Heart and Soul in Ha

[Gina]

Thank you   Possibly.

[happy-kat]

Love this image and using the vintage lens. Can I request the jelly fish at some point please

[Knight of Clear Skies]

If I zoom in I can see triangular stars in the corners but its good enough to fill my screen without them being a distraction. In a way it would be a shame to stop it down and lose that f2.5 magic. I have the same lens and got triangles as well when I tried it wide open.

There is a slightly later 6-element version of this lens (SMC version 2, also available on a Pentax-K mount) that supposedly gives better corner stars but they are quite difficult to find. They are also difficult to tell apart from the version 1, the later one has 8 instead of 6 aperture blades and slightly different markings on the aperture ring.

[Chris]

3 hours ago, ollypenrice said:

To quote the Nobel Laureate, The Times They Are 'A Changin'...

Your numbers are interesting. F2.8, nearly 200 subs, exposures of just a minute. Result convincing. I wandered over and had a word with a couple of mates, Brontosaurus, Pterodactyl and Triceratops. After a few beers we concluded that we dinosaurs may not be long for this world!

Olly

Maybe you dinosaurs can climb aboard the CMOS train and wave good bye out the back of the carriage to guiding  

[Gina]

1 hour ago, happy-kat said:

Love this image and using the vintage lens. Can I request the jelly fish at some point please

Thank you Yes, I'll see what I can do 

[Ouroboros]

6 hours ago, Gina said:

The noise is random and accumulates as the square root of the number of subs whereas the signal is purely additive.  This is the maths of the matter.  The result is that the signal is all added up resulting in N times the single sub for N subs whereas the noise becomes √N.  When the final image is normalised by dividing by N the signal is 1 whereas the noise is reduced by √N to 1/√N. 

Putting numbers into this algebra :- if the number of subs N is 100 the signal accumulates to 100/100 whereas the noise becomes 10/100 (√100 = 10) thus the noise is reduced to a tenth of what it was in a single sub.  Hope that makes sense.

Yes, I'm familiar with the maths. It's the read noise that benefits by the the root √N isn't it. Truly random dark current is the same whether it's a 10 x 60s exposures or 1 x 600s exposure.  I guess the high QE also allows for shorter exposures before the sky glow limit kicks in.  I see there has been a lot of chatter about this on this site and elsewhere. I'm just coming to it a bit late. Thanks for drawing my attention to it.

[giorgio_ne]

Beautiful image Gina! Could you please post for us a picture of your nice widefield rig? Are you using motorised focus? How are you fixing the lens on the mount?

[Uranium235]

There really is some spiffing images coming from 135mm lenses at the moment.... this included.

[Stub Mandrel]

7 hours ago, giorgio_ne said:

Beautiful image Gina! Could you please post for us a picture of your nice widefield rig? Are you using motorised focus? How are you fixing the lens on the mount?

I have taken up your radioactive gauntlet! It took me about five attempts to get good farming (set up on gamma Cass then slew...) but I appear to have heart, soul and double cluster all on the sensor using my Zeiss Sonnar 135mm 3.5

[StuartJPP]

Good work Gina...

[Knight of Clear Skies]

I'm really impressed by this image, I wouldn't have thought it possible to get this much detail out of a 135mm lens. It's got me wondering what I could achieve with my own equipment. I'm pleased with my recent Heart & Soul but it's a sketch by comparison, there is just no definition in the nebulosity. Looking at the quoted pixel sizes on my DSLR they aren't much larger than those on the  ZWO ASI1600MM-Cool, but does this make allowances for the bayer filter? Should I be multiplying the DSLR pixel size by 4 for a real comparison? This would help me work out what the limiting factor is in my setup. I tend to get a sub-pixel to actual pixel trailing in my unguided subs, which would tend to blur out small features, but with better tracking how much closer could I get to the above? I suspect I'd need a lot more data to exploit the resolution of the sensor fully, I usually downsample my images quite heavily for presentation.

[Gina]

Yes, the real resolution is the quad of four pixels (unless you've debayered it).  Bear in mind that the ASI1600MM-Cool is a very different beast from a DSLR even if it does have the same type of sensor.  These are specially developed sensors designed for very low read noise and also with the dual stage Peltier TEC cooling the thermal noise is vastly lower.  Cooling down to -30°C makes a huge difference.  The other thing that makes a big difference is using the very narrow 3nm narrowband filters.  Apart from the lens (a great bargain considering the quality of the glass, secondhand from the well known auction site) the rest of the gear is pretty expensive and built up over the years.

A DSLR is seriously limited for astro use.  The main nebulosity is from hydrogen gas which emits in the deep red where only one of the four pixels in the quad contribute to the image.  Also the Bayer filter further reduces the light getting through.

[Knight of Clear Skies]

I'm not expecting to match your effort with my rig Gina, just wondering how much further I could push it.  Also, further down the line I'd certainly consider upgrading to a setup like yours.

Checking the math, it does seem that the quoted pixel sizes for Canon sensors do take the bayer matrix into account, so the 5.2 µm for my 1100D is for a colour pixel. Seems I've partially answered my own question, I'm limited more by lack of data and perhaps tracking than by the pixel size. Stretching harder and looking at my image at 100% there is a fair bit of grain in the nebulosity, so there is plenty to be gained by simply taking more subs.

Unfortunately debayering isn't an option to me due to sheer terror, so I'm stuck with Ha responsiveness over 1/4 of my sensor. Throwing a fast lens on the front does help to compensate somewhat. Despite all its limitations (lower QE, bayer matrix, lack of cooling, well depth) a DSLR is still quite a capable instrument.

[Gina]

Yes, you can do a lot with a DSLR - I didn't mean to dismiss them out of hand for astro imaging - I did a lot of astro imaging with a DSLR in the past.  If you're good with your hands and technically fairly proficient, cold-finger cooling with a Peltier TEC to 0°C (or even just to ambient) can help with the noise.  Removing the red-cut filter helps a lot too.  A decent mount and guiding helps with longer exposures.  Yes, you need lots and lots of data and I'm talking about several night's worth - umpteen hours.

[Knight of Clear Skies]

No offence taken, I'm mindful of my camera's limitations but I expect E. E. Barnard would have thought it the work of the devil. I'll be watching out for your results with interest. 

[Gina]

I have a set of 330 60s OIII subs of the Heart and Soul - some good, some pretty mediocre - which I'm about to process.  I had over 500 from last night with lots of duds due to patchy cloud and then the breaking dawn before the run completed.  I left it running overnight.  Going through all those one at a time was pretty tedious but certainly helped by Blink in PixInsight which auto-stretches each image for display without altering the file.