Speed up reload from cache?

[jriley]

Is there any way that I could speed up my loading of the panorama from cache? Here is what I mean: a visitor has looked at the pano and clicks on a hotspot that takes them to another URL. Then, they hit the "back" button in their browser. It takes 5+ seconds for my pano to load back in from cache (and I know it is coming from cache, not downloading again- I can see that from my menuMeter plugin on my Mac.)

Any tips on speeding that up? One method of loading images versus another? ( e.g. - faces versus strip?) My client is having a problem with it taking what they claim is up to 9 seconds to load back in (though I measure it to be more like 5 seconds.)

The total size of the faces is around 4MB.

Thanks,

John

[jaaaab]

4MB per face?

hum you should better thinking in optimizing your images!

;)

[jaaaab]

oups sorry, 4MB total..

i think it's still too much, you should be able to get top quality around 1,5 - 2MB

++

[jriley]

Well, you should try telling that to my client 8-( They will be displaying this full-screen at 1280 x 1024 on a kiosk and have been extremely demanding regarding image resolution. This is as small as I have been able to go without them being unhappy about resolution; of course they now are unhappy with loading time from cache.

[kevin]

jaaaab,

You mentioned:

you should be able to get top quality around 1,5 - 2MB

Can you share your average image dimensions and file sizes? I am still trying to determine what looks best.

I have been starting with an Equirectangular image of about 4000 px X 2000 px that I then convert to cubic with Pano2QTVR. The starting image of 4000 px X 2000 px I optimize in photoshop by down sizing with Web Presenter Pro (http://www.fredmiranda.com/WP_Pro_Plugin/) from a large image (output from PTgui). I then save for the web at about 60% compression to reduce size.

Am I at all close to what you do to optimize?

In other forums and with other applications I know many advise your panorama images should be 2 1/2 times the size of the viewer height.

[kevin]

If it helps, my work flow, in general is as follows (I am open to feedback!):

1 ) Capture images in RAW Canon Rebel with Sigma 8mm

2 ) Open the 4 RAW CR2 images in Adobe Camera RAW Select temperature and retype (so if it reads 4050 As shot, I retype 4050 so the temperature is now a Custom setting. Under Lens I add -40 for Fix Red/Cyan Fringe, Vignetting Amount 100 and Midpoint 12.

3 ) Select all images as press Synchronize so temp, Fix Red/Cyan, Vignetting, and Midpoint become the same for each image.

4 ) Save as Tiffs.

5 ) Open in PTgui and stitch.

6 ) Output in Max Size – Tiff (or sometimes I output at 4000x2000), Blended Panorama Only, Stitch using PTgui, Blend using Smartblend.

7 ) Open final image in Photoshop and change the dimensions by changing image size via File>Automate>WP Pro>Light Sharpening, usually to 4000 x 2000 once this is done "Save for Web" File>Save for Web>2 Up>Quality 50 to 80 (Looking to compare).

8 ) Open Equirectangular image in Pano2QTVR and convert to Cubic - jpeg.

9 ) Edit the nadir and zenith in photoshop.