digitalFAQ.com Forum

digitalFAQ.com Forum (https://www.digitalfaq.com/forum/)
-   Capture, Record, Transfer (https://www.digitalfaq.com/forum/video-capture/)
-   -   Sanity-check my VHS home movie digitizing workflow? (https://www.digitalfaq.com/forum/video-capture/15573-sanity-check-vhs.html)

matt314159 02-23-2026 09:49 AM

Hi folks! I think I've done enough reaching and research, and I ordered the equipment this weekend, but I want to sanity-check this setup and workflow while there's still time to make adjustments.

Background:
My folks are sending me all of our family's VHS home movies. Somewhere between 30 and 50 tapes, but I haven't gotten a count yet. They were recorded on a Sears LXI shoulder-style VHS camcorder (manufactured by Hitachi) in the late 1980s and early 1990s. They're all first-generation camcorder recordings that were recorded on new fresh tapes, mostly in SP mode. They've been stored in my parents bedroom closet in the High Desert of Southern California, so I think the dark, climate-controlled low-humidity conditions probably work in our favor. They were pulled out and played maybe a dozen times at most over the years. So I'm not really expecting too many problems with the tapes.

Equipment
Here's my planned equipment chain: JVC HR-S5400U > Panasonic DMR-ES15 > IO-DATA GV-USB2 via S-Video.

I'll be using my home desktop PC for this which is an HP ProDesk 400 G5 with i5-8500, 32GB RAM, 1TB NVMe SSD main system disk, an internal 8TB WD Red Spinning metal HDD, and an external USB3 enclosure housing a 12TB Seagate enterprise HDD.

Workflow
I plan to ingest the footage using VirtualDub to interlaced masters with HuffYUV onto the 8TB internal HDD, then bring them over to Hybrid for deinterlacing and save the "finished masters" as H.264 MP4s with a CRF of 17 onto the external Seagate.

Am I thinking about this the right way? At first I was planning to just get a used VCR-DVDR combo unit and dub them but that felt like I'd be leaving too much quality on the table going to MPEG2 at a low bitrate over what I think is probably a composite internal connection.

But I'm on a fairly tight budget, so getting a VCR with built-in TBC was beyond what I was looking to spend. All in, this took about $300 for me to get the VCR, the ES15 for line TBC passthrough functionality, and the GV-USB2 and two sets of Markertek S-Video+LR Audio dubbing cables.

I'm hoping this is a budget setup that made the right intentional compromises.

Have I gone astray anywhere? Is it wise to just leave the ES15 in line for all the tapes, or should I only use it on problem tapes? Anything else I've not thought about that I should consider?

-- merged --

Actually I just realized this should probably go in Project Planning, Workflows. My bad, if a mod wants to move this, please do!

BarryTheCrab 02-23-2026 02:22 PM

Your JVC 5400 is quite old and although it has video calibration and stabilizer it lacks picture control and as noted no TBC.
A newer JVC and even a second VCR would help if you run into problems. How many tapes?

matt314159 02-23-2026 02:34 PM

Quote:

Originally Posted by BarryTheCrab (Post 106374)
Your JVC 5400 is quite old and although it has video calibration and stabilizer it lacks picture control and as noted no TBC.
A newer JVC and even a second VCR would help if you run into problems. How many tapes?

It's somewhere around 30-40 tapes. The Panasonic ES15 would be stepping in to give sort of a poor-man's TBC, and the GV-USB2 is supposed to be pretty forgiving. Would picture control be necessary for ingesting footage? I would think I'd want it to be as transparent as possible.

I am keeping my eyes out for something newer like a 4800U if a good deal pops up on one. But man feels like the market for these has gone insane, and even that one still doesn't have TBC built in.

billct97 02-23-2026 08:57 PM

Hi Matt, there are probably 100 things you might run into getting this going but I'll offer up the first hurdle you have to get over before anything else matters. Without a good TBC (in the VCR or external) you are likely (but not guaranteed) to get Frames dropped or Frames inserted. VirtualDub will display that on the right side of the GUI during capture. If you do get a high number of dropped or inserted frames you need to determine why and fix it or you will have disappointing results like stuttering video and out of sync audio. Other reasons that might (or might not) cause this... not using Windows XP, not using VirtualDub 1.9.11, listening to the audio while capturing, displaying the histogram while capturing. Want to know how I know this? Yep, had problems with every one of these things. I've also had success with a Hauppauge card and Windows 10 so this isn't a hard and fast list. Just one guys experience.

30-50 home videos? LOL, I like to digitize broadcast TV (commercials, newsbreaks, unique shows, etc..) from the 80's and up. I was lucky to find a couple on Facebook Marketplace recently who were cleaning house. They didn't want to dump their VHS collection in the trash and were thrilled to give them to me. Over 200 tapes from 1982 and up. Then came the favor, could you convert our wedding video? How could I refuse? Home video from 1990 can be best described as performed by someone who thinks continuously sweeping the room is entertaining. Lamps in a dark room at full brightness and a wedding in a room with huge windows on a gray and rainy day. I'll take 100 TV broadcast videotapes recorded at SLP on a noisy cable TV over 1 home video any day.

You might want to save yourself some headache and money and have a professional place convert them for you. Maybe try one tape first to see what kind of quality you get. Or dive into this fun hobby!

matt314159 02-23-2026 09:35 PM

Quote:

Originally Posted by billct97 (Post 106377)
Hi Matt, there are probably 100 things you might run into getting this going but I'll offer up the first hurdle you have to get over before anything else matters. Without a good TBC (in the VCR or external) you are likely (but not guaranteed) to get Frames dropped or Frames inserted. VirtualDub will display that on the right side of the GUI during capture. If you do get a high number of dropped or inserted frames you need to determine why and fix it or you will have disappointing results like stuttering video and out of sync audio. Other reasons that might (or might not) cause this... not using Windows XP, not using VirtualDub 1.9.11, listening to the audio while capturing, displaying the histogram while capturing. Want to know how I know this? Yep, had problems with every one of these things. I've also had success with a Hauppauge card and Windows 10 so this isn't a hard and fast list. Just one guys experience.

30-50 home videos? LOL, I like to digitize broadcast TV (commercials, newsbreaks, unique shows, etc..) from the 80's and up. I was lucky to find a couple on Facebook Marketplace recently who were cleaning house. They didn't want to dump their VHS collection in the trash and were thrilled to give them to me. Over 200 tapes from 1982 and up. Then came the favor, could you convert our wedding video? How could I refuse? Home video from 1990 can be best described as performed by someone who thinks continuously sweeping the room is entertaining. Lamps in a dark room at full brightness and a wedding in a room with huge windows on a gray and rainy day. I'll take 100 TV broadcast videotapes recorded at SLP on a noisy cable TV over 1 home video any day.

You might want to save yourself some headache and money and have a professional place convert them for you. Maybe try one tape first to see what kind of quality you get. Or dive into this fun hobby!

Ha thanks, yeah I've always been a techy tinkerer and into vintage technology no less (There's a Commodore 64 that sits behind me at my desk at work--the VERY same C64 I played on at my grandma's house as a kid--she gave it to me in 2016 after saving it all these years, and I restored it to working condition).

