digitalFAQ.com Forum

digitalFAQ.com Forum (https://www.digitalfaq.com/forum/)
-   Project Planning, Workflows (https://www.digitalfaq.com/forum/video-workflows/)
-   -   How to evaluate performance of device as TBC? (https://www.digitalfaq.com/forum/video-workflows/11227-how-evaluate-performance.html)

ehbowen 11-27-2020 06:05 AM

How to evaluate performance of device as TBC?
 
Some time back, I picked up a Snell & Wilcox Kudos standards converter (PAL/NTSC) off of eBay. I placed a bid for it since I heard on another forum that it could be used as a TBC. Shortly thereafter I heard that was incorrect, but it was too late to retract the bid and since no one else was interested I got it for the minimum bid ($85).

I've used it a couple of times playing around between PAL and NTSC, but I don't have any really good PAL equipment (just a basic player, no TBC or S-Video). It seems to work well enough in that role, but I'm curious as to how well it might function as a TBC going from NTSC (S-Video) in to NTSC out (S-Video to an All-In-Wonder). I don't have any experience in evaluating TBC function, although I've got a partially functional TBC-4000 to compare it to.

I do have a library of (expendable) pre-recorded tapes, including many Disney ones. What would be a good functional test of this device as a TBC, and if I uploaded short clips of the output would the experts here be willing to help evaluate them?

latreche34 11-27-2020 06:47 AM

I think the first test is Macrovision protection and I don't believe such devices are made to remove it.

ehbowen 11-28-2020 08:02 AM

Test Clips
 
3 Attachment(s)
All right, I've made a couple of quick test captures. The audio in capture chain #1 is a little scratchy, but audio doesn't go through the TBC in either case—I've probably got a noisy connection in one of the cables. The test tape is Universal's Liar Liar, approximately 10 minutes into the tape just after the opening credits. Catalog #82926, ©1997, "Digitally Recorded."

First, here is the test as captured WITHOUT TBC, except for the DNR/line TBC on the source deck (Mitsubishi HS-HD 2000). Capture chain is source deck >> S-Video >> ATI AIW 8500DV running on a Dell tower; capture software VirtualDub 2 on Windows XP: Direct Baseline


For the test captures I took advantage of the fact that the Mitsu has two S-Video outputs and dual audio jacks; both of the following captures were made in the exact same pass. First capture is the same chain as above, except that a Snell & Wilcox Kudos+ CVR600UD operating in the NTSC (525 lines) IN to NTSC (525 lines) OUT mode is inserted as a TBC: Source deck >> S-Video >> S&W CVR600 >> S-Video >> ATI AIW 8500DV with computer and software as above: Snell & Wilcox CVR600UD as TBC

The second capture was made simultaneously using my normal capture chain for the Mitsubishis: Source deck >> S-Video >> TBC-4000 >> ATI AIW x1800 on a homebuilt tower PC (ASUS P5QPL-AM motherboard) running Windows XP and capturing with VirtualDub 2: TBC-4000 Capture

To my untrained eye the two test captures look almost identical. What say you?

Editing Again: Looking at it again I do see a change in brightness on the S&W clip...but I saw the same thing when I was going through the TBC-4000 during setup. I suspect I may have an issue with the capture card itself. I'm going to make a new pair of captures, with the TBC itself as the only variable between them. Catch you later.

ehbowen 11-28-2020 09:35 AM

4 Attachment(s)
All right, I made two more pair of test captures. This time I used the same capture chain and components, with the sole exception of the TBC device. One of each pair uses the Snell & Wilcox, the other uses the TBC-4000. So the capture chain is: Mitsubishi HS-HD2000 >> S-Video >> TBC Device >> S-Video >> ATI AIW X1800 on the homebuilt ASUS P5QPL-AM motherboard tower computer, running Windows XP and VirtualDub 2.

Just to be mean I'm not going to say up front which of each capture was recorded with which TBC device. Who's the best guesser?

The first pair of captures is again from Liar Liar, same tape as above, about 45 minutes into the tape:

Liar Liar Test 2A
Liar Liar Test 2B

The second pair is from the Goodtimes Video release of the original Airport, in EP mode, about an hour and 39 minutes into the tape:

Airport Test A
Airport Test B

latreche34 11-28-2020 11:23 AM

I think the right way of doing this is get an SDI to USB3 adapter or SDI/PCIe card and transfer the raw AVI 4:2:2 to the hard drive, Converting back to analog will always change the luma and chroma levels since 3 conversion happens in your above workflow, A-D, D-A and A-D. With SDI there is only one A-D then TBC/FS and out to SDI. I would like to see another sample for this method.

hodgey 11-28-2020 02:09 PM

Do the tapes have macrovision? If not you can maybe use the captures to evaluate whether they have any impact on video quality, but a stable non-macrovision clip won't tell you much about their ability to handle crappy material, for that you would need something else. Something quick you can test is how they respond to searching, and what they do with a break in the recording on a tape (The signal you get from a SVHS VCR with TBC when there is no signal tends to especially confuse capture devices) . You can also test with the TBC in the deck turned off.

