April 18, 2014, 06:27:57 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 ... 87 88 [89] 90 91 ... 216
1321
EOS Bodies / Re: Canon 70d RAW Samples
« on: August 08, 2013, 10:32:13 AM »
Also, if you enjoy a good laugh here's the newest crop sensor vs the good ol' ff from the 6d :-> ... the larger sensor nearly has a 2 stop edge, though probably not at the same dynamic range:

FF also has 2.56x time the light collecting surface area so it's no surprise it does much better when you are not reach limited.

As for the probable modest gain at high ISO as some of us have said the cameras are already so good there that it's very hard to make them a ton better without making using active cooling or something, some totally new tech, and even then the mid-tone SNR can't be improved much although they shadows and DR could go up a lot I guess, but with today's tech that seems tricky/expensive/heavy/messy (active cooling pipes burst :D). People forget how amazingly efficient they are at photon capture already, more than half way to perfection (according to laws of physics) so there really isn't any room for 3-4 stops better or anything at all like people are asking for. If they could radically reduce read noise even more you could expand DR which is pretty low at high ISO, but it seems, as I said, that it might be tricky without costly/heavy stuff like active cooling or something. Not sure.

I guess does have a little room there for SNR as they are at the top but with more color-blind color filters, if they get a bit more efficient they could strengthen color filters again so you could maybe tighten those up a bunch and get another 1/2 stop better SNR.

At low ISO Canon has a LOT of room to get better though and it's perfectly possible on that end since everyone else has already done it. You have to be willing to invest in a new design and sensor fab, so far Canon seems to refuse.

Active cooling doesn't necessarily have to involve a liquid coolant. A TEC (ThermoElectric Cooler, or Peltier) could be used as an electronic heat pump, along with a copper heat plate and heat pipes to draw heat off and release it at areas along the body. It would add some weight, but it could also improve SNR at high ISO by a useful amount. I am not saying you could cool the sensor to -80C or achieve 90% Q.E...but you could probably improve Q.E. into the 70% range, and reduce dark current noise by an order of magnitude or two. The difference wouldn't be immense, but I think it could be useful, and visibly improve IQ at high ISO. I don't think that would mean we suddenly see usable ISO 102k or anything like that...but ISO 25.6k and 51.2k might become viable for real-world stuff, publication online and in print, etc.

1322
Lenses / Re: Dxo tests canon/nikon/sony 500mm's
« on: August 08, 2013, 10:21:24 AM »
Sorry about the delay in replying, the weather has been (almost too) good in this week.

What it comes to slanted edge testing, this is where I disagree (partially). If we consider a slanted edge test with a body+lens setup, there are several issues in that what I'd think as a deal breaker for recovering the real point spread function as I know it.

First, the pixel pitch typically does not actually support sufficient sampling.

It does but not of the PSF directly - of the PSF convoluted with "something", derived from the pixel size and the AA filter. That "something" is a known quantity, depending on the angle as well. From there, you can get the PSF. Again, the computation is NOT the same as deconvolution but still, it is not going to be too accurate if the PSF is too concentrated compared to one pixel but the instability is far from the (exponential) one for deconvolution. The lens PSF convoluted with the effect of the AA filter however is not "too concentrated" (the reason AA exists in the first place), and can be well reconstructed. Factoring out the effect of the AA filter itself if trickier but if you keep the same sensor, you want to keep that effect in place. 

If you are reverse engineering a PSF which is the convolution of lens + AA filter, then that would be the exact issue I was trying to point out before. The AA filter is designed to limit the resolution of the image that reaches the sensor plane by filtering out higher frequencies, while leaving lower frequencies in tact. If you are reverse engineering the image post-AA, then it has an intrinsic upper limit on resolution. The lens could very well (and in the case of a good lens like the EF 500/4 L II, most likely does) resolve more than what the lens+AA convolved resolve.

If you knew the exact nature of the AA filter, you could probably exclude its effect from the PSF, and arrive at a result much closer to what the lens itself is actually capable of. If you leave the convolution with the AA filter in, then you haven't really reverse engineered the lens MTF, you've just reverse engineered the lens+AA filter MTF. That might be useful for comparison, but it really doesn't tell you all that much about the lens itself.

