Camera properties DOF grid focus on not working.

uncle808us

Active member
#2
OK I had the camera locked so now it's working. but I still can not figure out how to focus on one object that is in front and have the back ground out of focus every thing gets blurred. I need a tutorial on how to use the DOF in the Cheetah renderer. If any one can help I'd appreciate it. Thanks
 
Last edited:

Helmut

Well-known member
#3
* In the panel for the renderer, activate the DOF checkbox.
* In the property panel for the camera, experiment with the parameters ; Aperture size being the obvious.
* It may help if you have some knowledge of operating a camera / lens manually.
:unsure:I thought it is midnight across the Big Pond?
 

uncle808us

Active member
#4
* In the panel for the renderer, activate the DOF checkbox.
* In the property panel for the camera, experiment with the parameters ; Aperture size being the obvious.
* It may help if you have some knowledge of operating a camera / lens manually.
:unsure:I thought it is midnight across the Big Pond?
I'm usually up till 2 or 3 am. I have no knowledge of operating a camera / lens manually. I will mess with the Aperture size.
Thanks
 

Helmut

Well-known member
#5
* The basics are quite easy:
* Form memory, I set up a trivial test array of 3 simplish individual 3D objects on a straight -Z axis and just tweaked the available settings to get a feel for the logics and the bokeh (see post scriptum) resulting. Make a few notes on the console (or outside of C3D). I would have liked to send you my own notes from the DOF experiments, but I cannot locate them.
:poop:Probably a victim of ill-advised spring cleaning :mad:

* There are 2 modes of focussing:
1 focus on an object in the browser (by pop up or drag & drop)
2 manually select the focussing distance (this is only enterable if no object has been defined)
* I use DOF quite sparingly, obviously it kills any rendering times for animations. For experimentation you can reduce the sampling rate in the render panel at the cost of grainy JPEGs.
* BTW: All aperture parameters can be animated. So, you can start with a totally fuzzy image and gradually focus on the distance / object of interest (or vice versa) or you can simulate a sliding focus.

PS: bokeh is a Japanese term for the out-of-focus areas of an image, mainly used in photography. I don't know a short and concise term for that in English (nor in German).

* Enjoy your experiments, Uncle Bob, and have a pleasant day :)
 
#8
Uncle, if you're really into photoreal renderings I would recommend getting a book or two about photography. There are quiet a few excellent photographers who share their knowledge, and it really helps to get better results in 3d if you fully understand what you try to recreate. It's not just about understanding stuff like lenses, dof, angles and so on. In 3d you sometimes have just to do the opposite of what photographers do to correct their images (like putting in some achromatic distortions and vignettes instead of getting rid of it).

In a world of bad photos made with cellphone cameras it's also helpful to see some real good, composed pictures that will give you lots of ideas (the composition is another important part that's mostly neglected in 3d tutorials). And there is a lot to learn about lighting, colors, crop-factors, enhancements through photoshop-like programs and so on. Most of this stuff is far more easy to accomplish with 3d (a phtographer would kill for blacklights or better ways of getting rid of reflections). And, at least for me, it sometimes is refreshing to look at 3d from another angle.

(In may experience it works the other way, too. My knowledge of 3d helped me to get better photos over the years).
 

Helmut

Well-known member
#9
* I fully agree with Hasdrubal.
* Using a camera provides not only an understanding of some principles of optics (perspective, wide angle vs tele, depth of field, etc), it also makes you aware of "where do I (the camera) stand" or how does light / shade / shadow work in a scene. Most of all, as Hasdrubal implies, it makes you aware of the importance of careful composition of a single frame or a dynamic animated take. There may be some rules, but the basic rule is that they should be broken whenever needed.
* Studying movies is another great way to learn new tricks. In my case - a slightly post-middle-aged canine :sick: - it certainly works. From Fritz Lang to Orson Welles to Bunuel, Kurosawa, Stanley Kubrick and numerous others: Steal their ideas and deploy them in your own work.
* If interested, also look at the work of great painters. I find that studying the composition of anything from Renaissance via Impressionism back to Altamira is creative bliss. Once more, steal their ideas and deploy them in your own work. Most of them are dead, anyway, and can´t sue you :devilish:

:unsure: It is much more fun to stand on the shoulders of giants than to sit on dwarfs.

* Of course, I have the advantage of modelling largely scientific stuff and have a much wider degree of freedom compared to others on this forum. I also have accumulated a small cluster of clients who (pretend to) like my idiosyncratic experiments.
 
