January 31, 2015, 02:35:25 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 ... 3 4 [5] 6 7 ... 327
61
EOS Bodies / Re: Canon Confirms Development of High Megapixel Camera
« on: December 28, 2014, 05:09:52 PM »
It is funny I have been saying this same thing for years, nearly 7, when I first compared the 7D to the 1Ds MkIII and I used to get no end of crap for saying it. Probably the most vocal maths oriented poster here used to crucify me, he was a 7D owner and insisted that his camera had vastly more 'resolution' than mine, he has since got a 5D MkIII and done the tests, guess what? His estimation of the crop cameras 'resolution advantage' has gone from >60% to around 15% at best on manual focus bench tests.


You are taking so many things I've said way out of context there, it's unbelievable. First, I've never said the 7D had a mere 15% advantage over the 5D III. I did say it might be about 25% on average for the average use case (i.e. no tripod, less than ideal focus, etc.) The 60% statement was qualified with the fact that ignores the bayer array or AA filter...i.e. it's the raw, monochrome mathematical advantage of the 7D's pixel size.


I believe under more ideal conditions, the 7D can realize about a 45% advantage over the 5D III. That is just a resolving power advantage, which when were talking about micrometer sized pixels, isn't something that jumps out of the screen at you...that would be something more like a 200% or 300% advantage, which at the moment only small form factor sensors have with 1.0-1.2 micron pixels.


You and I see different things, which is why subjective comparisons are useless. Maybe I sit closer to my screen than you do, who knows. I see the advantage of the 7D, you do not. Neither of us is right until someone actually does a proper test with proper testing tools and gets some actual resolution numbers. However we all know how well real numbers go down here on these forums as well...so again, it's all entirely pointless.


Simple fact: smaller pixels resolve more detail. I think that has been demonstrated thoroughly well over the last decade, throughout the continual march towards ever smaller pixels paired with frequently improving optics.
I agree that smaller pixels can resolve more detail if the lens has sufficient resolving power.

In practice, does pixel size affect low light performance for the same sensor area?


By "for the same sensor area", I assume you mean when your subject fills the same absolute area of sensor (i.e. you would need to crop the FF to get the same FoV as the APS-C). In  that case, the pixel size isn't really going to affect things much. Assuming same lens, same aperture (which would be necessary for identical subject size and DoF at the sensor), then your going to gather the same amount of light in total for your subject, regardless of which sensor you use. The FF could have bigger pixels, or it could have the same size pixels as the APS-C. The pixel size doesn't really matter...it's just an arbitrary factor. In the end, for an APS-C sized crop of an FF sensor, you gathered the same amount of light as the APS-C itself. Therefor, noise should be the same once the results are normalized. If the FF has the same pixel size as the APS-C, then simply cropping would be enough to normalize. If the FF had larger pixels than the APS-C, then downsampling the APS-C to the same image dimensions as the FF would average pixel data together, resulting in the same noise.


Now, if you framed the subject the same with both cameras, and used the same aperture with both, then the FF camera, regardless of pixel size, is going to perform better.


Pixel size, ultimately, affects resolving power. Smaller pixels, higher resolving power. If you move from an FF sensor with 10 micron pixels to an FF sensor with 5 micron pixels, your going to resolve more detail. Strait out of camera, the image made with smaller pixels will appear noisier...unless you downsample it to the same dimensions as the 10 micron image. Then noise will appear the same, however the 5 micron image will still be sharper and more detailed.


Smaller pixels, assuming same or better technology, can never be a bad thing. On a normalized basis, smaller pixels mean more detail (all else being equal.)

62
EOS Bodies / Re: Canon Confirms Development of High Megapixel Camera
« on: December 28, 2014, 01:13:56 AM »
It is funny I have been saying this same thing for years, nearly 7, when I first compared the 7D to the 1Ds MkIII and I used to get no end of crap for saying it. Probably the most vocal maths oriented poster here used to crucify me, he was a 7D owner and insisted that his camera had vastly more 'resolution' than mine, he has since got a 5D MkIII and done the tests, guess what? His estimation of the crop cameras 'resolution advantage' has gone from >60% to around 15% at best on manual focus bench tests.


You are taking so many things I've said way out of context there, it's unbelievable. First, I've never said the 7D had a mere 15% advantage over the 5D III. I did say it might be about 25% on average for the average use case (i.e. no tripod, less than ideal focus, etc.) The 60% statement was qualified with the fact that ignores the bayer array or AA filter...i.e. it's the raw, monochrome mathematical advantage of the 7D's pixel size.


