Decisions on TBC, VCR and capture card setup?

[lordsmurf]

For varying reasons, I'm answering a question originally asked of me on another site in a new thread on this site.

I've selectively quoted all of the questions, bulleted them for easy reading and answering:

Quote:

- planning to rip a collection of some personal vhs-recordings (i.e. rather important ones to me), of which most are about 10-15 years old with a few ones looking pretty worn-out. Quite a few of these tapes also suffer from excessively saturated colors (mostly reds) which is due to a fault in our camera that became more notable over time, so this could also be something to take into consideration.
- JVC HR-S85000. I can get this from someone for about 210 dollars (converted from euros). Judging from an online fact sheet it indeed seems to have all the nice features I could want for optimizing my video capture. Most notably it's said to have a Digital TBC, which I think would justify it's relatively high price.
- Related to this issue I was wondering if an external TBC (currently I'm looking into the TBC-1000) could add any additional stability to the setup even when the jvc supports the function already. I've understood from one of lordsmurf's posts that integrated TBC's and external ones do different things to improve the video quality altogether, but does this mean that if I want good image quality and a stable signal to prevent out-of-sync issues I'm theoretically better off with a setup in which both types of TBC's are active?
- portable (used) JVC BR-6200EG
- good-functioning consumer-grade Sharp (like the one I have) can't do already?
- don't own a dvd-recorder
- planning to archive my homemade movies to harddisk in the highest (storable) quality possible (for this purpose I'll probably end up capturing in huffyuv and compressing the material to h264 or dv-avi)
- me wonder whether it wouldn't be better to leave out an external TBC altogether and instead stick to the dynamic audio resampling function in vdub or even buy a HDD recorder that can keep things in sync without 'softening' the image? In any case, I'd still feel safer having a TBC around should the need for it to be used arise so I'm still looking for one.
- TBC-enhancer by Electronic design
- For.A FA-310P, which is also said to have options for color color correction and adjusting black and white levels
- on the subject of what a TBC can do for me, I was curious how it could affect the need for certain software-based timing corrections when capturing to a PC? During my previous capturing attempts for example, I noticed there was a constant need for audio resampling, and at every 'unsmooth cut' (which are relatively abundant in my recordings) virtualdub had to insert a few frames in order to keep the estimated error minimized. With two of my tapes (from 1997) this indeed seemed to yield perfect sync (at least where it mattered). However, with one particular video (which dates from half a year later, but is more played) resampling and inserting frames would likewise occur to keep the error fluctuating around a value of 0 ms, but on a rewatch the audio was obviously lagging regardless (also delaying the audio with a negative value would just cause vdub to add another 50ms to the error which it would then proceed to correct). This is why I would like to know if I still need to rely on such shaky timing corrections when including a TBC in the setup. Alternately, in consideration of tapes for which the TBC may turn out to be too detrimental for the image quality (so for which I'd prefer the TBC to be switched off), I wondered if adding a decent capture card that supports audio capture would be likely to prevent this type of miscalculation error by virtualdub? I guess the question here is whether it's the tape or the seperate capture devices that are apparently causing the program to screw up on resyncing.
what capture devices would be good recommendations for me. As for requirements, I simply would like one that keeps things well-timed, doesn't force compression, and doesn't add any saturation or other flaws to the capture. In other words, a good one. Or perhaps I should consider a HD recorder after all?
- Panasonic AG7350
- still a little concerned about the fact that the machine dates from 1995, and I'm not sure if this would practically classify it as a worn/outdated device, for example because it's framebuffer is considered too small by today's standards. To be more specific about the latter, according to the manual it's only 1,5 MB large which allows it to store 3 half frames in memory, whereas (if I'm not mistaken) most of the recommended vcr-systems support 2 or 4 MB TBC-memory

I read through all of this, and while I'm sometimes inclined to try and help people make decisions on gear I would not necessarily suggest, I don't really want to do that this time. You have important VHS tapes (your own words), you're clearly wanting to find the best gear, and not have to waste a lot of time experimenting on your own (you'd have already done that, if you wanted to). I think you would be better served finding some suggested gear, rather than finding gear and seeing if it is workable.

Let's start first with the VCR...

It seems to be understood to you already, but I just want to re-iterate that professional- and broadcast-grade S-VHS VCRs are excellent for playing VHS tapes meant for digital archiving. These machines will pull out more of the good signal and create a higher grade DVD (or other digital format video) than would a cheap consumer VHS VCR. Moving on...

Those older JVC and Panasonic decks you've come across are not necessarily ones I would suggest. The 8500 might be okay, but you're more or less gambling. Unlike the NTSC line, the PAL series rarely had updates or new models. For every PAL model, it seems there were 2-3 equivalent NTSC ones. I prefer the final line of PAL JVC S-VHS machines, either the HR-S7965 or HR-S8965. These were the PAL equivalents of the well-liked JVC HR-S7900 and HR-S9900 NTSC decks, which were themselves re-releases of the well-liked HR-S9600, HR-S7900 and HR-S9800 decks. Like other JVCs, it will play most tapes in excellent quality most of the time.

