|07 July 2018|
Los Angeles , USA
Week 1 Hardware, Software, Equipment, Experience
This week was spent trying out a few scans from simple to more complex without much prep.
The Idea for every exercise it to do it wrong and go back and do it right! Why start out reckless?
It's better for me to experience first hand why things go bad and how to manage them on the fly.
This is pretty easy to do with photogrametry.
Check back to this thread next week for a summary of my first weeks trial and errors.
I will also share all assets or attempts at assets rather for those that would like to see comparisons etc
Last edited by Travis : 3 Weeks Ago at 08:17 PM.
|07 July 2018|
Los Angeles , USA
Software Week 1
Agisoft Photo Scan
1st quick, test scan.
Set to medium for speed and using image size scaled down to 1886 pixels.
Even for a quick test with low res images and medium setting the results for
the first point cloud are a lot of fun
I will update with ultra High and full res images over the weekend and do a
test comparison to see how Recapture goes
I will list equipment and go into detail in the next post. Here is a preview of a quick test shot.
Up Next i
Autodesk Recapture ]https://www.autodesk.com/products/recap/overview<br[/img]
Last edited by Travis : 07 July 2018 at 05:45 AM.
|4 Weeks Ago|
Los Angeles , USA
1886 WAS just to get file size down to do some fast test scans.
The software adjust your cameras for you so no worries there.
Mainly just try to shoot to the most perpendicular areas you can
and more images seems to typically be better
|3 Weeks Ago|
Los Angeles , USA
Part 1 of week 1 getting familiar with shooting . predictions going in and updated opinions after a week of experience.
If you are getting into this first things first which is looking at what your gear is going to cost if you want to do this frequently. Having shot the first week I can honestly say it’s a lot more work than anticipated but also extremely addicting and the idea of building a collection of assets that make scene creation especially for quick concepting one of the most motivating and rewarding areas of CG I have experienced so far. It becomes obvious quickly that the tech and process are the among the few things holding this back from taking over a considerable amount of the creation process.
The following is not a guide or tutorial for what you should use do or charge just my estimate for what I would charge, What I am using, and What I would consider a value for time and return on work put in.
Starting with Gear and Tech from a hobbyist or noob working towards professional I will break down where I am, and what I am using, including some of the cost you may consider as well as alternatives that I would suggest.
What are the current Value of Scans and Recruiting talent to capture for projects.
My Idea of pricing prior to any research or experience and without interviewing artist and studios
If I were to base my price point for shooting scans off of my first days scanning efforts. I would say that a fair price for my time would be in the $600 to $800 or essentially 100 dollars a scan with each scan of 100 to 300 images. Time taking a lil over, or under an hour, if everything goes well. If everything goes south then 4 to 6 scans. All of this would depend on scale of each scan and travel needed for captures. Every hour of travel would be an additional 30$. A typical day to shoot I found was 9 to 5 so figure and additional 2 to 3 hours on each side for drive time to locations and back. Any earlier or later tends to create long shadows adding more time to clean up etc.
After 5 days of scanning I realized that realistically shooting decent scans in the size range of 6 foot by 8/10 foot areas would take about 1.5 hours to do correctly. 100/300 shots per obj.
The scanning software can be very picky on the order or way you shoot. Overlap and can be drastically picky if not shot precisely. This means if you travel to remote locations or any fair distance you better shoot each scan subject at least 2x I found on more than one amazing scan that a section of bad shots or just shots that don’t register can leave gaping holes in your subject. Sometimes the Fault is yours others everything looks right but the software just doesnt like it so take my advice if its a hero peach and make sure to take your time and shoot the asset 2x minimum. I will be making a second 5 hour drive to sequoia to reshoot just 3 obj for that very reason. Nothing is worse than rendering out one of the coolest assets you have ever seen for a game only to discover you are missing huge chunks from bad data.
After Week 1 my Idea and price point changed drastically which is exactly what I wanted as I am jumping in blind and pushing through to discover as much as I can.