1323
EOS Bodies / Re: EOS 7D Mark II Basic Specs [CR2]
« on: August 07, 2013, 07:21:27 PM »
I still am not buying the rumor that the 70D sensor is the same as the 7D2 sensor.  That only would have made sense if the 7D2 came out considerably in advance of the 70D, and that ship has sailed.   Why a lower trimline body would get such a vital core IQ component that a far better product will get a year later makes no sense at all.

Ditto. It doesn't seem logical for Canon to release a 7D II that didn't bring with it an entirely new set of technology, sensor included, continuing the trend started with the original 7D (do we start calling it the 7Dc for Classic, now? ;P) I also think that, whatever megapixel count is has...18mp, 20mp, or 24mp...the 7D II sensor really needs to shine. It needs to produce IQ better than the 7D and 70D that preceeded it. It needs to demonstrate Canon is still interested in their customers needs, and still capable of competing in a sensor fabrication environment riddled with far more competition and innovation than ever seen in the past.

(Not to mention the fact that a Canon employee mentioned in a DPR interview that they would be doing something interesting and innovative with the 7D II sensor not long ago.)

1324
EOS Bodies / Re: 70D what do you think?
« on: August 07, 2013, 07:16:29 PM »
First, I think you forgot an open quote tag. ;P

Quite well put IMO.  The problem I have (and the reason I jump in to these) is that they tend to generate needless FUD (as you more or less point out).  I think that is actually the goal of some of the protagonists, they think that if they keep beating the drum long enough Canon will be shamed into improving their performance in this particular specification.  But in the meantime, people need to know what the DR really means to their photography and that doesn’t come through in many of these dogfights.  Good comments.

Thanks. I agree about the latter stuff...would be nice to explain DR in a more realistic context. For a lot of shooters, then probably rarely ever see an ISO setting below 400, in which case the DR debate is entirely meaningless to them.

On the flip side, I kind of doubt Canon actually reads these forums enough to matter, and if they do, I doubt they redirect their product map as a result. I think most of the real-world reviewers have noted an improvement in dynamic range from recent Nikon products, but few have mentioned it as the thing they really, truly think Canon should improve with their next DSLRs...at least, not like they did when they were calling for better high ISO performance and an end to the megapixel wars. I think Canon hears those guys far more than they hear anything on these forums. Which just further invalidates any reason to debate someone like ankorwatt every time he decides to drop an antagonism bomb on a thread.

I think something far more useful would be to create an online petition to Canon, and work our asses off getting as many signatures as we can from DR-hungry Canon shooters online, offline, on CR forums and anywhere else on the web, for delivery by the end of the year (and before Canon decides to actually finalize and ship a Big Megapixel camera). ;P

1325
EOS Bodies / Re: EOS 7D Mark II Information [CR2]
« on: August 07, 2013, 07:05:12 PM »
That would only be the case if the CoC was larger than what I chose as acceptable. With such small pixel pitches of sensors these days, the CoC from ANY sensor is appreciably smaller than what my CoC needs to be that it no longer matters. When you need a CoC that is at least 50x larger than what you get from the 1D X, it doesn't matter if you photograph something with the 1D X, 5D III, 7D, 7D II, or anything else...its already far smaller than you need it to be for the vast majority of circumstances. Most of my work goes online. Even scaled to "HD" image dimensions, the necessary CoC is 20-30x larger than any sensor today is already producing. Even if I enlarge my photos by 3x, maybe even 4x, to print at 32x48 or 40x60...I'm not going to be standing within a foot to view them. I'll be standing back ten, fifteen, twenty feet, so the increase in distance is effectively the same as NOT enlarging the photos in the first place. My required CoC then is still large enough that 4µm, 5µm, even 7µm pixels are going to resolve more detail than I really truly need unless I am cropping out the middle 10%.

CoC (in that context) has nothing to do with pixels (unless you are extremely picky and good sighted). It is your tolerance for what you accept. My point was that the 1.6x factor stays, regardless of what your tolerance is. It goes back to what you said earlier, and you are changing the topic now.

Quote
So yes, there is a difference between FF and APS-C DOF. Its just that the difference doesn't matter in the real world in the vast majority of reach-limited cases...and even when it does...no one actually notices. The DOF benefit of FF shows up when you have the ability to frame identically, but in those cases it is either the focal length or the subject distance that changes by the crop factor, and that, not the physical size of the sensor itself, that changes the depth of field.