JVC's weakness is the ability to track long-play tapes, those recorded in EP or SLP mode, with perfect stability. It usually works, but not as often as the comparable Panasonic NV-FS200 (PAL version of the NTSC Panasonic AG-1980P). However, Panasonic's weakness is the extent to which the image is cleaned, and Panasonic gear is known for its over-processing of the signal, be it extra fake sharpness, more grain noise left viewable, or temporal compression or posterization. In all, just remember the goal here is to clean up the VHS tape as best as is possible, and sometimes that will mean trading a worse error for a mild one.

It's for these reasons that I suggest both machines, if you're looking to transfer a large volume of tapes, for an extended period of time. The other option is to pick on machine, and then set aside any tapes that just don't cooperate or look as good as you want. Outsource those to a service that specializes in restoration, at a later date.

The age of the VCRs should not be a huge concern, most of the best machines were manufactured, marketed and sold in the late 1990s. This was a time when tape-based (VHS-based) formats were in full maturity, in heavy use, and before our market was flooded with so much cheaply-made Chinese/Taiwanese crap electronics. This pre-DVD era had quite a few nice machines made and sold, and its why some of them tend to be hard to find -- owners refuse to sell! Indeed, I would be more wary of something produced in the mid-2000s than something made in the late 1990s!

Concerning the RAM buffer, trust that the manufacturer made good choices. There differences between 2MB and 4MB is more or less negligible. I've looked for differences in these machines, but barring the worse instances of certain kinds of uncommon timebase errors, I've never been able to make the machines act too differently. The type of RAM, how its used, and speed of the RAM in use is almost more important. There does not need to be a lot of it, for it to be effective at its DNR task. I would not worry or put much thought into this.

Now let's move on to the timebase corrector (TBC) ...

Older rackmount professional TBCs just don't perform the same as the TBCs made specfically for the "prosumer" (serious consumer hobbyist, lower-budget professional) video equipment out there. While the professional TBCs can work fine, a lot of them simply do not. In many cases, its due to excessive age. But it can go a little deeper than that, when you consider its reason for existing. Those TBCs were made for broadcast or duplication scenarios, and they perform well at those tasks. However, this doesn't necessarily mean anything to analog>digital conversion, which is a use kept in mind when prosumer TBCs were made. It's not unusual for a broadcast TBC to perform worse than a prosumer one -- in fact, that's the norm.

You really have two choices for a TBC.

1. The CTB-100 is base model assigned by the Asian maker of the device (made in Taiwan, I believe, if memory serves correctly). This is imported worldwide, and re-branded, among them the AVT-8710 imported by AVToolbox and sold at video speciality stores like B&H or (at one time) VideoGuys.
2. The other option is the DataVideo line of products, be it the TBC-1000, TBC-3000 or TBC-100 (which is a PCI-powered version of the TBC-1000). Just be careful of some of the TBC-1000's, there were various quirks to these, causing herringbone or softness issues. It's very random.

Again, I would go with a known-good device, unless you just want to gamble on other easier-to-find local gear.

I have a DataVideo TBC-1000 available for sale, for $350 USD. It performs perfectly, it was barely even used, and is essentially brand new. No herringbone or excess softness detected, a good copy! This model TBC can currently fetch $500 new, and even then you can end up with a bad copy!

I also have an AVT-8710 available for sale, for $190 USD. Again, good copy, works nicely, like-new condition, and brand new it runs at least $215 or more. You'd save at least $25 on this.

These were bought for some special projects in the past, and were just not needed after those projects were completed. I'd rather sell them to somebody else who could get good use from them.

When it comes to TBCs, don't worry about the extra features crammed onto the boards. The distribution amps can be decent, but a dedicated distribution amp might work better, depending on the scenario. The one I see most discussed is the proc amp functions, to adjust colors and contrast quality. Honestly, these are so limited in range that it's almost useless. You'd do better to buy a dedicated proc amp, even a $50-75 range vintage 1980s/1990s Vidicraft unit will be more effective for you.

Now about capture cards...

I want to specifically nix the Canopus DV converter boxes. For highest available quality, I just don't suggest these. Those boxes are good for lazy consumers, or consumers with touchy computers. The DV compression is just not best for VHS sources, DV is best left as a shooting format in cameras. A lot of the Canopus marketing (and the salesmen at B&H especially) get on my nerves with the fluffy marketing they use, especially as it relates to the "where-is-it?" TBC or the ADVC-300 processing that tends to make an image worse, not better.

If you have a high budget, I would take a strong look at the Matrox series of NLE cards. Then again, those are mostly useful for NLE work in Adobe Premiere CS3 (and CS4, I believe).

Another really good option is the older AGP-slot ATI All In Wonder Radeon cards, especially the latter 9000 series cards (9000, 9200, 9600, 9700, 9800), and using ATI MMC 8.7, 8.8 or 9.02. While these were "consumer" cards, they were marketed to higher-end computer/game/video hobbyists, and had surprisingly professional-quality video processing/capturing abilities. The only downside here is you need a slightly older pre-PCIExpress computer. I purposely keep several AGP slot computers, because these cards are an essential part of my workflow. I actually have some extra ATI AIW AGP cards that are no longer need, and I specifically have a really nice PAL 7500 series card that might interest you. I'm willing to sell.

