The April PhotoWorks contest has closed and the images are posted for voting. Visitto vote for your favorite image. While your there download the model for the May contest.
17 years ago
The April PhotoWorks contest has closed and the images are posted for voting. Visitto vote for your favorite image. While your there download the model for the May contest.
Your own image is quite good but without a shadow it doesn't quite work. The 3d Studio render is very nice....but I voted for...well we will just have to wait and see when the tally is done :o)
Yes, a shadow is essential to make it look right. That is what 'studio plastic' is supposed to be for - you can make the material clear so you can see an image in the background, yet the clear model will show any shadow cast on it. Rob could have made a simple model of a floor for the phone to sit on, set it to studio plastic, and he would have gotten that shadow on top of the book without doing any compositing.
HOWEVER... (there's alwyas a however with SWx, isn't there?)
Studio plastic has a bug when used with indirect illumination - the clear part reflects back a splotchy white no matter what you do. It is still broken in 2007.
If people want to easily cast shadows on pictures AND use indirect illumination, please report this bug so it rises in the list!
However, if you are just us> Your own image is quite good but without a shadow it doesn't quite work.
hmmm ok I don't use PW so I wasn't aware it was a bug rather than a missing capability or an omission on Rob's part.. BTW why doesn't an obviously important bug like this just get fixed? I have never understood why users should need to report the same bug again and again to get action - several releases after the event if they are lucky. I hope some attention has been paid in 07 to the old PW interface deficiencies...
I'm not sure Neil. I have to agree though, its seems like some bugs just never go away.
Agreed! As I understand it, SWx rates bugs based upong business impact (this is a big one - instead of just taking a snapshot, now I have to model the world) and frequency of reporting. That second part 'sort of' makes sense - if I was the only one in the world who minded, I should get at the end of the line. However, this thinking is flawed because sometimes its not clear its a bug, and even when it is as clear as this example its such a pain in the ass to re-report a known bug (often having to educate the tech support guy about it) that we users just don't bother (we have jobs to do). So we get bugs that languish for release after release because they just never climb that high on the list. This one seems so simple that it shouldn't take much to explain. The whole point of studio plastic is to have a clear material that shows a shadow - heck, if i didn't want it clear, I would just use PLASTIC!!! (duh). Good luck getting your tech support guy on board about that, though. They will likely each have to have it explained to them (it will help if you can point to the exact item in the pWx manual that describes 'studio plastic - I regret I do not have the manual on me right now to give the page and item), and then they need a clear, reproducable sample to submit upstream just to add another tick to an already known problem. What a waste of time. I have grown to let many of them go - but this one saves SOOO much time its worth the fight for once. Ed
Well I don't know how SW is structured internally but I imagine there is a small group of people in a corner somewhere with a permanent assignment to PW matters? It ought to be possible for PW to be maintained to a high standard independent of the main program. I don't see that they need to be using the same priority system as the rest of SW. Possibly there are only 25 or so bugs in PW that could do with attn - aside from the things which don't work especially well ;o) It can't be that much of a strain to fix them surely. I don't understand either why for simple issues like this SW won't come on board with reporting via the SW forum. For add-ins it would be ideal and users would be heartened to be able to communicate with the boffins directly instead of jumping the support hurdles to no apparent end. Perhaps you should whisper in the ear of a contact of yours at SW that they really need to do something in the instance of this bug. Honestly I think if you are going to offer a renderer then you should have shadows working...what is the point of a render without shadows...may as well just capture a SW window.
"Perhaps you should whisper in the ear of a contact of yours at SW"
It was more than whispering, and it was around eight months ago. I have done some more 'whispering', but getting anything to happen in PWx is slow. Case in point - It took me over three years just to get the defualt gamma set to 1 instead of 1.5 or whatever stupid value it was before, and that is just a simple change of, I presume, one or two bytes in the code. Remember all those people who's renderings to file would be darker than renderings to screen? It was because the default gamma unnaturally ligtened up the screen. I cringe when I think of the thousands upon thousands of hours lost worldwide because someone set a default to a value that made no sense. BTW - I loved the counter argument from PWx guys - 'but you can set it to 1 yourself'. OF COURSE I CAN, and of course I did. But what about everyone else?? I used to work in an office of 12 people rendering stuff, and I had to go around and teach each one of them about this. Every new installation of SWx the deault would get set back to 1.5 and I would get burned by it when my final rendering didn't match my tests to screen. Now multiply that wasted effort by 380,000+ users (or, more reasonably, the 10K-60K who I imagine use PWx). And to add insult to injury, no one could find a single good reason for it to be set to 1.5, but they were terrified to change it. Some of the guys at SWx are great and fight for us. But I get so frustrated dealing with a handful of them, and unfortunately the most frustrating of the bunch are involved with PWx. Trying to get them to change anything, even stuff they can't justify, is a nightmare. If I was a betting man I would say that they will not fix studio plastic, but in two years will instead create a new material that finally does it, using the justification that some poeple MIGHT already have their 'studio plastic' material set up and 'like' the blotchy white effect. Yeah, in the land of imagination their 'might' be two people in the world like this while tens of thousands suffer not getting this great function, but these PWx folks can be paralyzed with any improvements or fixes to the software. I don't know if its just easier to not do anything and rationalize it instead of trying to help out the bulk of the populaiton, but inaction is their usual speed. BOY I HOPE THEY PROVE ME WRONG, and look forward to eating my words. But I seriously doubt it.
