Blurry video...

[Delora]

I've taken some video at 1920 × 1080 and I'd like to make the resolution much lower. E.g. about 700 by something. (no calc on me right now). I note that the pixel ratio is 1.0 and it's all "HD" and widescreen and all that. I've got Adobe Premiere CS5 and a nice slinky PC for it, but when I try dropping the video into a smaller size it looks rather blurry. Obviously I'd expect less resolution but I'm not sure why the blurriness or if there's something I can do - even on still frames. I've got the same frame rate and the same aspect ratio. Any other ideas? Is it a good idea to drop resolutions by even numbers instead of arbitrary numbers? e.g. drop to 960 from 1920 rather than dropping to 900? I could just do with some general advice really at this stage. I'm trying to make files for a Flash based subscriber website idea I'm working on so outputting my lovely video clips to flv or f4v seems most desirable to me. Thanks D x

[Delora]

Oh and if it helps, the videos contain a lot of close-up work (- of the handwriting variety - no dirty minds!).

[jagabo]

Is your source interlaced or progressive? In any case, small text will definitely get blurry when reducing the frame size from HD to SD resolutions. You should also stick with frame sizes that are a multiple of 16 or 8. For example, 720x400.

A crop of small text from a 1920x1080 video file:


After reducing to 720x400 it's no longer readable:


Even if you enlarge it back to the same size:

[Delora]

Thanks for your reply Jagabo. My original video is interlaced (an MTS file).
Video
ID : 4113 (0x1011)
Menu ID : 1 (0x1)
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.0
Format settings, CABAC : Yes
Format settings, ReFrames : 2 frames
Codec ID : 27
Duration : 5s 160ms
Bit rate mode : Variable
Bit rate : 22.2 Mbps
Maximum bit rate : 22.7 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 25.000 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Top Field First
Bits/(Pixel*Frame) : 0.427
Stream size : 13.6 MiB (92%)

And my output video is currently:
Format : MPEG-4
Format profile : Adobe Flash
Codec ID : f4v
File size : 57.9 MiB
Duration : 5mn 14s
Overall bit rate : 1 542 Kbps


Video
ID : 1
Format : AVC
Format/Info : Advanced Video Codec
Format profile : Main@L4.0
Format settings, CABAC : Yes
Format settings, ReFrames : 3 frames
Codec ID : avc1
Codec ID/Info : Advanced Video Coding
Duration : 5mn 14s
Bit rate mode : Variable
Bit rate : 1 406 Kbps
Width : 720 pixels
Height : 400 pixels
Display aspect ratio : 16:9
Frame rate mode : Constant
Frame rate : 25.000 fps
Standard : NTSC
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Bits/(Pixel*Frame) : 0.195
Stream size : 52.8 MiB (91%)
Title : MainConcept
Language : English


I'm not sure why it says NTSC actually - this one was output from Premiere Elements using one of the (adjusted) PAL formats. And I do get that it'll obviously look more blurry when you have fewer pixels to represent the same thing, but I just get waaay more bluriness than I expect. (i.e. if i lower the resolution of a JPEG of a frame from one resolution to the other I still get a pretty sharp second image, albeit with less detail because it's smaller. I just can't get my head around the blurry nature of it. I know that years ago when I grabbed video from a Video 8 camcorder using an Amiga 500 with a basic capture card I was still getting clear images, so I just wonder if there's something I'm not aware of when it comes to resizing video or changing the output format... D . . . .

[poisondeathray]

For some reason mediainfo says your 720x400 f4v export is interlaced ?

Maybe you're blend deinterlacing ? Your original video is interlaced and you are exporting for web. If you disable clip blend it will use the other deinterlacing mode (leaves jaggies and aliasing artifacts). Neither is ideal with Adobe software. You can use higher quality deinterlacers through avisynth, but there is a bit of a learning curve.

If you have an option to shoot native progressive on your camera , use it. Each field (not frame) essentially has 1/2 the information in interlaced mode. When you downscale it text becomes even less legible.