I believe under more ideal conditions, the 7D can realize about a 45% advantage over the 5D III. That is just a resolving power advantage, which when were talking about micrometer sized pixels, isn't something that jumps out of the screen at you...that would be something more like a 200% or 300% advantage, which at the moment only small form factor sensors have with 1.0-1.2 micron pixels.


You and I see different things, which is why subjective comparisons are useless. Maybe I sit closer to my screen than you do, who knows. I see the advantage of the 7D, you do not. Neither of us is right until someone actually does a proper test with proper testing tools and gets some actual resolution numbers. However we all know how well real numbers go down here on these forums as well...so again, it's all entirely pointless.


Simple fact: smaller pixels resolve more detail. I think that has been demonstrated thoroughly well over the last decade, throughout the continual march towards ever smaller pixels paired with frequently improving optics.

63
EOS Bodies / Re: Canon Confirms Development of High Megapixel Camera
« on: December 28, 2014, 12:48:10 AM »
Quote
There is no problem of pixel size.

<physics>There is. </physics>


The only physical limitation on pixel size is when they become as small or smaller than the wavelengths of light you need to image. Outside of that, there are no physical limitations on how small pixels can be. We can always benefit from smaller pixels...although beyond a certain point, at common apertures (f/2.8 and smaller) you enter the realm of severely diminishing returns.


At diffraction limited fast apertures, lenses will always resolve more than the smallest physics-limited pixels, which is around 800nm (0.8um)...manufacturers are already wary of implementing 900nm pixel sizes...the smallest so far are 1000nm, which are already smaller than the 1100mm infrared limit for silicon based sensors.

64
Landscape / Re: Deep Sky Astrophotography
« on: December 27, 2014, 11:23:16 PM »
Thanks. :)


So, there is noise. Lots of noise. It's only 2 hours of integration, and I probably need a minimum of 5 to really do it justice. It's just not easy to get that data, since there are only a handful of dark nights a month, and Orion is really racing towards the western horizon. I hope to gather another three hours at some point, which should help.


The 5D III, as I've tried saying so many times on these forums, is one craptastically noisy sensor! It is NOT, buy today's standards, a low noise sensor, at all. Which is a little sad, for a camera barely three years old. That's where the color noise comes from.


As for electronic noise overall, it's actually fairly low. I used ISO 1600 specifically to get read noise low. Its around 3.x e-. I was also imaging at around 3C (it's the heart of winter here, nights are 15-18F), so the dark current is very low. The reason the darker regions look noisy is they have been very significantly stretched. I had 21.3mg/sq" skies where I imaged this, which is getting pretty close to the darkest possible 22mg/sq" skies on earth. That was necessary to even get a reasonable amount of photons on those dark areas. Still, on a per-sub basis, the darker areas probably only had maybe 5-8 photons/pixel/minute tops! :P


So, yeah...there is noise. There is always noise, and when you do a ludicrous stretch like I did, that noise can present a bit of a problem. The only solution is to expose long enough to swamp read noise, and integrate more and more. I need three and a half more hours of integration for my minimum, and I would really prefer another 7 hours.


Regarding dark subtraction, you have to match the dark frame temps to the light frame temps. That can be a major PITA, so I stopped bothering and now use dithering instead. Along with Winsorized Sigma Clipping integration, that takes care of the hot and cold pixels, sat tracks, etc. I still use biases and flats, though...and flats actually tend to increase noise a bit as it removes LP and vignetting.

65
Landscape / Re: Deep Sky Astrophotography
« on: December 27, 2014, 05:06:32 PM »
Bad weather has basically shut me down lately. Decided to start spending my clear nights at a dark site, as it's vastly more efficient to gather photons without LP than with. Here is my latest:





Orion's Sword - Wide Field with additional Dust and Reflections


Canon 5D Mark III
Canon EF 600mm f/4 L II
Orion Atlas EQ-G (Belt Modded)


Acquisition with BackyardEOS, processing with PixInsight and Photoshop. Total of 2h 20m integration for the background, grand total of ~4h total image data for bringing out the core (manual blending, this target has MASSIVE dynamic range.)



66
EOS Bodies / Re: Canon Confirms Development of High Megapixel Camera
« on: December 27, 2014, 05:01:14 PM »
Nope - Zeiss will never, ever make anything to compete with the Big Whites.

Really? 



