SD Widescreen Resolution

[Mike Kujbida]

How long was the program?
What bitrate did you render at?

[John Rofrano]

Yes, it is probably softer. Think about it. You’re using the same sensor to record more resolution. How could that possibly work? Since the sensor doesn’t magically grow new pixels, the same number of pixels are being used to represent more data so the information still must be packed into the same number of pixes and therefore the data gets “fuzzy” and so does the picture.

What probably happens it that it records anamorphically thinner horizontal pixels which means that less information is stored about each one to make them all fit. In NTSC this means that 873 pixels are being recorded with only 720 physical pixels to do the recording. That is not going to come out as sharp as if you physically had 873×480 pixels on the sensor. It’s the same argument with recording 1440×1080 HD vs 1920×1080 HD. Having a big enough sensor to record all of the data affects the quality of the recording.

~jr

http://www.johnrofrano.com
http://www.vasst.com

[Rich Kutnick]

Mike: Using Mark’s DVD Bitrate Calculator for a 2:06 (Hr:Min) video, I did 2-pass VBR and set the min to 2608, avg to 4352 and max to 7624. GSpot states that the Sys Bitrate is 10080 kb/s VBR and and Intervideo WinDVD software reports that the video bitrate is 7.728 Mbps.

John: This makes perfect sense and corroborates what I was thinking all along. None of my colleagues who use this method ever warned me about this resolution/detail hit (it appears that they either never noticed it or it did not bother them), as I was offering SD widescreen for the first time as an option to a Client thinking that if they did not want HD then at least I could provide them with a widescreen SD DVD that would fill their screen instead of them having to view the video with pillarboxing. The best idea, then, is that if a video is to be produced in widescreen SD that it be shot and edited in native 16X9 HD (1080i) and rendered to an SD-compliant MPEG-2 file ready for authoring–correct?

Rich Kutnick
VIDEO IMPRESSIONS

[Nigel O’neill]

Rich

I used to record live events in SD and HDV, delivering as a SD DVD. I found the SD footage recorded at 8 mbit/s to be a bit softer and muddier after I colour corrected, added FX and transitions and so forth. The HDV footage which was recorded at 24 mbit/s, which has a lot more data to work with from the start, was not impacted by the same colour grading process.

However, if you are going to record and store ANY event over 70 minutes on DVD, you WILL suffer image loss as you try to cram more and more video data on the DVD. I don’t generally store more than 90 minutes on a DVD as video quality tends to suffer above that.

My system specs: Intel i7 970, 12GB RAM, ASUS P6T, Vegas Pro 10e (x32/x64), Windows 7 x64 Ultimate, Vegas Production Assistant 1.0, VASST Ultimate S Pro 4.1, Neat Video Pro 2.6

[Dave Haynie]

[John Rofrano] “Yes, it is probably softer. Think about it. You’re using the same sensor to record more resolution. How could that possibly work? Since the sensor doesn’t magically grow new pixels, the same number of pixels are being used to represent more data so the information still must be packed into the same number of pixes and therefore the data gets “fuzzy” and so does the picture. “

Well.. 16:9 mode isn’t necessarily “more” resolution. It’s actually just a different interpretation of the same resolution: it’s 720×480, no matter what. If you show this on a 4:3 display with letterboxing, the horizontal resolution is identical, the vertical resolution will be cut down (except for a very few analog sets that could adjust and deliver full resolution in 16:9 letterbox mode — actually fairly common in PAL countries, but not here). If you display on a 16:9 screen, you’re covering more horizontal area, but less vertical area, than a TV of the same diagonal size in 4:3.

But actually, the real problem here is that the PD-150 had a crappy 16:9 mode. It doesn’t actually deliver full resolution in 16:9 mode, it crops the 4:3 image to 16:9, then scales up the vertical resolution to 480 pixels to go on tape.

This wasn’t uncommon in the early days of digital video. My first DV camcorder did exactly the same thing: it would either crop and zoom or letterbox, my choice, but it didn’t have a real 16:9 mode. I had a slightly less old Digital8 Prosumer class camcorder that addressed this: it expanded the horizontal by using the spare imager pixels normally used for digital image stabilization when in 16:9 mode. Once we all (well, most of us) moved to HD, we got real 16:9 resolution. And actually, the sensors are usually still 4:3, it’s just that they’re designed to be cropped to 16:9 in normal operation (and of course, some of the earlier HD camcorders used lower resolution chips with pixel offset to approach to full HD without killing light sensitivity as much as other camcorders did, but that’s another tale).

-Dave

[Nigel O’neill]

I think I had that crappy Panasonic camcorder too that cropped 4:3 to get “16:9” 🙂

My system specs: Intel i7 970, 12GB RAM, ASUS P6T, Vegas Pro 10e (x32/x64), Windows 7 x64 Ultimate, Vegas Production Assistant 1.0, VASST Ultimate S Pro 4.1, Neat Video Pro 2.6