Eta Carinae, as captured by an unguided el-cheapo Nikon DSLR atop an 8" f/5 Newt. 90 x 30-second subs of three different ISOs (800, 1600, 3200).
My experience with autoguiding has sucked. Not sure whether the problem is my (very old) laptop or the fact I'm using a far-too-big finderscope (600mm refractor). Or quite possibly something else entirely. I fumbled with it for several months late last year and gave up since 2017 began. I've just been experimenting with processing larger stacks of shorter subs since then.
My best results seem to come from several stacks of different ISOs into several final TIFs, stacking those again, and editing in Lightroom. Occasionally I'll combine several of those different TIFs in Photoshop, but I try to avoid that.
I've come to the conclusion that darks, flats and bias do nothing good for my images. I've tried them all, many many times, in different combinations, and the results are simply better without them literally every single time.
Thoughts, suggestions? Has anyone else had similar experiences with DSLRs, ie discovered that you're better off just using lights on their own?
Hi Tony, I had the same issue with guiding not happening time and time again, but after I got a 50mm finder guider everything started working. I still have the occasional night when nothing goes right but I have managed to get some good 5min subs at 1200mm and 1350mm fl. I think that with longer focal length guide scopes the guider just works too hard?
Darks just help to remove hot pixels, so if your sensor was really good you might not notice the difference, I can certainly notice the difference with my 1100d. Flats help to eliminate dust etc in your imaging train that shows up in your pics. I have never bothered to take any flats because you have to take them with everything set up on your scope exactly as you have been imaging and with a pure white light source.
Rick
Chris, that's a nice rendition Eta Carina, very psychedelic. Were you using narrow band filters or just weaved some magic in Lightroom? I like the effect. Regarding flats and Darks, I don't always use them, it depends on the quality I'm after and the conditions I was photographing in. Darks definitely remove hot pixels and flats help with vingnetting and unclean camera sensors, so for me they do improve the overall result.
Its really a matter of how pedantic you want to be. If you're happy with just capturing some memories of the night's session for viewing on your mobile phone then yes, you're right, darks and flats aren't necessary. I know what you mean though, sometimes they don't seem to make an ounce of difference and if flats aren't made correctly, then you're better off without them.
Now I am not an imager, just an occasional viewer, but I find that this thread is one of the most encouraging to date.
When you read almost daily of the effort seemingly required and the expensive gear needed and that you need to do this that and the other to get results, well I might just get a little, just a little, interested in exploring imaging.
You've got some really nice detail in the centre so those 30s subs look to be working quite well. Also look at what Kevin (cometcatcher) is doing with his DSLR, he spent years photographing out of his kitchen window with a HEQ5 and 8" newt Hundreds of 30s unguided exposures.
As has been mentioned, the amount of calibration needed can differ. On a bright target with longer focal length I could probably get away without darks on my DSLR but when stretching to show up the fainter stuff I get quite a bit of background banding, this is what the darks remove. Some cameras show this up more than others.
As for flats, if your sensor is clean (not a lot of dust motes) then flats are mostly just for fixing vignetting which purely allows you to accurately stretch out signal in the corners of your frame. If you don't mind that it isn't technically accurate in the corners then it isn't too bad
To sum it all up, if you're not wanting to push your system to the edge of its limits, extracting every whisper of faint signal and having everything as technically accurate as possible, then yes, you can get away without any calibration
To give an actual example, both of these images have been processed in exactly the same one. One is fully calibrated and the other isn't. After stacking both have had their black point set, colour calibrated, stretched and hit with the same curves/saturation change.
I am lucky in that I only get 3-4% vignetting in the corner of my images so it isn't noticable. it is also a high enough signal area where the banding that I mentioned earlier is buried until strong SNR.
These are both 36x180s shots at ISO800 taken 5 days ago.
So in answer to your original "question", yes, you can still do a lot without any calibration whatsoever I do dither ~3 pixels between each frame though which really helps mitigate some issues that comes from not calibrating.