Do you think Canon will ever make anything to compete with the Zeiss 1700mm f/4?   ;)


I wouldn't say there is any "competition" going on between that Zeiss and Canon lenses at all. Look at that Zeiss thing...that's a funktastically wackadillyc monstrosity of a lens there...and it was a special order design that required very special focusing mechanisms and only works with one particular Hassy 6x6. That isn't a mainstream product that has regular availability. If Canon wanted to, they could special-order something up like that as well (that's basically what the 1200mm lenses were.) I think the comparison here is irrelevant.

67
Photography Technique / Re: EC - adds or subtracts light?
« on: December 26, 2014, 09:00:25 PM »
Three digits of precision was about all most of us could expect using slide rules before the advent of pocket calculators in the 70s.  Still have mine.  Much of the world was built on that basis.  I might have a greater appreciation of the big picture as a result.

On the other hand, I computed Pi to thousands of digits for fun back in the day.  Current record is over 13 trillion digits by a couple of college students using an algorithm one of them designed. The hardware involved is a not as large as you might think:  2 x Xeon E5-4650L @ 2.6 GHz, 192 GB DDR3 @ 1333 MHz, 24 x 4 TB + 30 x 3 TB.  It takes a couple of hard drives to hold one copy of the result.  http://www.numberworld.org/y-cruncher/

Unsolvable problems like this are a good way to get started in serious numerical computation, particularly with the creation of new algorithms - there are many ways to compute Pi.  It also forces you to figure out ways to get the most out of your relatively limited hardware. Advances in systems and algorithms result from these types of efforts and spill over into other areas. My own research involves machine learning and combinatorial optimization problems where compute time grows exponentially with N - all of the computers on the planet working together cannot put a dent in the problem, my focus is on newer and better algorithms.

I prefer to keep my photography simpler.


However if you understand that much about PI, I highly doubt you would ever confuse what EC does, or how it works, or what the side effects of aperture, shutter, and ISO are, or any other technical facet of your camera. ;) You already understand your camera, there probably isn't much, if anything, left to learn about it. Utilizing it's technology would be second nature, and therefor quite simple. Learning the technology of a DSLR isn't some kind of P vs. NP or NP-complete problem.


That's the thing about learning...once you have, it isn't complicated anymore.  8)

68
EOS Bodies / Re: High Megapixel Camera Coming in 2015 [CR3]
« on: December 26, 2014, 08:58:18 PM »
If that is the case why SONY made a7S? They could stick to a7 and even more to a7R and that's it!


The A7s has more advanced in-camera noise reduction. Exmor is a good sensor, but it wasn't good enough to produce ISO 400k with the low noise levels that the A7s has. The Bionz X DSP is responsible for reducing the noise in each pixel as they are read out. Not much different than what Canon is doing now with DIGIC 6 in the 7D II. If it were not for the more advanced in-camera NR, I would agree...the A7r when downsampled to 12mp would have probably looked very similar to the A7s, primary difference being the downsampled A7r image would be sharper.
So in A7s they... "cook" the raw files?  :o


They do, just as much as Canon "cooks" the RAW files in the 7D II with DIGIC 6. ;P Everyone's cooking pixels today, it's all the rage.

69
EOS Bodies / Re: High Megapixel Camera Coming in 2015 [CR3]
« on: December 26, 2014, 12:30:08 PM »
If that is the case why SONY made a7S? They could stick to a7 and even more to a7R and that's it!


The A7s has more advanced in-camera noise reduction. Exmor is a good sensor, but it wasn't good enough to produce ISO 400k with the low noise levels that the A7s has. The Bionz X DSP is responsible for reducing the noise in each pixel as they are read out. Not much different than what Canon is doing now with DIGIC 6 in the 7D II. If it were not for the more advanced in-camera NR, I would agree...the A7r when downsampled to 12mp would have probably looked very similar to the A7s, primary difference being the downsampled A7r image would be sharper.

70
EOS Bodies / Re: High Megapixel Camera Coming in 2015 [CR3]
« on: December 26, 2014, 12:27:59 PM »
Thanks Lee Jay.  Is per pixel noise in this case dependent also on available light?

Of course.  SnR = sqrt(photons captured) [to first order] so more light = higher SnR.


I think it is better to say SNR is relative to SQRT(Electrons Released), as that would factor in quantum efficiency, which is an important factor. So you basically have: SQRT(PhotonFlux * QuantumEfficiency), with full SNR equal to (PhotonFlux * QuantumEfficiency) / SQRT(PhotonFlux * QuantumEfficiency). Not every higher resolution sensor, when downsampled, will produce a less noisy result...it depends on each sensor's Q.E.