Unless it's the file that's broken, judging by the direct clip it looks like it does at least buffer and re-create the video signal seeing as the clip is 90% green (thinking there is no signal or something?) and the ones with each device do not.

ehbowen 11-28-2020 02:36 PM

4 Attachment(s)
Quote:

Originally Posted by hodgey (Post 73071)
Do the tapes have macrovision? If not you can maybe use the captures to evaluate whether they have any impact on video quality, but a stable non-macrovision clip won't tell you much about their ability to handle crappy material, for that you would need something else. Something quick you can test is how they respond to searching, and what they do with a break in the recording on a tape (The signal you get from a SVHS VCR with TBC when there is no signal tends to especially confuse capture devices) . You can also test with the TBC in the deck turned off.

Unless it's the file that's broken, judging by the direct clip it looks like it does at least buffer and re-create the video signal seeing as the clip is 90% green (thinking there is no signal or something?) and the ones with each device do not.

File isn't broken; it's a Macrovision-protected tape and the blank screen (with just a few lines of video visible at the top) is all I got on the AIW when capturing. (And I did try it with both computers/cards.) So it's clear that the Snell & Wilcox can/does overcome Macrovision. Unfortunately, I don't know of any good (bad!) crappy tapes to test its performance on a lousy VHS signal; most of what I have on the shelf is family tapes and while we never had top of the line equipment we didn't scrape the bottom of the barrel either. I do have one of those EP "cartoon collection" tapes which is non-Macrovision and, at the same time, pretty lousy; either the Mitsus or the AG-1980P will play Tape 1 fairly well but only the Panasonic gives a good output of Tape 2. Think I should give that a try?

Quote:

Originally Posted by ehbowen (Post 73072)
Think I should give that a try?

Well, I did. The tape is the worst one that I own, at least of those that I've tried to capture. It's a StarMaker 100 Greatest Cartoons collection, tape 2 (of 2). Recorded in EP mode and as cheaply as possible; the tape says "Adjust your tracking control for best picture quality." Both the Mitsubishis and the Panasonic (AG-1980) can do a credible job with tape 1, but tape #2 has a kind of tear along the top of the screen which only the Panasonic's line TBC can deal with. As you shall see.

The cartoon is an M-G-M Tex Avery short, "Jerky Turkey". Each clip includes a normal segment, a fast forward, and then a few more normal frames. First up is a direct capture with no TBC; capture chain Mitsubishi HS-HD2000 >> S-Video >> ATI AIW X1800. Just to be cruel I turned off the Mitsubishi's DNR/TBC at the end of the clip: Mitsu Direct Clip

Next is with the Mitsu connected through the TBC-4000 (all connections S-Video). Please note the glitch in the middle is where I fumblefingered the VCR remote going from fast forward to play: Mitsu TBC-4000 Clip

Last with the Mitsu is with the Snell & Wilcox Kudos+ CVR600UD being used as a TBC: Mitsubishi Snell & Wilcox Clip

Here's a look at the same tape in the Panasonic. Much, much better correction in the VCR, although even it has difficulty with the first few seconds after you hit "play". The first one is through the TBC-4000: Panasonic TBC-4000 Clip

Last clip for now has the problem child tape playing through the Snell & Wilcox. Not half bad if I do say so myself: Panasonic S&W Clip

So, overall it looks to me as if the Snell & Wilcox unit does in fact do a serviceable job as a TBC. I'm going to keep it as a backup, or possibly incorporate it into a new capture chain if I get a couple more pieces (such as a decent PAL VCR).

Thanks for any input from the "old hands" here!

lordsmurf 11-28-2020 05:26 PM

A TBC must pass several tests, various timing scenarios, not just the false errors creating by anti-copy (Macrovision is just one anti-copy). However, MV is the easiest test.

- If MV passes, making the transfer fail, then the TBC is junk, not made for consumer sources.
- If MV is defeated/removed, then the TBC can then be fed into subsequent tests.

Realize that tapes that pass the MV test can still fail miserably at other tests, thus fail miserably in real-world usage.

Liar Liar
is a decent test for luma corruption, not just MV.

