View Full Version : Capping newbie, but i'm doing ok, i think...

19th June 2003, 20:02
Hi folks, I've just started capping, here's my system:

Athlon 1700xp (stock speed 1466mhz) @ 2127mhz (11.5*185)
256mb corsair pc3200
Asus a7n8x-deluxe
Gainward 650xp Golden Sample ti4200

I'm capturing from digital cable, using the vivo on the geforce.

I'm capturing at 720 * 576 divx5 @ ~8mbit/s, uncompressed pcm stereo 16bit 44k1 wav.

This is hovering at about a meg a second, with cpu use at about 80%.

i'm getting excellent results, i was just wondering if anyone else was doing similar? a lot of people say to cap using a lossless codec, but the filesizes are crazy, and i really can't see much, if any, difference in quality doing it my way. Even if i were to transcode to dvd standard mpeg-2, it's still looking excellent.

If i were to capture a fairly decent vhs source, would i be best going for svcd or dvd-standard 1.8mbit/sec mpeg-1?

if anyone wants to see samples, gimme a shout, i'm putting samples up for a few friends to see.

20th June 2003, 08:43
VHS is about 1/4 the resolution of DVD. There are many options. Saving VHS source as DVD-resolution is using a lot of space for a small amount of information. MPEG-1? Why? Look at 1/2 D1, commonly called CVD (China Video Disc) but with 48 for audio so you can move the files to DVDR eventually without a hitch.

20th June 2003, 12:47
you questioned using mpeg-1 for storing vhs caps. Didn't tell me why though. I'm new to this, and i don't see why it ain't a good idea.

You told me to use CVD, so i could transfer to dvd. how compatible is CVD? Surely if i were to capture, then transcode to dvd-standard mpeg-1 in the first place, that would suffice, especially for vhs, with the extra bitrate afforded by being dvd-standard mpeg-1, and it'd go straight onto dvd.

20th June 2003, 18:09
MPEG-2 has better data compression than MPEG-1 so, given the same amount of storage space, MPEG-2 should look better than MPEG-2. Should because there are lousy MPEG-2 encoders and good ones. Also, MPEG-1 does not support interlaced video. If your source has lots of motion, leaving it interlaced can be a very good idea.

One of the formats that is part of the DVD standard is 1/2 D1, meaning, 1/2 the full resolution of 720 pixels. (It's actually 704 because of allowance for overscanning. Digital camcorders and some systems DO record in those "extra" areas, however.)

A VHS or S-VHS tape has a limit on the resolution of the information encoded on it. S-VHS works out to about the same as 1/2 D1, also called CVD. Obviously, the tapes are analog but they do have time-based information so there is an analogy. All the digital formats have their roots in the analog formats.

SVCD is 2/3 D1 width, 480 pixels. However, given VHS or S-VHS source, it really means you'd be storing the video in a higher resolution than is necessary. It also means, given a set amount of storage space, a lower ratio of data per pixel compared to CVD. Frequently, CVD from VHS or S-VHS will look BETTER than SVCD because of this.

CVD was the first format and SVCD, as we know it today, was the result of a lot of money changing hands from competing companies and China. China was looking for a method of DVD-like video without paying the DVD licensing fees. S3 had been selling consumer video-compression over there and got trumped by their larger competitors.

As it turns out, SVCD resolution is pretty good and is used for digital satellite. SVCD is NOT, however, part of the DVD specification.

XCVD is what I use for VHS and S-VHS encodes. The "X" means it's outside the CVD spec but is very much akin. CVD calls for audio to have a rate of 44.1, like a CD. DVD calls for 48. The difference in space is very minimal. However, any time you convert numbers from one scale to another there will be some distortion.

I encode analog videotape captures to CVD visual resolution but make the audio at 48 and have a target size of 770M per hour or hour of broadcast with the commercials removed. This way, the file can be copied from the CVD disc and directly used to make a DVDR without having to re-process the audio and 6 will fit on a DVDR with room left for menus. The vast majority of the time, this is ok. Sometimes, if the source is real noisy and/or has lots of motion, I will target 30 minutes into 770M.

There are ways to change the header of SVCD files so they will play in most DVD players. If you have SVCD and want to move it to DVDR, this is the best option. "Best" is a dirty word in these forums. However, you can't do any better than NOT changing the data, can you?

I use XSVCD for things I've captured from satellite. SVCD resolution with 48 audio.

20th June 2003, 18:59
thanks, i'll look into cvd.