Messages

1

EOS Bodies / Re: Patent: Canon Foveon Sensor

« on: May 24, 2013, 11:39:56 AM »

Each pixel of the new Canon prototype sensor can capture all the colors.

This is made possible by the physical fact that long-wave (infrared) light penetrates deeper into Silicon than shortwave blue light.

Red light penetrates most deeply into Silicon, green light penetrates only up to the middle and blue light reaches just below the surface of the Silicon.

With the sensor informations the camera system can calculate, which color the pixel see.

The sensor filter is no rasterized like the Bayer pattern. You don´t need a anti-aliasing filter that blur the image. And you don´t need to sharpen the picture to get an usable image. The full performance of the attached lens is available in the image after taking the picture.

Not quite accurate. "Full performance" is a non-linear factor when it comes to lenses. A high quality lens, such as one of Canon or Nikon's super telephoto lenses, offers maximum resolution at their maximum aperture. With say a 600mm f/4 lens, at f/4, the maximum potential spatial resolution is 173 lp/mm. Currently, no sensor on earth offers comparable spatial resolution. The highest resolution DSLR sensors, Canon's 18mp APS-C and Sony's 24mp APS-C, offer (with the most optimistic measure) 116 lp/mm and 129 lp/mm. Neither of those sensors bring out the "full performance" of any lens, especially not top of the line professional gear.

Why make the new sensor design sense?

You get sharp images, fine details, perfect realistic colors and no moire.

All of these facts, except the "fine details" one, are true. Simple fact of the matter is, with three sensels at every pixel, you have a lot more raw data points to process in any given read than with a bayer type sensor. That affects how fast you can read, which limits how how far you can push spatial resolution with a foveon design. Currently, the fastest DSLR's process at ~500mb/s. A 36.3mp Foveon would need 200mb/s just to read out ONE SINGLE FRAME in a SINGLE SECOND! You DO get full color information at every pixel, but there are bayer sensors capable of resolving nearly as good color information, just as sharply, with FAR more fine detail.

This is one of the fundamental drawbacks of foveon-style sensors. They are sensel dense, but not pixel dense. From a spatial resolution standpoint, foveon is about 2/3rds LESS capable of resolving fine colored detail than a the top bayer sensors (factoring in the spotty spatial resolution of the red and blue channels). From a purely luminance standpoint, Foveon is about half as good or worse than the best bayer sensors.

The argument for Foveon is often made in the form of: "You can always scale Foveon up and get just as good IQ as a bayer". To some degree, this is true...with the exception of the spatial resolution (fine detail) factor. However the inverse is most definitely true: "You can always scale bayer down and get just as good or better IQ than Foveon."

The question is, what you want to shoot. The new sensor is very good, but not good for all. If Canon put the new sensor on the market they have also products with the normal sensor design on the market.

I would agree with this...assuming Canon ever puts the sensor on the market. I think MCS designs will probably ultimately take over in the long run. Increasing SNR will be critical to achieving higher and higher usable ISO settings, and I think any sensor design that relies on the filtration of light will eventually fall in favor of designs that maximize how much light is used, eliminating the need to filter at all.

It´s a little bit like AF-systems. the phase-AF is faster but the contrast-AF has a bigger hitrate. I see that the future of the AF-systems is a combination of both passive AF-systems.

Sorry, but the AF system comparison is no longer true. According to LensRentals, modern Canon equipment (newly released lenses paired with a new body like the 5D III or 1D X) is capable of producing just as high a hit rate with PD-AF as with CD-AF. PD-AF is now just as accurate as contrast methods, but considerably faster. For the types of photography that rely heavily on high speed AF lock and tracking, contrast detection will NEVER be sufficient, even when paired with phase detection. The raw performance of PD-AF with a dedicated sensor will never be surpassed, and I do not see AF-critical photographers ever choosing a camera that combines PD-AF with CD-AF.

