August 31, 2014, 12:46:55 AM

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 - sarangiman

Pages: [1] 2 3 ... 16
EOS Bodies / Re: EOS 7D Mark II & Photokina
« on: August 29, 2014, 04:53:22 PM »

And Sony? Compare the the 36 MP Alpha A7r(esolution) and the 12 MP A7s(ensitivity), then say again that more MP does not mean more noise if you dare. At base ISO maybe, but try going at 800 and beyond...

I'll take your ISO 800 and beyond and raise you a comparison between the A7r and A7s at ISO 3200:

See any difference between the two when normalized to 12MP? I don't think so. Everything was controlled in this test, down to using the same exact lens, same shutter speed, aperture, ISO, etc.

In fact, I don't see any appreciable advantage over the A7R until you hit ISOs of 25.6k and beyond. Because at that point, the crumbs start encroaching on the slices, in the pizza analogy. For deep deep shadows, the A7S' takes over at ISOs of 6400 and beyond. Really depends on the tone you're looking at, b/c different tones are affected differently by the different sources of noise (shot noise, read noise, etc.).

But ISO 800? 1600? You're unlikely to see any appreciable difference.

LTRLI also made an interesting point about finer-grained noise of higher resolution sensors. So there's also that to consider.

EOS Bodies / Re: EOS 7D Mark II & Photokina
« on: August 29, 2014, 01:11:21 AM »

Also, since smaller pixels hold less charge, the chance of blowing highlights is higher than a larger pixel sensor.

But for the same focal plane exposure, the smaller pixels also receive less light because, well, they're smaller. So total sensor DR does not necessarily have to suffer for higher resolution sensors.

   As a result, sensors with larger pixel have better dynamic range than sensors with smaller pixel even if the total sensor size is the same.

If all else is held the same, yes. In reality, this doesn't generally always turn out to be true. For one, sometimes manufacturers are able to decrease per-pixel read noise with the smaller pixels of the higher resolution (but same size) sensor. For a sensor of n times higher resolution, all you have to do is decrease per-pixel read noise by a factor of sqrt(n) to achieve equivalent normalized noise performance.

This may be why the Sony A7R has just as much DR as the A7S, and normalized noise performance is similar. But, admittedly, I'm just guessing here.

The point is, there are a number of variables here one must consider. It's not always straightforward. For example, the A7S with its 12MP has demonstrably lower DR in RAW than the A7R with 3x as many pixels. Probably resulting from increased downstream read noise of the architecture.

Nope. I haven't forgotten noise or SNR. The 5µm pixels will have twice the noise. However, a 2x2 matrix of 5µm pixels equal one 10µm pixel in terms of area. Average those four pixels together, and you reduce noise by SQRT(4), which is? Yup. A factor of two. The 10µm pixels have four times the area, which again, reduces noise by SQRT(4), or a factor of two. There is more noise per pixel, however the noise per absolute area of the subject is the same. The sensor with smaller pixels has twice the image dimensions. Downsample the 4460x2980 pixel image to 2230x1490...and there will be no difference in noise.

The only difference you WILL see? The downsampled image is SHARPER!  8)

Have a nice day.

No, jrista, the noise per absolute area is not the same. It's pretty close, but it's still worse. It'll only be the same if the n-times higher resolution sensor has its pixel-level read noise reduced by a factor of sqrt(n) compared to the lower resolution sensor. Let's take the case of a 10µm x 10µm pixel, vs. this pixel divided up into four 5µm x 5µm pixels. Let's do some math:

For a 10µm pixel that receives 200 photons, QE=50%:
  • Signal = 100 (50% of 200 = 100 electrons)
  • Read noise=2
  • Shot noise = sqrt(100) = 10
  • Total noise = sqrt (10^2 + 2^2) = 10.198 [yes, this is an approximation, but it'll suffice]
  • Per-pixel SNR = 100/10.198 = 9.806

