Freelance Rendering Fees
-
Guys,
It looks like Kyle posted a question and has not checked for answers! I would not be inclined to bother any further until he checks out what has been advised, if ever
Mike
-
This thread just exploded into life! This is some good advise, too.
I'd ended up contacting my client with a proposal a week ago, with a fee we were both happy with.
What I've struggled the most with is estimating time required. In the Architectural field, everyone wants more work done in less time. This mindset always throws me off. I pissed off my boss a month ago when I seriously overestimated my abilities. A 1 week job ended up taking 3 weeks.
I see you folks have taken into account the kind of things that will extend your required time, but what do you find wrecks your time estimate the most? And what did you struggle with the most when you started doing this stuff? -
mmm, good questions. I have a little note stuck on to my cubicle at work, a little diagram I heard someone explain somewhere that I've found to be undeniably true. The diagram is a triangle with a word at each of the three points; Cost, Speed, and Quality. The idea is that you can have any 2 you want but never all three. The more you think about it and experience it, the truer it becomes.
Whether or not it's a good idea to share that explicitly with a client, I can't say but depending on the situation it's probably a good idea that they have some idea of the concept. You'll need to feel them out and see which of those they're most willing to sacrifice and use that to make decisions.
Time estimating is reallyhard, and I'm still quite bad at it. I was bad at it in architecture school when I'd get home at 6am after being in studio all day so I could get a couple hours of sleep and clean up before a big presentation. And I'm bad at it now when that point comes in every big project when a program fails or Logmein isn't working right or whatever. It's difficult to put a schedule on art, and when you combine that with software, anything can happen. And yet, you do have to put a time limit on it, nevertheless.
I've heard various advice. Some people have the discipline to first work on the stuff that you know NEEDS to get done (modeling/texturing the building, working on the site, etc.) and only THEN do the stuff that you want to get done (find a new way to render grass, put a 3d person sitting in that car, etc.). I'm rarely able to do that though. That's why I'm personally more in favor of charging by the job rather than by the hour (I think I'm in the minority here though). I find it much easier to determine how much a productis worth rather than how many hoursit will take me.
I find the Speed part of the diagram is typically set by the client anyway. So it doesn't usually come down to you saying you can do a job in x amount of days but rather you can give them x amount of Quality in the amount of days they give you.
What wrecks my time estimate the most? If I have to be really creative it's usually that. But that mostly applies to personal projects and contests. In my work renderings the design is already there so most of the decisions are already made. So my main time wrecker is software. It's always towards the end of the project too, when you've got all those polygons crunched in to one big file and your folder structure has gotten a bit messier than it should be and your interacting more between modeler and renderer and photoshop and BOOM, something doesn't work. On my last project I had my building and site all modeled wonderfully in SU and exported to Maxwell with a new and improved plugin which I'd been using for a few weeks and all the sudden after it exported the model wouldn't render, wouldn't save, nothing. So I'm at home trying to solve it and might have a solution but I've got to transfer my files to work to render it so I'm trying to transfer huge files over through Dropbox and I get it over and I'm using Logmein.com to work from home on my work computer and Logmein isn't recognizing double clicks or alt clicks which eventually brings me to a stand still. Long story short I drove into work (40 minutes each way) on Labor Day to get the work done after I'd found a work around to my problem. It's stuff like that that just sucks up your time.
By the way, if you use multiple computers get to know Logmein and Dropbox really well. They'll be your new best friends.
What did I struggle with most in the beginning? For me it was learning the stuff at the same time I was expected to be producing presentable quality renderings. I was given no real learning curve. That's why I ended up working in Maxwell. The render times are longer (as compared to a biased renderer) but it's easy enough to hit render at the end of the day and hit Stop in the morning. The advantage is that the learning curve was WAY less than with a biased renderer and you can get excellent results really quickly.
So now some questions in return. What renderer are you using? How did come to a decision on a proposal? How much experience do you have with modeling? Rendering?
-Brodie
-
One of the biggest things that can add time is the customer changing their mind mid way through but still expecting the timing of the project and cost to NOT change. What I have started doing recently is storyboarding the job and having the client literally sign the boards as a "This is what I am looking for" cover my ass measure. This way if they change the job or timing I can say "remember this document you signed?". I have a storyboard for:
Initial/rough modeling and camera positions
Color/texture
Animation (if needed)
Final rendersThe client signs each board as the job reaches that process. This is something that works for me after being screwed out of $12,000 (yep my fault for trusting a new client to pay all at the end). I now have ALL new clients put down a "good faith" deposit (typically my cost, without profit). Then once the job is complete they pay THAT DAY. No net 30 or 90 days. I drop off renders or a CD and pick up a check. No check....NO CD. Easy. If they bitch....they probably were going to give me a hard time anyways and I do not want that type of client. After I know a client will pay on time I will ease up on the good faith and possibly the payment terms.
Hope this helps.
Scott
-
Just snagging this thread for future reference.