2

EOS Bodies / Re: Patent: Canon Foveon Sensor

« on: May 23, 2013, 06:58:21 PM »

I think the Foveon+Sigma story is an excellent example of how camera BODY and its functionality overall is significantly more important than just the sensor.

That might be true, but then that's why I like to purchase the sensor in Sigma's smaller and less costly DP series body (or I should say "camera").  Certainly the initial price of the SD-1 was absurd, and a bit of a fiasco...and the current SD Merrill is still not enough camera for the money.

And certainly Sigma makes very few lenses that are capable of making full use of the Merrill sensor's resolution.  Perhaps the new 35mm f/1.4, and a couple of their superteles...Again, that's why I like to purchase the DP series camera, because their lenses can and do impart the full resolution onto the sensor.

But I thought this discussion was really more about the sensor itself, rather than a convenient opportunity to slam Sigma for producing less than competitive DSLR's...So it's kind of sad that it has suddenly gone in that direction.

Please, don't assume you know my intent, and don't put words in my mouth. I was not being opportunistic or gleeful about the option to slam Sigma, I was simply stating a fact. The FACT is, they produce an inferior DSLR. It isn't a slam, I am not sadistically getting a rise for bringing the point up. It's just a fact (even according to DPReview, the SD-1 is a real mixed bag, and has some very glaring flaws, quirky dial functionality, etc.) I know certain people like them, but numbers speak loudly, and if Sigma's cameras were better, the numbers would speak to that. The people I know who own Sigma DSLRs own them for the sole purpose of having Foveon. Few ever really bring up the body features or functionality unless the discussion takes a turn for the worse, and they enter full blown defensive mode. Interestingly, but not really surprisingly, nearly every single person I know who owns a Sigma DSLR with a Foveon is a landscape photographer, with one who does portraiture.

In the current flow of discussion, a point was made about why Sigma's Foveon isn't selling because the technology is inferior for one reason or another. I simply wanted to point out that it is less likely that the technology is inferior in general (it most certainly has its strengths, and it excels where it is strong), but Sigma isn't really the company that can bring Foveon to full bear on the market against giants like Nikon, Canon, and Sony. The truly SAD thing is that it is Sigma who owns Foveon, and that they can't seem to execute it's success. Again...I'm not opportunistically trying to bash Sigma here...its just an empirical fact (something based on years of observing Sigma fumble around with their priceless Foveon football, and never really making it, nor the camera bodies that house it, into the end zone...to my own dismay, I might add.)

3

EOS Bodies / Re: Patent: Canon Foveon Sensor

« on: May 23, 2013, 02:06:35 PM »

well MTF diagrams  gives you good information about the lens you are picking, so does this curves about problems with 3 layers of filter, there are other constructions  , read earlier answer
IF there had  been only minor problem with a Foveon or similar construction you can be sure there had been  sensors  out on the market since  years back
And Foveon is not the first with a construction like this.They are the first to do a commercial product

Got any diagrams on how those problems are/should be solved?
I think that scientific method is the best, except when people start using it religiously, like "... this is the only way, now and forever, amen" or "... my book says X, so your book is wrong, because my religion is the right one".
Perhaps one of the reasons why Foveon is not very popular, is that it requires more in-camera processing power, which results in slow shooting speed and short battery life.

I would say the biggest reason Foveon doesn't sell is they are stuck in Sigma cameras. Sigma is NOT known for producing a high quality camera body, nor is it know for high quality or high end DSLR features and functionality. Their menu system is a joke. Foveon has some EXCELLENT strengths, and for types of photography that do not require high ISO (i.e. landscapes), it is an excellent design. The problem is that Sigma owns it, and they just plain and simply don't make a very good camera. Personally, I find that to be a sad state of affairs. I think Sigma purchased Foveon thinking the sensor itself would bring in the sales.

I think the Foveon+Sigma story is an excellent example of how camera BODY and its functionality overall is significantly more important than just the sensor.

