CompuServe Thread

Pricing Mesh's

18 messages in this thread
#108027From: Paul LindJun 16, 1994 2:01 AM
Sam, I agree with John. Call Viewpoint! Nobody else sells meshes like they do. Since my company does a little bit of custom digitizing, I'll add a few remarks about the mesh you uploaded. A.) Since it seems that you've defined a specific resolution for the mesh you're most likely going to have to target a general marketplace. We've done a few scans for various members of the medical profession and from looking at the C2 you uploaded you're mesh is too detailed for alot of clients and "way" too simplified for alot of others. B.) The mesh appears to have been done with a stylus digitizer instead of laser scanning, because your triangulated mesh looks like it is derived from "pick points" not a split beam read. If you did use a laser scanner and have the original data, it looks like you're throwing away alot of relevant data. You may want to generate a number of seperate skeletons by decimating the original data to various levels. Make sure you offer the original scan data in one of your offerings because a fairly large protion of the medical profession enjoys the ability to be able to use true scan data. C.) 5 people, 2 years? Do you have all of the muscles and organs in addition to the skeletal structure? Give us more detail about the rest of it. D.) How ever you decide to market it, keep in mind that there are a number of digitized skeletons in the medical profession already. They range from resolutions far above the C2 you uploaded to lower res ones like the Viewpoint skeletons. Paul
#108062From: SyndesisJun 16, 1994 8:32 AM
Wouldn't life be great if we had vocabulary and metrics for a common language to describe the resolution of a mesh? Raw triangle or polygon count or point count won't do it. We get calls all the time regarding the 3D-ROM: "I need a baseball. A high-resolution baseball." How high is high? What makes "high" different from "low with a great map"? Some people want 100 triangle humans. Yesterday, a guy wanted flat 2D human outlines. Another wanted human arms *with detail down to level of hair*. At one thing can be stated with certainty: you can't get blood from a turnip. For this reason, perhaps the best way to ship polygon-based models is to ship the highest resolution, then recommend a decimation tool. But no decimation tool is perfect; some hand-tweaking is always necessary, no? Another insight from 3D-ROM sales: People will spend three days looking for a model, because finding the model will save them time. They also want the modeling company to psychically determine the suitability of purpose, to match the model to the animation that their client wants.
#108117From: Paul LindJun 16, 1994 11:57 AM
John, That's the exact reason that we will never get into the mesh selling business. Way too much crap to deal with. Everyone needs a very specific mesh and no-one is ever settled with a generic middle of the road mesh. Paul
#108185From: Jun 16, 1994 8:42 PM
I ceratinly would be very interested in seeing the skeletons that are of MUCH higher resolution. Since I am in the medical profession I have done alot of research into this and have not come across anything that is higher resolution. There have been some things done with workstations and put out to videodisc such as slice of brain, but the data is not available in mesh files. Even the 3d work that is being done in CT and MRI does not have the resolution, if it did I would have pulled the data from that. I appreciate your input because you are the first person that has ever mentioned there are high resolution skeletons available. In regards to viewpoint see my earlier message. Thanks Sam
#108197From: Paul LindJun 16, 1994 9:37 PM
Sam, I have personally seen a number of skeletal structures that are phenomenally high. I'm not sure what part of the medical profession you're in but several medical research facilities on this coast have skeletal structures that are truly phenomenal. The military research departments also have quite a collection. I doubt that any of them are for sale but that's probably do to a lack of a market. As to workstations, we run quite a few of them here. I pulled you're mesh into Alias to view it. We recently scanned a skull and the data was at least 10 times the resolution of the C2 that you uploaded. The scan showed all of the various ridges and bumps that exist on bone, it even showed the tiny ridges in the teeth that occured from lack of brushing. Not to mention that it took all of about two hours to do 12 scans, mesh them together and deliver the final mesh. That's why I asked how you went about building your skeleton. Exactly what type of laser scanner are you using? We're running a Cyberware 3030 on a MS motion platform attached to an Indigo2 with 384 megs of ram. And it doesn't take anytime at all to digitize data. An 8 second scan of a human face provides such a high level of detail that it will pick up a light 5 o'clock shadow. The first time I scanned myself I found a scar that I thought had disappeared over 10 years ago. Not to mention that it simultaneously generates 24 bit color texture map info. We do have a host of proprietary tools to work with the data, but nothing beyond what most laser scanning facilities run. I understand your point regarding CT and MRI data, we've used the machines on a number of occasions to get data from museum pieces. That's why we run the Cyberware unit now. What hardware platforms are you running on? Paul
#108206From: Jun 16, 1994 10:32 PM
I certainly would like to talk with you in more detail regarding what you are doing. In terms of high resolution of the skeleton, after critically looking at them from a medical standpoint they are not ANATOMICALLY accurate. Let me give you an example. You mentioned that your skull is high definition with all the ridges, can you accurately discern where the nasal bone is?, can you see where the inferior orbital fissure is? looking directly into the eye socket can you see the 1)posterior ethmoidal crest? 2)inferior orbital fissure? Is your skull scanned as one whole object? Did you know the skull consists of 22 bones? Its these little things which make the difference. In terms of how I know its because I'm a Neurologist and have access to many of the high tech toys and the availability of neuroradiologists. Also, I am working with the University of Washington and they are one of the top schools in the country when it comes to visualization, and they dont have any high rez skeletons. I am not directly doing the scanning so I cant give you all the technical details. One last thing and I hope this isnt too sensitive an issue. You mentioned you have seen skeletons from low rez of Viewpoints to high rez. I looked critically at viewpoints and felt it did not meet my needs, the last price I have of there skeleton high rez is 5K. You can purchase, if you can find one, a HUMAN skeleton for about 3K and from thoughts on the future of VIRTUAL REALITY etc, is that virtual reality will still be for people who cant afford the REAL thing!. This gets back to my original question about how do you fairly price these things. Lastly do you have a number or address? Thanks Sam
