While I wouldn’t necessarily refer to myself as a ‘Pixel Peeper’, I do try to provide the highest quality videos and images to my audience when uploading to my social media outlets (Youtube, Flickr, etc).
Recently, I began uploading 26 performance videos that I’d taken on the Aventura Dance Cruise to my Youtube channel and had noticed that the quality was much lower than expected. I know that you have to wait a bit for Youtube to fully compress and process your video (even after it says it’s complete), in order to see the full HD version, but I noticed that the final output was vastly different from the quality of my videos viewed on my PC after being rendered through Sony Vegas. After some hours of tinkering and test uploads, I realized that my primary issue was that:
- The quality of my Youtube videos were inverse to the amount of plug-in filters (Track FX) I’d applied to the original video: So more corrections to levels, brightness, contrast, and overall color, resulted in worsening video quality on Youtube.
I found this to be completely infuriating as it takes quite a long time to color correct 26 video, only to have Youtube downgrade the videos to an unacceptable quality. Here’s a sample of a video that I’d adjusted, but uploaded in 1080p (1920×1080). Even when the bright spotlights are shining directly on the performers, you can still see how pixellated the video appears. This occurs even when rendered with a realitively high bitrate (which is one of the most important aspects to factor in when rendering a video). If you pause the video at any point, you’ll see how jagged the contour lines are on the performers, and can notice a very ‘boxy’ separation in areas of high contrast.
After making a few adjustments, I’d noticed that if I only applied a sharpening filter, the quality wouldn’t be as greatly affected, but unfortunatley, that still left my videos too dark and desaturated (I generally shoot in a flat color profile). I realized that the problem was with Youtube’s method of compression.
**Please note that the following numbers will be used as examples and are not necessarily exact figures**
If I upload a video at 100% (from my PC), Youtube may drop the quality down to 75% of the original quality after compression (this is why the video on your PC looks vastly superior to your upload). Well, what happens if my video upload is greater than 100%? This is where upscaling comes in. Generally you don’t wan’t to upscale an HD video all the way up to UHD/4K as there’s not enough information in the image to support such a degree of upscaling…but what if you upscaled the original video just enough above 100% where the 25% compression degradation drops your video to 80% instead of 75%? Though, it’s still not at 100%, I noticed that the slight increase in quality can help to significantly decrease the noticeable pixellation and jaggedness of my videos.
To test my theory, I upscaled my color corrected videos to 2048×1152 using Sony Vegas and this was the following result:
If you compare the two videos, you’ll notice a significant decrease in pixellation and improvements to sharpness in the overall video. Skin tones are also much smoother as the ‘boxiness’ is not as present. I don’t recommend you upscale all your videos as videos shot under optimal ‘bright’ lighting conditions generally don’t have the same problems since you wouldn’t have had to make as many adjustments to your video, but if you have any work that requires a significant amount of adjustments, try rendering your video in 2048×1152 instead of 1920×1080, and you may be pleasantly surprised in the increased output quality of your video on Youtube.
*If this has worked as well for you, or if you have any other tips on increasing the quality of your Youtube videos, please leave a comments below!