For the four 5µm pixels that also receive a total of 200 photons, or 50 photons each, QE=50%:
  • Signal = 25 (50% of 50 = 25 electrons)
  • Read noise = 2
  • Shot noise = sqrt(25) = 5
  • Total noise = sqrt (5^2 + 2^2) = 5.385
  • Per-pixel SNR = 25/5.385 = 4.6424
  • After averaging those 4 pixels, SNR increases by sqrt(4)=2, so SNR for that area is now 9.285
  • Another way of determining the SNR of the four averaged pixels is to calculate out the noise: total noise after averaging will be sqrt(4*sqrt(29)^2) = sqrt(116) which you can already tell is going to be more than sqrt(104), the total noise of the 10µm pixel. But let's continue: sqrt(116) = 10.77. Adding together the signal of the 4 pixels gives you a signal of 100. So SNR = 100/10.77 = 9.285.

Point is, SNR of 9.8 > SNR of 9.3. This is generally a bit academic of a difference, which is why high resolution sensors generally do so well. But for extremely high ISOs, where you have so little signal to begin with, it can make a difference. Or when the higher resolution sensor has many, many more pixels.

The take-home point is that when you average the smaller pixels of a n-times higher resolution sensor, you completely equalize shot-noise performance per-area. But not read noise performance. Why? Simply b/c you have n times as many read events.

Third Party Manufacturers / Re: Nikon D810 Product Advisory
« on: August 21, 2014, 02:41:31 PM »
most of what this D810 has over the D800E actually done with FW updates, which Nikon seldom does.

Like the redesigned mirror & shutter mechanism, the complete lack of OLPF, faster processor, and true ISO 64 along with the shot noise and DR benefits it can bring?

And for high resolution required work, we do not need or want the mirrorslap anyway, and thus the A7R would be much better choice for most of us.

Nikon is seriously doomed.

That's ironic, b/c the A7R's shutter shock is literally unavoidable (at longer focal lengths), while the D810 at least offers an option for an electronic 1st curtain that - when combined with mirror up+delay - can completely get rid of camera-induced shake. So, despite Nikon's poor implementation of electronic 1st curtain (only available in Mirror Up mode), at least it offers a way to completely eliminate shutter shock. The only way to mitigate it on the A7R is to add a ton of mass to your camera (which kind of defeats the whole point of the little FF wonder), and even then you may or may not completely eliminate it.

Third Party Manufacturers / Re: D810 users are seeing spots
« on: August 21, 2014, 05:35:10 AM »
They don't learn quickly: back in the day, when I shot Nikon, a solution to the D200's horrible banding problem had to be dragged, kicking and screaming, out of the buggers - and it never did really fix it (for me anyway) on either of my D200s.

That camera is why I'm a Canon shooter now.

Having issues with banding, and then being satisfied by switching to Canon... now there's something I'm going to have to scratch my head over for a bit.  :o

EOS Bodies / Re: Exmor vs DualISO
« on: August 18, 2014, 12:18:06 AM »
Intresting. Thanks for your effort. Therefore, ML would also provide to high ISO nightshots?

No, ML Dual ISO will not help with high ISO performance. Dual ISO simply helps to deal with the noise resulting from the architecture I mentioned earlier. The 'downstream' read noise this architecture is well known for - and what dual ISO helps to avoid - becomes pretty irrelevant at high ISOs. It's swamped by other sources of noise - namely shot/photon noise and 'upstream' read noise. Shot noise contributions are determined by how much light is collected, so roughly correlates with sensor size & QE. Nothing ML can do here. Upstream read noise for these Canon sensors is already quite good (especially for the 1Dx), and difficult to do anything about without actual hardware-level adjustments. So there's really nothing for ML to do here either.

EOS Bodies / Re: Exmor vs DualISO
« on: August 18, 2014, 12:17:51 AM »
Yeah, exactly what raptor3x said. You can probably safely assume a similar benefit across most Canon DSLRs.

