Laserdisc S-video or composite capture?

[digicube]

Here are the capture samples from DVL-V888.

[DB83]

To quote another Bananarama song "It's not what you do, it's the way that you do it. That's what gets results"


Logic dicatates that s-video should be better and gets my vote. Although I could not detect any appreciable difference between the two.


PS:


I came into close contact with the Rama's when they opened the local branch of HMV way back when. Having queued to get in, I bought their single (pre CD days this) which they duly signed with what we call biro pens, which have a hard point. The net result was that the record was scratched and un-playable. Idiots !!!

[dellsam34]

Unlike tape formats, Laserdisc is a composite format in nature and should be captured that way, Converting to Y/C serves no purpose except if the capture card's digital comb filter separation is crap and the one inside the laserdisc is better, but I don't think the Y/C separation board inside a laserdisc player outperforms a decent capture card form 2000's and 2010's. The ideal way to capture it is to use the laser-decode method from the doomsday project.

[Cornucopia]

From what I've seen, the domesday project is still not ready for prime time. Composite is the best bet for the forseeable future.


Scott

[jagabo]

Since what's stored on the disc is composite video, as dellsam23 said, it depends on which device has the better comb filter, the laserdisc player or the capture device (and any other processing done by the devices). First of all, most laserdisc players don't have s-video outputs so you have not choice. Most of the players that did have s-video out had crappy composite to s-video converters that left most the chroma carrier in the composite signal. That wasn't very noticeable on most TVs of the day because the frequency of the chroma carrier is higher than the TVs could display. But it's a disaster most capture devices that assume there's no chroma carrier in the luma channel. A small number of players had decent comb filters and produced better output than most composite capture devices.

But why are you posting deinterlaced AVC files? For what it's worth the s-video cap appears to have slightly more resolution in both the luma and chroma, but also slightly more noise.

[digicube]

But why are you posting deinterlaced AVC files? For what it's worth the s-video cap appears to have slightly more resolution in both the luma and chroma, but also slightly more noise.

Because I want to post videos online and everything is progressive now. Most likely I'll choose composite because that's what the voting seems to lean to.

[dellsam34]

From what I've seen, the domesday project is still not ready for prime time. Composite is the best bet for the forseeable future.


Scott

It is complete, just not for everyone, It requires building the hardware and know how to troubleshoot the software, The VHS decode is the one that is not ready for the prime time yet.

[jagabo]

But why are you posting deinterlaced AVC files? For what it's worth the s-video cap appears to have slightly more resolution in both the luma and chroma, but also slightly more noise.

Because I want to post videos online and everything is progressive now. Most likely I'll choose composite because that's what the voting seems to lean to.

But you're asking us to evaluate video capture. Showing us highly filtered and lossy compressed video makes that impossible.

[oln]

Because I want to post videos online and everything is progressive now. Most likely I'll choose composite because that's what the voting seems to lean to.

If the original is interlaced, any reason you're halving the fps to 29.97?

[DB83]

I have one Q. too.


This is from a UK tv show - The Brits Awards. (The most infamous one was when Jarvis ****er 'mooned' Michael Jackson )


Howe come the source is NTSC ? And could that have affected the recording/result ?

[DB83]

Lol. Nice text recognition there

[digicube]

It's a NTSC LD, so i assume the publisher converts it from PAL. I don't capture uncompressed, I capture in DV AVI which is also considered lossy. I'm assuming DV AVI vs MP4 is like WAV vs MP3, humans can't tell the difference. I want to keep the file size small so more people will be willing to download it and vote. I'll uploading the videos as MP4 so that is the final product I want people to evaluate. It seems there is little to no difference between the two files.

[dellsam34]

If the mp4 came from lossless AVI with generous encoding parameters then you can have decent quality for your viewers with pretty small file sizes, But DV to mp4? Holly cow.

[digicube]

I'm not in the 1%.

[dellsam34]

I think you are, Canapus devices are too expensive for most of us compared to a lossless composite capture card from the 2000's that can be had for as low as $50 used.

If you don't have one get it before it's gone:
https://www.ebay.com/itm/Pinnacle-500-USB-Video-Capture/193834482756?hash=item2d216f7c...wAAOSwHF1f8oAE

[DB83]

Surely the confession by the OP that he typically captures as DV is not relevant to this topic since it is a compare between non-DV s-video and composite.


And the OP probably knows that a DV 'capture' of a NTSC source will be shite in current standards (PAL is ok since it equates to the DVD spec)


But I would be interested to know the hardware used for the current comparison.


Am I correct inasmuch that LD was NEVER a consumer recording format which, if correct, questions how that UK broadcast even ended up on a LD.


What intrigues me, and probably the OP, is that the actual votes in the poll do not in any way equate to the unique replies in the topic.

[digicube]

ADVC300. Vocal minority, silent majority? I like to keep the DV capture files. Lossless files are just too big for me to keep.

I meant to say I'm not in the 1% of people who captures lossless, 99% probably capture in lossy format like DV, MP4, MPG.

[dellsam34]

You don't have to keep the large files if you don't want to, Just encode to h.264, keep the chroma sub @ 4:2:2, the files will be the 3rd of a DV file in size and quality is 10 times better than DV itself let alone converting it to mp4.

[oln]

I'm assuming DV AVI vs MP4 is like WAV vs MP3, humans can't tell the difference.

For sufficiently high bitrate no, deinterlacing will have an impact though if the original was interlaced. (How much depends on the method used.)

Most run of the mill lossless capture dongles have a pretty mediocre/primitive Y/C separation filter (which if you look at laserdisc forums has led to a big quest to find the best device for the job.). I don't know what the AVDC300 has though.

[digicube]

H.264 MP4 is not frame accurate for video editing, not a good archival format.

[dellsam34]

H.264 MP4 is not frame accurate for video editing, not a good archival format.

Frame accurate editing software for h.264 have long existed, maybe 10 years ago, such as Fame Ring SmartCutter, So that argument doesn't hold water anymore.

[digicube]

Thanks. I didn't know that exists. I use AVIDEMUX.

[jagabo]

H.264 MP4 is not frame accurate for video editing, not a good archival format.

Frame accurate editing software for h.264 have long existed, maybe 10 years ago, such as Fame Ring SmartCutter, So that argument doesn't hold water anymore.

It also depends on what h.264 settings you use. All I frame, or short GOP with only I and P frames are easily edited. Even 10 years ago!

[digicube]

I use Handbrake, no advanced settings to mess with.

[jagabo]

I use Handbrake, no advanced settings to mess with.

You can specify advanced x264 settings in the Extra Options box. Six frame GOPs with no b-frames:

keyint=6:bframes=0

[digicube]

Thanks. Is it better to have the MP4 interlaced? I prefer not to have horizontal lines showing during streaming however.

[jagabo]

Some people prefer to keep interlaced video interlaced in their final encoding. You can do that in Handbrake by entering tff or bff in the Extra Options box. But Handbrake isn't designed for interlaced encoding and it may screw up the chromra before handing it to the output encoder.