4

EOS Bodies / Re: Patent: Canon Foveon Sensor

« on: May 22, 2013, 01:17:58 PM »

A layered, Foveon-style sensor is definitely intriguing. It has a lot to offer, but at the same time, it also has it's limitations. From a raw spatial resolution standpoint, Foveon will probably never compete with Bayer...at least, not with the same low levels of noise. Foveon can be upscaled, but the results are not quite the same (remember, bayer has limited color resolution, but it still has full luminance resolution.)

The most intriguing technology I've seen to date that maximizes resolution and light gathering/quantum efficiency is Panasonic's MCS, or Micro Color Splitting. Instead of FILTERING colors, red light is split from the incoming light at each pixel, and directed to the appropriate neighboring pixels. The result is White + Red and White - Red pixels that, when interpolated, contain all of the color and luminance information that reached the sensor (there can still be losses due to IR Cut and transmission losses, as well as due to Q.E. losses).

Unlike either Foveon or Bayer, an MCS sensor achieves high spatial resolution as well as low noise. I'd really like to see Canon develop something like this, instead of a layered sensor...I think the end result is even better high ISO performance as well as more color-rich low ISO performance. (And, theoretically, by gathering more light overall, one would suspect dynamic range to improve as well...but I don't really know how an MCS design might affect read noise...which is really the fundamental issue Canon faces these days...noisy high frequency downstream components adding gobs of read noise.)

5

EOS Bodies / Re: Patent: More DO Supertelephoto Lenses

« on: May 21, 2013, 09:35:13 AM »

However, as with all patents, its easier said than done, so I'm not overly optimistic about seeing a low cost DO implementation soon.

Very true...however the entire optical engineering world told Canon that making a viable diffractive grating lens was impossible, and yet, Canon persisted and eventually succeeded. ;P I suspect Canon will figure out a way to make a diffractive particle dispersion lens at some point as well...or something else sufficiently superior to their grating lenses as to solve the IQ problems, and still support the construction of smaller, lighter weight lenses for long focal lengths. Canon is nothing if not persistent.

6

EOS Bodies / Re: Patent: More DO Supertelephoto Lenses

« on: May 20, 2013, 08:11:08 PM »

I owned the 70-300 DO lens for a couple of years. While it wasn't a bad lens, the DO did take a bit of a toll on IQ, particularly in situations where lens flare was likely (strong light source striking the front element). Bokeh was a little awkward/choppy at times also. I understand that the 400/4 DO is a very good lens.

I'd like to see a refinement in the DO design that would produce an even better optical quality. The light weight and relatively small package makes a strong selling point for this technology.

Canon has a number of recent patents for new DO lens technology, but it seems to be exceedingly difficult to manufacture.  It is based on small particles immersed in a resin compound.  The resultant radial dispersion of particles can produce a superior lens element that varies its properties radially.  The issue is getting the spacing of the particles right in a consistent way, and that may never happen.  You can't just pour a powder in a vat of resin and mold a bunch of identical lens elements.  Every one will be different.
 
They did not specifically mention this technology in the patent, but did talk about a resin compound, so its in there somewhere.

It may not be as easy as simply mixing "power" with resin an getting ideal results every time...but I would be willing to bet they could mix lightly charged particles with slightly varying charges in resin, and before it cools use a magnetic or electromagnetic field to disperse them properly.

7

Animal Kingdom / Re: Show your Bird Portraits

« on: May 19, 2013, 11:35:13 PM »

American Kestrel brunch

Nice shot. Might want to heal/clone out the bits of the leaf that encroach on the bird...otherwise, he's a beauty.

8

Animal Kingdom / Re: Show your Bird Portraits

« on: May 19, 2013, 11:34:35 PM »

Everglades Osprey dining.

Awesome shot! Love the sharpness and clarity of the bird...I love Ospreys, beautiful creatures!

9

EOS Bodies / Re: How will Canon respond to Magic Lantern's raw video hack?

