January 28, 2015, 01:22:02 PM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - jrista

Pages: 1 [2] 3 4 ... 326
16
Landscape / Re: Deep Sky Astrophotography
« on: January 07, 2015, 11:45:37 PM »
Stacking is increasing both read and photon shot noise:


TotalSNR = S/N = (S * n)/SQRT(n * (S + R^2))


If you are imaging under light polluted skies, then the light pollution is added into the signal:


TotalS = ObjS + SkyFogS


So, yes, the background sky is going to increase in level, and it is going to increase in noise. That's not a huge deal, you simply offset to bypass the skyfog. Problem is, that eliminates the part of the signal from light pollution, but it leaves behind the added photon shot noise that additional part of the signal introduced into the total signal. You tend to get more photons from light pollution than from the object in or near the city. So for any given pixel at 1/3rd histogram (the recommended exposure level when using a DSLR), you might have, on average, 100 photons from skyfog and 10 from your object. Your background sky IS going to be fairly bright.


When processing, you can simply offset to reduce the background sky back to an acceptable level. That is easily done with the Levels tool in Photoshop. After stretching, you will have a lot of noise to contend with. You can either use more subs to solve that problem, or simply use more advanced noise reduction tools (PixInsight is packed with them), and get extremely skilled at using them.


Once you offset for skyfog, the tail should still stand out. You can use a more aggressive stretch to bring it out more, but again, that is going to reveal more noise. The best solution for that is to shoot at a dark site. See my previous answer about imaging with light pollution for the reasons why.


I can still give processing a try. I can figure out how to bring out the image details, and share my steps with you.

17
This is a HiFi Audiophile product. For people who spend two hundred and fifty grand on a pair of speakers to go with their three hundred grand fully analog, vacuum tube based audio system that plays records from a turntable separated by five or six degrees from all possible vibration sources. :P $1200 is a drop in the bucket.


(BTW...go listen to an honest to god purely analog (i.e. records, not CD or digital in any way) high fidelity audio system in one of those now rare stores that still sells that stuff...the pure, unadulterated quality and absence of system noise is utterly mind blowing. I'd love a proper hifi system, but yeah...$1200 is NOTHIN!)

18
Landscape / Re: Deep Sky Astrophotography
« on: January 07, 2015, 09:19:45 PM »
I don't know if it applies to RAW or not. All I know is that's what Roger Clark uses, so maybe he stacks JPEGs. Regardless, in camera, or with ACR/LR, it is not the best way to go about correcting your field, not for astrophotography. It may be simple, but it is going to diminish the quality of your results.


Generate and use a proper master bias, master dark, and master flat for the most precise results. To really get the best results possible, use PixInsight to do everything...calibrate, register, integrate, and process.

19
Landscape / Re: Deep Sky Astrophotography
« on: January 07, 2015, 07:32:25 PM »
Most of the references to lens profile correction are done in-camera. For example, Roger Clark uses the built-in camera dark subtraction and lens corrections to avoid having to calibrate his subs. That is effective, to a degree. It does not produce the best results. In-camera processing power is limited, so the algorithms are lower precision. In-camera processing isn't going to use the more advanced algorithms we have today to optimally calibrate your light frames either. Plus, single-frame dark subtraction can fix hot pixels, but it tends to increase random noise.


I still recommend generating and using a proper flat, and either dithering or using a proper master dark, for calibration. None of the astro integration tools support lens profiles or anything like that. So if you didn't take the frames with in-camera calibration on, then you can't do it after the fact.



20
Landscape / Re: Deep Sky Astrophotography
« on: January 06, 2015, 11:55:40 PM »
Dropbox, Microsoft OneDrive, Box, etc. Tons of free online space. Gigs of it. I use OneDrive myself (I can just drag and drop into my OneDrive area in Windows, and it automatically syncs to the cloud drive.)

21
Landscape / Re: Deep Sky Astrophotography
« on: January 06, 2015, 11:34:40 PM »
Nice comet, East Wind! I haven't had the opportunity to image that yet (weather :'( ).


How are you stacking? There are some specific techniques to stack the comet separate from the stars, stack the stars separate from the comet, then combine the two. DSS can actually do it for you, it's decent. PixInsight has comet stacking capabilities as well...more manual, more complex, but the results can be amazing.


I would download DSS (DeepSkyStacker, free) and try that first.