I said repeatedly that the difference is for those who care, and most people would be happy with either format. But you are downplaying the benefits again and distorting them. What prevents you from framing identically, how does the sensor size changes the distance to the subject, etc. Stop lying to yourself. A larger sensor alone does not automatically mean the ability to get shallower DOF but the FF system built by Canon does have much greater potential for that than the Canon crop system. 

I will stop here because I do not see anything coherent in your posts. You are jumping from topic from topic, each time forgetting what you wrote before and what I replied to.

I'm not downplaying any benefits. You claim that the simple act of cropping (which is all a crop sensor does) changes the depth of field. If you hypothetically had both FF and APS-C sensors with the same pixel pitch, and you cropped the FF image in post to match the APS-C image, would that change the DOF? Of course not. The two images would be identical.

Pixel pitch puts a lower limit on CoC, which is why I bring it up (and I mentioned this in a previous response, so apparently neither of us can either read nor remember what we've read). A 5µm pixel would limit your CoC to 0.01mm. If you only need a 0.5mm CoC (and I'm being EXTREMELY generous here) for a 2x reduction for publication on the web, then it doesn't matter if you are using a 6 micron pixel, a 5 micron pixel, or a 4 micron pixel. All three would produce results similar enough that no human being could tell any difference for your reduction. You could even enlarge by a factor of two, and still have more than enough "CoC headroom" that the differences in pixel pitch wouldn't matter enough to produce an appreciable difference in DoF that anyone would notice it. A cropped FF with 6µm pixels is going to be perceived the same as an APS-C with 4µm pixels.

Mathematically, purely theoretically, you are correct. A 0.12mm CoC (6µm pixel) will produce a different DoF than a 0.08mm CoC (4µm pixel). I'm not solely trying to debate the simple mathematics of the issue. I am trying to point out that the difference between 0.12mm and 0.08mm CoC has no real-world relevance such that, assuming a focal length limited scenario where an FF and APS-C are used to photograph the same subject, with the same lens, at the same distance, cropping the FF image to match the APS-C image, DOF would be perceptually different. I'm arguing the point that DOF has nothing to do with crop factor, and everything to do with the lens and subject distance. That means you can get closer with the FF, or swap to a longer lens, to frame it the same as the APS-C...and the change in focal length or subject distance is what actually changes the depth of field (and does so in such a way as to give FF a significant DOF advantage over APS-C, yes!) I am not denying anything, and my argument has been coherent and consistent...I am denying the notion that crop factor in and of itself has any impact on DoF.

1326
EOS Bodies / Re: EOS 7D Mark II Information [CR2]
« on: August 07, 2013, 05:55:07 PM »

"to be projected on a full-sized movie screen"

That would be an enlargement by a factor of 20x, 30x, maybe more? Of course CoC is going to matter with such an enlargement, despite the fact that you sit back by 30 to 50 feet. I don't think most photographers enlarge much more than 2x, and the vast majority of photographers reduce their images (usually quite considerably) for publication online.

They post images smaller than 24x36mm? Don't say!

Look, it is really simple. Take any display size you like. Movie screen or a 4"x6" print. Whatever, fix it. Then take any CoC that makes you happy on that display. It can be as big as your head on the movie screen, or as small as a head of pin for a 4x6 print. Fix that, as well. Now, start with a 24x36mm image, and from a 15x22mm one. What blur radius (CoC) on each image, when enlarged, would produce the desired CoC on your favorite display size? You can compute it but without any computation, you can say that the CoC on the 15x22mm image has to be 1.6 times smaller. As simple as that. I do not really know what else to say.

That would only be the case if the CoC was larger than what I chose as acceptable. With such small pixel pitches of sensors these days, the CoC from ANY sensor is appreciably smaller than what my CoC needs to be that it no longer matters. When you need a CoC that is at least 50x larger than what you get from the 1D X, it doesn't matter if you photograph something with the 1D X, 5D III, 7D, 7D II, or anything else...its already far smaller than you need it to be for the vast majority of circumstances. Most of my work goes online. Even scaled to "HD" image dimensions, the necessary CoC is 20-30x larger than any sensor today is already producing. Even if I enlarge my photos by 3x, maybe even 4x, to print at 32x48 or 40x60...I'm not going to be standing within a foot to view them. I'll be standing back ten, fifteen, twenty feet, so the increase in distance is effectively the same as NOT enlarging the photos in the first place. My required CoC then is still large enough that 4µm, 5µm, even 7µm pixels are going to resolve more detail than I really truly need unless I am cropping out the middle 10%.