« on: May 19, 2013, 09:25:54 PM »

simples.

canon will sell a ton more 5Dmk3 bodies and a couple of tons more L glass to video people or all these new IS prime lenses
same thing that happened when the 5Dmk2 hit the scene
Sold like hot cakes

I agree.

10

EOS Bodies / Re: EOS not good for NASA?

« on: May 19, 2013, 09:19:07 PM »

well  you are wrong about high iso, if you shall have the same results from 1dx and D4 and brightness in the images you have to expose the Canon more/richer/longer  than Nikon, so overall there is no difference in high iso between the two cameras if you look at the raw file regarding signal /noise. About DR and low iso, base iso the 1dx has no chance against D4 due the signal/noise, go out and try d4 and 1dx side by side at high iso and you know/ se  what Im is talking about "real life shooting as some people called it"

Please, take a look at my "real life shooting":

http://jonrista.com/photography/

Everything here is taken with a lowly 7D with its utterly crappy sensor and excessively inferior technology. Put a 1D X in my hands, and I'll do and order of magnitude better. But, of course, my work can't possibly be better than anyone's work shot with a Nikon camera...oh, no! Not a chance!

Oh, and regarding "real life"...show me some real-world D4 photos that exhibit better high ISO performance than the 1D X. ISO 25600 and 51200 are both digital boosts on the D4...that results in more read noise intruding into the image, and it is usually quite apparent. I've seen near-noiseless photos taken with a 1D X at ISO 51200...stuff that compares to my 7D at ISO 800!!! I've never seen anything even remotely resembling that from a D4, D800, or any other Nikon camera.

11

Lenses / Re: Canon Tele lenses vs Nikon tele lenses ( both with converters )

« on: May 19, 2013, 06:05:56 PM »

Yes the 300/2.8mk2  from  canon looks very good in the MTF lab and Hasselblads measurements

and much more importantly it works well when you use it to take real photographs. 

Had a good time with mine today.

Do you use it with the 2x converter at times?

Also im close to ordering the 500 lens foot to put on the 300 for comfort!, i hope that it will fit....
Since the 500 comes with both a small and big foot , and the small one looks identical to the 300 foot I am kinda anticipating it will!

Will provide a bit more comfort for handholding with my hands

I use the 2x TC MKIII and find it very good indeed, that's what I was using today.

here is a shot taken yesterday with mkII 300 f2.8 and mkIII 2x TC. i underexposed by 1 stop that i pulled back in PP. It was overcast and raining but i was looking for a gritty look to match the wire.

I like it. The deeper contrast on the barbed wire definitely looks better.

(Note: Although I suspect you could have achieved the same thing (and with less noise) by exposing to the right, and correcting in post. Contrast of a RAW is really all about the default tone curve...fundamentally, digital sensors are nearly linear devices (there might be some slight attenuation due to the electronics). You never really lose anything with ETTR in a digital camera, and you'll always improve your SNR (and thus, reduce noise.) Think of a raw image more as a digital signal than as an image...it is very fluid and flexible in post.)

12

Lenses / Re: Canon Tele lenses vs Nikon tele lenses ( both with converters )

« on: May 19, 2013, 06:03:10 PM »

Yes the 300/2.8mk2  from  canon looks very good in the MTF lab and Hasselblads measurements

and much more importantly it works well when you use it to take real photographs. 

Had a good time with mine today.

Do you use it with the 2x converter at times?

Also im close to ordering the 500 lens foot to put on the 300 for comfort!, i hope that it will fit....
Since the 500 comes with both a small and big foot , and the small one looks identical to the 300 foot I am kinda anticipating it will!

Will provide a bit more comfort for handholding with my hands

I've used the 2x TC with the EF 300mm f/2.8 L II, and the IQ is phenomenal. Even with the 2x TC, it is still one of the sharpest lenses on earth. Sharper than my 100-400mm at is BEST, by a long shot. Background blur is also dependent upon the ENTRANCE pupil, so even with a 2x TC, your background blur is effectively the same as an f/2.8 lens...so it is a fantastic combination. You really couldn't go wrong with it.

