Best comb filter EVER? C2? Holo? Entech? Put on your gloves!

IMO the best is getting them onto a PC and processing them with AVISynth using TComb (or the older Guavacomb) or one of many other filters that can accomplish the task (lots of rainbow killers, and dedot, decrawl and some others that can help with dot crawl)

It takes a lot of effort to get things like this as good as possible, no matter what you use. And you will unfortunately never be able to get them back to original quality (though you seem to understand that already).

A good postprocessing can make up a lot ! I've captured about a hundred tapes a few years back using a Sweetspot (PDI Deluxe card). I re-captured one after I got your working CVSI-1. The difference was noticable, but I would not call it great.

**MUCH** more important is a **GREAT** VHS player. I've played around with about 10 different high-end (S)VHS Machines, ranging from Sony to Panasonic to JVC. I've tried all those "great" players and wasn't satisfied till I switched to a professional grade Sony Player (SVO-1500P). It's not even a S-VHS deck, just plain PAL & NTSC VHS, but the jump in quality was so huge, even compared to the best of the consumer decks.

For postprocessing I used NeatVideo (commercial NR plugin) and the results were really stunning. The final MPEG2 files easily surpassed the original VHS tapes.

C2 = Crystalio II ? Nice processor with a nice comb filter (nidi's using a X0 + C2), but it won't help you with capturing (*). I've watched so few LDs since I've sold my last Faroudja-based processor, but I wouldn't have thought about suggesting an Entech unit in addition to a C2...

I've just ordered my new capture device (BMD Multibridge Pro2) and I'm looking forward to do some capturing experiments from analogue sources with all equipment I've got on hand...


you tell me ! I had to laugh about uderman's statement in the new LD thread, when he suggested "simply buying a Entech SVSI-1"


[ (*) unless you have the C2 upscale the picture to 720p, capture the VHS in 720p and downscale them again in the postprocessing. Quite silly, but I've actually thought about this for 480p source since the BMD capture solutions don't capture 480p, just SD and 720p upwards. ]

What's your end format ? DVD ? I always had timing problems when I had a VP upscale and/or deinterlace the picture and then re-interlace it before capturing (using a Extron Scan Converter with RGBHV input and SDI output). If you want an interlaced end format like DVD, you should stay interlaced all the way. I don' think using a C2 for it's comb filter, then deinterlacing and then re-interlace the picture is a good idea. Of course it's a whole different affair if your end format is something like progressive AVI....

Unfortunately the hardware I do not have so much extensive experience with, so I decided not to answer that part. I'd tell you if I knew.

I think the signal path in the processors is fixed: deinterlacing before scaling and postprocessing. Therefore my guess is that the picture is deinterlaced to 480p, scaled to 1080p and then reinterlaced to 1080i.

I still have the Faroudja VP-100 in my closet somewhere that I use
occasionally with my LD player. The box does composite to Svideo conversion
http://www.sdinfo.com/volume_4_2/faroudja100.html

There is also the Faroudja VS50. I don't own one . $$$
http://www.hometheaterhifi.com/volum...oudjavs50.html

The Faroudja A1 is a video DSP with composite input
and SDI output. This is for broadcast use. I have not been able to find
any links on this one. I have one one these in storage somehere..

The HD Leeza has a very good comb filter, subjectively better than any
of those above. I use this one currently for the LD.

An unrelated thing that might be useful for transferring old VHS tapes
would be a time base corrector. As you are already using a broadcast
quality deck it may already have one.

Absolutly! If you are looking for top quality conversion a TBC is essential. Most consumer scalers/decoders will accept *non-phased chroma as they have to. But nothing beats a TBC for cleaning up the edges that are compromised with varying Y/C delays do to the time base errors.

The best TBCs for VHS are either the built in units or an external unit properly connected to an industrial VHS deck. The "proper" connection involves sync and subcarrier feedback to the deck to allow processing of the raw video signal. Some TBCs will work with a consumer VHS deck but there is a price to pay. In this mode, there is additional processing added to the signal so the sync feedback to the machine is not needed. But this processing does reduce resolution somewhat, especially in the chroma.

