Starship Asterisk*

I notice a lot of disparity in image sizes posted to APOD. Mostly this in't a problem, but posting the fuzzy (at full size) NGC7252 atoms for peace colliding galaxies at over 10 megabytes is going to slow down a great many of your users, versus a few moments it would have taken you to resize it more reasonably at source.

wmap wrote:I notice a lot of disparity in image sizes posted to APOD. Mostly this in't a problem, but posting the fuzzy (at full size) NGC7252 atoms for peace colliding galaxies at over 10 megabytes is going to slow down a great many of your users, versus a few moments it would have taken you to resize it more reasonably at source.

I wouldn't want it resized. The main page image is usually resized for quick loading, but this image links to the original, which is how it should be. In this case, the full sized image is the actual resolution as captured by the telescope, and contains a huge amount of fine detail. It is only "fuzzy" because at this scale you can see the resolution limits of the image, as opposed to your display being the limiting factor. Making the image smaller would throw away real information.

Sure, it's going to be a slow load if you don't have a good Internet connection, but that's just the nature of high resolution images. The wait is worth it!

I would agree with you, Chris, if APOD were a repository of full-size data images for research purposes like one of the digital sky survey databases. Rather, I believe APOD is an esthetic / educational media designed to reach the most people around the world, where ease of access is key. Following your logic, APOD would host the original full size images from the great observatories for all its posts - some of which run into hundreds of megabytes or more. I agree it's sometimes nice to post a large image, where there is an exquisit amount of fine detail for the lay subscriber to see, but the image of NGC7252 doesn't meet that criteria and would have been better delivered and viewed as a smaller-sized file. There's a maxim in writing manuals, where the author does the extra bit of work up front to save the large number of readers the time searching out acronyms, footnotes and endnotes. Same principle here.

wmap wrote:I notice a lot of disparity in image sizes posted to APOD. Mostly this in't a problem, but posting the fuzzy (at full size) NGC7252 atoms for peace colliding galaxies at over 10 megabytes is going to slow down a great many of your users, versus a few moments it would have taken you to resize it more reasonably at source.

wmap wrote:I believe APOD is an esthetic / educational media designed to reach the most people around the world, where ease of access is key.

The Atoms for Peace image displayed on the APOD page is not over 10M; it's 98946 bytes. That's a perfectly acceptable size even for dialup.

Yes, it links to a larger image; nothing wrong with that. Those who don't want to see a larger image won't click through; those who do, will.

wmap wrote:I would agree with you, Chris, if APOD were a repository of full-size data images for research purposes like one of the digital sky survey databases. Rather, I believe APOD is an esthetic / educational media designed to reach the most people around the world, where ease of access is key. Following your logic, APOD would host the original full size images from the great observatories for all its posts - some of which run into hundreds of megabytes or more. I agree it's sometimes nice to post a large image, where there is an exquisit amount of fine detail for the lay subscriber to see, but the image of NGC7252 doesn't meet that criteria and would have been better delivered and viewed as a smaller-sized file. There's a maxim in writing manuals, where the author does the extra bit of work up front to save the large number of readers the time searching out acronyms, footnotes and endnotes. Same principle here.

It's not purely about aesthetics. The images are also intended to stimulate an interest in the underlying science, and that's why the original images are linked.

It is very rare for the original images to be hundreds of megabytes. At the limiting resolution of the optical system, and the size of the object being imaged, very typical image sizes are a few thousand pixels on a side (as in the image under discussion), which means that good quality JPEGs at full resolution will usually not exceed a few tens of megabytes. These days, that is an acceptable optional download for most people.

I disagree with you completely regarding the NGC7252 image. At its full resolution, it is showing detail at the pixel scale. That means that any reduction in size of this image will be showing less real information. This is not an image I would want scaled down at all.