Beyond that, I'd really need to know more about your budget, and what sort of end-goal you have for your video work. Knowing that you want to archive in uncompressed (YUY2) or lightly-compressed "lossless" AVI (HuffYUV, for example) is good, but it might help to more.

About the audio sync issues...

These problems are often caused by the capture card and/or source output. By this I mean your signal is probably too noisy, and your capture card is probably too touchy. The addition of a TBC would definitely work to prevent dropped frames and sync loss on the analog hardware side, and then using a good capture card on the computer/digital side would also help. Most of the cheap capture cards are pretty terrible about good sync.
Note that there is a guide here on preventing audio drift or dropped frames, and some of this may apply. I think you should give it a good read. See http://www.digitalfaq.com/guides/vid...ped-frames.htm

What you've been doing in VirtualDub is mostly a crutch, and not a very good one. As you can see, it has side effects, and in some cases just won't work at all. Reading some of your comments, I think you've gone onto a bit of a tangent, trying to solve this sync problem. It's not as complicated as that.

... Well, I hope that has helped.

[Kereellis]

Hey there, and thanks for this great indepth coverage of my questions. Some of my recent replies which apply to this post as well can be found over at the original videohelp topic, but I'll quote them here for reference.

First I mentioned my main problem with the Radeon graphics card, which is related to my composite output-only V8-videocamera, that I still continue to use (and would also like to rip from) sometimes:

Quote:

Quick note though, if I'm not mistaken the Radeon graphic cards only has a S-video in connector. The problem with that would be that for some more recent camera tapes (which I inconveniently forgot to mention), I'll need to have the option to transfer composite video as well. In this scenario I guess it would make more sense for me to look for a device that supports both types of input instead, so any suggestions would be welcome here.

The following is in response to a user who then mentions the composite video input device that comes with the ATI cards. Both this and your recommendation of these cards somewhat threw me off because I actually own a Radeon 9600 XT/ Asus Vivo, with which I have only booked some disappointing test results thus far:

Quote:

That dongle would be the Asus Vivo right? I happen to have an Asus Radeon 9600 XT installed, but given my experience with it's capture abilities I was somewhat surprised to see it mentioned as a renowned capture device in lordsmurf's post. Admittedly I have only been able to test the Asus Vivo with composite input (as I lack any S-video devices), but I booked significantly worse results with it than with my easycap usb device. Not only it would add a lot of saturation (weird purple tones), but it would also slightly stretch the image horizontally making it look blurry. Well, for the first problem I suppose I could somewhat do some tweaking with the capture settings (if I'd see a reason to favor the Radeon over the usb stick that is), but the second one is pretty unforgiving. Considering these flaws caused by capturing with the Vivo, it seems to me that either using S-video input with ATI MMC doesn't cause these problems, or I'm really better off just resorting to my easycap. By the way, if anyone's interested I could post some screencaps which clearly show the differences.

Then I asked for some more detailed advice on proc amp devices, which by now I've decided I could really use as an inclusion in my setup (especially since I have so much color problems):

Quote:

Another key point for me in deciding on whether I should go for a CTB-100 rather than the TBC-enhancer would be whether I could find a dedicated, affordable proc amp device, since I wouldn't really want to give up the basic proc amp functions the enhancer offers. I've no idea what models to look out for though (or where to look for that matter, ebay only seems to have way too expensive proc amp gear up for auction right now), so any advice on this would be more than welcome. The extra budget I'm willing to spend on a proc amp would be about a 150$, which I hope would be enough for a non-pro, but still pretty decent device.

Since I can see how my budget would be relevant in this, the actual amount I'm willing to spend on a proc amp mostly depends on whether I'll have to buy a new capture device/card or not. An acceptable total for me would be about 500 euros for a good vcr, TBC and a procamp. Alternately I'm willing to spend about 600, for a vcr, tbc, procamp and a different capture device. I should note that other than the sound problems the easycap didn't perform that awfully as far as I could tell, or at least not compared to the Asus Vivo on default settings in vdub (like quoted above, I'm willing to provide the screens of my test results). Of course the easycap doesn't support s-video input, but currently I wonder if that's really such a big deal for this project since I'll only be ripping from mere VHS and V8 tapes (directly from cam with composite output). On a side note, the fact everything is pretty much archived on VHS is pure irony, because we did have a fancy s-vhs camera back then, except my younger self never thought of just keeping the original tapes, or at least copying them to S-VHS, instead of just ditching all that extra quality I could have had as source material :|

[admin]

I need to take some photos of some of the gear, in order to give a really good response to this post. However, it will be several days before I'm able to do that. Will post back soon!

[Kereellis]

That certainly sounds promising. I'll be looking forward to your next response then!

[zaq.hack]

Quote:

Now about capture cards...

I want to specifically nix the Canopus DV converter boxes. For highest available quality, I just don't suggest these. Those boxes are good for lazy consumers, or consumers with touchy computers. The DV compression is just not best for VHS sources, DV is best left as a shooting format in cameras. A lot of the Canopus marketing (and the salesmen at B&H especially) get on my nerves with the fluffy marketing they use, especially as it relates to the "where-is-it?" TBC or the ADVC-300 processing that tends to make an image worse, not better.