Hi Tony, I had the same issue with guiding not happening time and time again, but after I got a 50mm finder guider everything started working. I still have the occasional night when nothing goes right but I have managed to get some good 5min subs at 1200mm and 1350mm fl. I think that with longer focal length guide scopes the guider just works too hard?
Darks just help to remove hot pixels, so if your sensor was really good you might not notice the difference, I can certainly notice the difference with my 1100d. Flats help to eliminate dust etc in your imaging train that shows up in your pics. I have never bothered to take any flats because you have to take them with everything set up on your scope exactly as you have been imaging and with a pure white light source.
Rick
Yeah, flats - I get what they're supposed to do, but they just don't work for me. I've tried a sunny sky with a white shirt stretched across the OTA for flats, but nope - ruined the image something horrible. Tried even & odd numbers of them, big numbers, small numbers, slightly overexposed, slightly underexposed etc to no avail.
In any case, I'm so lazy with regards to collimation that the edges are all smeared from hell to breakfast anyway so I crop any vignetting out! I'm shooting 24MP and for widefield I've got my trusty ED80 anyway so no great loss. Just weird how they always destroy my images rather than improve them. Maybe it's the light source; maybe it's something weird I'm doing. No idea!
Chris, that's a nice rendition Eta Carina, very psychedelic. Were you using narrow band filters or just weaved some magic in Lightroom? I like the effect.
Thanks! No narrowband, all just the little Nikon. Its colour rendition isn't great, unfortunately. I've tried with vastly superior full-frame Nikons and the difference is dramatic, but I need the extra magnification - sorry, sorry, crop factor (don't want to get sucked back into that pointless debate) - so yeah the colours are often a bit weird. I like it in this instance though!
Quote:
Originally Posted by Mickoid
If you're happy with just capturing some memories of the night's session for viewing on your mobile phone then yes, you're right, darks and flats aren't necessary.
Are there people on this forum doing it for any other reason than that??? That's literally exactly why I do it! Well, that and the wonderful feeling I get thinking about a star that's 100 times the size and five million times the brightness of our feeble little Sun, and how it'll explode one day and burn brighter than the combined light of the other 100,000,000,000 stars in the galaxy.
I always use darks. Remember you have to temperature match them.
I used to always use them; I'd set the camera taking them immediately following a session as I packed up so they'd be as fresh as possible (for lack of a better term). They did sort of help with the noise and black point, but not significantly; one minute of tweaking in Lightroom created the same benefit. I'm wondering if a lot of the lack of improvement is because I'm using DSS instead of more advanced software that can make better use of the extra data...?
Now I am not an imager, just an occasional viewer, but I find that this thread is one of the most encouraging to date.
When you read almost daily of the effort seemingly required and the expensive gear needed and that you need to do this that and the other to get results, well I might just get a little, just a little, interested in exploring imaging.
Thanks to op and responders.
Thank you! I had a fairly successful musical / singing / performing career, very often focused around using what I had as well as I could rather than mourning what I didn't have. I'm applying the same principal here, I guess. Bad workman blames his tools etc.
You've got some really nice detail in the centre so those 30s subs look to be working quite well. Also look at what Kevin (cometcatcher) is doing with his DSLR, he spent years photographing out of his kitchen window with a HEQ5 and 8" newt Hundreds of 30s unguided exposures.
As has been mentioned, the amount of calibration needed can differ. On a bright target with longer focal length I could probably get away without darks on my DSLR but when stretching to show up the fainter stuff I get quite a bit of background banding, this is what the darks remove. Some cameras show this up more than others.
As for flats, if your sensor is clean (not a lot of dust motes) then flats are mostly just for fixing vignetting which purely allows you to accurately stretch out signal in the corners of your frame. If you don't mind that it isn't technically accurate in the corners then it isn't too bad
To sum it all up, if you're not wanting to push your system to the edge of its limits, extracting every whisper of faint signal and having everything as technically accurate as possible, then yes, you can get away without any calibration
To give an actual example, both of these images have been processed in exactly the same one. One is fully calibrated and the other isn't. After stacking both have had their black point set, colour calibrated, stretched and hit with the same curves/saturation change.