Those guys at ML are magicians, I have to say. Personally, I think it's much easier to just go for a Nikon D810 if you care about base ISO DR, especially as the D810 offers yet another 2/3 EV or so DR over the D800/A7 cameras. But if you have other constraints that force you to stick with Canon, this ML trick seems very useful. As others mentioned, the resolution tradeoff doesn't seem as severe as one might expect it to be.

EOS Bodies / Re: Exmor vs DualISO
« on: August 17, 2014, 08:19:47 PM »

Speaking of landscapes...high end professional landscape work is not produced by pushing Exmor shadows 4-5 stops. Landscape photographers bracket and HDR. Compare a HDR image to a heavily pushed image, even from Exmor, and the difference in tonality and fine detail will jump off the print at you. With AEB you can easily hand hold a 3 frame bracket.

Good point. And it's because of a subtle point. It's b/c even Exmor shadows are limited by shot noise - noise that comes from the inherent randomness of non-coherent light. This source of noise has increasing contributions with shorter exposures, even for an 'ideal' sensor (which the Exmor architectures are approaching).

You can think of it this way - if a landscape photographer won't use ISO 800, he/she similarly won't push a base ISO 100 camera's shadows 3 stops. Since it's essentially the same thing (ignoring quantization error).

All that said...I do find it puzzling that Canon went through the effort to make this possible in the sensor hardware but then never exploited it in the firmware. Are they afraid that it might be confusing to users, especially with the HTP mode option? Just add an Extended Dynamic Range (EDR) mode for RAW only and clearly state it's for pro users who are going to manipulate the tone curve in RAW.

Canon didn't go through the effort to make this possible in the sensor hardware. ML just found a way around the limitations of Canon's architecture, which pipes analog data off the imaging sensor to an off-sensor digitizing chip (ADC). This process introduces a lot of noise into the signal, which can be mitigated by pre-amplifying the image data (ISO amplification - raising the ISO) before sending it off the imaging chip. That's why Canon's architecture doesn't hurt its high ISO performance much, but severely limits its low ISO performance.

EOS Bodies / Re: Exmor vs DualISO
« on: August 17, 2014, 08:15:05 PM »
What makes you think they haven't used this in the 1DX?  ;)

Because the 1Dx doesn't have much more DR than a 5DIII... and shows the typical pattern of increased total input-referred read noise with decreasing ISO.

EOS Bodies / Re: Exmor vs DualISO
« on: August 15, 2014, 05:14:50 AM »
Nice work! Essentially, Magic Lantern is able to recover closer to the 'sensor DR' - by mitigating the effects of downstream read noise probably introduced en route to the off-imaging-chip ADC in Canon architectures.

The extra 0.8EV 'highlight headroom' (or sorts) is interesting. Makes me wonder if something similar is going on with the Nikon D810 & the 36MP sensor in there. DxO data suggest they're able to map a higher signal off the imaging chip to the same Raw value at ISO 64 -- essentially like extending the effective full-well capacity of pixels at this ISO.

Third Party Manufacturers / Re: DXO uh-oh?
« on: August 12, 2014, 03:08:40 PM »
I currently don't have any problems whatsoever selecting the AF point with the joystick. I can move it around without ever taking my eye from the viewfinder. The toughest part is hitting that little M.fn button next to the primary dial...there is probably a more accessible place for that button. But once it's pressed, actually selecting the AF point is not difficult at all.

Wow, you don't have the camera set up to automatically move the AF point when you engage the joystick (as scyrene points out)? I'm no longer surprised you don't have a problem with how slow the joystick itself is... When I shoot professionally (weddings) using the joystick, I'm always cursing myself for missing a shot b/c I didn't move the AF point over fast enough. Making shooting kind of like a FPS video game. This is with wide-angle fast primes, where it's critical the correct focus point is used.

There should be a better UI. Nikon's 3D AF tracking and 1Dx's subject tracking are formidable options, but in my experience the 1Dx fails quite frequently at sticking to the subject, and both cameras start to fail in low light. And Nikon's lack of cross-type sensors becomes limiting for portrait-oriented shots (has trouble focusing on eyes). No system is perfect, really.