@dcm: Hubble stuff puts most ground-based astrophotographer's work to shame. Having no atmosphere to contend with is a HUGE bonus for Hubble...it can resolve an incredible amount of detail. Seeing is the bane of all earth-bound imagers, although with cameras like the A7s, which is so incredibly sensitive, we may be able to employ lucky imaging techniques to solve that problem within the next few years. Lucky imaging (high speed imaging, allowing you to take tens of thousands to hundreds of thousands of frames, then discard any that aren't near-perfect, integrating only the best ones), combined with adaptive/active optics, is how the new Thirty-meter and Forty-meter telescopes from ESO will resolve more detail than Hubble (by a lot.) There are some adaptive optics options for ground-based imagers...their effectiveness has never been fully verified...but combined with lucky imaging, ground-based imagers with 16-32" scopes could produce some amazing results, for sure.

Yeah I have been working with DSS.  Got a nice stack.  background is bright due to the moon that night.  The issue is in processing using photoshop and trying to stretch the tail out of the background.  When I get done with the editing it looks like about 8 shades of grey and I give up to try again another time.  I'm certainly obviously missing something.


You want me to give processing the data a try?

Its a lot of data.  need to figure out where to stash it for you.  The TIFFs from DSS about about 236MB each.


How many files are you getting from DSS? If you use comet stacking, you should have just one...


DSS may not be saving the TIFF compressed, either. And, for best results, if you are saving TIFF, you probably just want to save it as 16-bit integer. I'd reintegrate and save to 16-bit TIFF, then open in PS, save it out again, and choose ZIP compression. That should reduce the file size.


I only need one integration, whichever one is best.


I also have support for FITS editing, and 32-bit FITS files are usually much better. I can load that into PixInsight. Either way, I still only need just one.

22
Landscape / Re: Deep Sky Astrophotography
« on: January 06, 2015, 11:22:41 PM »
Nice comet, East Wind! I haven't had the opportunity to image that yet (weather :'( ).


How are you stacking? There are some specific techniques to stack the comet separate from the stars, stack the stars separate from the comet, then combine the two. DSS can actually do it for you, it's decent. PixInsight has comet stacking capabilities as well...more manual, more complex, but the results can be amazing.


I would download DSS (DeepSkyStacker, free) and try that first.


@dcm: Hubble stuff puts most ground-based astrophotographer's work to shame. Having no atmosphere to contend with is a HUGE bonus for Hubble...it can resolve an incredible amount of detail. Seeing is the bane of all earth-bound imagers, although with cameras like the A7s, which is so incredibly sensitive, we may be able to employ lucky imaging techniques to solve that problem within the next few years. Lucky imaging (high speed imaging, allowing you to take tens of thousands to hundreds of thousands of frames, then discard any that aren't near-perfect, integrating only the best ones), combined with adaptive/active optics, is how the new Thirty-meter and Forty-meter telescopes from ESO will resolve more detail than Hubble (by a lot.) There are some adaptive optics options for ground-based imagers...their effectiveness has never been fully verified...but combined with lucky imaging, ground-based imagers with 16-32" scopes could produce some amazing results, for sure.

Yeah I have been working with DSS.  Got a nice stack.  background is bright due to the moon that night.  The issue is in processing using photoshop and trying to stretch the tail out of the background.  When I get done with the editing it looks like about 8 shades of grey and I give up to try again another time.  I'm certainly obviously missing something.


You want me to give processing the data a try?

23
Landscape / Re: Deep Sky Astrophotography
« on: January 06, 2015, 10:12:54 PM »
Nice comet, East Wind! I haven't had the opportunity to image that yet (weather :'().


How are you stacking? There are some specific techniques to stack the comet separate from the stars, stack the stars separate from the comet, then combine the two. DSS can actually do it for you, it's decent. PixInsight has comet stacking capabilities as well...more manual, more complex, but the results can be amazing.


I would download DSS (DeepSkyStacker, free) and try that first.


@dcm: Hubble stuff puts most ground-based astrophotographer's work to shame. Having no atmosphere to contend with is a HUGE bonus for Hubble...it can resolve an incredible amount of detail. Seeing is the bane of all earth-bound imagers, although with cameras like the A7s, which is so incredibly sensitive, we may be able to employ lucky imaging techniques to solve that problem within the next few years. Lucky imaging (high speed imaging, allowing you to take tens of thousands to hundreds of thousands of frames, then discard any that aren't near-perfect, integrating only the best ones), combined with adaptive/active optics, is how the new Thirty-meter and Forty-meter telescopes from ESO will resolve more detail than Hubble (by a lot.) There are some adaptive optics options for ground-based imagers...their effectiveness has never been fully verified...but combined with lucky imaging, ground-based imagers with 16-32" scopes could produce some amazing results, for sure.

24
Lenses / Re: Lens 'resolving power' vs sensors.
« on: January 05, 2015, 09:21:40 PM »
Happy to help. :) It's just a hobby, really...well, a hobby that assists me in my other hobby, really. :D I like to know everything about what I do, so as a photographer, well, I had to know how sensors and lenses worked. So I researched it.

