Eastern Veil in Ha and OIII with a Nikon D5300

[Xiga]

Hi all

So i've been waiting for a few weeks now to gather some OIII to complete my Veil nebula mosaic. It's been frustrating to say the least, with clouds, the moon, and life stuff getting in the way! I also have a limited timeframe to get it before it sets below the neighbour's house. So in desperation, last Sunday night (Sep 23) there were semi-clear spells forecast, and despite the fact there was a 99% moon about ? i decided to have a go nonetheless. These days, now that we have finally moved into our forever home (astro considerations were not on the shortlist unfortunately!) i now have to travel back to the family home in Crumlin (30 mins each way) to do any imaging. Luckily it takes more than that to put me off, even on a school night! 

I had thought the moon was going to be far enough away that i might be able to get away with it (from memory i think it might have been a little under 90 degrees away). When i saw the subs coming in, and i saw the Median ADU values (which were waaaay higher than normal) i thought the night was going to be a complete write-off. Factor in the clouds, which were annoyingly 'just-intermittent-enough' to affect pretty much every darn sub (they were 20 mins long) and i honestly didn't expect to get anything useful out of the night at all. But lo and behold, once i stacked them in APP i was sufficiently happy enough with what i saw that i thought it could actually be salvageable (especially once i had removed the gradient in the OIII stack). I know it makes sense in theory to just chuck sub-standard data and wait for a better night ( i really wish i had the privilege of doing so!) but living in the UK with a hectic job (and a toddler who calls ALL the shots, lol) i am forced to just take what little i can get. So i was chuffed just to see an OIII stack that could be used, as i was honestly expecting nothing but a smeared mess. 

So this is 15 x 20 mins of Ha and 11 x 20 mins of OIII (8 hrs 40 mins total). Taken with the usual gear: Nikon D5300 (modded); SW 80ED; HEQ5-Pro Mount

Calibrated with Flats & Bias (no darks, no need). Astro Pixel Processor used to calibrate, stack, and do gradient reduction. Photoshop used for everything else. Ha assigned to Red, OIII to Blue, and Carboni's action was used to generate the Green channel. 

As is the norm, i've been looking at it for so long today i've gone blind to it. So i need your input to let me know if i've gone OTT in any way. So all feedback welcome! 

I've attached the full widefield version (the framing of which will make more sense once the full mosaic is complete) and a cropped version.

Thanks for looking! ?

 

[MartinFransson]

Very impressive DSLR-work! Do you have cooling? I used 10 minute subs when I used DSLR, but never tried OIII. Have you tried using twice the amount of 10 minute subs?

[Allinthehead]

Wonderful image Ciarán. Looking forward to seeing the finished mosaic.

[Xiga]

9 hours ago, MartinFransson said:

Very impressive DSLR-work! Do you have cooling? I used 10 minute subs when I used DSLR, but never tried OIII. Have you tried using twice the amount of 10 minute subs?

Thanks Martin! 

No, my D5300 doesn't have any cooling, just a regular Ha astro mod. Thankfully the D5300 has very low thermal noise, which allows me to do long exposures. 

I need to do 20 min long exposures to get sufficient signal to swamp the read noise, so 10 mins wouldn't be enough. I could always increase the ISO instead, but then I would start to lose too much dynamic range, so I tend to just stick to ISO 200. 

[Xiga]

6 hours ago, Allinthehead said:

Wonderful image Ciarán. Looking forward to seeing the finished mosaic.

Thanks Richard. Here's hoping there's a clear night sometime this week, as the moon is on the wane. 

[tooth_dr]

You really have to click on the image to truly apprecaite just how good that is!  Pinpoint coloured stars, smooth background, fine filamentary details.  This is the benchmark for DSLR imaging.

[tomato]

That’s a remarkable result for an uncooled camera, it really stands up to close scrutiny. Well done indeed, a fitting reward for your perseverance!

[MartinFransson]

May I ask where you had your camera modded? I have a friend who wants her Nikon modded.

[Xiga]

7 hours ago, tooth_dr said:

You really have to click on the image to truly apprecaite just how good that is!  Pinpoint coloured stars, smooth background, fine filamentary details.  This is the benchmark for DSLR imaging.

Thanks Adam, you're too kind ☺️ 

You're right about viewing it full size, it does make a difference. I noticed as much while i was processing it, so even though i would still check the image as a whole every now and then, i was mostly focused on how everything looked at 100%. 