Third Party Manufacturers / Re: DXO uh-oh?
« on: August 12, 2014, 12:45:00 AM »
I really wonder why Canon doesn't bring eye control back. It certainly seems like fans of the EOS 3 ECF really want it. It would be interesting to have whatever it is your looking at in the VF be focused...that would just rock. I am guessing the system was expensive, at least that's what I'd read in the a film camera, it was probably one of the most expensive things. However in a DSLR, it's just one more expensive thing to add to the mix...maybe it pushes cost over the edge.

As much as I love the idea, eye AF never worked well for me on my EOS 3, or any of the Canon camcorders I used to use back in the day. I'd probably fare better with contacts, though.

IMHO there's a simpler, better way to do it. Of all cameras, the Panasonic GH4 actually comes close to what I'm thinking- if you enable an option buried deep in the menus. It's not ideal, but it's a good start. With a little creativity, one could iterate on it or something similar to design a UI for AF point selection that'd be significantly faster & easier to use than a D-pad, or a joystick, or the cumbersome 4-way controllers on most mirrorless ILCs.

Third Party Manufacturers / Re: DXO uh-oh?
« on: August 08, 2014, 09:39:24 PM »
The reason I prefer the joystick is b/c the D-pad requires too much travel for my thumb to change focus point. But neither the D-pad nor joystick are ideal. I can think of a much better way to select AF points quickly... can you? :)

But given how spectacular the D810 is at tracking the subject across the frame in '3D' tracking mode, the joystick vs. D-pad debate is less of a concern for me as I'm jumping ship. However, I'd still prefer a faster way to select AF point for those situations where AF tracking fails - e.g. in very low light, low contrast subjects, heavily backlit subjects, etc.

Interesting about the joystick breaking. Ultimately I don't care - that can be fixed so I'd prefer function over longevity. But that's just me.

Third Party Manufacturers / Re: DXO uh-oh?
« on: August 07, 2014, 05:28:09 PM »

Thing is, the D810 is actually a real competitor for the 5D III, as far as "general purpose DSLR" goes. It lacks in the ergo department, and is missing that 1 FPS, dunno about the AF system, although that was always pretty decent, but it is a LOT closer, and STILL has the better sensor IQ.

First of all, 'lacks in the ergo department' is very, very subjective. Yes the grip on the D810 is still too small & not fat enough for my hands, and the D-pad is not as good as Canon's joystick. OTOH, there are many more customization options on the D810 than on my 5DIII. So there's a greater chance I can configure the D810 to suit my needs over the Canon. And let's not forget the complete lack of EC in M mode w/ Auto ISO when it comes to Canon. The 1Dx's implementation is so nonsensical it's almost lacking - you either have to use the LCD screen to use EC in M mode, or dedicate the Set button to activate EC in M mode. The latter removes one of the largest advantages of Canon ergonomics in my opinion - the ability to press 'Set' even in shooting mode to instantly check focus on your last shot. Why you can't adjust EC using the dedicated EC button in M mode baffles me to no end...

Also - if you consider this an element of 'ergonomics' - programmable Auto ISO itself enough reason to choose Nikon. When I'm switching primes during a wedding shoot, I don't want to have to remember to go in there and change my minimum shutter speed (and Canon's choice of 'minimum shutter speed' is often unsuitable). With Nikon, I simply choose slower to faster in 5 increments based on if I'm shooting static vs. moving subjects. Game-changing for the types of photography I do.

I agree, the D800 wasn't really a competitor for the 5D III...but the D810 is.

Now this I'd love some clarification on. Most people are pointing out how the D810 is not much of an improvement over the D800. DxO's own scores on image quality show this. So what suddenly makes the D810 a competitor to the 5DIII, but not the D800? The half-a-stop extra DR? The electronic 1st curtain [EFC] that can only be activated with mirror-up? Just curious exactly why you feel this way.