Just wanted to point out that the lp/mm for each sensor posted way above are only correct if you have monochrome sensors with tiny pixels (low fill factor) without microlenses and without an AA filter.  If you have a Bayer sensor with real pixels and an AA filter, you're going to need at least 2.5 and maybe closer to 3 pixels per line pair.


If you actually read my post, I stated as much.

You made a passing reference to this in a different post and then posted wrong numbers and proceeded to reference those wrong numbers.  If you knew this, why not just use more correct numbers?


What wrong numbers? I stated that they assumed a monochrome sensor. Why did I assume monochrome, rather than something else? Because there are a lot of factors that greatly complicate "reality", and I preferred to keep things simple, as not everyone is a mathematician or an engineer. It's easy enough to approximate after the fact to account for other blurring factors. You may care that everything is 100% perfectly exact...most people don't. And when my goal is just to get across the concept, absolute exactitude doesn't matter.


BTW, I did not make a passing reference to the fact that I assumed mono in another post. I made a direct reference in the post I made in this thread:


Real-world resolution can differ a bit when you factor in bayer interpolation, low pass filters, bayer array layout, etc.

25
Lenses / Re: Lens 'resolving power' vs sensors.
« on: January 05, 2015, 04:54:49 PM »
Happy to help. :) It's just a hobby, really...well, a hobby that assists me in my other hobby, really. :D I like to know everything about what I do, so as a photographer, well, I had to know how sensors and lenses worked. So I researched it.

Just wanted to point out that the lp/mm for each sensor posted way above are only correct if you have monochrome sensors with tiny pixels (low fill factor) without microlenses and without an AA filter.  If you have a Bayer sensor with real pixels and an AA filter, you're going to need at least 2.5 and maybe closer to 3 pixels per line pair.


If you actually read my post, I stated as much.

26
Animal Kingdom / Re: Show your Bird Portraits
« on: January 05, 2015, 02:00:20 PM »
Thanks, everyone! :) Glad you like the shots.


I wanted to note, all of those were 1200mm f/8-10 on the 5D III. People worry about a TC reducing IQ...it really doesn't have to. They were all shot either on a tripod, or handheld with the lens resting on my deck railing. With proper stability, you can get the most out of even extreme equipment configurations.

27
EOS Bodies / Re: Canon Confirms Development of High Megapixel Camera
« on: January 05, 2015, 01:57:52 PM »

If Canon sensors suck so bad, why do I see so many white lenses on the sidelines of every football game I watch? Pretty sure they're not Sony lenses. :P


The needs of media are different to those of artists.

Last week I visited an exhibition in the Rijksmuseum called Modern Times about photography in the 20th century https://www.rijksmuseum.nl/en/modern-times
Art is not about equipment, it’s not about more resolution, it’s not about a couple of stops more dynamic range, it’s not about ISO, it’s not about how much you can push a file, it’s not about brand A, B or C, etc.
In a couple of years almost no one will know or care which camera and lens were used to take the few iconic photos made in this decade.

Better gear is nice but it will have little impact on how other people (the 99% non-gearheads we share the earth with) perceive our work.

I agree that better gear isn't going to help you out that much if your composition skills are weak, your post-treatments are tasteless, or you've got a variety of other skill related issues working against you.

However, in the hands of someone who is truly fantastic at what they do, better gear could mean higher quality or larger prints. It could also mean getting out strong images in tougher situations without having to make compromises.

Forest photography or sunsets are a great example I think, where if you were given a magic sensor that can capture 20 stops of DR without noise, you could have the ability to create much more beautiful pictures in those two shooting environments.

Better lenses and more MP = more beautiful prints...so those are always welcome upgrades too. Right now, even a D810 with a flawless lens only gets you a perfect print of ~26" @ 300dpi.


+1000


This is why so many pros (and avid enthusiasts like myself) have $30,000 - 50,000 kits. They are exceptionally skilled, and can maximize the potential of that gear.

28
Lenses / Re: Lens 'resolving power' vs sensors.
« on: January 05, 2015, 01:47:49 PM »
I agree that in the end it is mostly about the photograph and the technique. We are all after the holy grail of stunning, beautiful, compelling photography, and that's what we should strive for.


That said, people DO make new purchases as they find the need to expand their kit. You can either be educated, and understand the nuances of optical and sensor resolution and how they work together to affect your final results, and therefor be properly equipped to make the best decision FOR YOU.... Or, you can listen to all the myths and propaganda that permeate the web about resolution (i.e. that the diffraction limit is a hard cutoff, after which smaller pixels behave more poorly than larger pixels...complete and utter TRASH!) and poorly managed black-box "single number" testing (i.e. like DXO), and make the wrong decision based on bad or wrong information.