If you have a 15mp sensor with Q.E. of 60, and a 60mp sensor with Q.E. of 30, downsampling the 60mp sensor image probably won't produce better results than the 15mp sensor...you'll probably get a sharper result, but not necessarily less noise (you might possibly still have more noise.) Why? Well, for sensors of the same size and an identical exposure, the one with Q.E. 30% converted the same number of incident photons to fewer electrons, and it's the total amount of electrons gathered in the whole sensor that ultimately matters. The higher resolution sensor in this case gathered about half the light, for the same exposure of the same scene. It doesn't matter what you do, less light is less light, and less light means more noise (even on a normalized basis.)

71
Lenses / Re: Lens 'resolving power' vs sensors.
« on: December 26, 2014, 12:17:33 PM »
I realize that I'm replying to an old post that's part of an old thread with recent activity due to the high mp alleged confirmation by Canon.

I wanted to address one thing in particular though.  I question the assumption that recent lenses and mk ii lenses are being refreshed for higher mp sensors, including current aps-c 18mp crop sensors.

I was wondering earlier how the current EF 70-200mm f/2.8 IS II USM would perform on a higher mp camera, so I tried looking up its resolving power in lp/mm and ran across a review I hadn't read before, by DxOMark stating that for that particular lens its predecessor, the original EF 70-200 f/2.8 IS USM (mk I) had more resolving power.  Then in the comments, they stated that it was sharper on aps-c  bodies than it is on ff bodies (which I didn't understand).  I've not read any other review that made such claims.

Then I decided to look into the new 100-400mm zoom mk II and read that more or less it has about the same resolving power as its predecessor, and the true improvement is in the stabilization upgrade/update.  Was it already resolving well enough before?  I've never used one, so I don't know.

Any thoughts on this.  I am aware that many here are not fans of DxOMark and their results, but the 100-400 II still has me scratching my head.


Well, your relying on DXO. :P Guess that's enough said.  ;D


One thing I will say is, older Canon lenses often resolved very well in the center, and very poorly in the corners. For film, where huge enlargements were not as common (most of the time, it was smaller enlargements or contact prints, so maybe 8x10 tops on average), poorer corners were probably not as much of a problem. Newer Canon lenses and the Mark III TCs have all been addressing corner performance. Some of Canon's newer lenses did not improve center or midframe resolution much...but improved corner resolution a ton. Some lenses, like many of the new Great White Mark II generation, improved resolving power across the board.


For the best lens resolution tests around, check out Roger Cicala's LensRentals blog. He regularly tests new lenses, using an optical test bench (so he's testing LENSES, not camera systems), and his comparisons tend to be more accurate and generic than most. None of this "limited by the sensor" crap that skews and convolutes results. :P

72
Photography Technique / Re: EC - adds or subtracts light?
« on: December 26, 2014, 12:10:14 PM »
YEAH, it's kinda like PI ... 3.14 is close enough, but mathematicians have been trying to get the absolute forever.  Not sure what that will change in our world if someone ever gets that final decimal figure.  The only people that care are other math folks -- the rest of us grab the numbers we're given and go out and push a shutter button.

What do we have then?  An image.   What to the PI seekers have?   An enigma and no reward.   Which door do you choose?


LOL. Well, it entirely depends on the precision of the math your doing, whether 3.14 is enough. For a lot of mathematics work, that isn't close to enough precision. Now, we don't need precision out to thousands of decimal places, but neither is two decimal places enough for every use case.


I don't see photography as any different. Either you plateau and are simply ok with that, or you are always striving to do better, go farther, create more incredible art. At some point, you have to push the envelope, and part of that means learning more about your equipment. Just barely good enough (like Pi == 3.14) is just that, just barely good enough. It's limiting. It'll always hold you back and limit your capabilities if you simply settle for that and never strive for more. Your acceptance of the first limitations you reach will show up in your work as well, and you will forever be...ok. Never great, nor excellent, nor a "genious"...just ok.


Some people are ok with that. Which is perfectly fine. Other people are not. ;) Pi = 3.14 is simply not enough for some people, those who push the limits...