[Xiga]

4 hours ago, tomato said:

That’s a remarkable result for an uncooled camera, it really stands up to close scrutiny. Well done indeed, a fitting reward for your perseverance!

Thanks ? ! lol

The D5300 has very low thermal noise, for a DSLR at least, and noticeably less than Canon's models. This certainly helps, and thankfully here in N.Ireland we never have to worry about the high temps that some folks in other parts of the world have to deal with. 

[Xiga]

4 hours ago, MartinFransson said:

May I ask where you had your camera modded? I have a friend who wants her Nikon modded.

Certainly Martin.

I used the camera un-modded for about a year, as i too struggled to find anywhere that would do an astro-mod for a Nikon. I could have sent it to America for the mod, but i simply refused to spend the full cost of the camera again just to get it modded. I'm very much on a budget with my Astro gear, so i waited it out. Eventually i heard about a place in Amsterdam called JTW Astronomy. They did a top-quality job on the modification, i couldn't recommend them enough. My camera sensor was already very clean, nearly spotless in fact, so i was a bit concerned that it might come back to me in worse condition, but it didn't. I even did some test flats afterwards and they looked just as good as before. The guy i was in contact with was called Mark, drop him an email with the camera model and he'll give you a price. The cost is different for each model, and you will have to pay for recorded delivery each way, but even after including this i think the total cost for mine came in at around £175, which was basically half the price i could find anywhere else. Total turnaround time is about a week iirc. 

Hope that helps! 

[Xiga]

I was looking at the widefield version again, and noticed the uneven background colouring in certain places. Most noticeably in the lower right. At the time of the original image, i had thought it was just trace levels of Ha that was causing the excess redness. And no doubt there is definitely traces of very faint Ha regions all throughout the Veil, but on second inspection it's just too distracting to the eye. And in any case, i can't be completely certain that it's not just a gradient, so i ran it through APP to try and neutralise the backgound a bit better (something i didn't spend long enough on in the original). 

I do think it is an improvement, but interested to hear what you guys think. 

 

[tooth_dr]

I agree the background was also a little noticeable to me, and I did consider it was perhaps trace ha etc but it looked a little blotchy. I have to say that’s better now.

This will look epic printed out when finished. 

[alexbb]

Looking good!

[Xiga]

8 hours ago, tooth_dr said:

I agree the background was also a little noticeable to me, and I did consider it was perhaps trace ha etc but it looked a little blotchy. I have to say that’s better now.

This will look epic printed out when finished. 

 

2 hours ago, moise212 said:

Looking good!

Thanks guys.

I did struggle a bit with the sky background colour on this one. It's not easily neutralised at all. I might have one last go to see if it can be further improved.

[souls33k3r]

22 hours ago, Xiga said:

I need to do 20 min long exposures to get sufficient signal to swamp the read noise, so 10 mins wouldn't be enough.

Very impressive image Ciaran. Not only because you taking 20 minutes on a HEQ5 mount but also a cracking image with DSLR. How do you manage to get to 20 min subs? 

What you said above that you need to do 20 minutes exposures to get sufficient signal to swap the read noise. Mind if i ask how do you calculate this? 

 

[Xiga]

10 hours ago, souls33k3r said:

Very impressive image Ciaran. Not only because you taking 20 minutes on a HEQ5 mount but also a cracking image with DSLR. How do you manage to get to 20 min subs? 

What you said above that you need to do 20 minutes exposures to get sufficient signal to swap the read noise. Mind if i ask how do you calculate this? 

 

Thanks! 

I do love my wee HEQ5-Pro. It just does what it says on the tin (ok, box) and doesn't make a fuss about it. So these days i never have to throw away a single 20 min sub due to tracking, just clouds. Of course, my standards are no doubt a lot lower than others who use much more expensive equipment. I've never bothered to inspect my stars to the nth degree. If they look round to my eye without having to get a magnifying glass out, then i'm happy. No doubt they are not perfectly round, but hey, ignorance is bliss, right?! ?