I believe it is helpful and important to understand how the equipment in your hand works when it comes time to change or upgrade equipment. It may not matter in the field when your out there shooting, and your mind should be focused on all the other aspects. But when the time comes to actually buy a new lens, it's important to know whether you NEED to buy that $4000 Otus....or whether that $1000 Sigma would serve your needs perfectly well. There are some use cases where an Otus is probably the only lens that will give certain photographers the kind of IQ you want, but the Sigma would probably serve most more than well enough...and you should be properly equipped to make that decision with REAL FACTS, rather than myths and misunderstanding.


That's why I take the time to say the same boring things over, and over, and over, and over....and over again. :P Because these damnable myths just won't die!  :o

29
EOS Bodies / Re: Wait for it ... Revolutionary or Evolutionary in 2015
« on: January 05, 2015, 01:38:36 PM »
for Canon: evolutionary ... at best.  :P

many others: revolutionary.

2015 will likely bring mirrorless cameras with APS-C and FF sensors that finally surpass DSLRs in every single aspect of imaginge capture and imaging. It will be a revolutionary move in the history of photography. From pseudo-analog digital camera-besats (DSLRs) to truly digital, fully electronic, solid-state, mechanics-free imaging. Looking forward to it.  8)


I highly doubt were going to see mirrorless cameras that surpass DSLRs in every aspect. Were WAY too early for that, judging by how difficult it has been for manufacturers to put even moderately compelling full-package mirrorless products on the market. I think the Samsung NX1 has the ergos down, and it certainly has a nice sensor, but it is still lacking in some other areas. The A7 series has the IQ for sure, but it lacks in ergos and has that crappy compressed "raw" format (I can't even use an unquoted RAW to describe that format, as it's simply NOT. :P) Canon has struggled to even release anything even resembling "compelling" into the mirrorless space. Nikon has struggles with their mirrorless offering. The gap over established DSLRs is too large...they need to close that.

30
EOS Bodies / Re: Wait for it ... Revolutionary or Evolutionary in 2015
« on: January 05, 2015, 12:59:12 AM »
Didn't you mean to say: "Legen- Wait for it........... DARY! Legendary in 2015!" :D


Well, revolutionary or legendary, I'm not holding my breath for anything from Canon. Canon is the steadfast beast...they are only interested in what will sell them a million units, as so many members here have made clear. That isn't very exciting...but it certainly pays Canon's bills. So, not expecting anything more than evolutionary from the great steadfast beast.

I'm not sure we'll see anything like that from Sony either. The A7s was pretty impressive technologically, and I don't know if they can top that any time soon with anything other than a creepy-crawly evolution into something slightly better. Nikon, using Sony sensors, will probably be the same.

I've actually had my eye on Aptina from a sensor standpoint....they have some interesting technologies that they have only barely begun to exploit. They have a number of DR-enhancing technologies, including DR-Pix (which is basically dual ISO, only done right in hardware) as well as multi-bucket pixels (which has multiple applications, but one of them could potentially expand dynamic range by however many buckets are used per pixel..2x or 4x). With DR-Pix they already have sensors on the market capable of 120dB of dynamic range (that's 20 stops!) With multi-bucket combined with DR-Pix, we could co well beyond 120dB.

Sony has technology similar to DR-Pix. They used it in the A7s, which is part of the reason that camera has the single highest analog gain out there (ISO 25600 is a direct analog gain, after which secondary measures are employed to achieve even higher ISO settings. Canon tops out at around ISO 1600 or 3200 for direct analog gain, relying on a secondary amplifier for ISO's up through 12800, and secondary measures after that. I believe Nikon tops out at around ISO 1600 as well.)

All the technologies are there. They really just need to be tweaked, optimized and employed in actual consumer grade products now. I don't think Aptina sensors have ever been used in DSLRs, but they do have an APS-C sized sensor in their product lineup. Nikon might employ the A7s sensor in something. None of this would really be revolutionary, and certainly not legendary (the A7s might gain legendary fame among astrophotographers soon here, though...imaging with a resonable SNR at 10-20 second subs at ISO 25600 would be beyond impressive, and it seems the A7s can do it.)

I foresee basic evolutionary stuff (from all players bar Aptina, and their stuff usually goes into video and astro gear) unless Canon actually puts their layered sensor IP into an actual consumer product and puts it on a shelf somewhere we can all buy it. Given Canon's MO, I have to see it to believe it. God only knows what the chances of us actually getting a high resolution layered sensor with a resonably high framerate are, though. I suspect "reasonably high framerate" would be 3-4fps for studio work...

Pages: 1 [2] 3 4 ... 326