Anyway as I was going through my options, I decided that I was up for a fun and challenging project. Doing it on a shoestring budget made me make some intentional compromises. Yes, the deck I bought doesn't have internal TBC, but running it through the ES15 is going to function as that poor-man's line TBC (or at least that's the plan). Based on the reading I've done so far, it's supposed to do a good job at cleaning up the wobbly timing (albeit at the expense of doing an A>D then D>A conversion in the middle of the signal stream that at least in theory I don't like.) And I picked the GV-USB2 because that is supposed to be way more forgiving of timing issues in the first place. I'll find out soon enough once I get the hardware hooked up and do some test runs and see what kind of dropped frames I see. Helping my learning curve a bit is that I do have some experience with VirtualDub, even if it was way back in 2008. At least that basic interface isn't foreign to me, and I know what a lot of the settings do.

Around 2010, my late granddad used a DVDR-VCR combo deck to dub all their VHS home movies to DVD in low bitrate MPEG2 and the whole family was thrilled with the result, so the standard for them is pretty low and I will be my own harshest critic. But I'm pretty sure I can get an end result that is better than that dubbing process grandpa did, and honestly probably better than 80% of the places out there that do these transfers on the same type of equipment. I guess my goal would be to approximate what you guys do but at a rather basic entry point. But at least it's an SVHS player with S-Video out, through a device serving as a line TBC, then captured into interlaced lossless HuffYUV masters and deinterlaced well via software.

It might be optimistic, but I'm hoping I can get to a point where I ingest one or two tapes per evening after work, then queue up the deinterlacing and H.264 renders overnight while I sleep.

I think in the end, this will probably save money too, since I'm only about $300 deep into this project, and I should be able to sell the gear once I'm done with it to recoup some of that cost.

I'll be sure to check back in on this thread after a month to see if I eat these words.

aramkolt 02-23-2026 09:42 PM

Had to mention since a wedding tape was mentioned, but I actually just converted a semiprofessional wedding tape and was quite shocked to see color bars at the beginning of the tape so you could calibrate luma/chroma levels of any future dubs. This was standard practice for U-Matic, but I hadn't seen anyone do it on VHS. I'm not entirely sure it was utilized correctly as whites later in the tape exceed the white values of the bars (also not super uncommon to see on U-Matic either), but it could be that the bars were 75% white and not 100% maybe.

Agree that a more modern JVC machine could possibly yield better results, but since you've already got it, might as well see what you think of the quality before looking to upgrade. Main thing a newer one might do better with is noise, but if you're using "edit" mode (limited or no digital processing), it might be difficult to tell them apart - maybe. ES15 probably will do a little more softening of the image than say an internal TBC, but might stabilize some tapes better.

matt314159 02-23-2026 09:57 PM

Quote:

Originally Posted by aramkolt (Post 106380)
Agree that a more modern JVC machine could possibly yield better results, but since you've already got it, might as well see what you think of the quality before looking to upgrade. Main thing a newer one might do better with is noise, but if you're using "edit" mode (limited or no digital processing), it might be difficult to tell them apart - maybe. ES15 probably will do a little more softening of the image than say an internal TBC, but might stabilize some tapes better.

Was there a big jump in the VCR tech in the late 1990s? As far as I know the S5400U was 1997/1998 and the S3800U/S4800U's were 2000-2001 so they don't seem that far off from each other to me. Both have 19 micron heads, auto adjustment, Hi-Fi Stereo, S-Video out, and I think my gut would probably opt for a more transparent playback so DNR might not be super important (I can't say for certain though).

I do have a snipe queued up to execute in a couple of days, there's an S3800U with a mistitled listing over on eBay so I may be able to snag it cheaper than the market price. That would give me a backup as well as the ability to A/B test to see which one I prefer.

billct97 02-23-2026 10:04 PM

One more suggestion for the night since you mentioned deinterlacing and H.264. I capture with Virtualdub 1.9.11 on XP (old hardware, low CPU and RAM is fine) and move the files to a newer W11 machine with a good quality color calibrated monitor. For occasional YouTube uploads (@video-airchex) I deinterlace in Avisynth/Virtualdub (old habit) and convert to x.264 in Hybrid. Hybrid is only too happy to push all 8 virtual CPU cores to the max. If you are only going to view the files on a PC with something like VLC you can leave them interlaced (menu pick to deinterlace). Also remember if you capture 720x480 you might want to resize to 640x480 or even upscale to a multiple of that.

lordsmurf 02-24-2026 12:37 AM

Welcome. :)

Get
- a non-TBC JVC S-VHS, like the HR-S290x, 590x, 3800, 4800.
- a better capture card, GV-USB2 is largely meme'd junk, the "least worst" card from Amazon, lots of problems are commonly had with it

Low budgets are fine, but don't go too low. Video gear are tools, not a cheeseburger. It does have costs, generally ~$500 range minimum. Better tools jump into the $Ks.

The "buy it, use it, sell it" is also why you should just pony up some bucks for good gear. Quality gear has value, junk is your forever (or has low/no resell value).

3800 is fine. But you're not buying on eBay, you're gambling. Just realize that. (Tip: Open that VCR before sticking your tapes into it.)

Aya_Rei 02-24-2026 02:02 AM

That or use some mass produced commercial tape, ie a hollywood movie.

As I like to say, use a sacrificial tape.

matt314159 02-24-2026 11:21 AM

Quote:

Originally Posted by lordsmurf (Post 106383)
Welcome. :)

Get
- a non-TBC JVC S-VHS, like the HR-S290x, 590x, 3800, 4800.
- a better capture card, GV-USB2 is largely meme'd junk, the "least worst" card from Amazon, lots of problems are commonly had with it

Low budgets are fine, but don't go too low. Video gear are tools, not a cheeseburger. It does have costs, generally ~$500 range minimum. Better tools jump into the $Ks.

The "buy it, use it, sell it" is also why you should just pony up some bucks for good gear. Quality gear has value, junk is your forever (or has low/no resell value).

3800 is fine. But you're not buying on eBay, you're gambling. Just realize that. (Tip: Open that VCR before sticking your tapes into it.)

I really wish I'd found this forum first instead of reddit. They steered me toward the GV-USB2 since it's more forgiving of timing errors than the blackmagic Intensity Pro (non4K) PCIe card I also have access to (I work in the IT dept at a small college so I occasionally have access to surplus gear like that). It sounded like even with the ES15 that the blackmagic could still be cranky with analog ingestion. Since it was priced reasonably I went for it. Should I roll the dice with the blackmagic instead, or is there a different card that lives at a sweet spot for this kind of equipment chain with my goal being huffyuv lossless interlaced masters?

I'm still within the return window of the GV-USB2 so that's not a big deal if I need to return it, just need to find something superior to replace it with.

Is the HR-S5400U really so old that it would be deemed unsuitable for a project like this? I didn't think 1997/1998 was too far off from the newer 2000/2001 models. At least on paper, it looks like it ticks all the boxes for the features that matter.

Quote:

Originally Posted by Aya_Rei (Post 106384)
That or use some mass produced commercial tape, ie a hollywood movie.