If you have a high budget, I would take a strong look at the Matrox series of NLE cards. Then again, those are mostly useful for NLE work in Adobe Premiere CS3 (and CS4, I believe).

Another really good option is the older AGP-slot ATI All In Wonder Radeon cards, especially the latter 9000 series cards (9000, 9200, 9600, 9700, 9800), and using ATI MMC 8.7, 8.8 or 9.02. While these were "consumer" cards, they were marketed to higher-end computer/game/video hobbyists, and had surprisingly professional-quality video processing/capturing abilities. The only downside here is you need a slightly older pre-PCIExpress computer. I purposely keep several AGP slot computers, because these cards are an essential part of my workflow. I actually have some extra ATI AIW AGP cards that are no longer need, and I specifically have a really nice PAL 7500 series card that might interest you. I'm willing to sell.

Beyond that, I'd really need to know more about your budget, and what sort of end-goal you have for your video work. Knowing that you want to archive in uncompressed (YUY2) or lightly-compressed "lossless" AVI (HuffYUV, for example) is good, but it might help to more.

Curiosity. What do these cards do for you? Fewer dropped frames? Was there some sort of additional processing? Better quality analog paths?

The reason I say this is because I once ran into some fine folks trying to do video capturing on a Compaq server with a super-expensive professional capture board (nothing anyone here is likely to care about). They were getting dropped frames all the time. They could not understand how or why. It turns out it was simple PCI bus contention. If you set a capture board to 24-bit RGB, and they had, and you capture at 720x480, and they were, and you want all ~30fps of NTSC, your PCI bus is really busy. Now, those frames go to the RAM/processor and get ... nothing. They wanted perfect captures to store and work on. Okay, great. They were pushing it back down the PCI bus to a RAID controller to be stored. Even running "light" compression and capture as YUY2, it was more than the northbridge chip in the computer could handle.

An AGP capture card sidesteps this problem.

My solution was a CHEAPER PC with built-in storage controller (most of these run on their own PCI bus in the northbridge).

I still have a couple of AGP computers around. If those cards really produce a better capture, I may be willing to try one out. My Osprey has a clunky octopus on the back. So, VCR --> cable --> octopus --> card ... each connection is a potential for noise. I was thinking of changing the setup so that I had quality cable running via solder joints to a custom octopus so that there would only be one impedance "hop." Gold connectors are nice (I like shiny things!), but they are not solder ...

[admin]

The ATI All In Wonder series cards once performed just as good of quality as the Matrox professional NLE cards did (circa 2001-2003), but at far less cost. Matrox dd further advanced their cards in later years, beyond the ATI.

ATI AIW cards were able to capture video with excellent MPEG-2 (DVD compliant and higher quality), MPEG-1 and with various AVI codecs (including YUY2 uncompressed and lesser-compressed AVI like HuffYUV and MJPEG).

The image quality was very sharp, even at 352x480 (unlike a lot of soft-quality 352x480 that exists on certain cards and DVD recorders).

Comparing the AGP to the PCI versions of the same cards, yes, there were definitely some dropped frame errors at times, though mostly on older systems at that time. The CPU would need more juice on the PCI card version -- not an issue as time went by.

The ATI AIW does not have direct input either, it has a purple dongle box about 4 feet long, to stretch and plug into your a/v equipment setup (for input). It also has a squid of wires available for output.

Gold doesn't really mean much. It's actually not as conductive as copper -- just find yourself some nice thick gauge video wires. Sadly, yes, these usually are gold coated. For audio, it's not as big of a deal as it is for video. A really nice single cable (Monster, Philips or Radio Shack) and a generic pair for the audio, works well on a budget. Wires can get ridiculously expensive.

[admin]

You can buy some ATI AIW cards here right now!!!

9000 series 2006 edition AGP card for about $45 shipped - http://www.amazon.com/gp/redirect.ht...reative=390957

7500 series AGP card for about $50 shipped - http://www.amazon.com/gp/redirect.ht...reative=390957

128 PRO series PCI card for about $30 shipped - http://www.amazon.com/gp/redirect.ht...reative=390957
note - 128 PRO requires hacks to use ATI MMC 8.x series, as found on this forum

7200 series PCI card for about $155 shipped - http://www.amazon.com/gp/redirect.ht...reative=390957

9600 series AGP card for about $115 shipped - http://www.amazon.com/gp/redirect.ht...reative=390957

[Kereellis]

Quote:

Originally Posted by lordsmurf

Seriously, go

good VCR
> full separate TBC
> good proc amp
> ATI AIW capture to YUY2 uncompressed AVI
> clean-up in VirtualDub (ColorMill probably a good candidate)
> encode MPEG-2 for DVD in good encoder (MainConcept if $$$, TMPGEnc Plus and wait if only $) see guide here
> author in a non-encoding authorware to make DVD (TAW), see guide here

Be wary of reverse interlace off ColorMill, weird stuff. Had to Restream fix a few times!

Use s-video in 99% of transfers.

Asus above appears to be enhancing the chroma noise, nothing more.