If I were to venture a guess - I'd say the EFC? I do wish, though, that Nikon had an option to implement EFC in all shooting modes with a short delay to allow mirror vibrations to dampen out. EFC only working in Mup mode is a bit silly - especially in Live View.

My bigger point here is that the D800 was just as big a contender. Not only b/c of its superior image quality, but also b/c of Programmable Auto ISO, Exposure Compensation in M mode, spot-metering linked to AF point, and 3D AF tracking. The latter allows one to simply use the center AF point to initiate focus on a desired subject, and allow the camera to track that subject across the frame, as well as along the Z/depth-axis. This (1) obviates the need to select the proper focus point, which is time consuming, and (2) tracks moving subjects like running brides across the frame. With every Canon save for the 1Dx, I have to manually select the AF point when I can't focus and recompose (24/1.4 and 35/1.4). Try doing that with a 4 month old baby that constantly moves around. For this particular scenario, I believe my focus hit rate went from something like 10% to 80% simply going from a 5DIII to the D810.

Canon's complete lack of a separate sensor for AF tracking in all but the 1Dx is rather egregious. And I, personally, find it difficult to use a 1Dx b/c of its weight/size that, with serious glass, puts it north of what I'm willing to tolerate. The 5D III uses some tricks to track subjects to make up for its lack of dedicated hardware - e.g. I believe it cross-references data from AF sensors to check if a subject at some depth moved from one focus point to another, and I think it also uses some info from its 63 zone metering system to help track subjects. But none of these approaches come near the (lateral, X-Y plane) tracking accuracy of a dedicated 91,000 pixel meter, or the entire imaging sensor itself in Sony SLT designs.

It's funny, if anything, I think the Nikon system is somewhat less desirable now than it was a few years ago when the D800 was released. Why? B/c now Canon has some very, very fine lenses for it's system. The 16-35 f/4L IS & the 24-70 f/4L IS are great lenses for landscape photographers.

Third Party Manufacturers / Re: DXO uh-oh?
« on: August 06, 2014, 02:30:00 AM »

I think it's unfair to say it's "simply not true." It depends. I've shot landscapes that are perfectly still, with only a slight amount of motion in the sky. Now, usually I photograph at ISO 100. However, if your going to be integrating frames, you could get away with using a higher ISO, and taking more frames at a faster shutter speed. If you bumped up to, say, ISO 400, took a bunch of frames, then integrated them together (you can do that with Photoshop, but it's still better to use a tool like DSS to do it, as it can work on the RAW images themselves, rather than demosaiced results), you would very likely GAIN DR in the end.

Well of course. Image averaging increases SNR by the square root of the # of images averaged. You can mathematically derive this simply by knowing that noise adds in quadrature. Which is also why any additive or subtractive operations increase noise, as you mention later on.

But every camera benefits in this manner. And a camera that starts with a higher SNR (Nikon/Sony vs. Canon, assuming all else equal) will benefit just as much. So this is pretty irrelevant in the context of this discussion...

Actually, an easier way to understand all this averaging business is to think that - in terms of photon/shot noise - averaging 8 exposures that are each 1/8 shorter than one long exposure is very similar to just taking one long (8x as long) exposure to begin with. Or using a sensor w/ 8x the surface area. In reality, averaging n exposures is generally worse than taking one exposure n times as long (given you don't clip) b/c of the extra aggregate read noise of 8x as many read events.

Now, such a thing is completely unnecessary with a camera that uses a Sony Exmor sensor. However people who use D800's still do true HDR, and some of them will take a good dozen frames for an HDR merge. Firing off eight ISO 400 frames at a relatively high ISO is trivial in comparison.

That's correct. And, yes, even a D800/810 or A7R benefit from HDR or graduated neutral density filters even with scenes that technically fall within their Raw DR capabilities b/c HDR/GNDs allow you to expose shot-noise limited shadows more - thereby increasing their SNR. So even if using a GND flattens your image such that you have to darken your shadows in post, noise performance will still be better than underexposing those shadows - even for a sensor with no read noise (i.e. a theoretical shot noise-limited sensor). Now, whether it's necessary or not for any given scene/application is another matter entirely.