So yes, there is a difference between FF and APS-C DOF. Its just that the difference doesn't matter in the real world in the vast majority of reach-limited cases...and even when it does...no one actually notices. The DOF benefit of FF shows up when you have the ability to frame identically, but in those cases it is either the focal length or the subject distance that changes by the crop factor, and that, not the physical size of the sensor itself, that changes the depth of field.

1327
EOS Bodies / Re: 70D what do you think?
« on: August 07, 2013, 05:42:56 PM »
There is little difference in IQ for any DSLR at ISO 100.  You are buying features and hype.
 
For landscapes,  more MP will give you more detail, but as MP count goes up, so does the difficulty in actually getting that extra bit of resolution.  There are some who do very well with high MP cameras, but the average photographer is not obsessed with a incremental increase in resolution, the more important elements such as composition and subject, lighting, DR, etc put and IQ improvements far down the list.   In fact, I don't even know how to measure something hypothetical like IQ that only exists in the mind of a photographer and is different for each of them.  We can measure several parameters, and they trade off against each other, but IQ?  That sounds like something DXO would assign a number to ;)

There is a pretty huge difference at ISO100. That is actually where cameras differ the most since they are all pretty solid at high ISO now. But at low ISO some have one to THREE AND A HALF stops better dynamic range than others. For many scenes that doesn't matter at all, but for many potential scenes it could.

70D RAW file appears to show the same old DR as Canon has been stuck with since 2007 :(. I hope it is not their new process sensor!

My guess is Canon's noise problems are not really with their sensor, but with the downstream high frequency electronics. In other words, DIGIC. :( It seems most other manufacturers have moved to on-die processing of some kind...Sony with CP-ADC, and others with something similar...on-die ADC. Canon supposedly had similar technology with the 120mp APS-H (the press releases explicitly called out on-die "parallel image processing", which I can only figure is ADC)...why they haven't put the technology into practice is beyond me.

Yeah it depends upon your definition of sensor. If you are talking solely the photon capture and not a whit else than yes their sensors are very good, even the current ones, I believe, capture more than 14 stops DR, I forget the numbers but it might even be closer to 16. But the electronic readouts/ADC and such that are in some cases on and more or less a part of the sensor too and in some cases just off sensor are much worse and yeah that is where the problem lies. Exmor makes it all part of the sensor in a much different way.

From what I've found around the net, Canon's on-die circuitry is quite good. Their analog CDS is some of the best, which is probably what gives them the edge in terms of high ISO noise performance.

The ADC is slightly parallel (8-channel per DIGIC 5+ these days), but housed off-die in the ADC. They are also responsible for processing a considerably greater number of pixels each than the column-parallel on-die ADC that Exmor has. I think its the high frequency, bucket-parallel nature of Canon's off-die ADC that introduces the most offensive banding noise (the vertical stuff). I think there is some lighter banding, both horiz and vert, added by the non-uniform response of analog CDS transistors (which is apparently the reason why Sony moved to digital CDS).

If Canon could follow Exmor's lead....move the ADC onto the sensor die, go column parallel...even if they did not go fully digital like Exmor, that should improve things. Each ADC would have to work less hard as it would have to process fewer pixels. They wouldn't need to be clocked as high, so wouldn't introduce as much noise. Another trick Exmor employs is moving the PLL off into an isolated corner of the sensor die, away from the ADCs themselves, along with all other high frequency components...which apparently helps reduce noise introduced during ADC as well.

I'll grant ankorwatt ONE thing here: A smaller fabrication process, 180nm or smaller, would certainly help Canon achieve success in migrating more logic onto the sensor die.

Anyway, as I was saying though, the Canon bean counters maybe hear all the talk about video being the next big thing and figure putting the money into whiz bang stuff like dual-phase AF (especially since their mirrorless stuff was getting killed in reviews for AF and that market is apparently big in Japan) makes more sense since when they look into forums and see everyone who tries to complain about DR get driven away and the ones who stay getting mocked as DRip DRippers and figure they don't need to bother, nobody cares about DxO scores or trying to capture more types of scenes (as more DR allows) and maybe so long as people defend them to the ends of the Earth in the forums and beat down the Drippers and Ankorwatt and such and people don't get scared away and they keep decent sales they maybe figure safer to be conservative and not lose out jobs than push forward and spend more to improve DR no matter how small the risk just in case ROI takes some time. It's also harder to get fired for sales not improving a lot and not totally dominating compared to Nikon (as I suspect would have happened had they pushed top AF into 5D2 and top sensors into 5D3/60D) and easier to look bad for a big one quarter expenditure to push stuff like that forward? Who knows.

