I will be able to "get away with" the 1.25" Baader filters that are already in said wheel.

That will depend entirely on F-ratio of the imaging scope. 

I doubt that the auto focus issues are anything to do with the sensor your using. 

The ASI294mm is a relative pain in the bum to calibrate in comparison to the 694, it can suffer from some of the same calibration issues as seen on the 294mc version, probably sample dependent. Its the reason I have decided to stick with the ASI1600 until such time as a better 4/3 sensor is available. 

I assume you are binning the 694 with that scope? I think the bigger performance advantage of the 294 is in narrow band imaging due to the lower read noise but for LRGB I am not sure that you will see such a big advantage. 

Adam 

 

[DaveS]

I ran the parameters of the ODK and ASI 294 through the Astronomy Tools calculator and got a figure of 24.5 mm, while the Baader filters run 26 mm.

But I may have found one issue, the Sequence autofocus wants to bin 3x3 for the initial star find, and I had overlooked a bin 3x3 tick box in the SX driver, but won't be able to check if I've solved the problem until we get another clear night, and none are forecast this week.

[DaveS]

An update!

The sky has cleared just enough and for long enough for me to run an autofocus check.

And it worked! Yay! So that was the problem, as I suspected it was a driver issue, the "Enable 3x3 binning" box had to be ticked.

No need to blow £1500 that I haven't got.

Many thanks for all your input.

[nfotis]

If you still want to replace the old Trius with a modern sensor, I would suggest the IMX533 (available both in color and B&W models, cooled and non-cooled versions, from many manufacturers)

The IMX294 seems to be relatively fickle sensor, while the latest group (IMX533, IMX571 and IMX455?) offer lots of resolution and improvements.

The link below should give a comparison of the Trius to the IMX533:

https://astronomy.tools/calculators/field_of_view/?fov[]=3145||8677||1|1|0&fov[]=3145||35||1|1|0&messier=13

N.F.

 

[DaveS]

I've looked at the 533, but I'm not going to be buying any more kit for the foreseeable. With our rubbish weather there's no point.

In fact I wonder if there's any point in astronomy at all from our swamp of a country.

[barbulo]

Happy owner of a 294MM here. No flaws up to now. The amp glow calibrates out perfectly with darks and long enought flats (>3 secs) work well. High QE and DR are remarkable. I use it with 36mm filters (they came in the bundle) but 1.25" are enough for that sensor size. 

I´d recommend this camera until new 4/3 sensors come out.

[DaveS]

To expand on my previous post (From my phone), I have the 533 and the SX 694 in Stellarium and have been playing with the FoV.

However, if the 694 at ~0.46"/px is silly, then the 533 at ~0.38 "/px is downright ludicrous. I could bin it at 2x2, but that would be 0.76"/px, at which point I might as well use the G3 16200 at bin 1 and crop.

In any case, this is just a project to go after as many of the Hickson Compact Groups that are doable from here. It's not a major research project so probably not worth spending any money on.

Again I thank all contributors for their posts.

[Adam J]

7 minutes ago, DaveS said:

To expand on my previous post (From my phone), I have the 533 and the SX 694 in Stellarium and have been playing with the FoV.

However, if the 694 at ~0.46"/px is silly, then the 533 at ~0.38 "/px is downright ludicrous. I could bin it at 2x2, but that would be 0.76"/px, at which point I might as well use the G3 16200 at bin 1 and crop.

In any case, this is just a project to go after as many of the Hickson Compact Groups that are doable from here. It's not a major research project so probably not worth spending any money on.

Again I thank all contributors for their posts.

Yes exactly you can bin the 694 2x2 in hardware. 

Adam

[DaveS]

Yep, that would be a possibility, capture the RGB at bin 2 and 0.92 "/px, and the Luminance at bin 1 0.46"/px to pick up any detail that might be lurking below the bin 2 resolution..

The 533 does have some attractive features, but whether they are £1,149 attractive is another matter.

[andrew s]

I have had had a 694 and now use a 294mm on my ODK16.

I do scientific imaging so always use flats and darks with both.  The only issue with the 294 is you should not use scaled darks. 

On gotchas two come to mind.

1) The old ASI drivers had a memory leak which caused them to hang if you took a lot of images. The good news is it was fixed in the January release of new drivers.

2) the 294mm is prone to the formation of ice crystals on the sensor when cooled to below zero. The fix is to cool it very slowly.  I run at -10 and use a script that cools it over 30mins.

One final point small pixel cameras suffer from enhanced random telegraph noise. The 294mm is no exception.  If you run it at full resolution then a median filter will fix it without reducing the resolution of your images.