So, "simply not true"? Really?  ;)

Er, yes I still stand by that, even though we appear to be on the same page :)

You don't subtract it out. Of course not. ;P You AVERAGE it out! Averaging attenuates the standard deviation of noise. You actually literally CAN NOT subtract noise because subtraction actually enhances the standard deviation, making the noise worse (this is intensely obvious when you start doing astrophotography...I accidentally subtracted a master flat frame once, and the noise was terrible because both the flat and the light frame had random noise. You normally divide out flat frames to avoid that problem.)

Yes, b/c noise adds in quadrature. But my initial point still stands - it seems misleading to point out that image averaging can get you near Sony/Nikon levels of DR. B/c image averaging would also help the Sony/Nikon sensors. Each would keep pulling ahead, and we'll end up right where we began - with a base ISO DR advantage going to the Sony/Nikon architectures with low downstream read noise.

Shooting two ISOs is just a firmware hack to extract more dynamic range from Canon's whole readout pipeline. The best way to extract more dynamic range is averaging. Your basic noise reduction algorithm works by averaging. When you apply noise reduction to any single frame in Lightroom, or with Topaz Denoise, Nik Dfine, NoiseNinja, NeatImage, or any of those tools, you ARE increasing dynamic range. That's what noise reduction does. It increases dynamic range.

Actually, the higher ISO used when shooting two ISOs is just a way to get shadows well above the downstream read noise floor of Canon's architecture. It works, with the downsides of the resolution cost/artifacts, when image averaging is not an option. Also, with all this talk of image averaging, I feel compelled to point out that it's sometimes practically quite difficult to be thinking about image averaging when you're trying to shoot rapidly changing light, sometimes ND filters and long exposures to create motion, etc. It's essentially technology getting in the way of artistry, especially when you consider that there are better options out there for this particular purpose (base ISO DR).

Yes, you can technically say that NR can increase DR, but it comes at the cost of detail retention. Hence, IMHO, the best tests of DR are done on unfiltered data (or however unfiltered one can get it).

I know LTRLI disagrees, but what were really talking about here is not actually dynamic range. Were talking about editing latitude. I wish these two things, DR and editing latitude, weren't so intrinsically linked, but they are. Editing latitude, as in the ability to lift shadows, is only one benefit of having more dynamic range. Fundamentally, DR is about less noise. Not just read noise, which only exists in the shadows, but ALL noise, which exists at every level of the entire image...shadows, midtones, highlights, whites, blacks, everything. Denoise algorithms reduce noise, which means, by definition, they are concurrently increasing dynamic range.

Actually, since DR is defined as the range between clipping and some lower SNR threshold, DR is really not about ALL noise. Midtone/highlight noise is typically shot noise dominated (ignoring PRNU), but at this point the SNR is typically well above the lower SNR threshold people generally find acceptable. Save for very small sensors and/or very high ISOs.

While the ML Dual ISO technique is certainly one way of reducing Canon's nasty banded read noise, it's not the only way.

Just to be clear, I wasn't even talking about banding. I was talking about the downstream read noise that manifests itself as just random noise. The detrimental effects of this 'downstream' noise can be mitigated by amplifying the signal to the point at which the downstream read noise is irrelevant. This is precisely why Canon DSLRs can catch up in DR at higher ISOs... at these high levels of amplification, it's mainly sensor-level (upstream of the ISO amplifier) read noise that matters. And here, Canon is doing just as well as others.
There are a lot of ways of recovering dynamic range. Canon doesn't clip their signal, they offset instead, so all the image signal data is there. It can be recovered (which is actually NOT the case with Nikon and Sony cameras),

Again, this is no longer true. The D810 has an offset of 600.
Also, I haven't yet seen evidence that the increased DR at base ISO on the D810 is due to 'cooking' the Raw file. In fact, DxO's full SNR curves suggest some sort of nonlinearity introduced into capture at the lowest ISO. The SNR at clipping - where the noise is dominated by shot noise - remains the same for ISO 64 and ISO 100. That indicates to me that for an increased

Quote from: jrista
You should search some of the astrophotography sites. There is a lot if information about how all the various DSLR manufacturers mess with their signals.