As I like to say, use a sacrificial tape.

Yeah I'll pick up a couple of sacrificial tapes at the thrift store before I trust it with anything important.

RayNotes 02-24-2026 11:31 AM

Your workflow sounds like it make your audience (ordinary folks - not broadcast professionals) very happy. I haven't used the IO Data GV USB2 but the reviews I've seen place it at the top of cheap heap. Sometimes good enough is good enough, ya know?

I'm using a JVC S2901U -> DMR-ES15 -> RetroTINK 2X Pro in passthrough mode as the ADC -> Magewell USB Gen2 to capture via VirtualDub2. Your IO Data replaces both the 2X Pro and Magewell in this process. My opinion is that should be fine for your home project.

The ES15 absolutely will stabilize your video enough for good captures. Make sure you look-up the optimal settings first - there are some black level settings to get right and I think most people shut off the DNR.

IMPORTANT:
DO NOT turn on the VCR's video stabilizer. On a CRT it might have passed for "good" but in capture mode it just cuts the vertical resolution in half to "stabilize" the video and it looks dreadful.
DO turn on the Video Calibration. This appears to be mostly auto-tracking but it works. Use it
DO switch the picture mode to "EDIT" if you have that feature. This shuts off all the VCR's video smoothing and noise reduction that (again) probably looked great on a CRT but makes capture look too soft and mushy.

You might look into VirtualDub2 and the FFV1 codec (instead of HUFF) just to make sure your archival captures are what you want. Maybe the results are the same but it's worth a Google search.

Deinterlacing is where I spent the most time educating myself until I finally got versions of my vintage home movies that preserved the original aesthetic of NTSC video. I used StaxRip which wraps a lot of FFMEG processes with a reasonably easy-to-use GUI. It allows you to tweak the QTGMC deinterlacer to match your content and can output in virtually any codec. QTGMC also does a lot of heavy lifting in noise reduction with a lot of settings to get things right for your particular content.

Good luck and enjoy the process. If you don't enjoy the process, send it out and pay to have it done.

Quote:

Originally Posted by matt314159 (Post 106394)
Is the HR-S5400U really so old that it would be deemed unsuitable for a project like this? I didn't think 1997/1998 was too far off from the newer 2000/2001 models. At least on paper, it looks like it ticks all the boxes for the features that matter.

What I've found with models of that era is the video tends to bob up and down occasionally and you're right, the ES15 can't keep up with that. The HR-x901U series that came out in early 2000s seemed to have corrected this issue. BUT, they have notoriously crappy audio for EP mode tapes. If your tapes are from a camcorder, and not copies of that copy, they will be SP mode and sound much better.

But like LS said - if this isn't going to be a hobby of your moving forward then buy even better gear. Use it. Then sell it.

matt314159 02-24-2026 11:44 AM

Quote:

Originally Posted by RayNotes (Post 106398)
What I've found with models of that era is the video tends to bob up and down occasionally and you're right, the ES15 can't keep up with that. The HR-x901U series that came out in early 2000s seemed to have corrected this issue. BUT, they have notoriously crappy audio for EP mode tapes. If your tapes are from a camcorder, and not copies of that copy, they will be SP mode and sound much better.

But like LS said - if this isn't going to be a hobby of your moving forward then buy even better gear. Use it. Then sell it.


My income and budget is so tight that I really don't feel comfortable laying down $1,000+ on a really good VHS deck even if it's only for a couple of months before I sell it, because to do that I'd need to dip into credit card debt and I can't stomach the thought.

My thinking was/is if I could get a "pretty good" setup for $300 or so, I'd try to do that. I think I'll hook all the gear up when the rest of it arrives on Thursday, and see what a couple of test captures look like on sacrificial thrift store tapes and re-evaluate where I'm at with things. I'm kicking myself for spending $142 on the S5400U. I HAD looked at the lists, but in the moment got my wires crossed and got excited that I finally saw a unit that ticked all my boxes and actually, I incorrectly assumed that 5xxx series was higher than 3xxx and 4xxx so I hit buy-it-now at $120 + $22 shipping. I'm still hoping that it'll be "good enough" (and that it survives the shipping journey, and doesn't require too much work to get it ship-shape) but I'm definitely feeling pangs of regret.

And yes these are all camcorder masters/originals so I think they will be SP mode. Can't confirm yet as mom still hasn't dropped them in the mail.

RayNotes 02-24-2026 02:06 PM

My guess is if the S5400U shows up in good working condition you will be fine. No, they're not heavy hitters but if it's in good shape, your project will go smoothly. There are so many nonsensical model numbers to keep track of it's easy to get confused. Patience is the key. If the S5400 displays a lot of jitter, it will likely frustrate you. But maybe it's in good shape - and the money was well spent... fingers crossed.

The rest of your workflow looks fine to me - for the performance/expectation level you're describing. Camcorder originals (in my experience) tend to look pretty good if the tapes were stored in a reasonable environment.

For now - your plan sounds good: Get some test captures and evaluate the results. My suggestions: Look into that FLV1 codec over the Huff thing. Don't use the proc amp or other filters on input - capture a clean signal the correct things in post using software as you knock out the delivery files.

matt314159 02-24-2026 02:27 PM

Quote:

Originally Posted by RayNotes (Post 106402)
My guess is if the S5400U shows up in good working condition you will be fine. No, they're not heavy hitters but if it's in good shape, your project will go smoothly. There are so many nonsensical model numbers to keep track of it's easy to get confused. Patience is the key. If the S5400 displays a lot of jitter, it will likely frustrate you. But maybe it's in good shape - and the money was well spent... fingers crossed.

The rest of your workflow looks fine to me - for the performance/expectation level you're describing. Camcorder originals (in my experience) tend to look pretty good if the tapes were stored in a reasonable environment.

For now - your plan sounds good: Get some test captures and evaluate the results. My suggestions: Look into that FLV1 codec over the Huff thing. Don't use the proc amp or other filters on input - capture a clean signal the correct things in post using software as you knock out the delivery files.

This sounds solid. I'll look into FLV1 vs HUff. I was under the impression that Huffyuv was better for importing and editing later but if I can save a bit of space and still easily deinterlace, etc, then all the better.

BarryTheCrab 02-24-2026 02:32 PM

Do not self-kick! That model 5400 if it works well may be ok for you.
If it has trouble with some tapes then get a better unit and the 5400 becomes a decent 2nd machine.

matt314159 02-24-2026 09:21 PM

One thing I did decide today is that I definitely want to try out the Intensity Pro (non4k) PCIe card. My PC I was planning to do this with is SFF so the full-height card won't work but I nabbed an HP ProDesk 600 G5 micro-tower from work that's going on the recycling pile this summer. It's a 7th gen i5-7500, I goosed it to 16GB RAM because we had a lot of DDR4 laying around, with a 128GB SATA SSD for the primary system disk, and I've got a 5TB WD Red I plan to stick in it.

I'm curious to see if I can get through at least the bulk of the recordings using the blackmagic card with the VCR and ES15. If I can get by with minimal dropped frames, I think this is my preferred route, falling back to the GV-USB2 only if the BM card gags on a bad signal.