As for how i can do reliable 20 min subs, it's probably a combination of things, the most obvious would be:

1. Imaging with just an 80ED and a finder-guider, the mount isn't carrying too much. Less weight is always a good thing.

2. I'm imaging at a short enough FL of approx 520mm, which is not that demanding.

3. I've done the Rowan Belt Mod to my mount. I never did any testing to try and measure the before-and-after improvement, but the guide graph has definitely been better since the mod was done

4. I replaced the bracket that came with the SW 9x50 finderscope, with a small set of guide rings, which did away with the useless single springy connection point. It does still attach to the 80ED via the same hot-shoe as the original bracket, but i bought a replacement tightening screw off the 'bay for £1. This one has wings, which allows me to really tighten that [removed word] down good and hard. 

5. Cable management. Now i'm not OCD about my cables like some folk are, but i do keep them all bundled together, mainly to speed setting everything up (which i have to do each time). So even though i sometimes just let the bundle hang freely (mainly cos i've forgot) at least i know there's no chance of a cable getting snagged, or dragging. 

6. I use PHD2. It has a neat guiding algorithm called PPEC. After i started using it, i definitely noticed my avg RMS dropping by about 0.1-0.2". These days i usually get about 0.5-0.6" on a very good night, and about 0.8-0.9" on a very bad night. 

 

As for your other question on the read noise, the formula i use is:

Minimum ADU Level per sub should be no less than:  [Offset + (3 * r^2)/e] * (2^16)/(2^bits). Offset is Bias for a DSLR; r is read noise, e is Gain. 

My D5300 has an Offset/Bias of 600 ADU. I use ISO 200 exclusively, and at that ISO the camera is very close to Unity Gain, so approx 1e-. ISO 200 also yields a read noise of about 2.7e-. The D5300 is also a 14bit camera. 

So for me it is (600 + (3 * 2.7^2)/1) * 4 = 2,487 ADU. 

Ideally i would like it to be (600 + (10 * 2.7^2)/1) * 4 = 2,692 ADU.   But that would require slightly longer subs. So I should probably experiment with something like 22 or 23 mins tbh. 

 

[Xiga]

Ok, so i had one final go at cleaning up more of the blotchy red patches of sky background. The difficulty with them is the sheer size, no noise reduction routine was cutting the mustard, so i had to get the hands dirty in PS. 

I first extracted the stars into it's own temporary image, and then i duplicated the original image and completely de-saturated just the reds from it. I then pasted the stars back on top of this de-saturated image, flattened, and then copied it back on top of the original image and set it's blend mode to Color. I checked that the stars were not affected by all of this (they weren't). This gave me a much better sky background colour, but with the stars unchanged (of course the red nebulosity all looked wrong, so that needed fixed). So i did a Stamp Visible to make a new layer with everything from below, and then i deleted the Colour layer, as i didn't need it any more. I finally put the new sky background layer one layer down, and then used the Eraser on the top layer in just on the areas of sky background that i felt needed it. The effect is subtle, but definitely noticeable, especially when the two are compared directly. So it seemed to work ok! 

Edit - also cleaned up the noise in the sky background, and toned down the redness of the background stars, which were far too red. 

So i'm calling these the final final ones. Definitely. ?

[newbie alert]

That a crazy detailed image using a DSLR.. .beautiful...

[Xiga]

17 hours ago, newbie alert said:

That a crazy detailed image using a DSLR.. .beautiful...

Thanks newbie! ?

[Guest]

Fantastic management of the background stars - usually such an issue in this area of the Milky Way.

Really well done.

David

[souls33k3r]

On 03/10/2018 at 02:23, Xiga said:

Thanks! 

I do love my wee HEQ5-Pro. It just does what it says on the tin (ok, box) and doesn't make a fuss about it. So these days i never have to throw away a single 20 min sub due to tracking, just clouds. Of course, my standards are no doubt a lot lower than others who use much more expensive equipment. I've never bothered to inspect my stars to the nth degree. If they look round to my eye without having to get a magnifying glass out, then i'm happy. No doubt they are not perfectly round, but hey, ignorance is bliss, right?! ?

As for how i can do reliable 20 min subs, it's probably a combination of things, the most obvious would be:

1. Imaging with just an 80ED and a finder-guider, the mount isn't carrying too much. Less weight is always a good thing.

2. I'm imaging at a short enough FL of approx 520mm, which is not that demanding.

3. I've done the Rowan Belt Mod to my mount. I never did any testing to try and measure the before-and-after improvement, but the guide graph has definitely been better since the mod was done

4. I replaced the bracket that came with the SW 9x50 finderscope, with a small set of guide rings, which did away with the useless single springy connection point. It does still attach to the 80ED via the same hot-shoe as the original bracket, but i bought a replacement tightening screw off the 'bay for £1. This one has wings, which allows me to really tighten that [removed word] down good and hard. 