Yes, I know that manufacturer's can 'cook' Raw files to a certain extent, but my point was that there's no evidence so far that this 'cooking' is what is leading to the extra DR at base ISO for the D810. I was pointing out that the SNR curve is significantly more non-linear at base ISO compared to ISO 100 for the D810:

Almost like an emulation (albeit very tiny magnitude) of negative film's decreasing response with increasing exposure. I think that's quite interesting, and am trying to get to the bottom of it. It's literally like a roll-off at higher input luminosities. Would love to hear some thoughts from folks here.

Third Party Manufacturers / Re: DXO uh-oh?
« on: August 06, 2014, 12:43:10 AM »
The kicker is that RAW editors don't have to honor Canon's bias offset. The entire RAW signal is stored in Canon's files, and the offset is calibrated with a border of masked pixels. Who knows if editors like Lightroom, or DXO, or Aperture actually adhere to Canon's recommended offset. Even if they do, there is still negative signal information that can be pulled up, and the full noise signal is there. With Nikon RAW images...all that negative (deep noise) signal is simply discarded.

I use DeepSkyStacker and PixInsight to calibrate Canon RAW files for integration into a "stack". I use a 200-frame master bias image to subtract the bias signal from each light frame before integrating it. When the bias is removed from Canon RAW files, the dynamic range jumps by almost two stops...which puts it in the same range as Nikon files...

As Horshack mentioned earlier, this has changed with some recent Nikon models. In fact, the D810 has a bias offset of 600 in a 14-bit Raw file at base ISO. This might make it more suitable for your astrophotography, no, jrista?

Also: jrista you mention averaged dark frame subtraction as increasing dynamic range (DR) by 2 stops for Canon DSLRs - putting it in the same range Nikon/Sony sensors. I'm sure averaged dark frame subtraction to remove bias and some forms of FPN can help for certain use-cases (e.g. astrophotography), but I'm confused why you mention this here as if it would help any typical, say, landscape shooter suddenly get as much DR with a Canon DSLR sensor as you would with, say, the Sony A7R sensor.

That's simply not true. And I think you admitted this in a later post, but I do think it's important to stress the point.

Now, I know you know this b/c clearly you have a grasp of all this, jrista (stunning image, by the way :) ), but for everyone else - downstream (of ISO amplification) read noise essentially randomly varies the signal, so you can't simply 'subtract' out this random variation to reduce shadow noise (well, not without the usual costs typical NR software pay). Shadows suffer more simply b/c a constant source of electronic noise varies a smaller signal much more than a larger one; hence, shadows pay a larger SNR cost.

The only way I know of recovering the 'sensor DR' (without the influence of downstream read noise) is to simultaneously shoot two ISOs and combine results. For example, what Magic Lantern does. One can show that a Canon 5D Mark III is - for most practical purposes - 'ISO-less' above ISO 3200. That means that if you simultaneously shoot different rows of pixels at ISO 100 & ISO 3200, you can effectively avoid the downstream read noise effects and get more of the actual sensor DR (which is quite good for modern Canon, and Nikon/Sony, sensors). But then there are all the downsides this method brings...

Also, I haven't yet seen evidence that the increased DR at base ISO on the D810 is due to 'cooking' the Raw file. In fact, DxO's full SNR curves suggest some sort of nonlinearity introduced into capture at the lowest ISO. The SNR at clipping - where the noise is dominated by shot noise - remains the same for ISO 64 and ISO 100. That indicates to me that for an increased

Pages: [1] 2 3 ... 16