Equipment gets heavy fast even if you are in shape. Most locations I visited required a 1 to 4 hour drive. At the location I was looking at anywhere between 2 to 5 mile hikes as well as scout time to find the OBJ I wanted to scan. My gear solo was a backpack with tripod, Monopod, Gimball, Ballhead, Extra Battery, Water, 2 prime lenses all of this equals about 40 to 45pnds when all is said and done. Most of the first scans I did suffered more from impatience than lack of experience. I got tired of moving and re adjusting the tripod so handheld and occasionally hit the zoom, missed an overlap by to large of a margin or simply after shot 150 had shaky hands. I was also always aware I was losing time and wanted to jump to getting another great scan before running out of time. I would say this is the biggest lesson week 1 taught. How the rules and equipment were secondary to taking the time LOTS of time to shoot right and plan it all out and then do it again as a fail safe.
Having shot for 5 days my original idea was just to go to familiar spots with features I understood would make good scans. I was way off on my predicition of time invested on this. After the 3rd day shooting I decided scouting and shooting the same day was not ideal. My method going into week 2 will be to plan 2 scans per trip and one small environmental 35 to 100 shots.
I also added gear to my kit which includes a small hand broom and knife to brush off and clean up OBJ prior to shooting. This can add a good 30 minutes to each scan.
At the end of the week I would say that 4 to 5 hero scans (100 to 300 shots each ) or 8 to 10 environmental scans (35 to 100 shots) with an assistant would be a day rate of 800 to 1200.
My Camera Settings and Equipment.
Weight will likely be a factor even if you are in shape. Also it’s smart to wear pants and decent boots as you'll be getting down on the ground and in brush so play it safe. My original gear was about 40 pounds for Week 1. Going into Week 2, I will be able to shave that down by about 10 to 15 pounds.
Below is my original set up and suggested alternatives or changes as well as some thoughts on cameras. Keep in mind I am a noob and feeling my way around so research for yourself to get your kit in good shape.
Cannon 5d mkiii 23mp 5 pounds
Cannon 5dsr 50mp 816 shots per 64gig card roughly half of the mkiii shooting raw 5 pounds
RRS Ball Head 2 pounds
Fiesel Tripod and Gimbal 5 pounds
Sigma 24 to 35 mm 3 pounds
Sigma 50 mm 3pounds
Batteries and accessories 3 pounds
Fiesel Mono Pod 3 pounds
Total 29 pounds
+Backpack, Snacks, Water 7 pounds
Most of my gear going in I have for shooting portrait for character concept and videos. While all great quality I found the tripod far to heavy and cumbersome to constantly adjust to the terrain while trying to shoot. My suggestion on tripods is stick to anything in the 80 to 200$ range. Its going to take constant beating and needs to be light and adjust quickly. After shooting for 5 days the lighter cheaper the better because your going to be moving and adjusting this thing 100x in a day and banging it into rocks trees mud etc
Tripod and Heads
What I used week 1
What I would suggest as a better value
You will likely upgrade pretty fast but a tripod along these specs will always be good for an assistant or to shoot in crappy areas where your tripod is going to get banged up. It comes with a ball head which I found to be much faster and convenient than the gimbal head but also returns more errors or chances of losing a shot when it comes time to import into scan software.
There are cheaper solutions but this price range will get you a tripod that will support 17 to 20 pounds and as much as you’ll be moving around going any lighter will be regrettable I think
Camera In all honestly shooting all week long had me wishing for the Canon rebel or Sony swivel screen. Your going to get tired of bending over and have a lot of views that are tough to see when shooting so this would be a nice feature. I also thought having Sonys 40 mp would be a nice way to get those high res sharp images without the grain I get from higher ISO with the 5dsr. More on that later.
5dMKIII continues to be a great value compared to almost all the other cameras I have tried. Especially if you have installed Magic Lantern. The res is on the lower side but more than enough for good scans at 23mp. Shooting Raw leaves you with more than enough shots during the day with just 2 cards.
5dsr Resolution Monster!!! It comes at the price of being almost limited entirely to 100 iso. I shoot portrait with strobes so this typically isnt a issue but shooting outside with high g stop means its going to be really hard to keep the iso low and clean. I was originally excited to shoot with this beast but quickly found that the resolution for most of what I am shooting is more of a hindrance than help. You have way more pixel blur due to movement or shake so it becomes far more important to go slow use a shutter release and shoot on a tripod throughout the day. You are cut to about half the images in raw as the mkiii which means your only getting about 2 to 3 good scans per card. Way longer transfer times at home and the increase in resolution vs time it takes the software is nowhere near the value to make it worthwhile in my opinion for at least 2 to 3 more years.