I'd opt for a Dazzle box at Best Buy (the $80 one, not the $50 one), and use AVI captur in Pinnacle Studio 12, before I'd use an EasyCap. I have what appears to be a Tevion OEM of the EasyCap, thing barely works. Still checking facts on that, not sure it's OEM. Not given up on testing yet either. Still not eliminated chair>keyboard interface error as a problem. I know the Dazzle works. It won't clean anything up, but then again neither would an ATI (well, it could if you used ATI MMC 8.x with VideoSoap, which is generally best saved for grain noise -- VirtualDub still removes it better).

Well, all the time I have for this post for now.

Good luck, hope the info I have helps.

Thanks again for your input. I see you recommended the Dazzle as a capture device, but I don't really need the mpeg2 hardware encoding it offers, unless that's the recommended standard for pc-playback. Admittedly I am considering encoding to something like xvid, div-avi or h264 (at highest bitrates), because currently I can't see myself getting the amount of TB's required for archiving the hours of recording I have to do in uncompressed format. Are there any encoding types specifically useful for pc-archiving purposes?

What kind of problems are you experiencing with the easycap? Did the capture with it (vide: earlier post at the videohelp forums) look like it was causing any problems? The advantage in using one for me would be that I already got one. Audio capture needs a work-around though. Rather than using composite, it only works when I connect a line-in cable to the sound card.

The borders on the capture of my asus are significantly smaller than the easycap's. thus the horizontal:vertical ratio should be off right? It's exactly what looks like is going on too, which becomes notable if you download/switch between the images I posted. Also focus on the osd-timing indication, it's easier to tell the problem there than from the video-reproduced image. In any case, I've decided not to stick with this card. It just causes unnecessary and most likely incurable problems from what it looks like.

Do you know of any proc amps that can correct chroma shift like the tbc-enhancer? Moreover would it's ability to only shift RGB-color values collectively be useful? I do remember reading that bleeding mostly occurs on all colors, even though it's mostly the reds that become notable. If that theory is correct wouldn't the basic ('collective') chroma shift correction of the tbc-enhancer sufficient? The point is, I probably don't want to see blues and greens shifted to a different position if only the red is out of place, but I'm not sure if I should be concerned about it.

[admin]

I've been testing the Dazzle DVC-107 (Dazzle Video Creator Plus currently sold new in stores). It does NOT do hardware MPEG encoding, it's purely encoded in software when using the included Pinnacle Studio 12. Is that the software you've been using to capture with, on your Dazzle box?

This Dazzle box is not "bad" necessarily, but it's not what I'd consider one of the best devices for transferring. It's a good solution for folks who insist they buy new in stores, using a current product line -- but that's about it. Quality of a converted VHS will look at best the same as the tape, but more likely worse.

Capturing to streaming video formats is a bad idea, unless you're using one of those $1,000+ dedicated professional streaming encoder boxes. The quality you'll get from software-driven home solutions will just butcher the video, especially interlacing.

I'm not really sure what you mean by "proc amps that can correct chroma shift" -- I've never seen chroma really addressed like that in hardware before. When you say "shift", I imagine that one color is moved on screen. Maybe you mean chroma "adjustment", meaning that you can ..... I don't know. Maybe let's drop the tech terms, because I'm not following you. Are you wanting to make colors richer? (More saturation?). Or less? Or what?

If you really were talking about physical location on screen, that's more of a software fix, often requiring the use of some heavy AVISynth scripting.

[Kereellis]

Sorry, I get the impression you're not familiar with the background to my questions.

If you'd like to assist, maybe you could read some of my posts over here:
http://forum.videohelp.com/topic370277.html#1988124

In the topic it says what equipment I've tried and what I meant by chroma shift (and indeed, it's a matter of color repositioning, the tbc-enhancer can do that, albeit not for seperate rgb-values). Lordsmurf requested that I ask some of my questions here, but I guess the downside is I have to explain certain things twice.

I was mostly asking questions in response to this paragraph of his first post here:

Quote:

When it comes to TBCs, don't worry about the extra features crammed onto the boards. The distribution amps can be decent, but a dedicated distribution amp might work better, depending on the scenario. The one I see most discussed is the proc amp functions, to adjust colors and contrast quality. Honestly, these are so limited in range that it's almost useless. You'd do better to buy a dedicated proc amp, even a $50-75 range vintage 1980s/1990s Vidicraft unit will be more effective for you

Actually, I couldn't find a device going on just this information. In particularly I wondered what kind of proc amp I could use to cure color bleed/ chroma shift if I don't go for the tbc-enhancer. I also would like to know whether Lordsmurf still has the AVT 8710 for sale.

As for the directshow compression, I intended to do it during post-processing, so I hope that won't cost me thousands of dollars. I just wondered how different types of compression (mpeg2, div-avi and directshow types) match up since leaving everything uncompressed is the only thing that's going a bit to far for me (as I doubt it's necessary).

[admin]

No worries, I've read everything.

The only way to negate chroma bleed/shift is by complex specialized AVI Synth scripting. In hardware, your only real option is to desaturate the color with a proc amp.

