Works wonderfully in FF 4b7. It's a pretty simple model and I'm on a high end machine, so that isn't all that incredible (like a side-scrolling canvas demo), but it's neat.
I have the exact opposite experience actually in FF 4b7, and though I'm not running some crazy awesome machine, its still a decent box. The page basically crashes the browser.
Huh. I thought this was something completely different. I thought this was a robot to help cross-references submissions about the same topic, and find duplicates that the built-in dup detector doesn't find. As such, I thought it would be valuable.
If it's a bot, I can see the value in it. It would be better as a built-in feature, but nevertheless it makes sense. I just hope that it doesn't inspire less useful accounts which would be more akin to Reddit's novelty accounts.
This is so cool. Very well done. A feature request would be to add layering for muscles: I can't pick out the hip rotators and Piriformis. Ditto for Transversus abdominis.
This looks awesome and runs fast on my quad core system using chrome 9.
I got carried away dragging and zooming around but missed one thing. The search box is really powerful. You can locate each and every organ using it. I can imagine this being incredibly useful for medical students etc.
What's the upside of fully exposing the skin? I don't think skin exploration is the goal of this project, and this way they aren't alienating part of their audience.
this runs a little slow on my mac, anyone else having issues?
Also, what are those things on her nipples? You can see it clearly when the opacity of everything is set to 0 except for a slight opaqueness of the body.
Because Chrome has a multiprocess rendering architecture, and OS 10.5 doesn't provide a good path for hardware-accelerated rendering across processes. This means Chrome on 10.5 has to render each frame on the GPU, read it back to main memory as an image and then repaint it to the screen elsewhere. The bigger the window, the bigger the image and the longer this readback/copy takes.
You know what I would like? Google for my house. As in, I get a set of rfid tags that I can put on my keys, my remote, checkbook, or whatever. In the event I can't find something, I have a little device I can use to track down the missing item.
This is interesting in terms of its potential. However, right now this is really the bare minimum implementable unit of functionality and is a little too limited to be of practical use in teaching anatomy.
Features I'd like to see:
1. Arrows from the labels to the anatomical structure so you can tell which label refers to which structure without having to interactively click each label to highlight the structure.
2. I assume they've built their 3D model from the publicly available Visible Human Project data. It would be nice to be able to flip through axial, coronal, and sagittal stacks of the Visible Human images and see the image plane superimposed on the 3D view for localization and have anatomical structures labeled on the cross sectional images as well. It would also be nice to be able to select which cross sectional imaging modality is displayed (Visible Human has MRI, CT, and photographed cryosections).
Works great on latest chrome beta 64bit linux, GPU Accelerated Compositing, nvidia card, core 2 duo.
Uses negligible amount of CPU and seems as fluid as a local OpenGL application. The other webgl examples from http://googlecode.blogspot.com/2010/12/webgl-now-in-beta-her... work pretty well too, although the "previous page" button doesn't seem to work properly, so I have to close the tab.
Even though Google has a consistent color theme in all of its interfaces, I strongly believe that black background here would make it much more usable. Pretty awesome though!
64 comments
[ 6.4 ms ] story [ 90.6 ms ] threadVendor ID: 10de
Device ID: 06fd
Adapter RAM: 256
Adapter Drivers: nvd3dumx,nvwgf2umx,nvwgf2umx nvd3dum,nvwgf2um,nvwgf2um
Driver Version: 8.15.11.9038
Driver Date: 7-14-2009
Direct2D Enabled: true
DirectWrite Enabled: true
GPU Accelerated Windows: 1/1 Direct3D 10
Running on windows 7, Xeon quad core 3.20 GHZ, 6 Gig ram
http://news.ycombinator.com/item?id=2009346 - 4 comments
http://news.ycombinator.com/item?id=2010046 - no comments
Also, from Google's blog:
http://news.ycombinator.com/item?id=2009300 - no comments
http://code.google.com/p/threedlibrary/
Here are other examples from Google using ThreeD which are not minified:
http://code.google.com/p/webglsamples/
I got carried away dragging and zooming around but missed one thing. The search box is really powerful. You can locate each and every organ using it. I can imagine this being incredibly useful for medical students etc.
Good job Google.
The typical high-Cpu usage dropped like a rock.
Also I don't have the greatest video card so the auto-search is really irritating.
What's the upside of fully exposing the skin? I don't think skin exploration is the goal of this project, and this way they aren't alienating part of their audience.
Also, what are those things on her nipples? You can see it clearly when the opacity of everything is set to 0 except for a slight opaqueness of the body.
Im getting: To use Body Browser, you'll need a Web browser with WebGL support.
what am I missing?
If you could file a bug at http://code.google.com/p/chromium/issues/list with the contents of your about:gpu page, we'll do our best to figure it out.
Features I'd like to see: 1. Arrows from the labels to the anatomical structure so you can tell which label refers to which structure without having to interactively click each label to highlight the structure. 2. I assume they've built their 3D model from the publicly available Visible Human Project data. It would be nice to be able to flip through axial, coronal, and sagittal stacks of the Visible Human images and see the image plane superimposed on the 3D view for localization and have anatomical structures labeled on the cross sectional images as well. It would also be nice to be able to select which cross sectional imaging modality is displayed (Visible Human has MRI, CT, and photographed cryosections).
Uses negligible amount of CPU and seems as fluid as a local OpenGL application. The other webgl examples from http://googlecode.blogspot.com/2010/12/webgl-now-in-beta-her... work pretty well too, although the "previous page" button doesn't seem to work properly, so I have to close the tab.