Authoring at 1/2 D1

[stevendm]

I capture television at 352x480 with my ATI 8500DV using the PicVideo MJPEG at Q19 or Q20. I have been authoring them to full DVD resolution and have been very successful. With proper filtering I can get the same quality or better that I see straight from the TV.

I have heard about 1/2 D1 resolution and I have some questions. WIll I lose any quality by authoring to 1/2 D1? Since this is the same resolution that I am capturing at I would think that I would not lose quality. The main benefit I see is that I can store more on a DVD-R.

What do you think about using 1/2 D1 instead of full DVD resolution?

Is it worth it? Will all DVD players play these DVD-Rs?

Thank you

Steven

[Vejita-sama]

First, the legal DVD resolutions are 720x480, 352x480 and 352x240. ALL DVD players should play video at these resolutions.

Second, if you capture at 352x480 then resize to 720x480 some quaility is lost (as you increased the resolution from your source). It might not be visiable but it does happen. So yes, just encoding at 352x480 might be better.

Third, resolution has no effect on file size. That is you don't fit any more or less video on a DVD by using a lower resolution. File size is dependant on the source runtime and the bitrate. If the bitrate is 4000kbit/s then that's how big each second of video will be, period. However, video quaility depends on resolution and the ratio of resolution to bitrate.

That is a 80x60 jpeg that's 30k vs. an 1600x1400 jpeg that's also 30k. Both are the same size but quaility will be different.

[ImaWeTodd]

Hey Vejita-sama,

Is there an effect if steve is resizing interlaced source?

When I made svcds, I had trouble resizing my dv source and keeping the interlacing looking good. I cap at 1/2 DV and encode for DVD at the same. For old cartoons, this works well because I can keep the bitrate lower and still get a good encode. I might just run a test to see what a resized encode looks like.

very interesting. I assumed you could not resize well and that you could not get better quality than you started with, so what's the point.

[Vejita-sama]

It's true that you can't get better quaility then you started with. The geneal rule of thumb is to capture at the max resolution and bitrate that your system supports then when you encode resize/add filters/patch etc. as necessary.

When it comes to interlaced source you have to choices. Either capture at ???x480 to capture both fields, or capture at ???x240 and you'll only capture one field getting an instant 'de-interlace' (both of course with a lost in quaility as well).

For a true interlaced source you really don't want to de-interlace it. You should capture both fields at ???x480 and encode to MPEG2 (which supports interlaced source). You really only want to deinterlace if it's a telecided source (or if making DivX files as PC don't support interlaced video).

As for what gives the best results... that's a matter of personal taste. I normally capture VHS/LD source at either 720x480 then resize to 352x480, or just capture in MPEG2 at 352x480 and use that. Most of my VHS tapes are in only in 'ok' quaility and just capturing direct to MPEG2 is good enough 'for me', YMMV...

[stevendm]

You said that file size is only dependent upon runtime and bitrate. That makes sense.

My thinking was that if I kept the video at 352x480 I could use a lower bitrate to get the same quality. I was thinking that this was the same as encoding to CVD instead of full SVCD. I found that my CVDs looks better than my SVCDs with relative noisy (tv and vhs) sources.

Am I all wet here? If so, why is there a 1/2 D1 resolution in the first place?

Thanks

stevendm

[furball6969]

I really can't fathom why you'd capture at 352 x 480 and then re-encode to a higher resolution. You gain nothing by doing this. but you do lose space on the disc because you need to use a higher bitrate for the higher res to keep the quality.

Something else I don't get is why you're re-encoding to begin with. The Radeon 8500 DVD capture presets are DVD compliant.

I bought the Radeon because of it's abilities not because I want to spend a bunch of time re-encoding all of my caps before I burn them. But then again I do cap a lot of video; maybe you don't cap as much and don't mind the extra time... To each his own.

[stevendm]

I capture at 352x480 to save space over 640x480 or 720x480. I don't think the higher resolution captures gives significant improvements over the 352x480.

I think the runtime encoding done by the ATI card sucks. I encode from avi to mpeg2 so that I can apply filters to do some temporal smoothing and a tiny amount of sharpening. I also usually capture b&w films so I force greyscale.

I don't really mind the reencoding time if I get a superior finished product. By reencoding, I get the good output.

stevendm

[reed1374]

stevendm:
How are you capturing an AVI at 352x480 with MMC? Assuming you are using MMC. I can only get 640x480, then -> VDub resize to 352x480. I would really like to go directly to 352x480 capture if you have any suggestions.

AIW 7500, MMC7.7

R

[stevendm]

I don't even have mmc installed.

I am using ShowShifter. It is very good and can automatically change the channel on my cable box before recording.

Steven