This idea may crash and burn, but I kind of like the idea that my main PC doesn't have to be the one tied up with this project since it also happens to be my plex server and channels DVR server. Now it will have its own dedicated project PC running Win10 Pro and with nothing else installed but the software for this digitization project.

RayNotes 02-24-2026 09:43 PM

Quote:

Originally Posted by matt314159 (Post 106407)
One thing I did decide today is that I definitely want to try out the Intensity Pro (non4k) PCIe card. My PC I was planning to do this with is SFF so the full-height card won't work but I nabbed an HP ProDesk 600 G5 micro-tower from work that's going on the recycling pile this summer. It's a 7th gen i5-7500, I goosed it to 16GB RAM because we had a lot of DDR4 laying around, with a 128GB SATA SSD for the primary system disk, and I've got a 5TB WD Red I plan to stick in it.

My guess is that the HP ProDesk will get the job done, easily - as far as getting a clean capture. I've captured NTSC video on much wimpier PCs without dropping frames. Of course I've never seen a dropped frame in my life - from an in-tact tapes, and I take their existence on hearsay.

lordsmurf 02-25-2026 02:36 AM

Quote:

Originally Posted by matt314159 (Post 106394)
I really wish I'd found this forum first instead of reddit. They steered me toward the GV-USB2

Yeah, Reddit skews young and cheap and broke, and some subs have members that pride themselves on spending the least money possible, even if it's a terrible idea. GV-USB2 is just cheap, though not the cheapest/worst out there.

Quote:

since it's more forgiving of timing errors than the blackmagic Intensity Pro (
That's not saying much. Does a cat turd stink less than a dog turd? :laugh:

Quote:

It sounded like even with the ES15 that the blackmagic could still be cranky with analog ingestion.
Correct.

Quote:

Should I roll the dice with the blackmagic instead,
No.

Quote:

or is there a different card that lives at a sweet spot for this kind of equipment chain with my goal being huffyuv lossless interlaced masters?
Yes.
... and that is precisely why I have cards in the marketplace subforum here, at/near original price (no real markup). Capture cards should be the easiest aspect of capturing video, and that then leaves VCR and TBC to figure out. But the capture card market is flooded with junk, mostly Chinese, hence my market place listings. I try to make this easier for people like yourself.

Quote:

I'm still within the return window of the GV-USB2 so that's not a big deal if I need to return it, just need to find something superior to replace it with.
I would.

Quote:

Is the HR-S5400U really so old that it would be deemed unsuitable for a project like this?
I didn't think 1997/1998 was too far off
I'm not sure where you're getting that date. The 5400 was more like 1994. The second number, in a model number, had correlation to the date. Not always, but usually. Nothing in 1997 (NTSC), and 1998 was the 3800, 4800, 7800, 9800.

It's not useless, but it's not going to do you any favors. JVC had 3 effective generations of gear, and the x400 was the last models from the 1st/worst generation. The 2nd/3rd is the sweet spot.

Quote:

Originally Posted by RayNotes (Post 106398)
I'm using a JVC S2901U -> DMR-ES15 -> RetroTINK 2X Pro in passthrough mode as the ADC -> Magewell USB Gen2 to capture via VirtualDub2. Your IO Data replaces both the 2X Pro and Magewell in this process. My opinion is that should be fine for your home project.
The ES15 absolutely will stabilize your video enough for good captures. Make sure you look-up the optimal settings first - there are some black level settings to get right and I think most people shut off the DNR.

I don't agree with the capture card, but you get results that aren't bad. That 2901 and ES15 are definitely good, for what they are.

Quote:

You might look into VirtualDub2 and the FFV1 codec (instead of HUFF) just to make sure your archival captures are what you want.
FFV1 has really bad support in video software (especially NLEs like Premiere/FCP), and the codec was designed with HD progressive in mind. The SD interlace handling is lousy. Huffyuv has been here for 25+ years, and it's not going anywhere at this point. All other lossless SD codecs have fallen into obscurity over the years. Laragith has byte bugs, Ut has too many versions, Magic is ignored. Huffyuv "just works", no friction.

Quote:

Originally Posted by matt314159 (Post 106399)
My income and budget is so tight that I really don't feel comfortable laying down $1,000+ on a really good VHS deck even if it's only for a couple of months before I sell it, because to do that I'd need to dip into credit card debt and I can't stomach the thought.

Understood. We have to start high, but we can take shortcuts to being down gear costs, although it does come at the cost of quality.

Quote:

I'm kicking myself for spending $142 on the S5400U. I HAD looked at the lists, but in the moment got my wires crossed and got excited that I finally saw a unit that ticked all my boxes and actually, I incorrectly assumed that 5xxx series was higher than 3xxx and 4xxx so I hit buy-it-now at $120 + $22 shipping. I'm still hoping that it'll be "good enough" (and that it survives the shipping journey, and doesn't require too much work to get it ship-shape) but I'm definitely feeling pangs of regret.
Yes, model numbers, and serial numbers, can be very deceptive. The 5400 for $142 wasn't a deal of any kind, but it could have been worse. Let's at least test it now. Was this eBay? Almost noting good comes from that site, in terms of video gear. But it can be returned, with hassles, if needed.

Quote:

And yes these are all camcorder masters/originals so I think they will be SP mode. Can't confirm yet as mom still hasn't dropped them in the mail.
Be very sure that she puts ample padding around the tapes, and double-boxing is suggested. Sturdy boxes, not reused Amazon boxes, not USPS Priority Mail boxes.

Quote:

Originally Posted by RayNotes (Post 106402)
If the S5400 displays a lot of jitter, it will likely frustrate you.

But too much is a problem.

Quote:

Don't use the proc amp or other filters on input - capture a clean signal the correct things in post using software as you knock out the delivery files.
Good tip.

Quote:

Originally Posted by matt314159 (Post 106403)
This sounds solid. I'll look into FLV1 vs HUff. I was under the impression that Huffyuv was better for importing and editing later but if I can save a bit of space and still easily deinterlace, etc, then all the better.

Huffyuv is better,

Quote:

Originally Posted by BarryTheCrab (Post 106404)
Do not self-kick! That model 5400 if it works well may be ok for you.
If it has trouble with some tapes then get a better unit and the 5400 becomes a decent 2nd machine.

Or this. That's why I have certain VCRs. Although, to be honest, all it's done is made me a gear hoarder, I rarely use some of those things.

Quote:

Originally Posted by matt314159 (Post 106407)
One thing I did decide today is that I definitely want to try out the Intensity Pro (non4k) PCIe card.

You'll have to carefully scrub your capture. Errors are common. Never just start capturing, walk away, do a cursory quick look at the captures, and assume it's all fine. You'll see the flaws later, sometimes years later (and sometimes when it's too late).

Quote:

I'm curious to see if I can get through at least the bulk of the recordings using the blackmagic card with the VCR and ES15. If I can get by with minimal dropped frames,
Dropped frames almost always caused random audio drift, in addition to the "black frames" issues specific to Blackmagic. You're adding friction to the process. You already have enough problems with a non-TBC frame sync in the DVD recorder, and a not-suggest VCR. Why compound that with a 3rd problem?

