Time spent on illustrations

[Jon Berntsen]

Hi.

 

I've been searching back and forth for this, but was unable to find a decent thread. There are a lot of questions and answers about general project pricing on this forum, but not much about the actual hours spent in different kind of projects. It's kinda critical when pricing. (As well as I'm just curious). This must be a better global reference for pricing, right?

 

As I live in a country where everything is expensive, I have to be very careful about pricing too high. I don't want to loose clients to outsourcing, but I neighter want to work unreasonable hours for too little money.

 

To the point: I'd like to know how long time people offering low price renderings spends on their illustrations, and I'd also like to know how long time experienced guys in typical western studios use. - And in the end compare these numbers to my own time spendings.

 

I know it's not one correct answer to this, but I'll still go fishing.

 

How many hours would you typically spend on making and finalizing:

1) Exterior daytime with one house, 3d vegetation?

2) One interior from a house or apartment?

3) Commercial exterior?

 

Also leave more options if you have.

 

I'm typically using around 20 hours on a single house daytime exterior from start to finish. That is not included final rendering, (which usually runs when I'm off-duty) and waiting for feedback from the client, but includes initial brief with client, sketcing on paper, scene setup, modeling, lights, texturing, test rendering, photoshopping. I obviously don't do the CAD models myself, but I clean them up before importing to max.

 

Interior about 7-8 hours.

 

Commercial exterior about one - two weeks depending on the project size. That includes scene setup and preparations for more than one still.

 

Am I way off? I've been in the business for four years now, but as the market changes and we're facing tough financial times (again), I'd like to review my own pricing policy to remain with any customers at all. If I'm doing things slow, I'd like to know any tricks available for speeding up.

 

I'm using my own built material and objects lib, but perhaps spending too much time on test rendering to get the feeling that I want. However, that part will always be an important aspect for me. The devil is in the details... But not if I'm calculating my prices based on generally too many working hours.

 

 

EDIT: As far as I can see, the only thread with somewhat similar content is this one from 2008, http://forums.cgarchitect.com/30675-time-spent-projects.html. It's neighter recent nor within my primary focus area, exteriors. Just mentioning

Edited August 15, 2011 by chroma

[Jonathan Sanchez]

Hey Jon

I don't know how to give you an exact answer, dont' think it's possible since there are so many factors that can affect time such as the sizea and complexity of structure, number of shots needed, amount of detail needed ect. I've had residential exteriors which I've been able to finish in 2 days, and others that have taken me 2 weeks. So, you have to examine the details of your project and use your past experience and best judgement to figure out a timeline. Also, remember that as you get more experience you become more efficient and therefore take less time. I've seen my work process become more efficient and faster as time has gone by.. especially as my library of stock objects and vegetation becomes bigger.

Edited August 16, 2011 by sancheuz

[Brodie Geers]

I do hospital exteriors and little else so I can't comment on the residential stuff. I'm usually given design elevations to work from. I week is a pretty good time frame to be able to model the building and site, apply materials with a bit of customization, light the scene, render it out in Maxwell, and do some photoshop work. That's a very rare scenario for me however as my workflow typically involves some back and forth w/ the sketchup model as it's in the design phase. It'll typically go through a few iterations where only the existing structures and site remain static. Then it'll go to the client and eventually there will be more tweaks if the job develops. I still haven't closed the books on the first archviz hospital I worked on which began about 4 years ago (being built soon Lord willing) so my projects can be quite lengthy.

 

Often times it's the deadline that dictates things more than anything on my part. The boss tells me when it's due and I tell him what I can get him. If it's a few days I can put in some extra hours and get a good product. If it's a few hours I can show him a decent sketchup model and maybe even a crude site.

 

-Brodie

[Scott Schroeder]

For a typical single family exterior shot, I'm typically at 8-10 hours start to finish. Including rendering, including post. No change time for the client. If the client wants to make last minute changes, they are charged extra. But keep in mind, these are very typical around 1,500 2,000 square foot basic designed single family homes that go in a master planned community so no real architectural marvels here.

 

The larger custom homes, it honestly depends on the architecture, the elements, and the size. But 20 hours doesn't not seem too bad, but it all goes into what you are showing. IE if you only have one camera that's looking at the front and left sides, there is little need to spend time modeling the right and back sides.

[Jon Berntsen]

Thank you for your replies, much appreciated! VelvetElvis, you must have a render farm(?), my final renders usually take somewhere from 2 to 4 hours, 3500px width. But I'm not counting final render times in, of course. I'll post an example tomorrow of what I typically do. It would also been cool to see what you can do in 8-10 hours, Elvis.