I am lucky in that I only get 3-4% vignetting in the corner of my images so it isn't noticable. it is also a high enough signal area where the banding that I mentioned earlier is buried until strong SNR.
These are both 36x180s shots at ISO800 taken 5 days ago.
So in answer to your original "question", yes, you can still do a lot without any calibration whatsoever I do dither ~3 pixels between each frame though which really helps mitigate some issues that comes from not calibrating.
Hmmm... Not a lot of difference there! Thanks for the demo.
Having recently changed careers - and having two small girls under 4 - I'm pretty time poor, so currently enjoying the quick easy results of stacking short exposures. Such an approach lends itself nicely to my camera too. I'll get back to the more challenging targets eventually and will likely get back into proper calibration.
Hi Chris, I feel your disappointment with the calibration failing you however be sure that once you master it you will be greatly impressed with the improvement. I had plenty of trouble implementing flats but after researching and experimenting I found the right mix. Firstly I made a light box producing a nice even light suitable for the task and the key for me was firstly not to move the camera from its imaging position( this enabled proper orientation for the flats to be applied) and equally important was to ensure the histogram was around 40-50%. I had to try different exposures to achieve this. With these two aspects done correctly you will see the benefit of applying flats.
You most probably know these tips anyhow, if so I wish you better success with them. Great images none the less.
Anth
Following this thread with interest as apart from my efforts to get as much detail on my sensor as possible, I'm then faced with the challenge of processing it to present a true result.
But just what is the definitive outcome? Is there one? Post processing is highly subjective and I guess the end result is a balance between reality and personal appeasement.
My main photographic interest to date has been birding and trying to capture all the minute feather detail, and I'm hoping that a few of the things that I've learnt from that can help me with this new challenge.
Chris, I'd be pretty well chuffed with what you are producing and can tell you that it is as good as anything I've seen with similar, and even much more expensive gear. As I said above, PP is very subjective and unless someone else is paying you for your work, the only person you have to please is yourself.
Colin, I hope you don't mind me using one of your posted shots to further my 'education' and would welcome your comments on it.
I've downloaded your #1 shot above that I assume was the non-calibrated one (not even sure what calibration means in this scenario) and had a fiddle with it. All I've done is some adjustments in CS6 with the Gamma and De-haze tools, a slight boost of the Yellow with the White Balance tool and a mild sharpen. No other colour adjustments were made.
To my tired old untrained eyes I find the extra detail in the core quite appealing, however in my ignorance I may have missed the point of the exercise in removing it's milky haze.
. I assume you are referring to your D3300. Have you had it modded?
I'm in the market for a FF Nikon to have modded and I've pretty much decided on a D600/D610. I had a D800 but was less than enamoured with it's low light performance as I found anything at ISO800+ yielded noise in dark backgrounds, so not so good for shooting birds in shady environments. The D800E/D810 seem much better in that regard, possibly due to the removal of the OLPF.
I'd recently done the below set of calcs and added in your D3300 for comparison. Interestingly the D3300 had the highest pixel count per cm². The trade-off was that it had the smallest pixel size, so less light gathering ability per pixel. It seems to me to be a 'swings and roundabouts' scenario whether more detail is captured with more pixels, or more light.
In theory, in good light, the D3300 should out-resolve the D810, all other factors, such as glass, being equal. It would be nice to be able to test the theory. I suspect that the D3300 has the same PCB as some of it's more illustrious peers, with certain functions disabled.
If you peruse the figures below you will note that in DX mode the D810 has an acceptable pixel count of 4.22 MP/cm² but falls below the D600/D750 in pixel size.
At this point I should ask if you've tried your D750 in crop mode. You may get a pleasant surprise. It would be interesting to see how it stacks up against the D3300.