Quote:

Originally Posted by RayNotes (Post 106408)
Of course I've never seen a dropped frame in my life - from an in-tact tapes, and I take their existence on hearsay.

There reason here is simple -- they're baked in. You're using the ES15 + RetroTINK, which effectively hides dropped frames from you. Not removed, just hidden, not counted. GV-USB2 + AmaRecTV does this as well, being a mediocre card with a analog-era streaming software (not analog videotape capture software). You've somewhat abdicated your ability to control dropped frame monitoring, and just trust that ES15+RetroTINK is doing a good enough job not to be noticed. It handles audio sync decently, usually, but visual dupe/stutter is still noticeable.

matt314159 02-25-2026 08:59 AM

Thanks lordsmurf! Since the Intensity Pro is effectively free for me, as was the capture rig I put together with spare parts from work, I think I lean toward trying this first. And yeah I bought the S5400U on eBay so we will see what I get when it arrives tomorrow.

Among the capture cards you currently have for sale, what would you recommend for this workflow I'm putting together? That'll probably be my backup plan if this crashes and burns.

lordsmurf 02-25-2026 09:47 AM

Yep, free is always worth trying. But from experience, it rarely works. Same for cheap items.

Given lack of frame TBC, and valid concerns over dropped frames, you really want the specific Pinnacle I have. It's a resilient card, more than ATI, more than GV-USB2, vastly more than Blackmagic. When you cut corners on TBCs, the capture card quality becomes even more important.

RayNotes 02-25-2026 04:24 PM

Quote:

Originally Posted by lordsmurf (Post 106411)
I don't agree with the capture card, but you get results that aren't bad. That 2901 and ES15 are definitely good, for what they are.


FFV1 has really bad support in video software (especially NLEs like Premiere/FCP), and the codec was designed with HD progressive in mind. The SD interlace handling is lousy. Huffyuv has been here for 25+ years, and it's not going anywhere at this point. All other lossless SD codecs have fallen into obscurity over the years. Laragith has byte bugs, Ut has too many versions, Magic is ignored. Huffyuv "just works", no friction.

Dropped frames almost always caused random audio drift, in addition to the "black frames" issues specific to Blackmagic. You're adding friction to the process. You already have enough problems with a non-TBC frame sync in the DVD recorder, and a not-suggest VCR. Why compound that with a 3rd problem?


There reason here is simple -- they're baked in. You're using the ES15 + RetroTINK, which effectively hides dropped frames from you. Not removed, just hidden, not counted. GV-USB2 + AmaRecTV does this as well, being a mediocre card with a analog-era streaming software (not analog videotape capture software). You've somewhat abdicated your ability to control dropped frame monitoring, and just trust that ES15+RetroTINK is doing a good enough job not to be noticed. It handles audio sync decently, usually, but visual dupe/stutter is still noticeable.

OK - cool, I'll give the HUFF codec another look. I tend to go over my content reasonably well before posting it, but perhaps things are slipping through I'm not noticing. For my vintage TV commercial YouTube channel, the results are probably above and beyond expectation. I got the HR-S2901U for $14 at a flea market - looked like it was a part of some boardroom VCR as Eaton corp, that hardly got used. That was an extremely rare find. I just don't like the audio from EP mode.

Thanks for all the insight and help. It's appreciated.

matt314159 02-25-2026 07:15 PM

1 Attachment(s)
Quote:

Originally Posted by lordsmurf (Post 106411)
I'm not sure where you're getting that date. The 5400 was more like 1994. The second number, in a model number, had correlation to the date. Not always, but usually. Nothing in 1997 (NTSC), and 1998 was the 3800, 4800, 7800, 9800.

It's not useless, but it's not going to do you any favors. JVC had 3 effective generations of gear, and the x400 was the last models from the 1st/worst generation. The 2nd/3rd is the sweet spot.

Oh and this one is on me for trusting the stupid AI answer results at the top of a quick google search before I bought it. I know better and still fell for garbage AI answers anyway.

The 5400 arrived today, by the way. Since it's slightly below freezing in Iowa, I'm going to just let it set overnight before I open it up tomorrow morning and check it out, just to let the VCR normalize temperature and avoid any condensation risk.

I'm optimistic, the box is large and good quality and it looks like there's probably at least 3-4" of padding around the VCR inside. Probably not double-boxed, but a little better than I'd expect from an average eBay seller.

-- merged --

This morning, I got the VCR unboxed and it's in great shape! Looks like new, operates perfectly on every function I've checked so far. No funny sounds, video problems or smells that I can tell, and the chassis is literally like new, not a single scratch or scuff on it which impressed me for a device that is ~32 years old. I tested a couple of throw-away tapes and I'm very happy with the results! I think this will work quite well for my needs.

Then I walked down to my local thrift shop and picked up a little 32" 1080p Samsung TV from 2018 for $22.50 that should make a semi-decent monitor for this secondary computer setup at my house for this project. (I know some of you will cringe pretty hard at that, this is not a video editing display by any means)

Cat Tax:
Attachment 20204

billct97 02-26-2026 01:34 PM

Quote:

Originally Posted by matt314159 (Post 106442)
Then I walked down to my local thrift shop and picked up a little 32" 1080p Samsung TV from 2018 for $22.50 that should make a semi-decent monitor for this secondary computer setup at my house for this project.

That looks like my Samsung!

Do you know how to set the inputs for PC? It makes a huge difference.

Select Source / highlight the HDMI input.
There should be a down arrow below the source at the top of the screen.
Press the down arrow on the remote.
Edit Device Type and choose DVI PC.

matt314159 02-26-2026 01:51 PM

Quote:

Originally Posted by billct97 (Post 106443)
That looks like my Samsung!

Do you know how to set the inputs for PC? It makes a huge difference.

Select Source / highlight the HDMI input.
There should be a down arrow below the source at the top of the screen.
Press the down arrow on the remote.
Edit Device Type and choose DVI PC.

Thanks, I knew I was going to need to do that but not the exact menu sequence, so that's super helpful, thanks! This one didn't have a remote, but it's not so bad navigating the menus with the little joystick button on the TV itself.

Right now I just hooked it directly to the VCR with composite cables because I need to grab a DisplayPort to HDMI adapter dongle from the office before I can attach it to the computer tower.

RayNotes 02-26-2026 01:59 PM

Quote:

Originally Posted by matt314159 (Post 106442)
I tested a couple of throw-away tapes and I'm very happy with the results! I think this will work quite well for my needs.

Very nice! I'm glad the VCR is working well. That's half the battle. Good luck with your project!

matt314159 02-28-2026 07:13 PM

Still waiting for the ES15 to arrive, as well as the breakout cable for the Intensity pro. Before sending the GV-USB2 back, I decided to do a quick test run on a secondary tape I didn't care much about, and honestly? I think if I used this for the whole project, I'd probably be happy. I think I like it enough to start this project with the S5400U>ES15>GV-USB2 as the hardware chain.