[Scott Schroeder]

Thank you for your replies, much appreciated! VelvetElvis, you must have a render farm(?), my final renders usually take somewhere from 2 to 4 hours, 3500px width. But I'm not counting final render times in, of course. I'll post an example tomorrow of what I typically do. It would also been cool to see what you can do in 8-10 hours, Elvis.

 

No render farm, just one box. I've cut down the needless settings to render lighting fast and put detail back in post work using render elements or just good old fashioned Photoshop work. However, I also render no higher than basic web resolution which is maxed out for me at 720p or 1280x720. If the clients want anything bigger, I charge extra. I refuse to render large images for basically free. Strangely enough, no client has ever really fought back. They all agree, so it works out in the end.

[Scott Schroeder]

[ATTACH=CONFIG]44398[/ATTACH]

 

This is an example of a 8-10 hour job. Normally, I don't show open garages but this client wanted it to show an EV along with the solar panels on the roof. But, since the house was really basic it took no less than an hour to model the thing.

[mi75]

For a standard spec home, modelled, rendered (1-2hrs 3508px), Photoshoped with no changes I'm looking at about 3-4hrs not including rendering time. Normally I'll start a new house while the previous renders. It does depend on the design, those are quiet basic house.

 

 

 

I’m not proud of some of them, due to time, but that’s the time I’m allocated. After spending 6mth on an animation it takes a day or two to get back to that speed again.

[Jon Berntsen]

OK thanks again for input, and thanks for the example!

Here is one example of an illustration that I'm currently working on. It's not quite finished, but getting there. Certainly needs to be populated ++. So this would typically be a 20 hours project for me. Opinions? I'm getting anxious when you guys say you use below 7-8 hours. Do you really spend one working day or less from start to finish? Is that really healthy for the outcome regarding to client brief meetings, testing, composition, errors, giving your client more than one camera angle opportunity, etc? Well ok. Would you ever charge the client more, or do you charge the actual spent hours when going so low on hours? What surprises me is that none of you are from so-called outsourcing countries. I'm sure I'm getting faster by the time my library is bigger and the rendering farm is setup, but as I said I'm getting anxious.

Edited August 18, 2011 by chroma

[Dave Buckley]

[ATTACH=CONFIG]44398[/ATTACH]

 

This is an example of a 8-10 hour job. Normally, I don't show open garages but this client wanted it to show an EV along with the solar panels on the roof. But, since the house was really basic it took no less than an hour to model the thing.

 

so what took the remaining 6-8 hours?

[Dave Buckley]

OK thanks again for input, and thanks for the example!

Here is one example of an illustration that I'm currently working on. It's not quite finished, but getting there. Certainly needs to be populated ++. So this would typically be a 20 hours project for me. Opinions? I'm getting anxious when you guys say you use below 7-8 hours. Do you really spend one working day or less from start to finish? Is that really healthy for the outcome regarding to client brief meetings, testing, composition, errors, giving your client more than one camera angle opportunity, etc? Well ok. Would you ever charge the client more, or do you charge the actual spent hours when going so low on hours? What surprises me is that none of you are from so-called outsourcing countries. I'm sure I'm getting faster by the time my library is bigger and the rendering farm is setup, but as I said I'm getting anxious.

 

i'd charge extras for design/modeling changes or anything thats going to take a significant amount of time to ammend.

 

just quote an additional hourly rate for any changes made after a certain stage. and just to make you feel better I've never had a project yet that I've been able to get out the door and signed off in a day from receiving drawings to final render. No chance. Too many things, client comments, design changes etc - not if they want good quality anyway

[Jon Berntsen]

So my conclusion so far is that I'm touching the upper limit, but if I manage to deliver good quality there should be customers around willing to pay. Need to say that I've never had any big trouble with loosing clients before, but I want to adapt to the market if needed before I do.

Dave, you make me feel normal, thanks

[Brodie Geers]

so what took the remaining 6-8 hours?

 

hahaha

 

-Brodie

[Dave Buckley]

hahaha

 

-Brodie

 

I must stress I wasn't being sarcastic, was just interested to know how the remaining time was broken down.

[BVI]

For a standard house - 2.5 to 3 Days - 1 day to Model and detail, 1 day to texture and Light and 1 day for changes/tweaks and Post. We render at 3500px or higher if needed, 1024px is useless for anything other than email. For a more complex house it may take 2-3 days to model and so forth.

[Brodie Geers]

I must stress I wasn't being sarcastic, was just interested to know how the remaining time was broken down.