I don't personally have a problem discussing DR, even begging Canon for more. I've argued, here on CR, both sides of what I guess is a rather fine line...not long ago I got stuck in a DR debate and had to produce my own photos with 14 stops of DR to demonstrate how Canon lags behind the rest of the industry. I don't deny they need to work on it. I would love to discuss it more.

I do have a serious problem with ankorwatt, however. He is unfathomably narrow sighted, makes ridiculously inaccurate claims on a regular basis, has an unbeatable bias against Canon, and seems to do everything he can at every opportunity to bash and beat down Canon as an incompetent, incapable competitor that produces products unworthy of even being spoken about, rumormongored about by Canon fans, etc. He interjects debate into every single thread, makes bold claims that are completely false, obfuscates at every opportunity, directly insults and berates other members (particularly about their intelligence), and is otherwise the largest pain in the rear end I can imagine. I greatly dislike him, and for some reason, despite considerable effort to leave the issues be...I just can't seem to let some of his comments go. I think everyone, including myself, is sick and tired of hearing about how much ankorwatt thinks the D800 and Sony Exmor are the greatest thing since the Breath of Life itself woke man up out of the dust of the Earth, and yet none of us can get away from it. He's always there, either starting a new argument or berating some member, always over the same thing even when the topic has absolutely nothing to do with DR, sensors, or the D800. I think he is the root cause of 90% of the debates on this forum, the ever-present antagonist, and I think we could have far more civil discussions about dynamic range, sensor technology, etc. if he wasn't around pushing every Canon-fanboy button he can find all day every day. I try to exit debates with him as quickly as I can if I end up in one, and I try to ignore as much else that he says as possible. One thing I just can't let be, however, is anyone believing him when he says that Canon is incapable of competing in the marketplace, is incapable of building a new fab and therefor must use other manufacturers technologies, or that the only way to produce good photos is to get a D800. If someone wastes their money believing that crap, dumps their kit, and buys into Nikon only to find out they despise the ergonomics, or the menus, or any number of other Nikon features that are inferior to Canon's...well, I'd always feel I could have spoken up and given them a more realistic basis upon which to base such a decision.

I do think Canon needs to improve their fabrication technology and get up to date with modern sensor design. I don't think anyone here really thinks otherwise. I think everyone who needs it would LOVE to have the kind of DR other cameras have. I think people react to the antagonism they feel from ankorwatt specifically, and maybe a few other non-Canon members generally, so the reaction is to defend their chosen brand, rather than have a more civil discussion about the rumors and technology that interest them. It isn't really the DRivel...sometimes its just drivel...and people get sick of it. It would be awesome to have a normal discussion about DR and sensor technology without having to fight about it over idiotic points that don't really matter in the end. It would be awesome to read a Big MP or 7D II rumor thread without wondering on which page the pointless debate is going to begin, and whether I'm going to read something I just plain and simply can't resist responding to because its so asininely idiotic...and yet believable enough that some poor sap is going to read it, think the wrong thing, and end up wasting their hard earned money jumping brands without first having all the facts. Bah.

Anyway...I agree. Canon needs to enter the modern age. They need to improve their sensor technology. They need to improve their dynamic range. They need to respond to their customers requests and deliver. I think they can, too...they certainly did with the last round of camera releases (the big outcry before the 1D X was "fewer megapixels, higher ISO w/ less noise")...unless they really truly have become bean-counter central and despite their capability, are really only interested in the bottom line. (If that IS the case...well, at some point the competition's better sensor technology will eventually hurt their bottom line, forcing them to react in kind.)

1328
EOS Bodies / Re: EOS 7D Mark II Information [CR2]
« on: August 07, 2013, 01:59:29 PM »
.
Nothing ignites passionate argument like Religion, Politics, and ... DOF on APS-C

And DRoning about DRivel.  :P

1329
EOS Bodies / Re: 70D what do you think?
« on: August 07, 2013, 01:07:56 PM »
There is little difference in IQ for any DSLR at ISO 100.  You are buying features and hype.
 