Dear Sir,
Thank you for this posting! I really needed to hear this from someone using the EF 300mm f/2.8 L II together with the 2x TC.
I was reminded that Bryan C in TDP maybe for first time admitted that the 2x TC III from Canon together with this lens was useable.
Of course that is based in the top-notch optical quality from the lens itself because we all know that all flaws and weaknesses from the lens that is in front of the TC will be also magnified.
Thank you jrista for all your postings!
Wish you a very good weekend!
All the Best!

C

You are welcome. Glad to be of service.

To be clear, there is definitely a drop in sharpness when using either TC, however the 300/2.8 II with 2x was still sharper than my 100-400mm at f/7.1 or f/8 (the two ideal apertures). In my experience, ANY of the Canon great whites, at any aperture with or without any TC, are all sharper than pretty much any other lens in the Canon lineup, with perhaps the 70-200/2.8 II as an exception.

I had a long, hard debate about which lens to get next, between the 300/2.8 II and 600/4 II. In the end, I decided the 600/4 II offered me more for the kind of photography I do most (birds, wildlife, and some aerial stuff for model planes and air shows), than the 300mm with a teleconverter. I'll probably still pick up the 300/2.8 II, but it'll be a couple years, as I'll have to recover from the 600mm purchase this year. Regardless...whether you intend to use teleconverters or not, you really can't go wrong with any of the Great Whites.

13

EOS Bodies / Re: EOS not good for NASA?

« on: May 19, 2013, 12:01:54 PM »

It's not a secret that NASA uses Nikon cameras and lenses (http://www.nikon.com/news/2009/1221_NASA-D3S_01.htm) after using Hasselblad cameras during the Apollo mission.
Could we reveal the reason?

very, very simple reason:

Hassy has a bigger camera body ( more easy to manipulate the settings in a space suit)

Nikon sensors have better DR than Canon.... and in space this performance is crucial.

Yes it must be the DR. NASA has been awaiting that for years up until last year when Nikon introduced the D800...

Do you believe yourself what you're writing?

You.... again 

Well, factually speaking, DR is NOT the most important thing in space. NASA bought into Nikon a few years back because at the time Nikon was the king of HIGH ISO. You don't shoot the dark side of the Earth from space at ISO 100...you shoot it at ISO 6400, 12800 and at high shutter speeds to freeze the motion of 17,500mph! At High ISO, DR is physically limited. You lose about 1 stop DR per stop of ISO increase...you have only 7 or 8 stops at those high ISO, so the most important thing is the total electrons per pixel at maximum saturation. The higher the charge, the lower the noise.

Today, Canon rules the high ISO/SNR realm. By a relatively small margin compared to how much Nikon rules the low ISO/DR realm, but enough to give them an edge now. I believe the only reason NASA currently uses Nikon and has not changed to Canon is there really isn't any reason to. They are invested. They have the gear, have the lenses. Why change? They don't need the compelling features of the 1D X...it was built for sports, so it has an AF system and frame rate to match. I'd figure the most compelling Canon camera for NASA would be a 40-50mp FF monster with good ISO 25600 performance...which doesn't exist quite yet.

hmm what do you mean in all this text?

If you brush up on your English, you might understand it. Everything you need to know, my entire meaning, is in the text above.

14

Animal Kingdom / Re: Show your Bird Portraits

« on: May 19, 2013, 11:33:19 AM »

Wilson's Phalaropes, females in this case (which tend to be flashier than the males), out at Cherry Creek State Park in Colorado:

15

Animal Kingdom / Re: Show your Bird Portraits

« on: May 19, 2013, 11:30:54 AM »

Black-bellied Plover - Martha's Vineyard
60d + 100-400 mm

Nice shot! Like the lower angle perspective.