I can make the S&W box cry and hide in the corner. It has weaknesses that make it ill-suited for a general TBC.

Retail tapes (even the lousy EP "public domain" kind) can only be one test vector, and cannot be overly weighted in testing.

latreche34 11-28-2020 07:09 PM

From the clips it looks as good as the BE75, As I said above it can be used as a capture device as well not just a stand alone TBC.

ehbowen 11-28-2020 10:08 PM

Quote:

Originally Posted by lordsmurf (Post 73078)
I can make the S&W box cry and hide in the corner. It has weaknesses that make it ill-suited for a general TBC.

All right, I get that you don't like the S&W as a general TBC. Stipulated. But have you seen (rhetorical question) what your "approved" general TBCs are going for these days? Right now, in your Marketplace, there is an AVT-8710 offered for sale...at a price higher than I paid for my first three cars, combined!

You yourself have advised people to purchase 'TBC-ish' DVD recorders for pass-throughs if a 'real' TBC is unavailable or out of reach budget-wise. Are you sure it's wise to tell people to turn up their nose at some hardware which seems to show real promise with some problematic tapes? I picked my CVR600 up for less than $100. I found a similar unit advertised for sale...and not on eBay, but from a broadcast supply house...for under $200 just the other night.

LS, you're real generous with your 'do-s' and 'don't-s', but much less so with your 'whys'. This isn't the first time that I've brought this question up, but I have yet to hear from you a really good answer as to how to run this kind of a test and what, exactly, I ought to be looking for. At least for now, Papa, I'm keeping my baby.

lordsmurf 11-28-2020 10:51 PM

Quote:

Originally Posted by ehbowen (Post 73085)
But have you seen (rhetorical question) what your "approved" general TBCs are going for these days?

Yep. Buy it, use it, resell it.

Quote:

Originally Posted by ehbowen (Post 73085)
Right now, in your Marketplace, there is an AVT-8710 offered for sale...at a price higher than I paid for my first three cars, combined!

Really? My first car was almost $3k several decades ago (a new barebones base model was 2x+ that price at the time). Used, but also not a trashed-out clunker. Definitely not new, not with out quirks and issues. The block cracked a few years later.

Quote:

You yourself have advised people to purchase 'TBC-ish' DVD recorders for pass-throughs if a 'real' TBC is unavailable or out of reach budget-wise.
Yep.

Quote:

Are you sure it's wise to tell people to turn up their nose at some hardware which seems to show real promise with some problematic tapes?
Nope, didn't say that. But it does need to be said that the unit has a fail rate, and may required a TBC(ish) arrangement to function without surprises (which you'd only learn about later, post-capture, and only during content playback if not carefully scrubbing the capture).

Most formats converter boxes tend to only engage the TBC on the format conversion (PAL>NTSC), not like (NTSC>NTSC). Tenlab, for example, have a weak TBC that only works on conversion. Same for most Cypress (and al the rebadge/clones). There's others.

Quote:

I picked my CVR600 up for less than $100. I found a similar unit advertised for sale...and not on eBay, but from a broadcast supply house...for under $200 just the other night.
Deals are always awesome. Congrats.
But equally realize that finding deals on video gear in properly working condition is somewhat rare -- and gets more rare with each passing year. I know some people who insist on $X low-dollar budget for an item. Years later, they're still waiting, and still putting off the projects. (Ironically, those same people could have bought the gear that was needed years ago, and resold it later for more than they paid.)

Quote:

LS, you're real generous with your 'do-s' and 'don't-s', but much less so with your 'whys'. This isn't the first time that I've brought this question up, but I have yet to hear from you a really good answer as to how to run this kind of a test and what, exactly, I ought to be looking for. At least for now, Papa, I'm keeping my baby.
A big part of this is due to having the sort of test tapes needed for analysis. That goes back to all the kinds of timing-based errors that exist. I'm spoiled here, I have decades of tapes that have been acquired with certain errors. It's not something that everybody can get, either quickly or at all.

I've been capturing TBC samples for at least a year now, and hope to upload those to the site sometime in 2021. Perhaps that will help. While I can teach a lot online, there are still some issues that are hard to show.

hodgey 11-29-2020 09:55 AM

3 Attachment(s)
You would have to evaluate if it can do what you need of it. The coveted datavideo and cypress/avt TBCs have their drawbacks as well so it's always going to be a trade-off. Personally I use DVRs for most tapes and only pull out my datavideo TBC on very problematic ones as it has some annoying side effects. Even on problem tapes the panasonic DVRs very often work better though and correct horizontal jitter unlike the datavideo ones.