For landscapes,  more MP will give you more detail, but as MP count goes up, so does the difficulty in actually getting that extra bit of resolution.  There are some who do very well with high MP cameras, but the average photographer is not obsessed with a incremental increase in resolution, the more important elements such as composition and subject, lighting, DR, etc put and IQ improvements far down the list.   In fact, I don't even know how to measure something hypothetical like IQ that only exists in the mind of a photographer and is different for each of them.  We can measure several parameters, and they trade off against each other, but IQ?  That sounds like something DXO would assign a number to ;)

There is a pretty huge difference at ISO100. That is actually where cameras differ the most since they are all pretty solid at high ISO now. But at low ISO some have one to THREE AND A HALF stops better dynamic range than others. For many scenes that doesn't matter at all, but for many potential scenes it could.

70D RAW file appears to show the same old DR as Canon has been stuck with since 2007 :(. I hope it is not their new process sensor!

My guess is Canon's noise problems are not really with their sensor, but with the downstream high frequency electronics. In other words, DIGIC. :( It seems most other manufacturers have moved to on-die processing of some kind...Sony with CP-ADC, and others with something similar...on-die ADC. Canon supposedly had similar technology with the 120mp APS-H (the press releases explicitly called out on-die "parallel image processing", which I can only figure is ADC)...why they haven't put the technology into practice is beyond me.


because they sell cameras and have only invested in 2 fab  lines which one is old and can not produce column vise ADC at the sensor = 500nm feature size, which means that it cannot be used for small pixels and =higher resolution with the accuracy needed
120mp is from fab line 2 = the compact camera line where they( Canon) can produce  smaller circuits but it takes resources from the compakt line
So it is costs vs better and newer solutions in APS and 24x36
People buy anyway Canon because the brand name is so  strong and  they think  Canon makes the best sensors
And Canon says , Who needs more than 21Mp it is the optimal amount of Mp
Right?

No, people buy Canon because they think canon makes the best CAMERAS. Canon DOES make some of the best cameras.

People who don't care about the camera and only care about the sensor buy Sony or Nikon...and when Canon comes out with their 40 or 50mp camera, many of those people will switch (again, or for the nth time), and when say Toshiba produces a better sensor in the future that Nikon uses in the future D900, those people sill switch AGAIN, etc. etc. ad. inf.

1330
EOS Bodies / Re: Canon 70d RAW Samples
« on: August 07, 2013, 11:44:55 AM »
Also, if you enjoy a good laugh here's the newest crop sensor vs the good ol' ff from the 6d :-> ... the larger sensor nearly has a 2 stop edge, though probably not at the same dynamic range:

Thanks for the comparisons. Pretty excellent demonstration of the value of larger pixels at high ISO. :)

1331
EOS Bodies / Re: 70D what do you think?
« on: August 07, 2013, 11:20:58 AM »
There is little difference in IQ for any DSLR at ISO 100.  You are buying features and hype.
 
For landscapes,  more MP will give you more detail, but as MP count goes up, so does the difficulty in actually getting that extra bit of resolution.  There are some who do very well with high MP cameras, but the average photographer is not obsessed with a incremental increase in resolution, the more important elements such as composition and subject, lighting, DR, etc put and IQ improvements far down the list.   In fact, I don't even know how to measure something hypothetical like IQ that only exists in the mind of a photographer and is different for each of them.  We can measure several parameters, and they trade off against each other, but IQ?  That sounds like something DXO would assign a number to ;)

There is a pretty huge difference at ISO100. That is actually where cameras differ the most since they are all pretty solid at high ISO now. But at low ISO some have one to THREE AND A HALF stops better dynamic range than others. For many scenes that doesn't matter at all, but for many potential scenes it could.

70D RAW file appears to show the same old DR as Canon has been stuck with since 2007 :(. I hope it is not their new process sensor!

My guess is Canon's noise problems are not really with their sensor, but with the downstream high frequency electronics. In other words, DIGIC. :( It seems most other manufacturers have moved to on-die processing of some kind...Sony with CP-ADC, and others with something similar...on-die ADC. Canon supposedly had similar technology with the 120mp APS-H (the press releases explicitly called out on-die "parallel image processing", which I can only figure is ADC)...why they haven't put the technology into practice is beyond me.