Here's a 2m 57s test clip I captured with virtualdub2 and deinterlaced with Hybrid.

https://photos.app.goo.gl/2nFuAhLmJ3ExMrWR8

Are there any good main discussion threads for properly using Hybrid? I'm sure I made plenty of mistakes that would make you guys roll your eyes, but at a basic level, I did QTGMC slow, trimmed 8px off the bottom to get rid of the tracking noise line (sorry I'm blanking on the proper term), then added 8px back in to correct the resolution. Black levels are a very pushed in this scene but I did that intentionally in virtualdub because the play that follows this 3 minute behind the scenes intro had some really challenging dynamic range on the community theatre stage shot with a 1990s VHS camcorder. So I adjusted the histogram to that part which was the bulk of the tape.

Heck, is Hybrid even the recommended tool for a capture workflow like this? Or is there a simpler tool? The main things I care to do to the tapes are to deinterlace, crop, and encode to H264 with a CRF about 17 for the files I share around.

I think I have decided to hold onto the masters as FFV1. I don't have any plans to take them into Premier or anything, so I'd rather see the improved compression ratio while hanging onto the 480i masters.

Edit - I just recaptured and reprocessed it with Hybrid, and uploaded it to Google Drive this time instead of Google Photos, because I think I have storage saver turned on: https://drive.google.com/file/d/1TNo...usp=drive_link

Edit2 - rewatching the new encode I still don't have the PAR right. I saw a YouTube video saying to set input PAR to 10 x 11 and output to 1x1, but it still seems to be coming out wider. If I set 4:3 in VLC then it snaps back tot he right aspect ratio.

Aya_Rei 02-28-2026 09:35 PM

Input par doesn't matter, at least I don't think it does. Output being 1x1 works if you resize the video to a 4:3 resolution.

Like 720x540 or 1440x1080

Also Google Drive link does not work, file is restricted.

lordsmurf 02-28-2026 11:43 PM

2 Attachment(s)
Quote:

Originally Posted by matt314159 (Post 106442)
Then I walked down to my local thrift shop and picked up a little 32" 1080p Samsung TV from 2018 for $22.50 that should make a semi-decent monitor for this secondary computer setup at my house for this project. (I know some of you will cringe pretty hard at that, this is not a video editing display by any means)

Samsung monitors are pretty awful in terms of color accuracy, so be very wary about tweaking color on it. I'd suggest saving that task for something better, something that can calibrate. Samsung is often very contrasty, with overblack black levels. Just use it as capture preview, understanding the actual video may look much different.

Quote:

Cat Tax:
Attachment 20204
Awesome. :laugh:

Back in 2015, I had to unspool a VHS tape, to repair another tape with the clamshell...

Attachment 20207

...and it quickly became something fun to play with... :laugh: :smack:

Attachment 20208

He's no longer with us, but this is one of my most fun memories with him. :angel:

Quote:

Originally Posted by matt314159 (Post 106465)
I think if I used this for the whole project, I'd probably be happy. I think I like it enough to start this project with the S5400U>ES15>GV-USB2 as the hardware chain.

Two concerns:
- How well did you examine for dropped frames, and audio sync issues? That will require a long sustained capture to showcase the problem, not quick test..
- Again, that Samsung monitor is almost certainly hiding color/contrast/darkness issues.

Quote:

but at a basic level, I did QTGMC slow,
trimmed 8px off the bottom to get rid of the tracking noise line
then added 8px back in to correct the resolution.
QTGMC tip: Slow blurs, as it's NR heavy. Faster can often be better. Slow or Medium only for videos with complex scenes that Faster chokes on. For example, wrestling matches, water skiing, race cars, marching bands. Lots of fine detail moving fast.

Quote:

So I adjusted the histogram to that part which was the bulk of the tape.
Again, that Samsung monitor will not be accurate. You'll calibrate your video to a monitor, rather than to any standard.

Quote:

Heck, is Hybrid even the recommended tool for a capture workflow like this? Or is there a simpler tool? The main things I care to do to the tapes are to deinterlace, crop, and encode to H264 with a CRF about 17 for the files I share around.
- Yes, Hybris is suggested.
- And Hybrid is the easiest tool. (Handbrake has a dumbed down interface, so some consider it "easier" at first glance. But it has lots of issues when it comes to handling interlaced video tapes. Handbrake was made for non-homemade DVD conversions, and it shows.)

Quote:

Edit2 - rewatching the new encode I still don't have the PAR right. I saw a YouTube video saying to set input PAR to 10 x 11 and output to 1x1, but it still seems to be coming out wider. If I set 4:3 in VLC then it snaps back tot he right aspect ratio.
- 720x480 capture file, master archive file
- deinterlace
- hard crop 16px total, making it 704x480. It must be 2px widths, but that can be 8+8, 12+4, 14+2, or 16+0, etc.
- resize 704x480 to either 640x480 or 720x540 for 4:3 viewing ---- or 1280x960 for YouTube as pre-process step. This is where GAN upscale matters in Hybrid. Resize/pad from there to the final 1920x1080 compressed 16:9 streaming file.

You seem to be doing fine.
- better card would help a % for quality and continuity
- don't calibrate to a monitor
- be mindful of deinterlace and resize settings

Video captures can easily be done vastly worse.

EDIT: I do see visual quality concerns from the attached clips. But I'd need to see more, longer, and bigger, for more analysis. Again, better card + monitor + be mindful of settings.

matt314159 03-01-2026 05:10 AM

1 Attachment(s)
Quote:

Originally Posted by Aya_Rei (Post 106467)
Input par doesn't matter, at least I don't think it does. Output being 1x1 works if you resize the video to a 4:3 resolution.

Like 720x540 or 1440x1080

Also Google Drive link does not work, file is restricted.

Woops, my bad, I changed permissions to "anyone with link".

Quote:

Originally Posted by lordsmurf (Post 106468)
- Again, that Samsung monitor is almost certainly hiding color/contrast/darkness issues.

Oh I definitely understand that, but that should be irrelevant if I'm going by the vdbub histogram, correct?

Attachment 20209

On that capture I was adjusting the video proc amp settings to just pull back some of the blown highlights that were very prominent throughout this tape of a community theatre play down into acceptable range. Looking at the footage again I shouldn't have pushed the blacks so much though.

Since my family home movies are going to be a vast array of scenes across a tape, I doubt I'll use this much in that scenario, unless there are some recommended defaults that apply across the board to make the capture safer or something.

Quote:

Originally Posted by lordsmurf (Post 106468)
QTGMC tip: Slow blurs, as it's NR heavy. Faster can often be better. Slow or Medium only for videos with complex scenes that Faster chokes on. For example, wrestling matches, water skiing, race cars, marching bands. Lots of fine detail moving fast.

duly noted, thanks. Is "Fast" better than "Faster" or is "Faster" truly the one I want to stay with?

Quote:

Originally Posted by lordsmurf (Post 106468)
- 720x480 capture file, master archive file
- deinterlace
- hard crop 16px total, making it 704x480. It must be 2px widths, but that can be 8+8, 12+4, 14+2, or 16+0, etc.
- resize 704x480 to either 640x480 or 720x540 for 4:3 viewing

I re-ran one test clip with 720x540 and that definitely fixed it. Another way I found that seems to work is to leave 720x480 as the target resolution, but add --sar 8:9 which gives the file the right flag for VLC to play it back properly. But I presume the former is preferred due to the variety of playback scenarios the file may encounter, some of which might not honor the flag?