Regards Andrew 

Edited February 15, 2023 by andrew s

[DaveS]

Well, the autofocus is now operating as it should, all I need to do is find out what's causing these spurious diffraction spikes and sort them out.

At first I thought they were due to blooming, but the ABG should have taken care of them. As this is from 1+2 hours I think diffraction from a telephone wire would be eliminated by Sigma stacking. I think it has to be something in the imaging train. I checked for spider webs in the OTA but it's clear, and those spikes are suspiciously orthogonal. I will have to take the imaging train apart again to look. It's not on the filter as I can see them in a blue and red image.

It's like playing blummin' whack-a-mole  .

Edited February 21, 2023 by DaveS
Typos

[The Lazy Astronomer]

The 294MM is a great camera, but fair warning: it can be quite finicky to calibrate properly. Sometimes it's a case of throwing salt over your shoulder before embarking on a flats sequence. 🤣

You seem to have been able to make things work with your existing gear (which is great, and obviously the preferable option for anyone), but if you still fancy yourself a 294MM, I'm considering selling mine within the next few months to fund an upgrade to an IMX571 based model. Drop me a message whenever if you can feel holes burning in your pockets! 😁

[CCD-Freak]

Player One just announced that they are coming out with a 294 camera.  I wonder if they have tamed calibration issues ?????

 

[Xilman]

On 15/02/2023 at 11:08, DaveS said:

However, if the 694 at ~0.46"/px is silly, then the 533 at ~0.38 "/px is downright ludicrous. I could bin it at 2x2, but that would be 0.76"/px, at which point I might as well use the G3 16200 at bin 1 and crop.

FWIW, I use a SX 814 Pro, but Terry Platt very kindly loaned me one of his personal 694s while my camera was being repaired --- I am familiar with the 694, in other words.

To the point: the pixel size on my 2614mm focal length scope with the 814 is 0.29"/pixel which might be thought ludicrous. However, the diffraction limit is 0.31" so it is actually a good fit in certain circumstances. I generally bin 2x2, so getting 0.58" per pixel and the seeing is generally in the range 2-3". Still oversampling, you may think. However it has advantages. For a start, the response is linear right up to a few counts short of 64K because a binned count of 64K spread over 4 pixels is only 16K per pixel, well within the linear range. Unbinned, linearity tends to drop off at around 50K. The extra 20% range is well worth it in my opinion.

Secondly, the Nyquist criterion states that at least 2 samples per feature must be taken for it to be detected. Twice 0.58" is a reasonable match for seeing-limited imaging.

Third: seeing need not be a limiting factor. Lucky imaging and speckle imaging monitor the instantaneous condition of the atmosphere, not its time-average. I have used 1ms through 10ms unbinned exposures to freeze the detail of close double stars, the moon and bright planets.

Fourth, and this is closely related to the second, deconvolution in software is also Nyquist limited. If you have a well-sampled point spread function, the image of a bright star for instance, it is amazing what maximum entropy, RL, CLEAN, etc, can do.

Fifth, if you want to do photometry it is generally better to spread the light of a star or asteroid over a number of pixels, thereby averaging out pixel-to-pixel variations in sensitivity as the image moves over the sensor.

Don't discount small pixel sizes, in other words.

Paul

[DaveS]

Thanks Paul.

The 294 is looking increasingly attractive, and would be a better fit for the ODK than the 533, plus it has a wider peak to the  QE curve.

However, the SX 694 has the advantage of being able to bin 2x2 in hardware (If I wanted 0.92 2/px) and only needing Flats and Bias frames.

But the 294 has a significantly higher QE, lower noise (At HGC gain), and wider dynamic range. But like all CMOS is finicky over calibration amd has Amp Glow that needs calibrating out.

Argh!

[Xilman]

I am a confirmed CCD fan for the reasons you give, though darks are needed as well as flats and biases if you want accurately measurable images. All can be taken during daylight or cloudy nights and a library of calibration images lasts for months, IME anyway. Dark response is linear and so dark frames can be scaled to sundry exposure times.

The SX cameras have 16-bit dynamic range, rather better than the CMOS versions with which I am familiar. That said, my knowledge may be seriously out of date.

Unless you live on a mountain top (I don't know Dorset very well but I believe it is fairly low-lying 😉) 0.92"/pix is a good match for the Nyquist criterion for your seeing.

[DaveS]

From my experience, I found that my images calibrated better without Darks, as they left more noise in the final stack then just Bias. I remember in the Trius manual Terry says the same thing.

The next clear night (Yeah, right) i will try getting some Bin 2 RGB data to see how it compares with the Bin1 Liminance I already have.

[Xilman]

Thanks, I will check this again.