Is this a good Video8 workflow?

[bradgranath]

Hallo. What a place y'all have here. I paid the twenty bucks to say: thanks for still being here!

I have a stack of ~60 Video8 Tapes from between 1990 - 2005. All shot with various Sony Video8 Camcorders. No Hi8. No Digital8. Mostly Sony MP120 and TDK HS120 tapes, with some maxell and Fuji tapes as well. All in very good condition and well cared for. No idea which tapes are recorded in SP and which in LP, or how to tell.

Step 1: stick a serialized label on each tape and make a spreadsheet with the ID number, tape capacity, what's written on the label, and what year I suppose it starts. Done.

I'm trying to set up a decent workflow. I've been reviewing them on a Sony TRV-7000 connected to my FireWire port. It allows me to review the contents of the tape, but the quality is dreadful. The DV compression is bad enough (not so bad that I didn't try it to see if it would work... lord it looks awful) but the main problem is that there are dreadful multicolored vertical stripes that make hash of the colors when I try playing most of the tapes.

Some tapes don't have this issue and play totally fine. Always the same tapes though, so it isn't a random issue with the VTR.

Anyhoo... I also dug out my old Sony CCD-V8AF, and plugged it in to the Sony ECR137 TV and these 'trouble tapes' play without issue.

So, back to square one: The TRV-7000 isn't gonna cut it; DV compression or no.

So...

Now I've got an Osprey 260e on order from Ebay. Planning on capturing to huffyuv60i AVI to a 6TB harddrive directly through the CCD-V8AF.

On the V8AF playing through the ECR137 the tapes look fine. No tracking issues, no wobble, audio sounds fine. Do I still need a TCB? When I used to capture Digital8 or miniDV tapes - back in the day - 'dropped frames' were a thing? but this is different in analog, right? None of the tapes was ever 'striped' ahead of time so the timecode is likely a mess, but I'm not familiar with what that mess looks like in the analog world and how/how much to worry about it. My current plan to deal with it is:
Monitor the playback and capture stats for problems, stop capture when they occur, try to restart capture from a previously captured point without issues, mark the file split in the log.

As it captures, I'll stand by with a pencil and make notes on what happens and where during playback. Definitely gonna space this out since I've got something like 100 or so hours of tape to roll.

Deinterlace with QTGMC and transcode to huffyuv30p. (Settings to tune QTGMC for erratic dad-cameraman who sometimes forgot that he'd pressed record?)

Transcode offline proxies.

Editing/magic/other stuff...........

Delivery to YouTube/iPhotos. This delivery part is getting to me. My parents are richie-rich techphobes. They have iphones, but can't manage their MacBooks or get their TV to change to correct input. So, I want this in a format I can throw on a thumbdrive whenever they ask for it and or, I can just dump it into their iCloud. Is it worth trying to upscale it myself, or should I just upload the highest resolution I have and let the chips fall as them may on a given platform?

[lordsmurf]

For Video8, you still want a Hi8 player with line TBC.

You'd better be careful about "reviewing" tapes. At this late date, the 2020s, there is a good chance that each play of the tape damages it. Too many times these days, tapes are "one and done". If you wasted that "one" from watching, or even doing that silly pointless FF/REW "trick", then it may not work well, or at all, for the actual capture.

Osprey is supposedly comparable to AIW, but more fiddly in usage. So the card may be fine.

Video8/Hi8 tapes are "drop-happy" -- and the timing can actually mess with capture in several ways, not just frame drops. Some of the issues you're describing can be from lack of frame TBC. (Not just any random "frame TBC", but specific models, such as TBC-1000, and others).

Retain your interlaced lossless masters.

Don't bother upscaling. If they're techno-ignorant, then give them a deinterlaced H.264 in MP4, with the black bars add to the side so they don't watch it stretched. Hopefully they know how to use a thumb drive. Everybody has "those kinds of people" in the family, and usually it's refusal to learn, intentionally being obtuse.

You're software workflow seems fine. Just be mindful of settings.

[bradgranath]

What is the advantage of the Hi8 deck? I can get rollin' on a TBC. You have one on offer now? Also saw a 'working' DPS 290 (leds and lcd lit up in the pictures) on ebay for $250, which seems too good to be true. Should I jump on it? Mebe also needs something to genlock to?

[latreche34]

Hi8 camcorder has the advantage of avoiding combining luma and chroma in a single wire called CVBS (yellow composite), S-Video gives sharper details and better chroma separation vs washed out details and faded chroma. Line TBC stabilizes scan lines.

Don't buy home decks, they have mechanical problems.

[hodgey]

It sounds like the TR-7000 has something wrong with it (haven't seen this particular issue) otherwise it ought to work well for playback.

The late model Sony Hi8 and digital8 camcorders has an excellent digical video circuit including an excellent TBC that stabilizes the image really well, so the playback quality on them is usually as least as good as the decks, the newest which are from the mid 90s.

There are some cases where decks can be useful as they are a bit more serviceable , and they can rewind tapes much faster. The main case where you do want one is if you have tapes with PCM audio, something found on the fancier decks and some of the "pro" cameras like the CCD-V5000. The consumer camcorders did not support this.