Just want to make sure I'm thinking about this correctly. I still have a decent bit of tinkering and testing to do before I decide I'm ready to start capturing the real tapes.

lordsmurf 03-01-2026 08:05 AM

1 Attachment(s)
Correct, H.264 flagging is rarely obeyed. It's mostly just MPEG-2 where flags get any proper attention.

If you open the QTGMC.avsi, you can see exactly what each mode does.
Hint: Look at the column names, not the rows.

Code:

#                                                              Very                                                        Very      Super      Ultra
        # Preset groups:                                    Placebo    Slow      Slower    Slow      Medium    Fast      Faster    Fast      Fast      Fast      Draft
        TR0          = default( TR0,          Select( pNum,  2,        2,        2,        2,        2,        2,        1,        1,        1,        1,        0      ) )
        TR1          = default( TR1,          Select( pNum,  2,        2,        2,        1,        1,        1,        1,        1,        1,        1,        1      ) )
        TR2X        = default( TR2,          Select( pNum,  3,        2,        1,        1,        1,        0,        0,        0,        0,        0,        0      ) )
        Rep0        = defined(Rep0) ? Rep0 : TR0 < 1 ? 0 : undefined()
        Rep0        = default( Rep0,        Select( pNum,  4,        4,        4,        4,        3,        3,        0,        0,        0,        0,        0      ) )
        Rep1        = default( Rep1,        Select( pNum,  0,        0,        0,        0,        0,        0,        0,        0,        0,        0,        0      ) )
        Rep2        = default( Rep2,        Select( pNum,  4,        4,        4,        4,        4,        4,        4,        4,        3,        3,        0      ) )
        EdiMode      = default( EdiMode,      Select( pNum, "NNEDI3",  "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3",  "RepYadif","Bob"  ) )
        NNSize      = default( NNSize,      Select( pNum,  1,        1,        1,        1,        5,        5,        4,        4,        4,        4,        4      ) )
        NNeurons    = default( NNeurons,    Select( pNum,  2,        2,        1,        1,        1,        0,        0,        0,        0,        0,        0      ) )
        EdiQual      = default( EdiQual,      Select( pNum,  1,        1,        1,        1,        1,        1,        1,        1,        1,        1,        1      ) )
        EdiMaxD      = default( EdiMaxD,      Select( pNum,  12,        10,      8,        7,        7,        6,        6,        5,        4,        4,        4      ) )
        SMode        = default( SMode,        Select( pNum,  2,        2,        2,        2,        2,        2,        2,        2,        2,        2,        0      ) )
        SLModeX      = default( SLMode,      Select( pNum,  2,        2,        2,        2,        2,        2,        2,        2,        0,        0,        0      ) )
        SLRad        = default( SLRad,        Select( pNum,  3,        1,        1,        1,        1,        1,        1,        1,        1,        1,        1      ) )
        Sbb          = default( Sbb,          Select( pNum,  3,        1,        1,        0,        0,        0,        0,        0,        0,        0,        0      ) )
        SrchClipPP  = default( SrchClipPP,  Select( pNum,  3,        3,        3,        3,        3,        2,        2,        2,        1,        1,        0      ) )
        SubPel      = default( SubPel,      Select( pNum,  2,        2,        2,        2,        1,        1,        1,        1,        1,        1,        1      ) )
        Blocksize    = default( Blocksize,    Select( pNum,  bs,        bs,      bs,      bs,      bs,      bs,      bs2,      bs2,      bs2,      bs2,      bs2    ) )
        bs = Blocksize
        Overlap      = default( Overlap,      Select( pNum,  bs/2,      bs/2,    bs/2,    bs/2,    bs/2,    bs/2,    bs/2,    bs/4,    bs/4,      bs/4,      bs/4  ) )
        Search      = default( Search,      Select( pNum,  5,        4,        4,        4,        4,        4,        4,        4,        0,        0,        0      ) )
        SearchParam  = default( SearchParam,  Select( pNum,  2,        2,        2,        2,        2,        2,        2,        1,        1,        1,        1      ) )
        SearchParam  = ESearchP && (Search==4 || Search==5) ? pNum < 4 ? 24 : pNum < 8 ? 16 : SearchParam : SearchParam
        PelSearch    = default( PelSearch,    Select( pNum,  2,        2,        2,        2,        1,        1,        1,        1,        1,        1,        1      ) )
        ChromaMotion = default( ChromaMotion, Select( pNum,  true,      true,    true,    false,    false,    false,    false,    false,    false,    false,    false  ) )
        Precise      = default( Precise,      Select( pNum,  true,      true,    false,    false,    false,    false,    false,    false,    false,    false,    false  ) )
        ProgSADMask  = default( ProgSADMask,  Select( pNum,  10.0,      10.0,    10.0,    10.0,    10.0,    0.0,      0.0,      0.0,      0.0,      0.0,      0.0    ) )

        # Noise presets                                          Slower    Slow      Medium    Fast      Faster
        Denoiser      = default( Denoiser,      Select( npNum, "dfttest", "dfttest", "dfttest", "fft3df",  "fft3df" ) )
        DenoiseMC      = default( DenoiseMC,      Select( npNum,  true,      true,      false,    false,    false  ) )
        NoiseTR        = default( NoiseTR,        Select( npNum,  2,        1,        1,        1,        0      ) )
        NoiseDeint    = default( NoiseDeint,    Select( npNum, "Generate","Bob",      "",        "",        ""      ) )
        StabilizeNoise = default( StabilizeNoise, Select( npNum,  true,      true,      true,      false,    false  ) )

        # The basic source-match step corrects and re-runs the interpolation of the input clip. So it initialy uses same interpolation settings as the main preset
        SourceMatch  = default( SourceMatch, 0 )
        MatchNNSize  = NNSize
        MatchNNeurons = NNeurons
        MatchEdiMaxD  = EdiMaxD
        MatchEdiQual  = EdiQual

        # However, can use a faster initial interpolation when using source-match allowing the basic source-match step to "correct" it with higher quality settings
        Assert( SourceMatch == 0 || mpNum1 >= pNum, "'MatchPreset' cannot use a slower setting than 'Preset'" )
        #                                                                    Very                                                        Very      Super    Ultra
        # Basic source-match presets                                Placebo  Slow      Slower    Slow      Medium    Fast      Faster    Fast      Fast      Fast
        NNSize  = (SourceMatch == 0) ? NNSize  : Select( mpNum1,  1,      1,        1,        1,        5,        5,        4,        4,        4,        4    )
        NNeurons = (SourceMatch == 0) ? NNeurons : Select( mpNum1,  2,      2,        1,        1,        1,        0,        0,        0,        0,        0    )
        EdiMaxD  = (SourceMatch == 0) ? EdiMaxD  : Select( mpNum1,  12,      10,      8,        7,        7,        6,        6,        5,        4,        4    )
        EdiQual  = (SourceMatch == 0) ? EdiQual  : Select( mpNum1,  1,      1,        1,        1,        1,        1,        1,        1,        1,        1    )
        TempEdi  = EdiMode # Main interpolation is actually done by basic-source match step when enabled, so a little swap and wriggle is needed
        EdiMode  = (SourceMatch == 0) ? EdiMode  : default( MatchEdi, ((mpNum1 < 9) ?  EdiMode : "Yadif") ) # Force Yadif for "Ultra Fast" basic source match
        MatchEdi = TempEdi

