advantages of capturing analog video to lagarith vs. dv-avi

[TrackingError]

I have many self-recorded/non-commercial analog videotapes (mainly VHS and Hi8) that I wish to digitize. I have settled on using Stoik Capturer to facilitate the capture-process itself, as per some recommendations found on this site, but I'm unsure as to where to go from there.

I want my end-result digital file to be of archival quality, so I am looking at capturing to either Lagarith or DV-AVI. DV seems to result in slightly smaller digital files, but I'm not sure if perhaps Lagarith might have some other advantages over DV which might make it the better choice despite the increased size.

Any thoughts/suggestions would be very much appreciated.

[manono]

One's lossless and the other isn't. Also, if the tapes are NTSC and if the material being captured is in color, DV AVI isn't kind to the colors. Me, I cap in DV AVI, but almost all my material is black and white and PAL.

[TrackingError]

All right, thus far it sounds like I should choose Lagarith. Thank you for your input

[jagabo]

If you zoom into any DV capture you'll see blocky artifacts caused by DV compression. Not a lot, but a little.

For NTSC VHS caps the 4:1:1 chroma subsampling isn't really an issue as long as you use a decoder that smooths the chroma channels rather than duplicating them. (VHS has color resolution more like 4:0.1:0.1 so 4:1:1 is more than adequate).

https://forum.videohelp.com/threads/289684-DV-capture-quality-sucks?p=1758114&viewfull=1#post1758114

Note the color bars in that post were from a high quality source. If it had been VHS tape the bars would barely have been visible. See the second image in this post (color from a VHS tape):

https://forum.videohelp.com/threads/319420-Who-uses-a-DVD-recorder-as-a-line-TBC-and-wh...=1#post1980652

Of course, this doesn't rule out the possibility that any particular DV device has it's own problems.

[themaster1]

I'm told that UTvideo codec is best now it compress like lagarith ( so smaller file) yet retain the same level of quality than huffyuv it's worth a try i guess (i still haven't)

Dv ntsc should be avoided it use a 4.1.1 yuv colorspace which can lead to problems if you don't have a good decoder

[poisondeathray]

I'm told that UTvideo codec is best now it compress like lagarith ( so smaller file) yet retain the same level of quality than huffyuv it's worth a try i guess (i still haven't)

No, it's still slightly worse than lagarith (in the same colorspace) compression wise, but encoding and decoding speed is faster than lagarith and huffyuv. Quality wise, lossless codecs are all the same when used properly (they are "lossless")

UT is more of a performance lossless codec. For HD streams, it's almost impossible to get decent editing performance with lagarith or huffyuv (seeks are slow in a NLE) , they feel "sluggish" and unresponsive . But for SD streams, any of them should be fine with a decent computer

[TrackingError]

Thanks for the input, everyone!

[2Bdecided]

If you have the space and the speed, capture to lossless. If you are "archiving", make sure you archive a copy of your chosen lossless codec (binary and source code).

I use DV-AVI, because I can't see what harm it does to (S-)VHS sources (even if you intend to post-process heavily). HuffYUV takes 2-4x as much space, and doesn't playback or edit smoothly on my old PC.

Cheers,
David.

[Tafflad]

<snip>

I use DV-AVI, because I can't see what harm it does to (S-)VHS sources (even if you intend to post-process heavily). HuffYUV takes 2-4x as much space, and doesn't playback or edit smoothly on my old PC.

Cheers,
David.

As I need to do the same ... what do you use for capture .. previously I used WINDV, basic but solid & reliable - is there something better.

[2Bdecided]

I thought you were going lossless this time Tafflad?

Yes, I still use WinDV with my ADVC110. To be honest, use anything that works. There's no timecode, no changes in audio format, and there should be no gaps in the stream. The capture SW is just copying the stream, and it's not processing it in any way. As long as it can confirm that it hasn't dropped any frames, it's fine.


(Whereas when transferring actual DV tapes, there are many issues which mean one program can be better than another in some circumstances).


Cheers,
David.

[Tafflad]

I have no way to capture Lossless .... have just purchased another ADVC unit, so capture will still be DV .... will capture again .. without Full Frame TBC in the path ... to see if the is improvement, there was possibility that TBC had 'over softened' the video.