SignVideo makes a great one: http://www.signvideo.com/single_dual...-processor.htm
Elite Video BVP-4 Plus (BVP4+) is the one I use. Love it.

I'll see if any of my contacts have available TBCs right now. They're always buying/selling gear, depending on projects.

For capturing, I'd use something less-compressed like HuffYUV, MJPEG or MPEG-2 15Mbps I-frame-only --- then convert to something more compressed (and maybe filter along the way!) like MPEG-2 @ 720x480 6-7Mbps bitrate. That's a high bitrate DVD spec. Lots of video, at this spec, will fit on a 1.5TB drive.

You can get Seagate 1.5TB drives for under $135 now, internal or external. Internal SATA about $10-20 cheaper, usually.
Great prices @ http://www.amazon.com/gp/redirect.ht...reative=390957

I have probably 12TB of space. Most of it is working/backup space, but video archives do occupy about 2TB, and there are probably 2,000 videos there (20-40 minutes each, most of them, on average).

[Kereellis]

That's been illuminating, thanks. So I guess mpeg2 isn't a bad way to go after all then.

I'll gladly hear about it if you stumble across a tbc I shall leave the option for notification of replies to this topic enabled just in case.

[Kereellis]

Hello again admin/smurf! If you don't mind, there are a few more tbc-related issues I'd like to inquire about.

Since I've gotten nervous about the project taking so long to kick off I ended up buying a rather pricey tbc-1000 off ebay. However as things stand I'm more than a little concerned about what it's doing to my video. First off though, I'd like to make a few things sure about how to use these devices in general. For the time being, I'm using vdub for my test captures in uncompressed HuffYUV avi format. For this I'd be helped a lot if I could get an answer to the following questions:

1) What options for timing corrections should I leave enabled when capturing video passed through the external tbc? Turning everything off would still result in a slight lag in the audio track during my attempt at capturing a tv recording, but this lag was notable from the start so maybe I just need to leave on audio block insertion? How about audio resampling and frame drop/insertion?

2) This one is kind of related of to the first one: what settings/ hardware setup can I try to make sure the tbc functions properly and outputs a constant signal? I found that when leaving vdubs timing correction options enabled, it would still insert about 6 frames during the start of capture (I haven't run this tape for over 15 minutes, but I didn't notice any more frame inserts after the initial one. I've never suffered the dropping of frames, but that was already the case when I didn't use a tbc). Aside from this, vdub indicates that it still constantly resamples the audio and calculates a minor estimation of vts error (between -10 and 10 ms). Is this most likely due to the clocks in the computer hardware as I'm suspecting or shouldn't this kind of thing happen at all when capturing from a tbc-signal?

3) This is pretty basic, but what's the best way to hook the tbc up to the vcr/pc? In particular I'm not sure if I need to pass the audio through the tbc as well, although that is what I did in the above tests. Furthermore, am I right in assuming the only way to get the most out of the signal output by the vcr is to use S-cables for both the input and output of the tbc?

Now for what's my current main concern. During the tests I also paid specific attention to the video quality of the image that was passed through the datavideo, and sadly the colors apparently come out far worse compared to the tbc-less capture as I hope is evident in the below images (not precisely of the same frame but the issue is most notable in the darks in the background anyway):

> Screen from normal uncompressed capture, without tbc
> Screen from uncompressed capture through the tbc-1000
> SEE NEXT POST

I was prepared for a minor trade-off error like slight blurriness or the hockey stick effect, but I haven't heard about the color range being affected by a tbc to this extent. Now what I personally see wrong with my untrained eye is that the colors don't blend together smoothly but instead are dithered similar to those in an image with a limited color range (kind of as though this were a gif). Am I overlooking an important capture setting in vdub that needs to be adjusted to the output of the tbc or something (like an interlace setting, somewhere?)? If not, would I be well advised to just politely urge the seller to give a refund and hope a better one comes along soon?

[admin]

Attaching images to thread.

Without TBC
tbcfaalgeentbcgeengif_.jpg

With TBC

You must be logged in to view this content; either login or register for the forum. The attached screen shots, before/after images, photos and graphics are created/posted for the benefit of site members. And you are invited to join our digital media community.

[Kereellis]

Ah, thanks for doing that, any ideas as to what the problem is though? It's not normal tbc-behaviour, is it?

I've contacted the seller and his first advice was to use the y/c ports as they would add less noise/grain, so I'll see if I can get my hands on a second s-cable now in any case. However I have doubts that it wil solve this particular issue, so more suggestions would be very welcome. I should add that after playing the video through the tbc to my tv this problem didn't occur (at least seemingly it didn't, my tv has a different callibration of course). I did notice the picture relayed through the tbc was quite a bit darker, possibly higher in contrast (it's harder to tell without being able to make screens), but only if I'd use the tv's freeze function I would notice a similar blending effect (like it has to do with bad deinterlacing. However this happens also upon freezing a frame on the tv without tbc). This kind of strengthened my suspicion that the problem has to do with how vdub handles the output of the tbc, but I can't see how as I'm not filtering the image during capture whatsoever.