        #                                                          Very                                                        Very      Super    Ultra
        # Refined source-match presets                  Placebo  Slow      Slower    Slow      Medium    Fast      Faster    Fast      Fast      Fast
        MatchEdi2 = default( MatchEdi2, Select( mpNum2, "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "NNEDI3", "TDeint",  ""    ) )
        MatchNNSize2                  = Select( mpNum2,  1,        1,        1,        1,        5,        5,        4,        4,        4,        4    )
        MatchNNeurons2                = Select( mpNum2,  2,        2,        1,        1,        1,        0,        0,        0,        0,        0    )
        MatchEdiMaxD2                = Select( mpNum2,  12,      10,      8,        7,        7,        6,        6,        5,        4,        4    )
        MatchEdiQual2                = Select( mpNum2,  1,        1,        1,        1,        1,        1,        1,        1,        1,        1    )


Feedbucket 03-01-2026 11:08 AM

You may want to crop the borders while adjusting the contrast/brightness since non-representative areas of the frame like the head switch noise on the bottom tend to force the waveform to spike at either end (measuring at Y=0 or 255) where it should really taper off before then. Useful if for no other reason than to get a better look at what the levels really are. (Don't forget to undo the crop before capturing.)

If you're going straight to delivery after capture, you'll want to make sure nothing's in the red.

matt314159 03-01-2026 03:00 PM

1 Attachment(s)
Would you be willing to glance through my virtualdub2 settings and see if anything stands out as horrifically wrong before I lock these in as my capture settings? Any other areas I should check?

You'll note in the screenshot I'm using FFV1 at the moment. I think I lean slightly toward the more efficient compression of FFV1 since I'm not planning to ever pull these into Premier, but I'm still weighing it against HuffYUV.

One thing that's kind of annoying with my GV-USB2 is that whenever I re-open virtualdub2 it defaults the Capture Pin framerate to 59.94 when I want 29.97 and I have to change it every time. Besides that, it seems like most of my other settings "stick".

RayNotes 03-02-2026 08:05 AM

Quote:

Originally Posted by matt314159 (Post 106477)
You'll note in the screenshot I'm using FFV1 at the moment. I think I lean slightly toward the more efficient compression of FFV1 since I'm not planning to ever pull these into Premier, but I'm still weighing it against HuffYUV.

One thing that's kind of annoying with my GV-USB2 is that whenever I re-open virtualdub2 it defaults the Capture Pin framerate to 59.94 when I want 29.97 and I have to change it every time. Besides that, it seems like most of my other settings "stick".

The FFVI is a modern codec with better a compression ratio than HUFF and enjoys current attention from developers. HUFF is more interoperable with editors and media players because it's been around a very long time. Bottom line is that I'm sure you can find online pros & cons to help you decide which suits your needs. I capture FFV1, then deinterlace with QTGMC to H.265 since my system can process it with hardware at over 100 FPS and YouTube accepts it the codec - which is my typical delivery mechanism. You needs may vary.

Note: I see you're capturing at 10 bit. This will increase the size of your files but it might help QTGMC in post - depending on the source quality. Most likely 10-bit is overkill for your final MP4/MOV delivery files. VHS itself doesn't carry that much color information.

Yes, the VirtualDub Capture Pin will need to be set every time you start VirtualDub. This is not unique to the GV-USB2 - it's what VirtualDub does. 720x480 @ 29.97 FPS is the correct setting. It will stick as long as you keep VirtualDub open and don't re-open the Capture Pin dialog. But every time you re-open the capture pin dialogue, it resets to its best guess again. Just set it again and leave it alone.

matt314159 03-03-2026 01:34 PM

Quote:

Originally Posted by RayNotes (Post 106482)
Note: I see you're capturing at 10 bit. This will increase the size of your files but it might help QTGMC in post - depending on the source quality. Most likely 10-bit is overkill for your final MP4/MOV delivery files. VHS itself doesn't carry that much color information.

Yes, the VirtualDub Capture Pin will need to be set every time you start VirtualDub. This is not unique to the GV-USB2 - it's what VirtualDub does. 720x480 @ 29.97 FPS is the correct setting. It will stick as long as you keep VirtualDub open and don't re-open the Capture Pin dialog. But every time you re-open the capture pin dialogue, it resets to its best guess again. Just set it again and leave it alone.

As I'm doing more reading, I'm thinking I won't miss anything as far as quality by capturing 8 bit masters. Thanks for pointing that out.

Now, I'm capturing 92.97 in Virtualdub, but when I do QTGMC I've got bob selected, so the output framerate for my H.264 playback files is 59.94fps. Is that generally recommended? I noticed the video looked a lot smoother but almost had a "soap opera effect" to it that I wasn't sure I loved. I want to capture as much as possible from the tapes and not leave any fields on the floor if I can help it.

RayNotes 03-03-2026 02:41 PM

Quote:

Originally Posted by matt314159 (Post 106505)
Now, I'm capturing 92.97(sic) in Virtualdub, but when I do QTGMC I've got bob selected, so the output framerate for my H.264 playback files is 59.94fps. Is that generally recommended? I noticed the video looked a lot smoother but almost had a "soap opera effect" to it that I wasn't sure I loved. I want to capture as much as possible from the tapes and not leave any fields on the floor if I can help it.

The 29.97 at capture is because VirtualDub recognizes the difference between Frames (two full interlace scans of video), and Fields (a field being one single scan from top to bottom, which only populates every other horizontal line of the screen). It takes two full scans of teh odd and even fields to properly fill the screen, thus creating a Frame. This *is* the correct setting for capturing in VirtualDub. Even your computer will declare the properties of the captured file to be 29.97 FPS - which is techinically (but not visually) true.

What QTGMC is doing is taking both sets of Fields to form one "Frame" of progressive scan, deinterlaced video. So from frame #1, the odd (top) and even (bottom) fields and magically blended together to form a single progressive frame of video. BUT then... it takes the even fields from frame #1 (the set that comes second in time in that frame) and combines them with the odd fields in frame #2 (which come later in time than the even fields of frame #1) to form a completely new frame of video that sits between the original Frame #1 and #2. This process is repeated over and over, thus double the frame rate. It makes 29.97 FPS (which is really 59.94 fields per second) video into 59.94 FPS progressive scan video - which is correct and what you want if you want the final product to "move" like NTSC does.

BTW - if you ever capture DV over FireWire, the bottom fields are first (BFF) instead of top fields first (TFF).


All times are GMT -5. The time now is 11:33 AM

Site design, images and content © 2002-2026 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2026 Jelsoft Enterprises Ltd.