#11
While I would say that photography is the most important, basic knowledge, Helmut is of course right when he goes that step further.

If you have a tablet, there are a ton of apps (very often for free) that give you access to the paintings of the masters, wherever they hang in reality. It's inspiring to watch, but a bit more abstract to learn from.

For me comics do the trick a bit more directly, from people like Hal Foster and (he is a 'must' if you're interested in the human form) Burne Hogharth (his Tarzan Stories are often plain stupid, but the drawings are always great). My favorite artist is an European, Jean Giraud, in the U.S. better known for his work as Moebius (I have to admit, I prefer the European Comic Artists over the Americans).

About directors: One of the 'others' is of course Alfred Hitchcock, who for artistic reasons (more control) preferred to work in a studio which is often very visible. I would further recommend Ridley Scott (he made some films I dislike, but even those are very well filmed) and Sergio Leone (not quiet the father of spaghetti westerns, but more influential than one would think for somebody who only directed a mere seven Movies). "Seven" remembers me of course of another director who deserves a mention: David Fincher.

Akira Kurosawa is of course the most known Japanese Director, but some others have very strong visuals, too (I like for example "Twiglight Samurai", not only for the cinematography but also for the story, from Yamada and "Sword of doom" from ... uh, well, I can't remember those names, but imdb helps: Kihachi Okamoto).

(As an aside Kubricks most important work is probably 2001: A Space Odyssey. The visuals were a standard that wasn't surpassed for decades, but I have to admit, I never made it through the movie itself. I like long movies like Lawrence of Arabia and I have seen quiet a few boring / bad ones out of curiosity or because they are important. But the Odyssey triggers something in me where it doesn't matter how fully awake I am. After a quarter of an hour I'm fully asleep. I tried several times. Even when I had a bout of insomnia (we don't have a tv in our bedroom, otherwise I'd have bought a blu-ray exactly for that).

All in all, wherever some artists create visuals, it's good for us too study what they do, at least if they are any good. I wouldn't go that far and recommend stealing, but if you understand what they do you can make their tricks your own. And the subconscious helps us too here; after a while we will incorporate what we learned from others without thinking, sometimes not even conscious about what we 'know' (for example, when I learned about the golden ratio a long time ago, I found that I had used it long before subconsciously).
 
#12
Meanwhile I got the book Swizl mentioned. I'm only half through but as he wanted to buy the new edition from 2014 I thought I do write about my first impressions. It could take some time till I finish it (other things to do, other stuff to read).

In short: I can't really recommend it (especially not if you already have an older edition).

It's not a bad book by any means. Some 15 years ago I would have highly prized it because back then it was invaluable. Brin certainly knows his business. But all in all it looks as if written in the beginnings of the 2000s, with a few paragraphs put in to modernize it a little bit. So the modern physically based renderers are mentioned in a short paragraph but the main text is clearly about the old fashioned ones, where phong shaders (& cie.) where the most used. The pictures shown also look like they are coming from the early 2000s. It's not said when the first edition was written but my guess would be 2002 – and the bulk of the information given seems to stem from that time period (reading it I got the feeling that Global Illumination was something quite new ...).

For example, in the second half is a chapter about camera matching. It's done by eye. There are references to digital cameras but the techniques used are for film (important stuff like crop factors is not even mentioned, at least not that I have seen it browsing through that chapter).

And the part about the lighting (where specular in the material is explained at length but not that you can't use it anymore in physically based renderers) is not usable in Cheetah's falcon renderer (in the part about lighting HDR is only shown as dome light and is explained somewhere in the latter half of the book).

Another gripe I have with the book, it's just about characters and animation (and it recomends to have lights that are used only to light the character or part of it, not the environment. With that you can get very unrealistic looking lighting). Product shots and stuff are not even mentioned. And especially beginners could get a wrong impression if they try to do in falcon what they learn from this book.

There is still enough content that's valid, but imho not worth the price for the book. I'd still recommend getting a few titles about photography (there are so many approaches and styles), especially about composition, color etc. Of course, it's not always easy to adapt this (new) knowledge to 3d, but it's probably the better approach to try to adapt it with that what's acually available in the 3d-software you use.

The 3d related stuff in this book you can learn from other sources, mostly for free, when (and if) you need it (for example, Cheetah doesn't support all the possibilities for compositing shown here).