-
@BRODIE:
I am far too familiar with the cost/speed/quality triangle. It seems to be a fact of life in this field, where you try as hard as you can to stretch yourself to the three points, but never really making it. How do clients typically react when cutbacks on the job must be made? I'd assume they'd always be happy once they see the final product?"What renderer are you using?"
Sketchup pro 8 and Podium v1.73 and v2. So far they're giving me very nice results (and the occassional headache as the project nears its end)."How did come to a decision on a proposal?"
I talked to a couple of people, and figured out an average hourly wage between their suggestions. The client was asking for a maximum of billable hours, which I was able to estimate based upon other hours she was suggesting.
40 billable hours @ $50cdn = $2000 maximum."How much experience do you have with modeling? Rendering?"
I picked up Sketchup a few months before my 2 years Architectural Technologies program. This left me constantly a few steps ahead in the rendering department during classes. I picked up Podium partway through, and have been playing with it off and on.
Since starting at my Architectural firm, we've decided to start offering renders to our clients (for a price). I've constantly been needing to improve my game. So.. uh... 2 years with Sketchup, and 2 months of (serious) rendering for clients.@Scottpara:
This storyboarding idea of yours sounds interesting. What happens during, say, the modeling phase when the client wants a whole new set of cabinets, after you built the current version from scratch?I like the idea of swapping the CD for the cheque. That's the way I want to approach this job. After working in this field for a few months, I can already tell the clients will get pissy when I suggest they ACTUALLY have to PAY for work (not necessarily the current ones, but certainly future ones). How do you approach the topic, and what does your contract look like? Have you ever had to sue anybody for ignored payment?
-
Testing testing
I swear this forum ate my last post -
Most of my experience with having to let folks know that they might not be able to get the best rendering in the world is at work rather than freelance. The closest and most recent scenario happened a couple weeks ago when a client wanted 4 elevation renderings and a new perspective rendering on a building that has been in the process for a couple years now. I had ample time to do all that I needed to do, the only hold-up was that they were changing the color and had no idea what it would be changed to. I had to let them know several times that quality-wise, the sooner they get me the information the better. And finally, if I don't get a color by this day, I'm going to have to start the rendering with the previous color and if I had to photoshop the color drastically the quality would suffer and we wouldn't have time for much back and forth between that and the deadline.
That's pretty typical for me. I let them know that I can make whatever reasonable requests they have (and even some unreasonable ones), but the more drastic the request and the less time I have to work on it the lower the quality may be. Sometimes that leads to compromises where they decide the change can be made later or wasn't as important to show as they thought, or perhaps the view angle can change so that you can spend more time on the change in place of an existing building you would otherwise have had to model. That sort of thing.
It all goes back to the triangle. Here at work sometimes I won't say anything at all. I'll just jack up the 'cost' aspect of the triangle by working OT. In freelance work, that sort of thing would have to be discussed up front though of course.
SU and Podium is a nice mix, especially to start out with. As you get more comfortable with it and you get better at rendering you'll likely want to move on to a software with more features but I still go back to Podium occasionally for quick and stuff because for hardly any effort you can get something that looks much better and more professional than native SU output. Good thing about doing rendering for a company is that you should be able to get them to foot the bill.
-Brodie
-
Which programs have you worked your way up through?
I'm using PodiumV2 now, and loving the look of things. I'm supremely jealous of people who're rendering with Maxwell and Vray, but their prices (and required learning time) seem steep
-
Is Maxwells learning curve steep? I only briefly looked at the demo once, as it advertices how easy it would be to set up physical accurate material etc ... but the speed ... or the lack of it to precise... un-installed.
-
I was doing SU and Photoshop originally because that's what our company just did. Then my boss came to me one day after seeing a photoreal rendering done with Revit/Mental Ray and said we had to start doing 'that.' I had a pretty open budget. The way my boss saw it, he'd rather spend money up front and try stuff than buy a program, try it out, buy another one, etc.
I'll list the stuff I tried and a brief review.
Podium - as I said, great little program and very fast, especially once it went to multicore. It was a great intro to rendering (omni lights, bump maps, etc.) but I quickly realized I'd eventually grow out of it and needed to find a more professional program with more capabilities while I still had a wide open budget.
Indigo - I found Indigo pretty confusing. It seemed to use a lot of 'rendering lingo' that I didn't know so I didn't really know how to get going. I knew I could create good stuff based on the renderings I'd seen but the learning curve proved too much for me as a beginner.
Revit/Mental Ray - I knew before I started that this was going to be a loser but our I.T. guy who was in charge of getting the software felt we'd better do it to cover our bases with the boss since that's the software used to create that original rendering he stumbled across. Revit just isn't made for doing modeling/rendering in the design phase of a project before stuff is nailed down. Why go with Revit over 3ds Max?
3ds Max/Mental Ray - If indigo was confusing, this combo was something else altogether. It's like walking into a Nasa space shuttle cockpit with all the buttons, except most of them are hidden behind 3 or 4 panels. And there was terminology galore that was way beyond me. Why's your render look bland, overexposed, and blueish? Good luck finding out.