If it takes you guys a while to get back at me, I'll try to get some advice from vhelp too if that's alright. I'm just really troubled by this issue

[admin]

In the past 9 or so years that I've used a TBC-1000, I've come to the conclusion that the quality control ranges a bit. Various errors are reported, and often re-reported by others, but almost impossible to duplicate.

These have included

• obvious softening of the picture
• obvious color shifts
• color augmentation (such the "cooling" effect described here: http://tangentsoft.net/video/tbc1000.html)
• noise bars (seen some videohelp.com posts describing these)
• electrical interference

I personally had no noticeable/obvious errors with softening, color shifts (or color differences of any kinds), or any kind of geometry or noise issues. Had I looked hard for something, I'm sure I could have found something. But let's be honest here -- it was being used on aged and damaged VHS tapes! It's hard to see a tiny imperfection from such a source.

I did, however, run into electrical grounding issues during a move in 2006. The new location was new construction, with a 126V average strength, and the TBC just did not like it, even when the TBC was on a UPS with AVR. I was forced to trade in my TBC-1000's for AVT-8710's, which did not have that issue. (Although the AVT-8710's are plagued by other issues -- none of these items is perfect, a perfect TBC does not exist!) The units worked perfectly fine for the buyers, as they had for me pre-move.

The color and softening is easily explained away by the fact that the TBC is digitizing the incoming video, correcting, and sending out an analog output. Anything that is digitized takes a hit in some way. These process in 4:2:2.

I don't really know of a way to test the TBC as you ask. That's honestly more the job of a video engineer, or a broadcast engineer -- the techies that build and repair the hardware -- not an end-user in the creative sector of the field. While I know a good bit about the theory, tech, hardware, maintenance, etc -- my knowledge does dead-end in a number of places. You've just found one.

Your VirtualDub error must be in the computer motherboard or power supply (PSU), capture card, hard drive speed, hard drive defrag level, audio card, or capture software. You seem to have eliminated the signal as the cause of dropped frames (which led to an issue in audio sync and/or dupe frames, as reported).

HuffYUV is a good capture codec. Just wanted to mention that.

Audio lag in TBC-100 is 1 frame, less than a second. Remember, 25fps or 29.97fps. Tiny portion of time. Most audio is off a tiny bit, even on broadcast TV or DVDs, after the best of care and handling. Sound and light simply travel at different speeds. It usually takes 5-6 frames or more to detect a sync loss of any kind.

Chain hardware like this:

• VCR >
• full TBC >
• optional proc amp >
• optional detailer >
• digital capture device

Your images show a color palette compression, or "posterization". A posterized image is usually the fault of the preview or player. Open videos in VirtualDub and copy frames to JPEG that way, to make screen shots. Some players mess up, don't play videos as they exist.

If this is really an error in the video, it appears to be caused by the capture card or capture software. I find it hard to believe the TBC would do this -- this would be a new error that I've never come across. I don't know where to begin. You need to eliminate the computer entirely, run a test.

See how this looks:

• VCR >
• TBC >
• TV

For sure use s-video through and through -- no composites.

[Kereellis]

Quote:

Originally Posted by admin

I don't really know of a way to test the TBC as you ask. That's honestly more the job of a video engineer, or a broadcast engineer -- the techies that build and repair the hardware -- not an end-user in the creative sector of the field.

Most likely I'm thinking on a much simpler level here, but for an end-user, wouldn't it suffice to keep an eye on the fps input shown by vdub, which should constantly be at 25 fps if the tbc is switched on? In any case, because of the trade off errors when using an external tbc (the several possibilities of degradation in image quality), which I definitely experienced (at least with composite > tbc > composite and even moreso with y/c-> tbc-> composite, more on that later), would I be well advised to first try and see how the a/v-sync holds up when capturing directly from a particular tape? I used to be a little cautious about resorting to the timing options in vdub, but seeing how I still need them even with a tbc, I wonder if I shouldn't be fine without the latter in certain cases. For example, a few months back I had ripped another tape which was pretty old as well (albeit not played back too often), and the captured file showed no notable sync errors. I usually leave all options checked except for the 'resample video to match audio' one, since that one sounds more harmful.

Quote:

Originally Posted by admin

Your images show a color palette compression, or "posterization". A posterized image is usually the fault of the preview or player. Open videos in VirtualDub and copy frames to JPEG that way, to make screen shots. Some players mess up, don't play videos as they exist.

While I would like to believe the problem indeed lies with playback, it's pretty unlikely since even during the capturing process itself this posterization effect could be easily seen in the vdub preview screen (it's where it was first noted). Besides that, I don't think that MPC would play two files in an identical format (huffyuv/avi) differently, each time showing a posterized image for the files captured through the tbc. To get 100% certainty though, I just opened one of the files captured through the tbc-1000 in vdub and the effect was still there, so I guess bad playback rendering can't be the reason. As for the test on the tv, I conducted that, and it was harder to tell if the effect was happening on an analogue screen. Image quality was definitely shown to be worse on there though, but mostly due to chroma noise and a higher contrast. After my last post, I started to wonder if I should select a different signal input for the capture device than PAL_B when capturing through the tbc-1000, although I kind of doubt that would help. My other guess would be that the capture device I used (the easycap) needs to be configured to capture from a digitized signal as the tbc outputs it (I suppose?). For a final test I'll probably try, JVC > tbc-1000 > Asus Vivo (not sure if I've tried that) and post the results later.