End of rant - its been eating me up for a long while Ed
Back in the early days of calculators (we're talking more than 30 years ago) if you made a mistake when entering a number, you had to clear out the entire number with the CLX key. On 41C the software guys introduced (or stole from TI, perhaps) the back arrow key, so you could just delete the mistaken digit(s) and continue. But when they started coding the 12C financial calculator, the marketing guys were worried that the business people would be confused by the new way of doing things and kept the CLX key. Generations of real estate and banking people have lived with that decision. I just checked on the HP website and found the HP 12c Platinum Financial Calculator. They've added and algebraic mode and programming, but they've still got that stupid CLX key!
Jerry Steiger Tripod Data Systems "take the garbage out, dear"
I see, well, I'm glad we could help you let off steam on behalf of all those frustrated PW users out there Ed :o) Maybe I'll dust off my Blender notes and Maxwell people will tout some too and we will see if the phrase 'dwindling market share' stimulates any useful activity ;o)
I have been following this thread with interest, PhotoWorks has been a constant source of frustration to me since I first installed it. Hence I went to Maxwell Render
Maxwell Render V1.0 is due to be released 26 April, tomorrow in Europe, after long delays. I'll keep the newsgroup informed of my tests with V1.
Below links to some images posted to the Maxwell Render forum. All these images were modeled in SolidWorks and rendered in Maxwell Beta or Maxwell Release Candidate 5.
Designed byDesigned by Gunnar
There are a few Maxwell Renders on my website for anyone who care to take a look.
hey some great images there John...and I think more people will pick up on Maxwell despite the long render time because frankly PW just can't match it for realism ~ esp. if it has no shadows -doh... ;o)
BTW a few Indigo users have been using Neat Imageto clean up their renders. Not sure if it relevant for Maxwell since they improved the code but I thought I would mention it. later -
The way I would have normally rendered my contest image I would have rendered just the phone and merged the background and phone image together in an image editor. I would have then been able to add the shadowing to the image. Contest rules don't allow the use of post processing so I had to "make do". The ID/"studio plastic" bug does suck but there are ways around it. Of course work arounds are counter productive and there should be no excuse for 'broken" functionality.
I agree the Maxwell images are amazing and I purchased a copy a few months back. I have been waiting until V1 was released before I did much with it and it looks like tomorrow I might have my chance to play with Maxwell some more. I'm not sure what has changed from beta to release but beta wasn't able to work with decals which is a big thing for me. It also seemed like it was a loooong wait to achieve a noise free image. John has some convincing images at his site (nice job John) though and I look forward to giving the Maxwell V1 release a go.
Clearly there are workarounds to the studio plastic bug - there is a shadow behind the cell phone rendering that shows the phone in the coin tray. But, as you said, why do workarounds when it would be better to jsut have stuff work.
Since this is a contest for learning, I'd like to take this opportunity to learn something. The bottom of my glass in my entry looked as if it had a light on inside it. The only lights in the scene were a light in the lamp shade where the bulb should be and the hdri image. The bottom of the glass was set .010" off of the top of the table to avoid contact effects. In later renderings, I was able to tone down the brightness of it by turning down the reflectivity and transparency of the glass. I arrived at that through my usual "random button pushing" technique in PW. How do you develop a sense for what is causing this? Why does it turn out to be reflectivity when the rest of the glass which has the same reflectivity value doesn't exhibit the same problem?
Ok, that was my first issue. The second issue was the lampshade. I want the lampshade to glow a bit the way lampshades do in the world. Not transparent, but I guess translucent. I made the buttons on the phone glow a little by using the "constant" illumination selection for the face material, but the lamp shade isn't the same kind of thing. Again, how do you develop the intuition for lighting and materials required for this kind of work?
Another issue was how to illuminate the base of the lamp, which had some interesting geometry which you simply can't see because it's too dark. In the end I wound up putting a dim directional light aimed at it, and turned off the shadows for the light. I wanted it to be illuminated by light reflected from the rest of the scene, but the number of reflections settings didn't seem to make it happen.