Kerkythea - As I understand this has now turned into Thea. When I tried it, it was one of the few free rendering programs and could do both biased and unbiased rendering. This was the first program besides Podium that I really liked and the unbiased renderer gave me great results right out of the box. And new Thea has a realtime preview that looks really nice. I haven't used it in a couple years now but I'm still confident enough to recommend it highly and I imagine it's less expensive than many other programs.
Maxwell Render - I would have kept going with Kerky, had I not found Maxwell. In terms of learning curve, their tagline is that it's 'as easy as taken a photo' or something like that. For better or worse, I've found that to be absolutely true. The reasoning is that there is very little 'rendering lingo' when it comes to Maxwell. Instead most everything is described in camera and physical language. I imagine those familiar with photography would have very little learning curve indeed. I didn't but the great advantage is that it's much easier to learn what a cameras 'f-stop' is and what it does (through wiki or any number of sites geared for photographers) than a 'falloff map.' Or when it comes to the environment instead of making environmental adjustments that have no basis in the real world, you can actually set things like how much water is in the atmosphere to get different looks in your sky.
In terms of speed, most people who have tried it and given up say it's slow. I'm not a fair judge because I'm still pretty new to biased renderers and have yet to produce anything comparable in Vray as what I was doing in Maxwell within a couple hours. That said, my view is that it's a trade off. With biased renderers there seems to be more time spent upfront. First the learning curve is higher with biased renderers because it's not their natural inclination to create something photorealistic, you must discover how to get it to produce photorealistic renderings. And second, there are many many more settings to fiddle with in a biased renderer.
On the other hand, an unbiased renderer will have a smaller learning curve and less settings. It is programed not to use 'tricks' so you don't have to worry about things like 'falloff maps' which dictate how light will fall off of an object at glancing angles. Instead, an unbiased renderer will just calculate that based on real life physics giving you a photorealistic fall-off every time without having to fiddle with that particular setting (one of many examples).
The other side of the speed equation is that Maxwell starts with a very noisy render and progressively get less noisy. I've found this gives me a nice quick preview which will let me know if something has gone wrong. Then for final renders I just start them at 5pm before heading home and hit stop at 8am the next morning. That's usually ample time for any final rendering and it doesn't decrease my productivity at all. On the other hand I've found biased renderers to produce renderings quicker (of course) but because it renders the scene in little boxes rather than all at once, I may have to wait for 1/2 the render to finish before realizing there was an issue in a given part of the rendering. That issue is going away with realtime though, and btw, Maxwell will be coming out with a realtime preview within the next month or so.
I should also mention that Maxwell (and Thea) has a 'studio' program that comes with it. Basically a separate program which you can open your model in and make tweaks there rather than in your modeling program. This is a HUGE advantage for SU users because it's the only way you'll ever get to place a number of high poly trees and cars in your model.
Vray - Vray didn't seem to be very big a couple years ago so I never even tried it. Now it's the arch-viz go to (in combo with 3ds max). I'm still learning it but it looks promising. Now that I've ranted about biased vs. unbiased, let me say that once you get to be a real pro and know all the settings and lingo and have a nice library of materials biased rendering is probably the way to go. Particularly as it opens you up to the option of animation as well which is just too slow in unbiased renderers. Someday I hope to be able to produce in Vray what I can now produce in Maxwell and if/when that day comes I'll happily switch over. Equal quality in less time is a no brainer. Plus you get multiscatter and fur which would be quite nice for grass.
All that said, here's my recommendation. Keep with Podium for now but download a demo of Kerky and Maxwell and play with both. Render something in those that you've rendered in Podium and try to get your boss to foot the bill for whichever you like better in exchange for a clear quality improvement over Podium. Then down the road, work on learning Vray and 3ds Max and all of the lingo that goes with them. To me that's a good progression.
-Brodie
-
Icedkasz, Did I understand that you allot 40 hours for rendering? I can hand render a 18X24 inch board between 8 to 16 hours. While $2,000.00 seems to be a OK price, 40 hours seems like a lot of time. On the other hand, once you have a model, and its render values, multiple, quality scenes are easy to knock off.
-
My understanding was that the 40 hrs. was sort of a maximum so he could charge per hour without the client being afraid it would go on forever and they'd end up owing $10,000. It also protects the renderer from having a client who knows they're paying $1,000 no matter what and might as well make endless changes and expect you to put 80 hrs into it.
-Brodie
-
@Brodie:
You'd be correct, 40hrs is the maximum. I doubt I'll need it all, unless the client suddenly expects more work near the end (almost guaranteed).Thank you for the info about the rendering software. Do you still use Sketchup for the modeling phase?
-
Almost exclusively. I'm not entirely sure why 3ds Max seems to be the go-to program for arch-viz. It may be a 'better' modeler in terms of features and flexibility but it's not even close to being as intuitive as SU and SU can do 95% of what 3ds Max can do in terms of Arch-Viz modeling. I could see myself someday getting to the point where I use 3ds Max for my site work, but it's hard to find a compelling reason to model a building in 3ds Max as opposed to SU.
-Brodie
Advertisement