I use a surplus broadcast decoder that has a FRAME BASED COMB FILTER. Early 1990s technology but still far superior to any consumer chip based decoder. This unit switches between the various filter modes based on analysis of the image over a frame. However this unit is very picky about input video. You can't feed raw VHS into it so a TBC is required. It also delays the video by two frames which now means the audi must be delayed as well to keep lip sync.

*In consumer VCRs. the chroma is time base corrected as you can't do color playback without some form of timebase correction. However the luminance portion is not corrrected as consumer TVs have more than enough lock in range to follow the jitter. This results in a "non-phased" signal inthat the sync and color subcarrier are no longer locked together. This is OK for viewing on a monitor but the signal no longer to broadcast specs. As the chroma is stable but the luminance is not due to the mechanical tolerances in the VCR, you have the luminance moving realitive to the chroma. This causes noise and softening of edges around saturated color areas.

Y/C is stored independently in the VHS format!!

A comb filter is unnecessary if using the S-Video output!!


You don't need the BEST COMB FILTER EVER!!! Just a S-VHS VCR!!!!!

..

wasn't this the difference between S-VHS and VHS systems ? I think it's correct for tapes recorded on S-VHS decks, but incorrect for VHS recordings or (most) original pre-recorded tapes. But somebody please correct me, if I'm wrong.

What I transfered to HDD/DVD were mainly pre-recorded tapes and I found not even the best S-VHS machine I had on hand (some high-end JVC) could touch the composite quality of a professional VHS-only player. With a self-recorded S-VHS tape is was the other way around - for obvious reasons.

I believe the VHS recording system is called over and under

luma and chroma are recorded seperately

from wikipedia:

VHS tapes have approximately 3 MHz of video bandwidth, which is achieved at a relatively low tape speed by the use of helical scan recording of a frequency modulated luminance (black and white) signal, with a down-converted "color under" chroma (color) signal recorded directly at the baseband. Because VHS is an analog system, VHS tapes represent video as a continuous stream of waves, in a manner similar to analog TV broadcasts. The waveform per scan-line can reach about 160 waves at max, and contains 525 scanlines from top of the screen to bottom of the screen in NTSC (486 visible). PAL variants have 625 scanlines (576 visible). In modern-day digital terminology, VHS is roughly equivalent to 320x480 pixels with a signal-to-noise ratio of the image at 43 dB.

JVC would counter 1985's SuperBeta with VHS HQ, or High Quality. The frequency modulation of the VHS luminance signal is limited to 3 megahertz which makes higher resolutions impossible, but an HQ branded deck includes luminance noise reduction, chroma noise reduction, white clip extension, and improved sharpness circuitry. The effect was to increase the apparent horizontal resolution of a VHS recording from 240 to 250 analog (equivalent to 333 pixels from left-to-right, in digital terminology). The major VHS OEMs resisted HQ due to cost concerns, eventually resulting in JVC reducing the requirements for the HQ brand to white clip extension plus one other improvement.

In 1987 JVC introduced the new format called Super VHS which extended the bandwidth to over 5 megahertz, yielding 420 analog horizontal (560 pixels left-to-right). For comparison DVD is 480 analog lines horizontal, so a Super VHS tape can capture a DVD or broadcast standard definition DTV video with only a small amount of blurring. The chroma resolution remained the same at approximately 0.4 megahertz bandwidth or 30 analog lines horizontal, as was common across tape standards from Umatic to VHS to Betamax. Even a live NTSC broadcast is limited to 120 chroma lines maximum. (For comparison DVD is 240 chroma horizontal.)


Michael

The only caveat here is the architecture of the internal TBC. I'm willing to bet they took the easy way out and built in a composite TBC as it's simpler circuitry. That means they combine the Y/C back to composite, run it thorugh the TBC, and then split it with a comb filter (or worse) on the TBC output.