 

My bad, I missed the part where he said the house modeling only took an hour or two. Would have been funny at any rate. My guess is besides materials, vegetation placement and post process a good chunk of that was taken up in render time. Looks like displacement grass?

 

-Brodie

[Brodie Geers]

Also, VE - you said you never seem to have to do anything over 720p? Are your clients strictly showing this stuff on the web or do they find 720p acceptable for printing as well?

 

-Brodie

[Jonathan Sanchez]

These are some samples of two projects that took me around 40-50 hours each. They each included two pics which means I had to model the entire house. This also included some revisions and changes. Render times were about ~1.5 hour each (3500px) with a 5 computer farm running i7s 2600s. Seems to me what takes me the longest is getting the landscaping just right. I like to not have to rush through my projects, rather sit back and meditate on how to improve the image. Makes me slower, and probably more expensive, but clients appreciate.

Edited August 18, 2011 by sancheuz

[Scott Schroeder]

What did I do the other hours? Watched a movie. For basic house renders like that, I have a flat fee and a flat hours set. It's a this is what you get, what it will cost, no-haggle sort of deal. It just happened that that particular house was about 4 hours total. 1 hour to model, 1 hour to get the site up (landscape, etc), 1 hour to render, and 1 hour in post. The other 4-6 hours were profit. Gotta make profit in this business.

 

I have a large library of prebuilt home pieces in Max. I realized I was modeling the same thing over and over again, so I created a sort of lego block set that allows me to snap into place a home in fast time. My landscaping is also in per-arranged plots so all I need to do is drop in the sections and do some minor adjusting. Ditto on the lighting rig. Everything is well oiled, well maintained, and fully parametric. I have scripts that auto apply materials and mapping based on names of objects, and most of this project is done with button clicks. If you search around ScriptSpot.com, your life becomes hugely easier. It also helps to know Python and Maxscript so you can script your repetitive processes.

 

The grass is not displacement. It's a 5'x5' square of geometry grass blades that I created and turned into a Vray Proxy. They are quickly placed around the scene.[ATTACH=CONFIG]44425[/ATTACH]

 

Though, I really have to stress this, this is only for your very typical cracker-box style homes. One-off custom homes take substantially more time, at least 24-40 hours depending on square footage or detailing. These renders are my Big Mac renders. Just like in McDonalds, you order a Big Mac and you have some slight customization options, but you pretty much know what you'll get for the price. It just happens to be that where I live, these are the homes that are being built. Personally, these are frustrating due to the lack of overall challenge, but they are more than paying the bills a this point.

 

The render size depends on what they want to do. If they are printing newspaper ads at 3"x5", then a 1280x720 image works just fine. If they are going full page, or larger brochure ads then they'll usually need a higher res. But I really make sure I'm in contact with them to know where this is going to go. Before, I would drop hours on 4000 pix renders only to find out that client was using the image on the web or a tiny newspaper ad. Talk about a waste of time!

 

Much of the inspiration for this process comes from the environment art process. They don't model the whole scene, they model major pieces then reuse them. My process is similar and it has allowed me to shave hours off of the budget for these very repeatable style homes.

 

Source: http://www.thiagoklafke.com/zestfoundation.html

http://www.thiagoklafke.com/modularenvironments.html

 

This process is also helped that I have a trust level with the developer and they stay out of my way while I render. I've done enough jobs for them that they know when they send me a request, they'll get it back in high quality. The noodling, back and forth, annoying client meddling has been removed from the process. But I also write that into the contracts. They have to sign that I have explained the process to them and any deviations from that will result in extra charges at 2x the normal rate. That pretty much stops any and all client side meddling and only the most important changes are addressed.

 

As soon as the CAD is received a general camera location is set and that does not change. For example, in the image I posted, they wanted to really show the solar panels which were on the garage side of the house. So only the garage side and the front were modeled. Heck, I didn't even model the front door. Why? Because you don't see it, as it's a deep recess. Why waste time modeling elements you never see?

 

How many of you have looked up on your past work only to see your 4000 pixel image used as a 240x240 thumbnail on a webpage? Makes all of that render time a waste of energy. This is why it's very critical I make sure I know what the client is going to do with the image. Trust me, by the time they contact us, they know they marketing uses for our images. Why not streamline the process a little?

 

If you are blindly rendering sizes, or allow excessive client meddling, you are not communicating your process properly. You need to have that dialog with the client.

Edited August 18, 2011 by VelvetElvis

[Scott Schroeder]

1024px is useless for anything other than email.