I can route using 'play thru' with a DVR to gain some Line TBC function and any required signal clean up.

[Brad]

Just so you know, the term is "passthrough"; not sure where you came up with "play thru".

[Tafflad]

Just so you know, the term is "passthrough"; not sure where you came up with "play thru".

my error, thanks for the education.

[sanlyn]

I have no way to capture Lossless .... have just purchased another ADVC unit, so capture will still be DV .... will capture again .. without Full Frame TBC in the path ... to see if the is improvement, there was possibility that TBC had 'over softened' the video.

I can route using 'play thru' with a DVR to gain some Line TBC function and any required signal clean up.

Except for less softening, you will likely get the same or similar results concerning digitally imbedded artifacts from VHS playback, and a somewhat etched or over-processed look to the color, and the beginnings of posterization effects. None of those effects were caused by using a frame TBC.

You would have to decode to lossless for major cleanup or visual alterations. And even without cleanup, you will have to re-encode if you want anything other than DV playback. There is no way to get around that.

Of course you would need the pass-thru tbc, but turn off Pansaonic's playback DNR -- it will soften video in a way similar to the that of the extra TBC earlier.

VHS and most other analog media are limited when it comes to attempting a complete filter-and-fix solution during capture. Noise reduction in most capture devices is really just primitive noise blurring, and such filters can cause dropped frames, ghosting, and motion smear. Brightness/contrast/saturation are fast-acting filters that are usually set to avoid crushed darks and clipped brights (the latter being rather serious with the current samples). But VHS output changes so dramatically from scene to scene that a kind of "worst case" adjustment is what you end up working with -- especially when the autogain controls on most cameras will kick your efforts in the pants. Trying to correct Hue with VHS is almost always an exercise in futility; what works with one scene will be disastrous two scenes later.

[crespo80]

If space is a concern, I can suggest to capture Xvid all-I (no B frames, no motion search) with target quantizer 2.0 and interlaced mode: it's really fast, it takes 1/4 of the space of any lossless format, it compresses better than Mjpeg and much much better than DV, while being easy for editing and playback, the image is absolutely blocks-free

[2Bdecided]

Except for less softening, you will likely get the same or similar results concerning digitally imbedded artifacts from VHS playback, and a somewhat etched or over-processed look to the color, and the beginnings of posterization effects. None of those effects were caused by using a frame TBC.

In case anyone missed it, here's a thread I started in another forum section, where I'm trying to find out if what you say is true...https://forum.videohelp.com/threads/360804-DV-vs-lossless-capture-of-VHS
...because I've never seen these issues with my captures.

Have you ever used an analogue>DV capture device sanlyn? I've used two: the Dazzle hollywood DV bridge (which was atrocious), and the Canopus ADVC110 (which seems fine to me, if the input levels are in-range, and you have an S-video signal, and you have more picture than noise, rather than more noise than picture).

Cheers,
David.

[sanlyn]

I never owned one, but an acquaintance and old movie fan I knew about 7 or 8 years back tried it. I don't remember what card he used, but it seems Canopus was discussed. It was a bitch cleaning tape noise out of that capture; there was always something weird about the color and gamma levels and odd edge effects that wouldn't go away, like bad interlace or something. AFAIK he never tried it again. He tried it with more than one tape, but I saw only one of the results. He finally went back to his AIW card to finish the projects. I do know that I envied one of his players, a Panasonic PV-S4670 SVHS with new heads and belts, etc., rebuilt by some outfit in Texas, and 8000 series JVC. I didn't care for the JVC, but it took me another few years to get one of those Pannies for myself on eBay that was in good condition. He helped me to locate a motherboard for my own AGP card, which I used to build a new PC about a year later and still use today.

As a result of your post I tried to contact him today, took forever to find his old phone number. I guess he has disappeared, moved, whatever. Another family owns the house. He was a retired tech from WNBC-NY at the time and I'll bet he moved to Florida -- that's all he talked about. Name was Harry Weinstein, if you wanna look him up. There must be 5 million Weinsteins in Florida.

Have you ever spent more than $50 for a capture card?

[Brad]

The Canopus devices aren't known for being inexpensive.

[sanlyn]

The Canopus devices aren't known for being inexpensive.

I didn't mean for that silly question to be taken seriously.