So if this is the case, then a good comb filter on the TBCed composite output may be an improvement. Now if they have a Y/C based TBC built in, then the Svideo output is best but I really don't think any VHS built in TBC was of a Y/C design. I could be wrong.

But seriously here, even SVHS is not very good image quality IMHO. Worrying about the effects of a comb filter with this format is akin to throwing deck chairs off the Titanic. With Laserdisk, a good comb filter makes more sense.

The ACEDVio's live DV encoder make things messy. One would believe that the DV bitrate would be sufficient for VHS capturing, but you'll get blocking in fast moving scenes. Better use the AVID with some kind of non-destructive codec.

I think so. I was under the impression that the ACEDVio ist the internal version of the ADVC-300. Just got the call that my Blackmagic Multibridge is ready for pickup *g*

Long ago I talked with a rep. of Camelot (which makes the Crystal Vision VPS-1 comb filter unit; composite to S-video converter) and he said it recovers all of the chrominance up and down the video spectrum.

Yet I found out a little later by experimenting that the VPS-1 must have a somewhat lesser color bandwidth (and therefore a somewhat lesser chrominance resolution) than others. Using a Snell & Willcox test diagram in the LD Video Essentials, other comb filters including one in my TV showed much more red in the chrominance resolution test bars compared with the VPS-1. I would have guessed the resolution to be no more than 0.8 MHz or 65 lines per picture height (4:3 aspect ratio) out of a possible 1.5 Mhz or 120 l/ph aka TVL.

Meanwhile the VPS-1 showed less dot crawl than any of the others (in a DVDO iScan Plus, in a JVC S-VHS VCR, and in a Sony RPTV).

>>> For comparison DVD is 480 analog lines horizontal ...
DVD should be 540 l/ph for 4:3 programs and 408 l/ph for 16:9 programs or 720 pixels left-to-right.

All of the consumer analog video tape formats (Beta, VHS, 8mm, also U-matic) recorded luminance and color separately, and therefore benefit from pure or direct S-video output from the VCR.

>>> Even a live NTSC broadcast is limited to 120 chroma lines maximum ...
For NTSC broadcasts, the 120 chroma lines applies only to some colors, namely reddish orange and greenish blue. Other colors resolve at as low as 40 l/ph. Technically correct NTSC is constructed from a component video complement you can vaguely think of as Y/Porange/Pgreen; the correct term is YIQ. It will decode acceptably with any standard Y/Pb/Pr color decoder and when the decoder has an 0.5 MHz chroma bandwidth (99% of them are that way) the result is indistinguishable from programs encoded from Y/Pb/Pr.

Many broadcasts, notably newscasts, used tape equipment with Y/Pb/Pr and 0.4=0.5 MHz chroma bandwidth throughout and therefore there was no content that took advantage of the wider chroma bandwidth for reddish orange and greenish blue.

Meanwhile, TV sets with 0.5 MHz (40 l/ph chroma resolution) arrived within a few years of color TV and very shortly thereafter 99+% of TV sets sold were that way. Today a technically correct NTSC color decoder is hard to find. It would have to either go from S-video to YIQ directly to 480i RGB or from S-video to YIQ to Y/Pb/Pr the latter which could be output to a different video device.

Video hints: http://www.cockam.com/vidcolor.htm

Wait, you mean that old CVSI-1 sitting in a box behind the bench on my sunporch is actually any good? If I recall, I bought it on ebay for <$100 years ago—back when the backbone of my video setup was component/VGA, not DVI/HDMI—to use with my NES before I decided to just RGB-mod the console. Hah. That's funny.

PM if you are interested in selling it!

So is there consensus that the ENTECH SVSI-1 is a worthy addition to a home theatre with laserdisc? I plan on adding one to my existing system consisting of:

Pioneer CLD-D704 (to ENTECH SVSI-1) to DVDO Edge to Panasonic PT-AE3000U.

Opinions...

If you're thinking about the SVSI-1 on eBay right now, BUY IT! It's the only SVSI-1 I've seen on eBay in years and it's reasonably priced.