 

I appreciate the input, but I have to contest this comment. This shows an lack of understanding of the basic fundamentals of end target resolutions. So let me explain. In my case, I talked with the clients. I was able to get a copy of the brochure template they use in every one of their sales centers. The brochure is 1/2 sheet of paper with roughly 4"x2" set aside for the elevation renders. The rest of the space is taken up by the floor plan and home information. Also in talking with the client, every sales center in every one of their locations uses the exact same TV, which is a 32" with a max resolution of 720p.

 

So, in doing the basic math of resolutions you should know that to find your target image size you take 300dpi times the final image size and get your render resolution. This also works in reverse, you take your render resolution sizes divided by 300 and you get your max print size. Well, a 720p image (1280x720) gives you a max print size of 4.2"x2.4", which as I mentioned earlier, is more than enough for the allotted space in the print brochure. And if all of the TVs that this will be displayed on max out at 720 resolution, it makes no sense to render at anything larger. Their website is also created at a fairly low resolution in order to open it up to a wider range of people, so any image at 1280x720 is more than sufficient for their website.

 

The most popular website size, as of 2011, is still a laughable 1024x768. http://www.hobo-web.co.uk/best-screen-size/

Sure, we all have massively high resolution monitors. But your average consumer looking to purchase a home does not.

 

So if a 1280x720 image works for everyone of their deliverables, why in the world would I as an artist render at anything higher and waste that time? Again, it really helps to understand your end target resolutions and not just blindly render out huge image sizes.

 

I only work with myself and use my main system to render. So I have to be fully aware of what I am doing as to not waste down time rendering. When you have a large amount of elevations to render over a span of a month or so, you can't be wasting precious render time for an image size you'll never need.

[Joel Gray]

Scott (VelvetElvis)

I like your thoughts on deliverable resolution sizes, Scott. It makes me think about my own process and how I might be able to lower the resolution size (and time/fees) on some of my projects. Overall though, most of my clients are architects and they want poster sized prints of everything. Web, mobile phones and ipad-like devices would all be well suited to your approach, especially if you're discussing with your client in advance exactly how they intend to use the image.

 

Jon (Chroma)

You'll get a better feel for the time it takes as you gain more experience. Simply stated though, more time usually equals better quality and higher fees. But if you're new to the industry then be careful not to assume that your average time spent on a project is on par with your competitors. You really have to be critical of your own work and the time you spend... Nicholas (BVI) gives some solid estimates for you on entry #15.

 

Good luck!

 

Joel

[Jon Berntsen]

To the resolution guys here, I want to add my experience. For me, eventually almost all illustrations that is being delivered lores are going to be used for print at a later time, and of course asap when you don't have time to render it. They just call you and want it for an example blown up on a construction site sign "today" and since they expect this service for free, they've already ordered print time. I have a workflow where I render final stuff when I go home for the evening, and then post process everything from that big image. I can still charge them extra when they want the big size stuff, but the advantage imo is that 1) I can deliver it pronto, and 2) I don't need to duplicate the photoshop post work (and perhaps spend time getting it right the 2nd time). But I guess we have different clients, different projects and different workflow.

 

To the #15, this is what I'd call a reasonable amount of time spent. I've got a post in another thread right now. I can attach it here as well, it's also a WIP, but an indication of what I'd spend about 20 hours with. If anyone else has some examples as well, I'd be interested just to see. Big studios, people from India/Argentina/Ukraine ++, freelancers, and basically everyone.

 

Have a wonderful new week.

 

[ATTACH=CONFIG]44457[/ATTACH]

Edited August 21, 2011 by chroma

[BVI]

I find the problem with supplying a low resolution image as a final is that the client may come back 6/12/24 months down the line and want a higher resolution. It's always a mission to retrieve the file, resetup the render then redo the post etc. Its better to render out the image as high as you can, I'd say 3500px is a minimum and then send that off - that should suffice for an A1/A0 etc.

[Brodie Geers]

Why would you need to redo the post work later on? Isn't it pretty common practice to do everything on layers and work nondestructively so if there's a minor update you can just replace the background layer with the updated rendering.

 

-Brodie

[Dave Buckley]

Why would you need to redo the post work later on? Isn't it pretty common practice to do everything on layers and work nondestructively so if there's a minor update you can just replace the background layer with the updated rendering.

 

-Brodie

 

 

That's what I thought, render small resolution for draft views but always work with a full res .psd?

 

For example render 2000x1500 then increase image size in PS to like 6000x4500 and do all the post work on top of that blown up render. Just shrink a copy-merged version for sending back to them at the original render size. This way when you come to render the full res version you can just drop it underneath all your post work that you've done previously?