Architectural Verified Views

[ollie_XYZ]

Hi,

 

I was wondering if anybody could help/comment on this subject.

 

I have heard quite a few companies mention they produce verified views for buildings. I understand the purpose of this (planning, environmental impact asessment), but I have no idea about the processes involved.

 

To Quote:

 

"Images that can be generated using supplementary survey data to create verified views which can be checked by third parties to confirm accuracy."

 

I found this text on a CGI companies site. So, as far as i understand it's a process of ensuring an architectural 3d model comped into a photo is accurate in terms of structure, and in relation to it's surroundings.

 

That's what i understand at least, dunno if that helps or even makes sense?... i'm really interested in the process if any body knows?

 

Thanks,

 

Ollie

[upshot]

I think you hit it on the head.

 

Whenever I create an image with a real background I know there are some serious fudge factors. I 99.9% of the time I cannot pinpoint the camera's position, FOV, Lense length, etc.... So, it's the ol' does it look right? Are the lines converging in the same position as the photo? Alrighty then!

 

I've always wished that I could get a surveyer to take site photos with specific data right down to the XYZ angle of the camera as well as the obvious stuff like it's position, lense type, and all that... Maybe that's what they are selling?

 

There is also software that you can pinpoint stuff in a photo as reference points and then specify those points in your model. I've never had much luck with this method but, it's suposed to be able to get the virtual camera correctly matching the real one. Nice marketing... ;-)

 

The bottom line is that you can make a super-big project look a little bit smaller or a somewhat small project a little bit bigger by playing with your camera. I wouldn't doubt that you could, with the right equipment, make the future project "perfect" or "as it would look" but, I'll raise the BS card on 90% of what you see out there...

 

any other opinoins?

[Ernest Burden III]

"Images that can be generated using supplementary survey data to create verified views which can be checked by third parties to confirm accuracy."

 

Why the need to mention a third party accuracy check?

 

What they are really talking about is--building an accurate model from site surveys. Matting into photos is another aspect of having good base data, but not a particular proof of truth.

 

And you can reverse-engineer photos to get semi-accurate to very accurate data.

[Jeff Mottle]

Hi Ollie,

 

Our London office at Smoothe routinely does what are called a certified views. Given the historic nature of Europe and London in our case, planning commissions are very particular about the imagery that is created for high-profile projects that have impacts on views etc. If I recall, we have actually had out processes approved by the planning authorities as the final images are in essence a legal document. To do a certified view we have a proprietary process, but it essientially involves getting a photoghapher, and surveyor involved. Along with the survey data and existing city site plans, we can ensure that our camera matches are identical to the actual proposed building.

 

We do have a full document that explains this process in-depth, but I'm pretty sure I'm not able to release it publically. sorry.

[3dp]

hayes davidson also in london have a verified view methodology

myself if working on planning images i'll insist taking the shots or being there when taken so as to note the surrounds camera etc

in the uk a good os map is also needed (although they can be out when compared to a local survey of a site or is it the other way around)

if i end up with a stitched background image then the process and settings are noted ie 3 shots at 35mm flattened out to a 115mm

basically the aim to get the model to sit in the background plate so there is no distortion that is glossed over in photoshop later

 

basically as long as you can illustrate EVERY stage of the process if it comes to appeal/litigation if the work is sound you are covered

[upshot]

Obviously I have some bias on the subject but, I think it's complete BS that people want to get litigious over these renderings. 'photo-reality' is in the mind of the beholder and If you were to hand sketch a view using the 3d model as a reference no one would even consider a suit over 'visual-tricks'. I can't tell you how many times I've beat my head on my desk repeating "it's not reality, it's a computer aided sketch you fools...". One trick I've used is the ol' photoshop filters to help the beholder to clue in...

[ollie_XYZ]

Hi,

 

Thanks to everybody for there comments and info... including Upshot But I asure you it's not BS if that's what your client requires!

 

Much appreciated,

 

Ollie

[samhutchison]

Hi I am hoping to get into producing verified views as a profession.

 

I am a landscape architect so already have an in-depth knowledge into how the method of these should be carried out. However, what I don't know is what equipment I need to do this?

 

- What type camera will be sufficient? Does it need gps built in?

- If a topo survey has been undertaken already, I assume I can just place the model and GPS location into that to accurately locate?

- Is the 3d Max software the best to use?

 

Any advice will help me tremendously.

Sam

[ronll]

I've done several projects involving potential view obstruction by proposed construction. In a few of these cases, I not only did the imaging and photography, but also had to testify in court as an expert witness on the subject. The method always involved taking very accurate legible notes on camera location, lens measurement and testing, methods of 3D imaging involved, and hiring surveyors to document juxtaposition of existing and proposed structures. In all measurements in the process there will be small errors which accumulate when combined in the final result, making it difficult to state what the margin of error is going to be. And obviously the closer an object is to the camera, the more important the error is, but also the easier it is to measure accurately. Around here, views of mountains are important. The locations of mountains can be pretty effectively measured from Google Earth, but if you are off by a couple hundred feet, they will probably still see the mountain.

 

In the cases where I testified in court, it had a lot to do with the skill and preparation of the attorneys involved. In one case the opposing attorney was the same attorney who had been on my side in the previous case. So on cross, he knew how well prepared I was so he had to resort to trying to show bias on my part because I had been hired by the defendant. But ultimately it comes down to a judgement call on the part of the judge anyway (none of these had juries).

 

BTW, my fee for legal testimony is triple my normal hourly rate and is written into all contracts as such and is pretty much accepted as normal in the legal community. In one case, my scheduled time was delayed by four hours while I sat in the courtroom, and then I was on the stand for five hours the next day. Altho I hate testifying, it pays well.

[ronll]

 

- What type camera will be sufficient? Does it need gps built in?

- If a topo survey has been undertaken already, I assume I can just place the model and GPS location into that to accurately locate?

 

I don't think GPS is going to be nearly accurate enough for camera location. It can be off by several feet. In my cases I try to measure the exact location, including height, of the film plane (ccd plane) relative to the walls, window frames, etc. A few inches off in the location of the camera can make a big difference. As to lens, I will use a fixed lens (no zoom) and test it in my office by taking pictures of a wall and then measuring the field of view and do the trig to the film plane. You cannot trust the manufacturer's stated FOV.

Edited February 15, 2012 by ronll

[ronll]

Obviously I have some bias on the subject but, I think it's complete BS that people want to get litigious over these renderings.

 

In my office, these are two different types of projects; renderings or "pretty pictures" and view analysis images which are usually not very pretty.

[mubb]

Hi there,

 

I thought of adding my 2cents here since I have been producing verified views for the past 7 years in London.

 

The process includes consultation between the architect/developer with the planning council to determine which views are needed for the planning application. Then a visualization company is commissioned to do the job.

 

In his turn the visualizer producing the Verified Views (also named AVR) will instruct a photographer and a surveyor team to do the following:

 

1-A professional photographer will take photos of the site (based on the views agreed by the council). He will use a special camera for this job.

2- The photographer will mark the exact position of his camera on the pavement.

3- The surveyor team will go on site to survey the camera position. Also the surveyors will survey strategical points of top of nearby buildings and street features (these points are normally specified by the visualizer).

4- The photos along with the surveyed cam location and strategical points are sent to the visualizer.

5- Using these information, a camera match can be done by the visualizer and a verified view is produced.

 

after this there is a process of inserting the architect building into the verified view and creating a photo montage, I wont go into describing this process since it is too long to describe

 

and this is how you get a verified views.

 

let me know if you need further help.

 

Maher