Martin Mojzis / Fine Art Photography and Graphic Art Atelier.

NIKON VERSUS CANON II.

ANSWERS TO QUESTIONS AND ATTACKS.

Ladislav Kamarad claims, that he has been dealing deeply with the issue of digital photography since its inception, and my article is allegedly based on a lack of understanding of deeper contexts:))
No need to go far for an answer. Just look at one astronomical forum, where the issue of AA filter is discussed. LK writes here, among other things: "... Otherwise, be careful with that LP filter - it could then cause moiré during ground photography (this is not very dangerous for stars and DS objects). I have the impression too, that the "1st LP filter" is correctly called AA for the EOS 5D mk2 (AntiAliasing) filter and some recommend removing it for general photos. But I also caught a pretty serious 5D test with and without of this AA filter with the result that without the filter the subjective sharpness is higher but it is caused only by aliasing artifacts on the contrasting edges, while the actual resolution was really lower! It is logical in the end."

Judge for yourself, how it is with the depth of dealing:) I explain the principle and function of the AA filter at the bottom of this article, so that even those who have not yet dealt with this issue will understand.

Next to the guest book LK - even though its owner declares himself untalented for marketing, this book is a good example of thoughtful media manipulation. All contributions are approved in advance and those that do not suit the author will simply not appear. And vice versa, some may even arise on their own, who knows :) So if someone insults you or otherwise damages your reputation, you do not have the opportunity to react or defend (of course, I wanted to respond first in the book, without a chance, to e-mail emergencies with LK pretending to have his head deep in the sand:) Well, it's a normal mess for me, and let everyone take a picture of it themselves.

Finally, I present a message, which I then tried to insert into the book, but LK did not publish it: You really amused me. My contribution and article mention only the final, marketing impact of the whole issue, I do not break down the technical side of the thing, which I am familiar with in detail, because everything important can be easily found, for example, on the Wiki. The stylistics of the article and the article are intended to amuse the reader and to lighten the whole thing. MM

I will not deal with the article with the title Petr at all, another cowardice. Such that when I wrote it off (LK did not allow me to enter the answer directly again :), the e-mail was returned to me from the server, stating that such an address does not exist. Either LK wrote it himself, or he was so afraid that I would answer him as he deserved ... who knows. The post already contained swearing and attacks on me. To laugh:))

Next "pearls":))
1. Manual selection of the lens according to the display on the camera according to the writer of the article is not possible, moreover, it is said that based on the lens selected in this way, the whole camera is bad.

I would like to assure the writer, that it would not really work on a display such as the Nikon D100 or Canon 300D. On the contrary, with today's 3" VGA displays (3.2", 1.15 Mpx at PhaseOne IQ walls) this is absolutely no problem if:
- you are not blind
- you are aware of what an aperture is and how it is used
- you have a tripod
- you know that the preview on the screen is jpg, even if you are shooting in raw.
If the writer wishes, I will choose any lens for him. Full warranty. Of course it will be expensive:)
Regarding the alleged finding that the 5DMII is bad - if the person in question was careful at school and was able to read better thanks to it, he would find out that I am writing at the end of the article: "By no means do I want to convey that I recommend for the Canon EOS 5D Mark II to be a bad camera. It's just not suitable for a long-time Nikon owner, who also has very high demands on the quality of the drawing details. In addition, this article does not want to be a test - it is only a summary of the subjective experience of a professional photographer." And I assure you, dear writer, that if Canon comes with a body over 30 Mpx, which will have an AA filter at least at the level of professional Nikons, AF from at least 7D, an intelligent price and Nikon will still pretend to be the year 2007, I will easily buy it to my nikon equipment and leave it for free creation and contracts, where high ISO is not needed:))) So.

2. Another writer states, that he "compares the drawing strange and aa filter is already very thin ice."
That thin ice got me - huh, this is about life!!! Bugaboooooo!! Behind chimney!!!:)))
As for the examples, I will add some more time to the original article, perhaps they will be more understandable. Regarding the ice, here is a link to an article that was published on the known Luminous Landscape server in the 5D Mark II test, Chapter Resolution. Professional wildlife photographer Nick Devlin tested it very conveniently on ice (perhaps not thin) in Alaska and writes, among other things: "... But even then, I can't help but feeling that a very aggressive anti-aliasing filter is at work. While the 5DII produced beautiful files, they lack the "bite" in the fine detail that a perfectly exposed Phase One files shows. ..."
Whole article you can find ►here.