1332
EOS Bodies / Re: EOS 7D Mark II Information [CR2]
« on: August 07, 2013, 12:23:51 AM »
Actually, DOF, by definition, is this:

Code: [Select]
DoF = (2 * N * c * f^2 * s^2) / (f^4 - (N^2 * c^2 * s^2))
The factor for CoC, circle of confusion, is c. It is effectively arbitrary. [...]


I suggest you read this first:

http://en.wikipedia.org/wiki/Depth_of_field

Do not miss this paragraph:

If the original image is enlarged to make the final image, the circle of confusion in the original image must be smaller than that in the final image by the ratio of enlargement. Cropping an image and enlarging to the same size final image as an uncropped image taken under the same conditions is equivalent to using a smaller format under the same conditions, so the cropped image has less DOF. (Stroebel 1976, 134, 136–37).

or this:

Note that the acceptable circle of confusion values for these formats are different because of the relative amount of magnification each format will need in order to be projected on a full-sized movie screen.


"to be projected on a full-sized movie screen"

That would be an enlargement by a factor of 20x, 30x, maybe more? Of course CoC is going to matter with such an enlargement, despite the fact that you sit back by 30 to 50 feet. I don't think most photographers enlarge much more than 2x, and the vast majority of photographers reduce their images (usually quite considerably) for publication online.

As for the quote from Strobel, in the 1970s, film was the primary means by which photography was done. CoC was usually considerably larger back then than it is today with pixels less than 10 microns in size. Particularly in the case of large format cameras. With such large CoC sizes, it was a more important factor, even for something as relatively simple as a 2-3x enlargement. More recent films manufactured with more modern technology have produced silver halide film grains on the order of a few microns in size (one film in particular that was used by Zeiss to test high grade fast optics apparently was capable of resolving 400lp/mm, more than any sensor that I know of as of yet), but generally speaking CoC sizes today are quite small when compared to the film of the 1970s (a CoC that is 2x the pixel pitch of the average pixel size today, which is around 5µm, would be 0.01mm...a CoC for medium and large format film from the '70s would be on the order of 0.2-0.3mm...a difference by a factor of over 20x). It takes a pretty significant enlargement (say projecting on a large movie screen) to make CoC a meaningful aspect of DoF for the majority of photographers today.

I would point out that I speak from experience. I print at 24x36, 30x40, and 32x48 on a fairly frequent basis. For those particular prints hanging on my walls, they are usually viewed standing back ten feet or so...no one has ever complained about my depth of field being too thin or too large, or that the inaccuracies in my 7D AF resulted in a horribly misfocused subject totally unworthy of such a large honor. ;)

Anyway...points have been made. The debate, once again, is going nowhere. I'm going to bed. It's up to you and Neuro now. :P

1333
EOS Bodies / Re: EOS 7D Mark II Information [CR2]
« on: August 06, 2013, 11:53:11 PM »
Keep in mind the timeframe. Five years ago FF was much more expensive to manufacture than it is today. The fact that the 6D, still relatively new, sells for as little as $1700, is quite telling here.

It seems that Canon's fab procedure hasn't changed, at least from what we can tell.  Has the cost of silicon wafers come down that much?  Or could it be that Canon was reaping very high profit margins on the presumed high cost of a FF sensor, and now they've decided to push more units at a lower profit margin, as an alternative strategy to drive the bottom line?

Just sayin'.  Not that I'm cynical, or anything...   ::)

I do believe 300mm wafers have come down in cost. I remember them still having some challenges five to six years ago with defect rates on them (this is pretty agnostic of industry...not specific to sensor fabrication). It is obviously a less serious problem for tiny chips like GPUs and CPUs or other ICs. Growing the wafer crystals has become more refined over the years, in no small part to some of the advancements made while trying to perfect the process for growing 450mm wafer crystals (which, as far as I know, has still not been taken up by any IC manufacturing industry...there is apparently a very high initial cost to jumping that has to be recouped, something no manufacturer seems willing to deal with as of yet).

As far as I know, it's not that actual growing of the Silicon ingots to 300mm (or 450mm for that matter) that is the yield/cost issue. Getting very high quality and uniform ingots has always been extremely important to high-voltage devices (think kilo-Volts), because spikes in dopant concentration can result in avalanche-type catastrophic breakdowns (leading to the magic blue smoke leaving the devices in spectacular ways).