BTW, there is no "final decimal figure" with Pi...it's an irrational number, which means it's precision is infinite...we can keep calculating Pi to ever more decimal places, and there will always be something left to discover. Beyond simply precision, Pi is an endless number, it contains discrete strings of digits that can, and do, represent...anything, possibly everything. There are mathematicians who have searched for keys to genetic code in Pi, rendered parts of it as images (some think the Mona Lisa is in a sequence in there somewhere), played out sequences of it's decimal places as music, and many other things. It is an infinitely complex number...all that precision isn't "needed" per-se, however if no one ever looked for more than two decimal places, we wouldn't have pondered the implications of such a complicated number, nor tried to play it back as music (and some sequences can be quite beautiful when rendered with the proper notes), nor rendered it as images (there have been some incredible images rendered from bits of Pi as well), nor any of the other things we've done with Pi.


Personally, I have a hard time simply stopping once I've achieved the minimal amount of skill required to barely succeed at something: Set everything to automatic, switch to Av or Tv, press the shutter. Hmm. You can be successful with that, but it will only take you so far. You are going to run into limitations, improper exposures, too much noise, too much blur, not enough DOF, something. Photography isn't a purely creative endeavor...it is a technical endeavor as well, and understanding the technical side if things can only make you a more effective photographer. It'll help you discover more levels of skill or talent within yourself, empowering you to do more, see more, create more. Just like all the things we've done with Pi...despite the supposed "lack of necessity" of having more precision than 3.14.

73
EOS Bodies / Re: Canon Confirms Development of High Megapixel Camera
« on: December 26, 2014, 11:39:50 AM »
Almost all Canon's current lenses would do just fine on a 50MP full frame sensor, and the better ones would excel.
If you are used to images from a 24MP sensor and accept that level of quality, I agree.
However, if you expect to take advantage of the 50MP sensor and see closer to 40PMP images you are dead wrong.

Pls help me learn: Many Canon/Nikon lenses are supposed to be on par. How does Nikon get away with 36mp cameras with it's lenses? Thx. (Genuine question)


Nikon doesn't really get away with it. Nikon lenses affect output resolution, and in many cases limit it. Unless your using a Zeiss lens on a D810, or maybe the 14-24 Nikon or their new 800mm f/5.6, it's unlikely your going to get close to the resolving power limit of the 36.3mp sensor.

74
EOS Bodies / Re: Canon Confirms Development of High Megapixel Camera
« on: December 26, 2014, 11:37:44 AM »
What I read between the lines of "there are two ways of roughly divided into high pixel and high sensitivity to user needs" and "up the number of pixels while maintaining the pixel quality" is -

Canon is going to cater to those who need high ISO performance.

Canon is going to cater to those who need high resolution, without hurting pixel level performance.

Canon is not going to improve DR at low ISO.

They didn't say that.

Sensitivity can mean also mean better DR - it is an ambiguous statement and because it is ambiguous it is perfect marketing material.


No, in the context of the interview, sensitivity means high ISO. That is very clear from what they were saying...that Canon already focused on high sensitivity, now they are focusing on high resolution. The only mention DR MIGHT come from their mention of "preserving pixel quality"...which to me does not sound like any kind of improvement, simply that they are maintaining current pixel quality while concurrently increasing pixel count.

75

FWC is dependent upon surface area...the surface area of the photodiode. With an FSI design, some of the pixel area is dedicated to wiring and transistors, which thus necessitates that the photodiode (the actual light-sensitive part of the silicon) become smaller. By flipping the die and etching on the back, the photodiodes can become larger, since the wiring is all on the other side. That increases FWC, which means for the same size sensor, with even the same size "pixel pitch", your actually gathering more light.

Thanks for the explanation.  For some reason, I'd thought the surface merely translated photos into electrons, but that the electrons were stored below the surface.


The electrons are actually stored around a specific potential layer within the photodiode. Exactly how deep that layer is, and how many electrons it can hold, depends on the specific design of the photodiode. Generally, the amount of charge is related to surface area, but research is has derived ways of changing that layer, deepening it, increasing the potential, etc. I don't know if any of those technologies have been employed yet, and if they have, it's probably only in sensors with ultra tiny pixels (1-1.2 micron), and we might not see them in APS-C or FF sensors for quite some time.


Generally speaking, though, it's the photodiode that has the charge releasing and holding potential, and that potential is realized when a photon (or few, depending on Q.E.) strikes the photodiode, releasing it's energy. The energy either becomes heat, increases the photodiodes potential for releasing another electron, or actually releases an electron. The larger the the area of the photodiode (or more specifically, the layer within the photodiode that is actually responsible for holding the charge), the greater the maximum charge the photodiode can hold. The increased area of BSI photodiodes increases the pixel's FWC (full well capacity), which in turn has the capability of increasing dynamic range.

Pages: 1 ... 3 4 [5] 6 7 ... 327