The 3rd writer experienced this: "- Comparing wheels is a mental disorder." People, think :) I had so much fun with it, I wrote it with exaggeration and I really wouldn't have thought in a dream that someone would take it personally and start insulting me for it. Plus, if the writer was a professional and held the camera in his hand sometimes even 20 hours a day, then he would solve not only the wheels, but also the structure of the coating rubber and the texture on the metal parts:))

The 4th writer determined the ortel himself, because the issue of Bayer's mask is already well known among photographers ("the bullshit about Bayer's mask and its influence on drawing really got me")
I do not recommend users with this level of knowledge to participate "highly professional and sophisticated" discussions - as they prove above mentioned contributions of authors 1 - 3:)) Rather, I would recommend something family-friendly, such as a digineff, the chapter "How does our cheerfully clicking digital switch actually turn on, cudliky budliky?" Read first and then gradually, slowly, spread over several years. Do not overload. And it's also about dogs and that's good.

5. This writer also took my photographs (finally someone!). They say they will do quite well, but in the post-production they allegedly "make colouring books".
Dear, colouring books are black and white pictures, that do not have filled areas, so that children can learn to master crayons and a little bit of color for the more talented. After reading your post on my website, I searched for these images several times, but in vain - they are all already colored, even if sometimes only grayscale: ( yes, I admit ... I change the color of my photographs, and also the tonality, cutouts, gradation ... do you know that for more than a century, the photographer has not been just a documentary? It can also be an independent creator, an artist (and that painters could go crazy not to do so), who can manipulate his work in any way ... In our country it was mainly enforced by Anna Farova ... nothing? It may be a scary finding for you, but there are other photographers too, than those who take photos for Blesk, Playboy and on a soccer match ... :))

So much fun reading over the letters of the audience. The next part will be only dry, austere and professional. For the purposes of this article, I deal only with the image, aliasing problems occur wherever the continuous signal is converted to discrete (digital).

What is an anti-aliasing filter (AA filter, blur filter, Low-Pass filter - all labels mean the same thing) and what is its purpose?
When resampling an analog signal to digital, if a frequency higher than half the sampling frequency (so-called Nyquist frequency) occurs in the original continuous signal, the signal is distorted due to a phenomenon called aliasing. Therefore, the number of camera sensor points per unit of length must be at least twice the maximum number of lines that will be projected by the lens onto the same unit of length. If this is not observed, aliasing occurs, which manifests itself in color moiré (irregular color patterns in the affected part of the image).

To avoid moiré formation, an AA filter is placed in front of the sensor, the purpose of which is to filter out frequencies higher than the Nyquist theorem. The filter is typically a two-layer, lithium-niobate compound with a so-called double-fracture effect: each optical point is split into a cluster of four points after passing through this filter. This, of course, significantly reduces the sharpness, with the fine details completely disappearing from the picture. Producers try to find a trade-off between this blurring and preventing moiré. For multi-sensor systems (Foveon, 3-CCD, etc.) moiré does not generally occur and therefore the AA filter is generally not used here.

Now remember the principle of the Bayer mask, i.e. a colour filter is placed in front of the colour-blind sensor, half the pixels now take the green colour to which the human eye is most sensitive, and the other half split between red and blue. To get an image on the output, consisting of almost the same amount of pixels as the color-blind sensor, the remaining pixels must be mathematically calculated = invented:)

If we now realize that the sensors, commonly used in cameras, combine the two, for the image, completely killing factors, it is perhaps now clear to everyone, what our photographs would be technically better, if manufacturers for digital cameras hadn't taken the Bayer mask sensor from their video cameras, or at least abandoned it after they'd made enough money, and the big players, i.e. Nikon, Canon, Sony, Kodak and Dalsa, instead went on to produce 3-CCD (which Sony makes, but only for video cameras) or Foveon sensors. Sigma's loneliness in doing so, and its SD-1 at a completely meaningless cost, seems brave but somewhat futile in this context.

© Martin Mojzis.
Prague, October 2011.

►To the beginning of the article.
►Overview of all articles.
►Homepage.
►Site Map.