Resolutions of embedded video on the internet

If you want the best quality, set your video card to ONLY output to the tv. Outputting to 2 different displays at different resolutions is hard enough for a desktop video card, let alone a laptop one.


Should be.


Depends on the video player you are using.


If using quicktime or vlc, the player does the scaling. If using stock Windows Media player, Windows itself is. However if you are using a codec pack like Sharky's, FFDShow will be doing the scaling, which I recommend because it does a really good job.


If the video is full screen, then its sending the video at whatever resolution the video card is set to output.

Suggest you try measuring the various resolutions using an optical comparator with a reticle, along with resolution-burst patterns. Suggested this technique in the Blu-ray Software forum , ideally using numerous resolution/frequency burst patterns stored on a test disc. (A link I provided in that thread shows patterns on the S-M test disc and the thread outlines how to convert frequencies into resolutions.)

With a PC already linked to a display, resolution bursts, such as those on AVS member dr1394's site , or elsewhere, could be downloaded and used instead of an optical disc. Ran the terms (optical comparators reticle) on Google shopping (low-to-high pricing) and found loupes for ~$6 and up. An Edmund Optics engineer recommended a costlier 9X model that accepts their 27mm reticles . The scalings of computer settings and display resolution could be worked out once the hardware/software is at hand. No doubt skill at measuring the effective resolutions of the finest details seen on screen would increase quickly. -- John

^^^Yeah, the S-M disc creators outline potential difficulties, too, (at my link). But, if you see differences between 480p and the best-resolution 1080i/p sources it should be possible to measure it (max effective resolution). The oft-repeated garbage-in/garbage-out. -- John

EDIT:
The link above to a 9X comparator and fine-ruled reticle shows what you'd see with the lens against a direct-view screen (sans image). You Tube etc. videos are indeed often fuzzy. Scaled to 1080p by your display, the finest visible details remain fuzzy; they're just larger on a screen meant for consumer motion video. While a detail, say a human hair, is fuzzy, placing the reticle over it gives you a reference measurement: how many reticle-line divisions wide.

That reference measurement could be compared with test patterns until a similar or identical width is found. (Referring here to fine vertically oriented details--a hair--to measure maximum effective horizontal.) The horizontal rule in the reticle (linked above) would be turned 90 degrees to measure vertical resolution detail, then compared with vertical-resolution bursts. Here are the vertical and horizontal bursts on the S-M test Blu-ray. My link earlier to the S-M test disc thread discussion mentions how to convert the frequencies given to lines of resolution.

For a 1080p display, of course, the 37.09-MHz vertical-lines test pattern (2nd row down, far right) should have 1920 B&W lines across your screen width--perhaps confirmed by counting lines per inch and calculating per screen width. But as the S-M test disc developers point out, that's not the purpose of their resolution-burst patterns. Some years back one AVSer mentioned he used a pro computer graphics program to create his own lined patterns to measure resolution. Also, an upscaled 480p image measurement on a 1080p display could be compared with similar measurements but wouldn't represent the original 480p resolution. Ideally, this technique would be used to compare mostly 1080 image details, factoring in upscaling if other capture/production resolutions are used.

But a measurement such as this could provide the equivalent (test pattern) maximum horizontal resolution, just for comparison purposes, of what might appear to be the finest detail on a You Tube image (for similar upscaled comparisons). A Blu-ray movie, even with all the resolution limitations mentioned in the S-M disc discussion, could be similarly measured for what appears to be some of the finest details--ditto for live or recorded 1080i broadcasts--although live programs would need frame grabbing, DVRing, etc. for measurements. -- John

Yes. Especially if you are running two resolutions, then the video card has to work twice as hard.

Flash Player, which outputs the video at whatever resolution the video card is set to.

Yes, to let the tv do the scaling, you would have to set the computer to output 720x480, but then you would be changing the resolution every time you want to watch a different video. But honestly, youtube has some of the worst quality of all the streaming sites. Letting your tv scale, or letting the computer scale, isnt going to make any difference. Its still going to look like garbage.