@Swizl : Actually I don't really know the differences between the three editions. Could be that some of the stuff I think was already in the first one got in there later on but the important things in the book are imho already in the first edition. The newer parts are not explained well (or mostly not at all, only what they are, but not how to use it). If you still work with Modo I'd recommend Richard Yot's tutorials about rendering (some older ones are for free, but even there a lot of it is still valid and valuable). For C4d or whatever you use are probably some other render-cracks around. Those people can maybe show you a few new tricks; this halfheartedly updated book probably can't (there are better ways to spend your money).

I do wish, though, that the author would write his book anew from scratch, with new examples and so on. It could be great.
 
#14
Sorry to say so, but that's the PR-pitch ... To me the book's content looks very outdated, like a voyage back in time. And the examples are of a very low quality for todays standards. Most people today don't really use bounce lights etc. Or, as said before, already some 15 years ago most 3d apps had the possibility to light scenes with HDRI, not only via light domes described in the chapter about lights. Stuff like this wasn't written 2014. Or shouldn't.

It's up to you, if you want to buy the book. I thought, I'd warn you, because I sincerely believe that there isn't much that you can learn from it anymore. And I don't recommend it to anyone.
 

Helmut

Well-known member
#15
* I mentioned the "study" of paintings because these are by definition a 2D model of some 3D reality. In photography you are recording an existing scene. In contrast, any random painting starts out as a blank canvas and the artist transfers a mental concept, even if they "copy" a landscape or the image of people.
* Comparing landscapes, portraits, what ever by different painters / different periods in time starkly shows how a different person or a different era stresses different aspects of a specific genre, be it by composition, by the technique employed, by the placing of light or by subtle symbolics hidden by the fog of history.
* For some noted paintings we have access to numerous preliminary sketches of tricky details. In some cases the master has even returned to the subject to revisit an existing canvas years later.

* Much can be learned about composition, the mastery of textures and the creative use of light and shadow / shade, from Altamira to Renaisance masters to Rembrandt to Edward Hopper / Lucian Freud / etc. Basically, I look at paintings to appreciate different ways of "seeing" things.

* Needless to say, it is fascinating and great - possibly slightly infantile - fun to generate a C3D model "in the style of Velasquez". But then again, I have decided to retreat from sanity in the twilights of my life. I don´t greatly care about dessert at first, but a diet of mild fringe lunacy is muchly satisfying. Lunacy contains neither gluten not evil cholesterol but ample quantities of ethanol and is suitable for vegans, vegetarians and cannibals :unsure::mad::devilish:
 
#16
Helmut, of course you are right about the old (and not so old) masters.

But the same is true about photography, illustrations and comics. It's the same process (even with the mental concept transferred to (photo-)paper.

You can (and should) 'study' good photographies (which you probably do) that are composed as carefully (incl. coloring) as any painting by Caravaggio, Goya or Monet. I can learn even more about textures from photography (which doesn't mean that I don't spend lots of time holding things in my hands to see how the material looks in different lights etc. to find out what's exactly the visual difference between transparent plastic and glass). Whenever possible I do use photographs as reference. And if one is interested in photo-real renderings: You shouldn't forget how a real photography looks, how far away even the most advanced and best renderings are from the much more complicated and richer reality.

Illustrations and Comics are the same, too. Good artists do compose their images meticulous. As in paintings, they are nowadays in full control of colors (not like the cheap 4-color-comics of our childhood, badly printed on cheap paper, poor renditioned (and translated) copies of their foreign originals). For me the advantage here is the simplicity (compared to complex or even abstract paintings), the way they tell stories and often have to convey action with a simple still. It's easier to grasp what the artist does here, what his intentions are. Sometimes it's amazing what a gifted draftsman can convey with just a few strokes, even in black and white. Yes, there is a lot of trash around (but the could be said of paintings. ; only the trash Rembrandt's contemporaries produced is long forgotten), but some comics are works of true art (more often just good craft). Another thing is, for good and bad, comics helped shaping today's culture very much (which can be seen in the plethora of new superhero-movies coming into worldwide cinema every other week or so (actually it's astonishing that there is still a market for this stuff). And there is a reason why story-boards are produced for movies.

Last but not least: No, in my opinion it's neither infantile nor even mildly lunatic to create a 3d-rendering (or model) in the style of Velásquez (the é and the à are on my keyboard, but ´and so I always have to look in the 'tastaturübersicht' (the little app on top with the keyboard-layout). Whole movies try to recreate some old art. Actually stuff like this is a good exercise you probably learned a lot from (I would like to see something like that, and I'm probably not alone *hint, hint*).
 
Top