5. Cable management. Now i'm not OCD about my cables like some folk are, but i do keep them all bundled together, mainly to speed setting everything up (which i have to do each time). So even though i sometimes just let the bundle hang freely (mainly cos i've forgot) at least i know there's no chance of a cable getting snagged, or dragging. 

6. I use PHD2. It has a neat guiding algorithm called PPEC. After i started using it, i definitely noticed my avg RMS dropping by about 0.1-0.2". These days i usually get about 0.5-0.6" on a very good night, and about 0.8-0.9" on a very bad night. 

 

As for your other question on the read noise, the formula i use is:

Minimum ADU Level per sub should be no less than:  [Offset + (3 * r^2)/e] * (2^16)/(2^bits). Offset is Bias for a DSLR; r is read noise, e is Gain. 

My D5300 has an Offset/Bias of 600 ADU. I use ISO 200 exclusively, and at that ISO the camera is very close to Unity Gain, so approx 1e-. ISO 200 also yields a read noise of about 2.7e-. The D5300 is also a 14bit camera. 

So for me it is (600 + (3 * 2.7^2)/1) * 4 = 2,487 ADU. 

Ideally i would like it to be (600 + (10 * 2.7^2)/1) * 4 = 2,692 ADU.   But that would require slightly longer subs. So I should probably experiment with something like 22 or 23 mins tbh. 

 

Thank you so much for your response Ciaran. 

So brilliantly explained.

There is however one question regarding your formula, where is this "3" coming from in your formula that you ideally like it to be "10"?

[johngm]

Hi Ciarán. 

First of all, a fantastic image, and as a previous poster has said, you certainly have set a benchmark for DSLR imaging ?. 

I have an astro modified Canon 600D, (and an ED80 to) and have recently purchased an Astronomic 12nm Ha Clip filter, and I am thinking of buying the OIII version. What filters were you using ? With the Winter Constellations rising, I would like to attack the Rosset Nebula using this Method.

Best regards and clear sky's 

John

[Xiga]

On ‎04‎/‎10‎/‎2018 at 09:05, David_L said:

Fantastic management of the background stars - usually such an issue in this area of the Milky Way.

Really well done.

David

Thanks David. You're right about the sheer number of stars. Deciding on the right amount of star reduction was tricky for sure.

22 hours ago, souls33k3r said:

Thank you so much for your response Ciaran. 

So brilliantly explained.

There is however one question regarding your formula, where is this "3" coming from in your formula that you ideally like it to be "10"?

That's a good question. The answer, simply, is they are taken from experienced imagers far smarter than me! Lol. The bottom one is just the bare minimum you should aim for, so you really want to be further towards to the top end value. Using a factor of 10 for the top-end is probably slightly overkill, so something like 7 should work fine too. Although for reference, as outlined above, in pretty much all of my images where I have used the Ha filter, I am just about hitting the bare minimum ADU values.

[Xiga]

2 hours ago, johngm said:

Hi Ciarán. 

First of all, a fantastic image, and as a previous poster has said, you certainly have set a benchmark for DSLR imaging ?. 

I have an astro modified Canon 600D, (and an ED80 to) and have recently purchased an Astronomic 12nm Ha Clip filter, and I am thinking of buying the OIII version. What filters were you using ? With the Winter Constellations rising, I would like to attack the Rosset Nebula using this Method.

Best regards and clear sky's 

John

Thanks John! ? If things go according to plan, I might get a chance this Sat night to get the OIII for the Western Veil too.

I'm using 2" Baader mounted filters which I got from FLO, link below. They are 7nm for Ha and 8.5nm for the OIII. I'm very happy with how they perform.

https://www.firstlightoptics.com/narrowband/baader-narrowband-ccd-emission-line-h-alpha-filters-2.html

ps - The Rosette makes for a great target with a modified DSLR. It's nice and bright so there's no problem picking up loads of Ha and OIII. I captured it myself last winter and was very pleased how it came out (so much so I used it as my profile pic). Best of luck!