[latreche34]

I always use tape rewinder, I'm not concerned about rewind speed as much as I am about head wear of my camcorders.

[bradgranath]

Any thoughts on this TBC? https://www.ebay.com/itm/32494051843...sAAOSwznhhnT8B

[lordsmurf]

Quote:

Originally Posted by bradgranath

Any thoughts on this TBC? https://www.ebay.com/itm/32494051843...sAAOSwznhhnT8B

No, don't get that.

That's a rackmount ("pizza box") unit, meant for broadcast sources. Your consumer analog VHS videotapes will confuse it, as it doesn't expect that sort of messy signal. It will usually work very badly, if at all.

Furthermore, these things are always beat to hell, not at all well kept. So even if you had broadcast sources, expect issues. Just look at the unit pictured in that auction. The rackmount is all damaged, probably from being bounced around on a concrete floor. Just old stuff in the way.

The seller is a recycler, and likely doesn't know a TBC from a toaster. "received as working" is code for "we have no idea if this thing works, but the person who dumped it off (random intern?) swore it did!" You have to know how to read between the lines.

No, no, no no no.

No.

[bradgranath]

I did read through your notes on choosing a TBC in that one thread. Prolly right, not worth it.

But... The rack ears don't look broken to me and all the edges and corners look sharp? So maybe not bounced? But maybe you're talkin about a different bit? And I'm using video8 tapes, not VHS. And this link to DPS marketing material describes it as being useful for transcoding consumer VHS and Hi8. http://www.broadcaststore.com/pdf/mo...99/DPS-290.pdf And the seller states that it's been tested and they accept returns. I'll admit to being skeptical, but it seems like it might be worth the risk of paying return shipping to find out.

[lordsmurf]

Quote:

Originally Posted by bradgranath

I did read through your notes on choosing a TBC in that one thread. Prolly right, not worth it.

But... The rack ears don't look broken to me and all the edges and corners look sharp? So maybe not bounced? But maybe you're talkin about a different bit? And I'm using video8 tapes, not VHS. And this link to DPS marketing material describes it as being useful for transcoding consumer VHS and Hi8. http://www.broadcaststore.com/pdf/mo...99/DPS-290.pdf And the seller states that it's been tested and they accept returns. I'll admit to being skeptical, but it seems like it might be worth the risk of paying return shipping to find out.

I've used multiple DPS/Leitch/Harris units, and they all disappoint. Perfect SP tapes from a prosumer camera may be fine. Or not. But anything more degraded, especially from a VCR, will falter signal-wise.

And again, that auction has code words.

And it does have obvious damage/abuse from the photos. The left ear is bent. Those rack mounts are thick metal, and you'd almost have to use a sledge hammer to beat it back. So when you see a bent ear, it was mistreated.

Rackmounts are also meant to be calibrated, not used as-is.

Beyond all that, I hate the massive size, and all the sharp metal edges. It's really not that easy to use outside of a rack.

There's also a very shady seller (group of sellers?) in Houston that needs to be avoided. The infamous "Houston VCR scammer" buys gear on eBay, and flips it for 2x+ the price. The problem with that is the original listing may be honest (not working, etc), and he'll list it as "working, tested" before he even has a chance to get his purchase in the mail. Too many members here have dealt with this a-hole. He's super rude, lies, and fights you for a return/refund. Expect your eBay hassle to be wasted time, and you won't get your money back for a month or more. He knows tricks to get his negative feedback removed, as members have have learned (their comments would disappear).

[bradgranath]

Hmm, perhaps it is bent. Was pretty sure that was barrel distortion from the wide angle as it appears fine in the other shots, but like ya said, better safe than sorry, 'specially with that geography.

[lordsmurf]

No, not distortion there, bent.

[bradgranath]

I'm back!

I'm trying a more serious build, but am still in the early days of setting it up and proving it, before I start running tapes through it (taking your advice re unnecessary playback).

I've put together a 2004 era rig to run an AGP AIW 9600 that I sourced from ebay (i KNOW, gimme hell if you think this is really an issue, but the card arrived thoroughly intact with every original piece of packaging and in very good shape - not dusty or oxidized or smudged)

THE BUILD

ASROCK 775DUAL-VSTA
Intel Q6600 (quad core 2.4Ghz) {Will likely replace with a dual core 3Ghz later, but the Q6600 works fine for now}
2x 1GB DDR2-533
ATI AIW 9600 PRO
Linksys Gigabit PCI NIC
M-AUDIO Audiophile 2496
128 GB Lexar SSD system drive
128 GB Lexar SSD scratch drive
WinXP 32 bit SP3

Did a careful visual inspection of each board's ECaps, none are bulging or leaking. They all perform as expected when booted.

Also a green AVT8710 from someone who was using it to do video transfer but had finished their project.

Help I could use:

- What ATI driver versions are best? You've talked before about the ATI driver cancer that can happen with WinXP. I don't mind reinstalling at this point, but I'd like to know where I'm trying to get to.