I've found the internal TBC in the JVC decks to be too unreliable for general use as it tends to cause vertical jittering, so I prefer letting the DVRs do the line-tbc job instead most of the time. The TBC in the older panasonic NV-HS1000 doesn't suffer from that so it pairs a bit better with e.g the datavideo. I don't know how good the one in newer panasonics (which use completely differenc ICs) do as I haven't used those.

Multiple A/D -> D/A conversions in the chain which the TBCs (both internal and external) and some other things like internal DNR and some external proc amps do will also have an impact on the final image quality.

Where the DVRs can run into issues is with macrovision as it can end up messing with the AGC in the unit and re-create a macrovision signal on the output which may mess with the capture card. The Sony/Pioneer DVRs have an option in the hidden service menu to disable "sync agc" (in addition to white agc which can be adjusted in normal settings) but I haven't gotten around to testing it with macrovision tapes and they don't handle tapes with noise/issues in the vertical sync area too well. Also need to test what impact it has to lower the input video level, which I do with the panasonic DVR to avoid clipping, has on macrovision tapes. I do know though that macrovision is more of an issue with NTSC, it wasn't as much used on PAL tapes.

Quote:

Originally Posted by lordsmurf (Post 73086)
Most formats converter boxes tend to only engage the TBC on the format conversion (PAL>NTSC), not like (NTSC>NTSC). Tenlab, for example, have a weak TBC that only works on conversion. Same for most Cypress (and al the rebadge/clones). There's others.

I don't think this one is like that, ehbowen posted the manual on the forums a while back. S&W made both a tbc only variant and one with standards conversion as well. They have a bunch of stuff that can be turned off and on, including what to sync the video to, AGC and various noise reduction. Based on the specs I would guess it's based around one of the analog devices chips but I could be wrong.

-- merged --

As for sample tapes, here are 3 I did a while back to illustrate something. You can pretty clearly see the difference between the direct capture and the DVRs in the break in the recording (it's an off-airr recording, so no macrovision). They don't show horizontal jitter correction too well, on a more dodgy tape you would see more difference between the LG and panasonic (ignore the VCR OSD as it doesn't sync up well with the video).

dpalomaki 11-29-2020 10:35 AM

I believe the point here is that what you have is not among the highly recommended TBCs because it has some drawbacks/weaknesses which may emerge with some source material but not other. The proof of the pudding is in the eating, and if it does what you need done (i.e., you are satisfied with the results with your source material) then be happy. If you are building a studio/service to deal with the widest possible variety of source material it likely will not meet your needs.

All TBC's should be able to deal with Macrovision unless specifically documented to respect copy protection. As I understand it honoring copy protection became mandatory for all consumer/prosumer video gear sold in the USA as a result of the Millennium Copyright Act. Prior to that it was hit-or-miss, with most major brands of VCRs and digitizers honoring it.

latreche34 11-29-2020 09:21 PM

2 Attachment(s)
The Snell & Wilcox TBS150D a.k.a Grass Valley CVR400 is also a good known TBC/Capture box for low quality consumer sources such as VHS, I have yet to find one of these to experiment with, Don't know if this one has a version with audio inputs like the TBS190 (CVR600) does.

User Manual.

http://www.digitalfaq.com/forum/atta...1&d=1606706185

http://www.digitalfaq.com/forum/atta...1&d=1606706185

Bogilein 06-27-2021 01:34 PM

Quote:

Originally Posted by lordsmurf (Post 73086)

I've been capturing TBC samples for at least a year now, and hope to upload those to the site sometime in 2021. Perhaps that will help. While I can teach a lot online, there are still some issues that are hard to show.

Half of the year 2021 is over. Is there any hope that anyone will see the samples or is this just another one of the many promises that no one ever gets to see, such as the "Hybrid" Guide or many others?

lordsmurf 06-27-2021 01:47 PM

Quote:

Originally Posted by Bogilein (Post 78331)
Half of the year 2021 is over. Is there any hope that anyone will see the samples or is this just another one of the many promises that no one ever gets to see, such as the "Hybrid" Guide or many others?

Unfortunately, when you live with multiple sclerosis, as I do, desires and plans are often dashed or delayed.

So the testing, photos, and video samples, are still ongoing, not cancelled.

latreche34 06-27-2021 02:46 PM

The moral lesson of this thread is the OP was using the device the wrong way and he decided to ignore my advice, If he used it the way it supposed to be used he would have much better results, Analog outputs on those devices are for monitoring purposes and they are not time base corrected/synchronized because a well calibrated pro monitors are used to monitor the video, The SDI out is time based and synchronized, that's where he should be taking the digital signal from.


All times are GMT -5. The time now is 02:01 AM

Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.