Output file size

[jetpat]

Hi

I am creating a Pano2VR tour, but just need a bit of advice about wether the file size of my project is right for when it goes online.
I have imported into my Pano2VR project 6 pano images of around 70mb each, this gives me a HTML5 Output file of 365mb. The final Project will have 60 Pano images, which i estimate will give a final output file of 3.7gb.

My image output setting is:
Quality: 90
Multi Res
Level Tile size: 510px
Levels: Auto

Does this look fine in terms of file size, or should i think about reducing its size, if so, how.

Thanks.

[360Texas]

While the project file size might be large.... HTML5 tiles are small and view port images are served quickly.

[Isaac Brown]

For 6 panorama images 360mbs seems pretty excessive though, can I ask what the panorama input image dimensions are in px?

[jetpat]

In the example of the test file i am doing above using 6 pano images, i have 2 Aerial Pano images at 25000 x 12500 px and 4 Ground Pano images at 20000 x 10000 px. These are the original image sizes i am bringing into my Pano2VR project before exporting.

Thanks.

[Isaac Brown]

Ok I see. I normally use 8192x4096px 3d rendered panoramas, so im guessing pano2vr doesn't generate as many multi-res tiles for my tours, hence the smaller output file sizes I am getting. Could you lower the input image resolution at all so you generate less tiles? (I think this is how the tile generation works)

[soulbrother]

My image output setting is:
Quality: 90
Multi Res
Level Tile size: 510px
Levels: Auto

All good, but in most cases you can reduce the quality to 70 and the result is still good enough.
I never use a better setting than 75 to 80 for panos and pictures, shown on the web.
In some cases I reduced to 60!
Additional you can downscale the pixelcount of your panos from 25k and 20k to 12, 14 or 16k (or whatever you like)
Dont forget to resharpen somewhat, after that downscaling.

[jetpat]

Thanks all for your suggestions, will try them and see.

[RockHunter]

In many programs you have the workflow "Maximum input" -> "set desired output". It's for example how audio, video and image workflow operate.

In Pano2VR your input size = output size. So if you put 25000px panos in, you get 25000px panos in the output, broken into (horizontally 4 cubes) = 6250px per cube face. Each cube face is then broken into tiles. So if you set you tile size to 625px, you will get 10x10 tiles per cube, with a total of 6 cubes.

Then for multires outputs, which is the default, you will get a complete copy of each pano at 1/2 the res (1/4 actually, as it's squared), so it'll be 3125px per cube face, with each tile being 625px and so forth, with 5x5 tiles per cube.

If you want to keep the high resolution, switch to manual multires and let it produce only two or three resolutions. You'll need to do the math how you want to break up your project. If the tile size and the size of each cube face doesn't divide up evenly, you'll end up with a few tiles ad odd sizes, which doesn't matter that much.

The only other way is to make a copy of your source panorama and reduce the resolution there to 12000px, or 8192px (that's the maximum for most headsets) or whatever you like. You don't have to remake your tour, just go to the Properties panel for each node, and pint it to a new input file in the "Image Input" section. Hotspots etc, will not move.

Hope that helps