starrynite
08-27-2007, 01:41 PM
Applied Lag Reduction:
I have discussed at length with a number of members this procedure, how it works, why it works, with the end result being a high-performance Club / Community area where avatars do not get stuck. Though there is some disagreement on the use of the term "pre-rendered", we all agree the end result is a lower resolution image which may not be what you would want. My focus is on performance.
The table I use is as follows and is relative to the size of the surface area the image is applied to:
Applied Surface Area - File Size
16 x 16, 256 sq ft, -> 200k max
8 x 8 , 64 sq ft, -> 50k max
4 x 4 , 16 sq ft, -> 12.5k max
2 x 2 , 4 sq ft, -> 4k max
1 x 1 , 2 sq ft, -> 1k max
I set my screen resolution to 1280x1024 and use FaxViewer to render image on screen.
I use the zoom function to do do the scaling
I use MWSnap, take picture of rendered image resulting in BMP of image
I save-as jpg and trim snap-shot image.
End result is 10:1 smaller file, same image relative to applied surface.
We would all agree the these converted images would not be suitable for printing or a photograph, but are very acceptable as an image re-rendered in 3d. Text is an issue as evident in the NASA Space Museum, and I have discussed this issue at length with most of you who agree that no matter what resolution you make your image, the end text resolved within Kaneva in 3d remains a problem due to anti-aliasing.
This is a calculated compromise to achieve performance within the current platform parameters that does in fact work. The consequences are lower quality on the 16x16 and 8x8 only and not noticeable on 4x4 and smaller.
This should help most significantly reduce avatar lag and immobility including those constructing large scale projects such as "Green Acres". For existing large scale projects, reprocess all images relative to the applied surface to begin affecting the avatar performance.
These changes will directly increase the performance of your project. Starrynite
I have discussed at length with a number of members this procedure, how it works, why it works, with the end result being a high-performance Club / Community area where avatars do not get stuck. Though there is some disagreement on the use of the term "pre-rendered", we all agree the end result is a lower resolution image which may not be what you would want. My focus is on performance.
The table I use is as follows and is relative to the size of the surface area the image is applied to:
Applied Surface Area - File Size
16 x 16, 256 sq ft, -> 200k max
8 x 8 , 64 sq ft, -> 50k max
4 x 4 , 16 sq ft, -> 12.5k max
2 x 2 , 4 sq ft, -> 4k max
1 x 1 , 2 sq ft, -> 1k max
I set my screen resolution to 1280x1024 and use FaxViewer to render image on screen.
I use the zoom function to do do the scaling
I use MWSnap, take picture of rendered image resulting in BMP of image
I save-as jpg and trim snap-shot image.
End result is 10:1 smaller file, same image relative to applied surface.
We would all agree the these converted images would not be suitable for printing or a photograph, but are very acceptable as an image re-rendered in 3d. Text is an issue as evident in the NASA Space Museum, and I have discussed this issue at length with most of you who agree that no matter what resolution you make your image, the end text resolved within Kaneva in 3d remains a problem due to anti-aliasing.
This is a calculated compromise to achieve performance within the current platform parameters that does in fact work. The consequences are lower quality on the 16x16 and 8x8 only and not noticeable on 4x4 and smaller.
This should help most significantly reduce avatar lag and immobility including those constructing large scale projects such as "Green Acres". For existing large scale projects, reprocess all images relative to the applied surface to begin affecting the avatar performance.
These changes will directly increase the performance of your project. Starrynite