Recommended/Best software capture settings?

[Vchat20]

Ok. Let me put the back story on here for you. I've got this really old aimslab vhx capture card here which does not do hardware capturing so any actual capturing being done is all loaded down on the processor. Even worse, for how long I have had this card I have not used it for any capturing at all. Just some occasional tv viewing at best.

So recently I have come across the idea to, for one of my favorite tv shows, take and use the card to capture off the episode of this show off of my DVR (cableco provided, bit for bit stream save to dvr's hard drive.) through the s-video port and audio of course through the composite connection.

First test run with this was just that: a rough test run. Used VirtualVCR to test. Was able to capture to 48khz pcm audio and 720x480 DV compressed video with my 2ghz celeron cpu idling around 50% or so and no dropped frames. I cant exactly remember what the 'video capture' and 'video preview' ''pins'' were set at.

But my question is anyone who has had to deal with full software capturing such as this has any recommendations or knows the best settings I can use for the best quality possible despite the connection that the video's being transffered over? (s-video was my best shot. would have done direct firewire, but the channel the show is on specifically has crippling 5c copy protection on it).

And FYI, this is specifically for private usage here. I prefer to have digital copies of most of this stuff I can load on my laptop and watch anywhere I please instead of being tethered to a tv so to speak.

[G)-(OST]

Haven't you heard!!! People round these parts don't like the use of the word "BEST" in the subject....LOL. https://forum.videohelp.com/viewtopic.php?p=1441969#1441969.

[Vchat20]

Haven't you heard!!! People round these parts don't like the use of the word "BEST" in the subject....LOL. https://forum.videohelp.com/viewtopic.php?p=1441969#1441969.

Hence the reason why I also tacked on the word 'recommended'. And on the other side of the fence, I do believe I provided ample details in my post to narrow down what im specifically looking for to those who read it. If not, let me clarify here:

Got about 60GB free drive space to use at one time meaning no more than that for half an hour of video. Am ok to start with uncompressed video and have been looking into the three primary uncompressed codecs (huffyuv, lagarith, and arithyuv) as long as I can find one that falls in the drive space requirement above as well as drive speed (will be storing to boot drive, 7200rpm, udma5). But my primary concern is the ambiguity of the different raw video formats like YUY, YUV, RGB, etc. listed on the capture program under both 'video capture pin' and 'video preview pin' and if I can get a better explanation on these, that'd be great.

[jagabo]

HuffYUV is fast but doesn't compress by a lot. Lagarith compresses better but is too slow for capturing. I don't know arithyuv but from what I've read it is somewhere between HuffYUV and Lagarith. MJPEG is lossy but looks pretty good at higher quality settings.

Video is broadcast in the YUV colorspace. Y is the grayscale component of the picture (used by black and white TVs), U and V are the color components. U and V have about half the resolution of the Y component. Capture cards capture in the YUV colorspace with half resolution for U and V. There are many different ways to pack this 4:2:2 (ratio of 4 parts Y, 2 parts U, and 2 parts V) data, differing in the order of the Y, U and V components.

http://www.fourcc.org/yuv.php

Use whatever works with your hardware and software.

[Vchat20]

Well, I wanted to bring this topic back up because im running into an eensy problem relating to this little adventure of mine.

Needless to say that so far I have picked up on this stuff real darn quick and have nailed it all down to where I am getting near dvd quality off the dvr (which has a pristine mpeg2 stream stored on the hard drive which is a bit-for-bit capture from the cableco's satellite feed and not just a basic re-encode of an analog catv channel) through s-video. That is I am capturing using virtual vcr. Full NTSC video, 720x480, 30fps, using the Arithyuv lossless codec. Audio is just standard 48khz, 16 bit PCM audio coming across the soundcard line in. with this setup, cpu usage hovers around 65-70% and any dropped frames if any are due to me forgetting to defrag the drive beforehand (easy to forget for me).

But heres the problem im running into: I wanna crop out the commercials. Just through my own figuring and confirmed by some threads here, I just use virtualdub in direct stream copy mode and clip out each section of the show between commercials and later rejoin via 'append segment'. all fine and dandy except it seems when I try to append the 2nd segment to the first and initial segment of the video, virtualdub gives me a 'video streams have different data formats' or something similar to that (sorry, been fiddling with this and forgot.). So I am currently stuck here with 3 clips I need to join and cant. And id rather not try and re-encode these cuz as it is with the lossless format, video rendering rate for direct stream copy tops out at like 35fps because of drive speed.

If anyone knows what im doing wrong here, I'd appreciate it greatly.

[jagabo]

What you're doing should work. The only thing I imagine might have gone wrong is that you saved one segment with a different codec (audio or video). In any case here's a much simpler solution:

Cut out the commercials rather than saving each program segment and appending them.

Mark-in at the start of a commercial segment, mark-out at the end, press Del. Move on to the next commercial segment, cut it out, etc. Finally Save as AVI.

[Vchat20]

Ah! Thanks. Didn't know I had that option. MUCH simpler that route. Major thanks for pointing that out.

And in any case, i figured out the original problem. Seems the first segment differed because I had to re-record the first section of the show due to getting the little timeshift bar on the dvr caught amidst the show intro during capture. so the first 1/3 of the show was technically a separate capture from the 2nd and 3rd sections. I re-recorded the whole show again from the beginning and kept it all in a single recording and it came out just fine.