#108217From: Paul LindJun 16, 1994 11:13 PM
Sam, Please don't take any of this stuff too personally. I'm definately not a medical person and have little or no knowledge of the medical profession or medical needs. Then again, by asking about pricing the skeleton on the Autodesk forum you're not exactly dealing with the AMA. Half of the stuff you asked about the skull is way beyond my knowledge. 🙂 I do however know a little bit about digitizing and meshes. With an investment approaching a quarter of a million dollars in hardware and software solely for scanning, we're fairly serious about mesh quality and usefulness. Our digitizing tends to be for only in-house visulization projects and we've been doing some custom work for several research facilities. The reason I popped into the thread was because when you posted the original message you mentioned laser scanning, custom code, proprietary software and unequalled resolution, 5 people/ 2years. While I'm not interested in yet another skeleton to add to our collections, I am interested in people/ organizations performing state of the art custom work. After downloading the mesh, I was struck by several things: a.) The mesh is almost definately not a laser scan product. The resolution isn't even in the ballpark. The circular patterns on the round areas are indicative of a stylus based digitizer. b.) The data is not as well defined or decimated as it could be. If custom tools were used to decimate the data, it is amazing that the final data is identical to triangulated pick point meshes. In dispersing the point cloud, the arrangement of the triangulated mesh could have been done alot better, leaving you with more data in essential points and less data in non-essential areas. Will Schroeder at GE Corporate Research and Development has written a number of papers and collaborated on a number of projects that could drastically aid your project in the area of point cloud decimation. [>> Continued in next msg]
#108218From: Paul LindJun 16, 1994 11:13 PM
[>> Continued from previous msg] c.) In comparison to the level of data that research facilities use, and the data that we see on a regular basis, the C2 is fairly low res. I was expecting to see something alot more defined. Although I didn't know how many bones there are in the human skull, I do know that an accurate scan will show every jagged edge of every suture and will reveal hairline cracks whereas a point stylus scan will not. The people that we've come into contact with always have a specific need and always request the ultimate level of digitizing. When I first read your message I thought that I could refer things your way, After looking at the C2 I realized that it wouldn't help most of the people we deal with. d.) At the resolution of the C2, a single person should easily be able to build an entire skeleton in under 2 months including decimation, touch up and translating into any format needed. Even breaking the skeleton into every imaginable piece shouldn't take a person any more than 6-9 months at the outside. That's why after viewing the mesh I asked about how the work was done and commented on the fact that the level was not what I was expecting. I definately didn't mean it as a slight and sincerely hope it didn't appear that way. The C2 is nice, it's just not what I expected. Paul P.S. – A good price would be under a thousand dollars, for wide spread appeal you'd most likely definately have to come under $500.
#108232From: Jun 17, 1994 12:44 AM
Thank you for all of that info, I have just learned a tremendous amount in just a few threads. I appreciate all of your input especially the pricing info. I would be very, very interested in seeing at least some portion of your high rez skeleton or skull. I did not take what you said as a slight, there is always some pain in the learning process especially when I find out that there is more out there that I certainly thought(G). Once again I would really love to see what you are talking about. Thanks Sam
#108238From: Paul LindJun 17, 1994 1:52 AM
Sam, No problem. What programs do you have to view it with? How do you need the data formatted? We actually have a human skull here in the studio, so I'll see if I can't throw it onto the digitizer this weekend. Do you have an Internet account? CIS e-mail can get expensive. Paul
#108396From: Jun 17, 1994 9:07 PM
The programs I have to view the data in are 3d Studio and any of many graphic viewers if you just want to send graphic files. You can send it to me directly at Samuel Coor 202 41 Ave NE Olympia, WA 98506 I look forward to it. By the way I now have an anatomically accurate vertebral column, sacrum, ilium, and coccyx. They WILL be made affordable thanks to all your input. Thanks Sam
#108399From: Paul LindJun 17, 1994 10:02 PM
Sam, We'll get a few things together and send them your way. Paul
#108241From: ALAN IGLESIASJun 17, 1994 2:03 AM
Wow, Paul, that's quite a rig. Maybe if I came to visit you that thing could find all the *hair* I had 10 years ago!<g> May all your new face wrinkles be smile ones, -Alan
#108247From: Paul LindJun 17, 1994 2:37 AM
Alan, We could certainly try. 🙂 Paul P.S. – I still haven't gotten over that horrible flying logo from the Luxor ride. I'm not sure about you, but it syill haunts my dreams. :-]
#108321From: ALAN IGLESIASJun 17, 1994 12:17 PM
re bad Luxor Logo: Hee Hee…It probably wouldn't have been so bad if we didn't have to watch it a hundred times in that stupid line! The actual rendered scenes in the ride itself weren't bad, though. Nice spotlight-in-darkness effects. I think I'll make a huge oblisk(sp?) and use Gary's Scatter routine to make my own virtual underworld… -Alan
#108389From: Paul LindJun 17, 1994 7:19 PM
I liked the actual show, I thought the promos and other hype was trashy. It almost seemed like they decided at the last moment to add the promo stuff. Paul
#108420From: ALAN IGLESIASJun 18, 1994 5:40 AM
I know — It *did* have un "unfinished" feel to it, didn't it? Hmmm -I remember thinking that- oops, they musta run out of money! -Alan
#108250From: John EllisJun 17, 1994 3:07 AM
Paul, You've made some really insight points as has John F. Looks like along with breaking new ground, new terminology needs to be developed as well. -JE