It's the yield of the manufacturing processes for everything you on the base wafer that matter (diffusion, implants, etching, photo lithography, metallization etc).

The problem for the jump to 450mm wafers is that you need the entire infrastructure of a fab to be in place: steppers, aligners, photo lithography, and so on. It's so expensive to develop, that basically the entire IC industry needs to coordinate around it; it's not enough that Intel or TSMC (or both) says "hey, we want to do 450mm now" and presto! they have it.

Agreed on the last part...it is an industry-wide resistance to migrating to larger wafers. New fabs are built at an increasing rate, however, as more IC devices are used in an ever increasing array of applications. I figure someone would have put the money into building a new fab capable of 450mm by now...

I did read an article some time back, maybe seven or eight years ago now, that indicated that growing the crystal did not always produce perfect, uniform growth, and that further refinements of the process could improve quality and render an increase in usable 300mm wafers and fewer defects per wafer. I don't remember if it was online or in a magazine.

1334
EOS Bodies / Re: EOS 7D Mark II Information [CR2]
« on: August 06, 2013, 11:42:14 PM »
I probably shouldn't even try, because it seems that you won't be convinced, but why not one last go?   ;)

If you do not change FL and distance, and keep it at f/2.8, DOF changes; cropping makes in shallower. AF errors are magnified.
Sorry, no.  Cropping doesn't change DoF.  Cropping then magnifying does change DoF,

Of course, you magnify when you crop (by crop, I mean use a smaller sensor). Do you use a smaller screen do display your smaller sensor photos? DOF by definition is relative to a reference print size. That size is kept the same regardless of the sensor. Do I need to spell out everything?

Actually, DOF, by definition, is this:

Code: [Select]
DoF = (2 * N * c * f^2 * s^2) / (f^4 - (N^2 * c^2 * s^2))
The factor for CoC, circle of confusion, is c. It is effectively arbitrary. One does not necessarily know what the final output size of their photo will be in the end, or even if there may be multiple output sizes. If one wishes to be as concrete as possible, the CoC is physically limited by pixel size. To be "safe" when using a bayer type sensor, one should usually use a CoC that is at least twice the pixel pitch to account for the uneven sampling. Because of the even sparser nature of red and blue pixels relative to green, and due to the fact that an AA filter is usually used, it is better to use a CoC roughly three times the pixel pitch. That would be the only truly concrete definition of CoC for any given sensor.

Assuming one uses the pixel pitch x3 for CoC, then that greatly simplifies the initial argument, and do away with the notion of a reference print size. One could assume that the pixel pitch for a FF sensor and an APS-C sensor are identical. If that is the case, then one could photograph the same subject with the same lens at the same distance with both sensors, crop the FF to the same image dimensions as the APS-C, and the depth of field will be 100% identical in every respect, regardless of what size the images are scaled to. Therefor, depth of field has nothing to do with crop factor or field of view.

Nor, for that matter, does it really have anything to do with a reference print size. I would also offer the argument that even in print, as print size increases, so too does the most comfortable viewing distance. If you have the luxury of 80mp of MFD goodness, you might be able to print a highly detailed photo in an immense 40x60" size at 360ppi, drawing your viewers to within a few feet to examine all the detail. It you are incredibly meticulous, careful scaling might eek out enough detail do do something similar from something like the D800 or a hypothetical 40-50mp Canon FF. Generally speaking, as print size increases ppi drops, and so too does the viewers desire to stand back farther and farther to take the whole thing in. As visual acuity is also a function of distance, CoC could, for all intents and purposes, be a constant...and therefor a non-factor when one needs to determine their depth of field. You can pick whatever CoC your "comfortable" with, use that same value every time you compute DoF...at which point the formula above proves the point once and for all.

I guess therefor that one could then state that DoF is, by definition, purely a function of the lens and relative to viewing distance of the final output, at the time the photo is taken. ;)

1335
EOS Bodies / Re: EOS 7D Mark II Information [CR2]
« on: August 06, 2013, 10:24:43 PM »
I probably shouldn't even try, because it seems that you won't be convinced, but why not one last go?   ;)




 ;) ;D

Pages: 1 ... 87 88 [89] 90 91 ... 216