I realize the easy answer to this is that if I were just a bit less muddle headed it might all come together for me, but barring that, does anyone have any suggestions to teach this old dog a new trick?
Rob Rodriguez wrote:
For reference, here's a link to the improved image, but still there are plenty of lighting problems here.\images\cellphone.jpg
The glass glowing: I assumed that this was due to caustics (which I have not used because, though they are real, they tend to mess with compositions and aren't often missed by the consmuer of the image) - I would experiment with the setting for the C photons for the lights in the scene. Actually, I would first render the scene with NO LIGHTS jsut to see what the indirect illumination was contributing, then add one light at a time. Usually I try to tweak a rendering in so it looks good with no lights (illuminated by the scene) then add a light or two to carve out shapes, add highlights, and define the composition. By doing things one element at a time you can find out what is causing the glow - random button pushing, which I did for years, makes it a little harder to root out the source of your issues.
"Another issue was how to illuminate the base of the lamp, which had some interesting geometry which you simply can't see because it's too dark. " This usually means the diffuse component of your material is too low, especially if you are using indirect illumination. You mihgt also choose to make your HDR scene lighter (are you on 2007 for this?)
Number of reflections, if its the setting I am thinking of, has to do with mirror bounces, not how much 'light' bounces around. Changing number of reflections ought not to lighten up your object unless its mirrored and is one or two bounces away from reflecting a big white thing somewhere else in the scene. Ed
Like Ed I tend to start with no lighting (all turned off) and use the image to light my scene. Sometimes I have to go through a number of images before I find one that achieves the desired effect. It always surprises me to see what a great effect just using an different image can achieve. Also, lately I haven't been using an HDRI image. I've been using spherical jpgs and they seem to work just fine. I actually like them better for a couple of reasons. The reasons being, they don't seem to be as intense (lighting wise) as the HDRI and they map to the sphere perfectly so if you have reflective materials you don't have wierd seems or gaps. Also like Ed I then start adding lighting (typically just one or two) to add the highlights and begin fine tuning the image.
In later renderings, I was able to tone down the brightness of
Once I have the lighting set up and I have an image that looks 75% -....? correct then I assume the lighting is good and I need to work on the trouble areas by adjusting the material settings. I think this is just from years of trial and error and learning over time what does and doesn't work in PW. Remember though you only want to change one parameter at a time so you can see the effect it has. If your changing a bunch of different settings at the same time you will be confused as to what did what and therefore your not learning what each setting does.
A number of factors could play into this, lighting locations, scene image, different materials in the proximity, etc.
Process of elimination for me. As I explained above, I try and set the lighting (based on past experience) first and then move to material adjustment from there. As different elements drop out of the equation I'm able to hone in on the image I'm looking for. Not to say I don't go back and adjust the lighting slightly but only after I'm sure I've done all I can with the material settings.
The number of refelctions setting won't help with this, it only controls the reflections not the light. The material settings of the lamp control this. Note: directional lights are always targeted at the origin. If you have geometry some distance away from the origin it can be difficult to control for light specific areas. Spot lights might work better for this.
Hope some of this helped. I think your second image looked nicer than the first. The real answer is something my co-workers say everyday. All "you" really do is just push the button, the computer is doing all the work. Right, that's it. just push the button.
I agree with Robs post except one phrase which probably deserves a 'little' more calrification. Directional lights are not targeted at the origin - all the light is parallel to the vector between the light and the origin. Its a minor quiblle, but can make folks arrive at bad decisions so it is worth a word or twenty. Unlike a spot light that starts at a point and is directed at a point, directional lights are from a source that is an infinite distance from the object (think of the sun, which for all intents and purposes is infinately far... any shadow convergence you see in the real world is due to perspective, not the large distance between object and the sun) so all the light it generates is parallel - think of an infinately large light source where all the light goes the same way.
Why use directional lights to target highlights? Because it is easier... in a scene where you want the left side of everything to be brighter for composition reasons (or to carve out an object) all you need to worry about is the vector, not whether the light is far enough that it lights the left side of everything and not just the front. In real-wrpld photo studios, they use large panel lights some distance away from the product to simulate this - photo studios tend to be pretty big because they don't have a simple button to creat parallel light like we do.
Remember, the goal is always 'photo-realistic', not realistic. Pick an object in your kitchen and take a snapshot of it, then look at how that object appears in a catalog - unless you are a genius (or lucky) it always looks better in the catalog because the photographer knew how to set the scene and give the product its best representation by controlling the light and reflections on the piece. That is the goal - tell the story you want about an object by recreating the environments used by professional story tellers, not get whatever you get by recreating the mundane real world.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.