- What VirtualDub version should I use?

- Which huffyuv codec is best?

- Is there a good way of verifying that the 8710 is OK? My current plan to "test" it is to try running a known good DVD through the SVideo circuit, checking for gross errors, and then to burn an NTSC test pattern to DVD and use a vectorscope plugin to verify that it's interpreting colors correctly. For that matter, is there a good source for education on how to use the procamp functions on a 8710? Manual? (I can prolly google that)

- I have a Sony TVR108 as a primary playback device, and also the old CCD-V8AF as backup. Am I missing anything here? Definitely planning to be careful with the tapes.

[aramkolt]

Usually the Sony models that end with "8" are mono audio output only, so that'd be a nonstarter for me, but I suppose if all of your tapes are mono, that might not matter.

The 8710 *can* be a nice unit from what I've read, but I'm not sure if there's a reproducible way to test if yours has issues or not. Generally the test is described by others to clean up certain defects from specific bad tapes.

Those drive sizes seem a bit small to me for storage since these video files will probably take up around 30GB/Hr.

You are unlikely to get any sort of capture performance boost from going from one quad core to another for purposes of video capturing, so I wouldn't bother with that personally. You really shouldn't have the computer connected to the network either during captures, but I suppose if that's how you plan to transfer your files off of the machine, 1Gbit Ethernet card is useful.

Inspection of electrolytic caps doesn't really tell the full story if the capacitors are good or not, but as long as the machine runs without errors, you're probably fine for now when it comes to motherboard stuff. I have a few of the AIW 9xx0 series ATIs and I'd say about half of those had at least some bad capacitors, so I fully recap them also, but that's probably not worth it for most people as it takes some time and skill, and there's the potential to damage the card if you're not careful.

I'd say ideal qualities of any TBC are:

-Doesn't make the video of a stable source look *worse* when passing through - DVD source is fine to test that - just capture with and without the TBC inline and see if the TBC passthrough capture looks worse in any way. To get more granular, a vectorscope/waveform monitor of the resulting captures should make it much clearer how much noise/detail is being removed compared to no TBC inline.

-Doesn't clip black or white levels on passthrough.

-Doesn't show any frame drops on 15+ minute captures. (I'm on the fence about whether an occasional frame insert is ok).

-Doesn't develop audio sync issues on captures that are 1+ hours long.

-Ideally doesn't change input levels in terms of hue, saturation, chroma, brightness, but most of those could be adjusted in post assuming that there's no clipping going on.

At the end of the day, going through any TBC is a form of mild generation loss since you're converting from analog to digital then from digital to analog. This essentially means that you are taking something with infinite possible values and then assigning digital values instead which are fixed numbers (similar values get rounded to the same value), then re-converting those back to an analog signal. The "value rounding" essentially is a sort of noise reduction, but that's a permanent data loss and also can result in detail loss as detail and noise are often two halves of the same coin. All analog tape captures require at least one analog to digital stage, but the more devices that go back and forth between analog to digital in the chain, the less like the original video you get. This is why I personally like TBCs that do not convert back to analog at the output so that you just have the one analog to digital conversion. Those would typically be TBCs that have SDI output, though there are some passthrough DVD players that will allow digital output via HDMI.

The last factor is the bit depth of the TBC itself. All standalone consumer TBCs are 8 bit, whereas many of the professional ones are 10 bit. The significance of this is how small of an increment that values can be rounded to, essentially adding more significant figures to each value.

A more real world example would be this string of numbers which you might think of as luma values for different pixels: 1.2531, 1.6212, 1.8991, 2.6803, 4.5472.

A lower bit storage of those values might be to one decimal place, so the stored data would be 1.3, 1.6, 1.9, 2.7, 4.5.

A higher bit resolution of the same list would be the same list being stored/saved at: 1.25, 1.62, 1.90, 2.68, 4.55.

Both options lose data, but the higher bit depth looses "less" of the original data. At some point, the human eye isn't going to distinguish between additional decimal places however. This visually matters more if you have scenes that are mostly close to the same value to begin with - such as very dimly lit scenes or sky scenes. Other thing to keep in mind is that there's really no advantage to higher bit depth of a TBC if you aren't capturing in that many bits since the values will get rounded at the point of capture Basically any 8 bit device in an otherwise 10 bit chain will reduce the effective bit depth to 8 bits.

A real easy test for the difference in 8 vs 10 bits used to be turning your graphics settings down from "Millions of colors" (10 bit) to "Thousands of colors" (8 bit) in video settings on your computer. Odds are you would say Millions of colors gives a better visual result, but that of course depends on the detail and if the original content was 8 or 10 bits to begin with. With a CRT, the difference would probably be even less noticeable though.

Does the bit depth always matter visually? - Probably not for most analog SD content, but in dark or sky scenes, I'd say probably yes. Is it worth the extra space that it takes up? For most people, probably not. I personally like the sweet spot of "visually lossless" ProRes422HQ which is 10 bit, but takes up roughly the same space as an 8 bit lossless Hufyuv capture (as a bonus, it's also more Mac and video editor friendly).