My best suggestion if I were to buy today would be the current Canon 6d full frame at 27mp which gives a nice balance of # of shots in Raw, is half the price of the MKiii I believe and resolution your computer and software can handle for typical 4 to 7k machines. Time wise my machine handled most 200 shots in about 45min to 2 hours for the mkiii on ultra and jumped to 4 to 7 hours with the 5dsr. You can always downres the files but doesn't make much sense if you have the mkiii
My Computer i7, 64gb ram, dual geforce 1070 from research it looks like one 1080 will do equal to slightly better than dual 1070. For the cost dual 1080 doesn't seem to give a significant enough increase in time to be worth it. I am looking to see what the online render solutions might be as paying 20 to 30$ to upload to the cloud and forget about it would be a great solution.
Gas is 4$ a gallon here so I'm looking at 40 to 80$ a trip for most of the places I and capturing. A lot of artist I know don't include the cost of their equipment or travel in their pricing so I list this as it is a relevant expense.
Hotel 100 to 150 I often travel 5 hours out and back in the same day but as mentioned above 1 day for scouting and planning and one day for shooting. The scouting is a part of your typical day rate and something I wouldn't pass to the client unless they specifically choose a destination. Your Knowledge and experience is a part of your cost so knowing where and what to shoot should already be built in.
Triple the time you think it will take. I will be far more efficient and fast no question but I would be curious to here how many scans other artist solo make in a 8 to ten hour day as for me my initial was far to ambitious to shoot correctly. I would also say that finding a client willing to do clean up and handing off scan images is worth a discount. You can easily double your time going through cleaning up and packaging the content
Week one proved very quickly the extraordinary value of Mega Scans.
During Week 1 I quickly realized that my original idea of getting 6 to 8 scans in a day was way off base especially shooting without an assistant. Prior to shooting Week 1 I bought the montly subscription at 40 after day 3 of shooting I decided the yearly rate was more than worth it especially considering all of the typical assets you need but don't want to take time to shoot yourself. Get your basics from Mega Scans and Keep your hero scans for your own collection to ensure original looking imagery. My price point would put a single hero scan from 1 to several hundred dollars where as an equal Scan at Mega scans packaged and ready might only be 8 to 12 credits> I do feel package discounts on entire collections at discount would be a good idea. I don't need 300 types of Grass but if they were packaged and a deeper discount I would be more than willing to buy the entire collection instead of the 25 to 20 I need especially knowing the value.
These are my thoughts so far this week about prep and experience. Later this week I will post some of the scans and go over the experience for each to share. Hopefully some of you can jump in and share your thoughts, scans and experiences as well. Maybe you have a different experience or better advice. This thread is all about starting blind tripping and taking notes along the way so feel free to offer a hand to everyone starting out or ask for one
Important additions I added to later but are now a must.
White Balance Card. Even after getting a color card and white balance I forgot to shoot them the last two days.
Shooting will save a lot of time. If anyone has a nice look dev to share on how to correctly use these for look dev please share for all of us to get familiar with.
Hand Broom. Most good rocks I found in the Forrest had a lot of small debris. Running a brush over them will save a ton of headache when it comes to clean up and render errors.
Small Knife or cutting tool for weeds and junk around Assets. As a side note I figured shooting in places like Huntington gardens would save time by putting lots of assets in close range of shooting. Issue is everything is far more crowded overgrown and leave little to no room to navigate around the subject to shoot as well as the inability to clean up the asset>
Next up Ill discuss Camera Settings and Lens as well as issues I ran into. Overall I have never got as much use out of my 24 to 35mm sigma as I did this week it really is the workhorse of my set up. 50mm gets cleaner sharper results but requires more room to move and more shots with a higher risk of loosing info or poor spacing from my experience so far.
I apologies ahead of time for difficult reading as I have to fly through this between other task. anyone wishing to help out and do a nice edit is welcome hopefully there is enough here to follow along and understand.
|reply share thread|