For now the results won't do much for me though. As I've found out after I got my new profigold S-cables, hooking up the tbc with s-cables would result in a blank image on both tv and pc. Further tests (conducted to eliminate the possibilities of other devices being at fault) showed that the tbc simply doesn't output a signal on any of the 4 y/c ports, although the composite output does work with y/c input. Actually I've just contacted the seller for a refund, so I'm anxiously hoping that will work out right now. Meanwhile, I seem to have little choice but to continue my search for another tbc, so if anything turns up in your network I'd gladly hear about it. I'd definitely more than welcome any offers for hardware from trustworthy sources this time. In addition, would it be wise to spend some bucks on something like a used ES10 as a line-level tbc in addition to a full frame one? The use I'd have for it would simply be to have another option left before having to resort to a full-frame tbc for certain tapes. Then again, I don't know if the dvd-recorder throughput isn't just as likely to degrade the image quality as the full frame tbc in the end. Of course, I might just be a little paranoid now because my first tbc was such a letdown.

Since it looks like I have to postpone the project again anyway, I also consider using the extra time to track down a dazzle box for capture as you suggested, but I wonder how it would theoretically match up to the Radeon 9600XT (non AIW) I already have? Are the colors and detail preserved a little better, or does it have an option like sharpening for example? I know sharpening isn't the same as enhancing detail but I still kind of miss it on the Radeon. Another consideration: someone on videohelp mentioned the FireGL videocards from ATI which are apparently these insanely expensive and professional devices. However, I happen to know of a Dutch auction site where even (slightly used) pro-grade equipment is sometimes put up for an affordable price, so I wondered if there would be any models to look out for, if not to avoid? Would the capture circuits of these cards likely produce better results at all?

Alright, sorry for another long post. I always seem to be full of doubts, but it's kind of the result of bad luck I guess..

[admin]

If you're still getting dropped frames past the TBC, then there's probably something wrong with the computer. At this point, I would refer you to guide on How To Avoid Dropped Frames and Audi Sync Problems.

For quality of video capture, it's hard to beat the ATI All In Wonder AGP Radeon series of cards, without going into the custom NLE cards from the likes of Matrox and some other higher-end pro/broadcast capture cards.

The ES10 posterizes for sure. When you have to use an ES10, however, you're just thankful to get a clean signal -- the compressed color palette "banding" is not a big concern. Present, but the lesser of evils. It's not really a "TBC" necessarily, either.

You could always try an AVT-8710, just in case that TBC-1000 you have is damaged! (And it sounds like it is.)

[Kereellis]

Quote:

Originally Posted by admin

For quality of video capture, it's hard to beat the ATI All In Wonder AGP Radeon series of cards, without going into the custom NLE cards from the likes of Matrox and some other higher-end pro/broadcast capture cards.

I'm looking at a few of these cards on ebay. As someone on videohelp had to point out to me, my Asus Radeon 9600XT isn't quite the same device as the ATI Radeon 9600XT AIW, hence why I currently consider switching to an AIW-product just to see if it performs any better.

I'm looking at a few of these at ebay. It would be nice for me if I could upgrade to a slightly faster card, so I get something out of this deal even if the capture results are the same. One of the cards that seem to match your recommendations would be the ATI X800XT AIW (it's for an AGP-slot as well), which appeared to me as more recent/capable than my 9600XT. However I'm not sure if it would also be a good choice for capturing since you may have excluded it from the recommendations in the first post for a reason? Perhaps I should mention, it's not more pricey in the ebay store where I'm looking at it.

Secondly, this has little to do with hardware (just tell me if I should start a different topic about it), but in the meantime I've been trying out Procoder on one of my earlier avi-captures. There are a number things I couldn't figure out on my own though:

1) What bitrate would be a safe choice? For my first attempt I used a variable bitrate of 12000-15000 kbps. There are quite a few still shots in my homevideos, and I wouldn't like detail to be lost on those, so I figured a high minimal bitrate would theoretically be best.

2) I found it difficult to compare the source avi to the outputted m2p, because I couldn't find a way to play the file back interlaced in mmc, and vdub MPEG2 wouldn't play it back properly (by which I mean the video looks like it underwent a mosaic transformation in vdub, kind of weird). Is there some method I could use to make screens of both files deinterlaced, or do I first need a different decoder since I can't seem to disable it?

3) Predictably, the top or bottom field first options.. I read that lot's of people have problems determining the right setting and sadly I'm no exception. As you may know Gspot doesn't read this information for huffyuv files. For my encoding attempt I went with tff (for both input and output of course) and so far I haven't notice any problems on playback. Then again I can only play back the file deinterlaced as told. If there's any easy way to make sure the right settings from the start (without trial and error) I'd like to hear about it.

[admin]

x800 cards were really just extensions of the 9800 cards, same difference. It was mostly different by requiring latter versions of AGP to run. Some motherboard did not accept latter-generation AGP cards.

I'll have to get back to those other questions a bit later.