D3300 - DX Sensor 23.5mm x 15.6mm 366.6 mm
Total Megapixels 24.2MP
Pixels per sq cm 6.62 MP/cm²
Pixel Area 15.13 µm²
D750 - FX Mode 35.9mm x 24mm 861.6 mm²
Total Megapixels 24.93MP
Pixels per sq cm 2.8 MP/cm²
Pixel Area 35.4 µm²
D750 - DX Mode 23.5mm x 15.7mm 368.95 mm²
Total Megapixels 10.3MP
Pixels per sq cm 2.8 MP/cm²
Pixel Area 35.4 µm²
D8XX - FX Mode 35.9mm x 24mm 861.6 mm²
Total Megapixels 37.09MP
Pixels per cm² 4.22 MP/cm²
Pixel Area 23.72 µm²
D8XX - DX Mode 23.5mm x 15.7mm 368.95 mm²
Total Megapixels 15.4MP
Pixels per cm² 4.22 MP/cm²
Pixel Area 23.72 µm²
Numbers are just that, numbers, and how they relate to reality is quite often a different beast entirely. I treat them purely as a possible indicator.
I think you are doing just fine with your . I assume you are referring to your D3300. Have you had it modded?
I'm in the market for a FF Nikon to have modded and I've pretty much decided on a D600/D610.
At this point I should ask if you've tried your D750 in crop mode. You may get a pleasant surprise. It would be interesting to see how it stacks up against the D3300.
Hi Kevin,
Nope - the D3300 is unmodded. I considered it for a while but then we started enjoying using it with the kit lens as a simple knock-around point & shoot for the beach / back yard (instead of the D750 bodies my wife uses for her studio). For the price it's actually surprisingly good, particularly with good lenses. It's also fine for AP because I don't feel nervous leaving it in the dew for hours or knocking it around in the dark (which I manage to do worryingly often), and the crop factor is very useful over full frames for AP. Having said that, I MUCH prefer full-frame for regular photography.
The D600 isn't bad... but even after 4 years owning one I still haven't gotten used to the stupidly small AF sensor size. Just such a sensationally dumb decision. The low-light is pretty good, yeah, but to be honest it's only about a stop better than the D3300. Better dynamic range and colour, though. And a few other benefits of course. Vastly better for video if you ever use that.
If you're looking for a new full-frame, OMG OMG get the D750. Save for an extra month or two. You won't regret it. It's SO good. I know a few photographers who were only semi in love with their D810, switched to the D750 and never looked back. We absolutely love ours. We loved our old D700s, but the D750 is easily our favourite. The low-light performance is insane. I've only used it once or twice for astrophotography (my wife uses it nearly daily and is reluctant for me to leave it outside in the dew for hours so I kinda have to steal it from her studio) but the colour rendition & low light performance is streets ahead of the D600.
Ah, see what happens when you get me talking about cameras!
Getting back to your shot of Eta Carina above, it would be interesting to see what the EXIF data shows for your White Balance setting.
re the D600/D610/D750 they all share the same Sony sensor, the difference being the D750 has the Exceed 4 Processing engine.
For my intended use I think that a modded D600/D610 will be fine as I will be focusing manually via a hook-up to my laptop, and I believe it's high ISO capability will be adequate for my needs.
When I switched from Pentax to Nikon I started with a D600 but in the quest for more fine detail for birding I was seduced by the D800. I've since replaced that with a D7200 which I'll probably hang onto as it's a great camera. The main reason for the D600/D610 is the extra real estate that the full frame sensor offers, particularly desirable for Milky Way shots.
If you can find the time I'd be interested to see a set of Eta Carina shots with your D600 in DX mode to see how it compares with the D3300.
Cannot have been easy doing any form of processing on my shot Kev, down sizing from 150MB down to <200kb has its issues
I think it was the first that was not calibrated, just debayered, registered and then stacked. I calibrate my flats with a bias then stack the calibrated flats. Then calibrate the lights with the master flat and matched darks. Then I debater, register and stack
The main difference between my two shots is that there is a visible dust mote around the centre left
Thanks for the explanation Col, although at this stage of my journey my level of understanding is probably on a par with a preschoolers grasp of Einstein's Theory of Relativity.
I've flicked between the shots looking for the 'dust mote' and am not at all sure I can see it.
And of course 'seeing' is not only a variable when gazing toward the heavens, it's also a factor in what we see on our different monitors.
I'm guessing that the translucent haze I removed is actually an integral part of the image.
Hundreds of 30s unguided exposures.
