About a month ago I thought I had it all figured out when I decided to capture my old Video8 videos on my Sony DCR-TRV460E using Firewire and WinDV. The captures were good and better than the ones I did a year prior using Elgato Video Capture and composite cables. Recently I read about S Video being better for analog video and all the info about TBCs.
Anyway I have gone through the VirtualDub guide from Sanlyn and Lordsmurf and attached my settings below. Just captured one of my old tapes and found that there are some weird lines on it that make it look worse than the Firewire capture. I believe its something to do with interlacing but I am not sure, interestingly if I watch both clips in VLC and go to Deinterlace mode > Yadiff 2x, nothing changes on the S Video capture but the Firewire capture does deinterlace and become smoother.
I have tried using Huffyuv and Lagarith and get the same result, though I am preferring lagarith as I can play it back in VLC easily.
Current setup is:
Sony DCR-TRV460E > S-Video into DVK-100 Cam1 in > S Video Out of DVK-100 to I-O data GVUSB2
I have tested with and without the DVK so I dont think thats the issue.
Below are some screenshots of VDub and a quick side by side sample, the left is the S-Video and right is Firewire.
I would attach a longer video but obviously the file size will be a problem, but whenever theres motion I get the pixelated/ jagged edges and a bit of a blur, like theres frames missing or something.
Any ideas?
They're not weird lines, that is standard interlacing, where the second field appears slightly after the first and appears as jaggies.
VLC Player is obviously automatically deinterlacing the DV file better than the AVI. In VLC, if you force deinterlacing (Video>Deinterlace = ON), you'll see the S-Video version will be much better and on a par with the DV file, if not better.
At the most basic level, you can deinterlace both (remember that that DV file-all DV files- is interlaced exactly like the analogue AVI file) with VDub's deinterlace filter (Yadif, double frame rate, TFF for the analogue AVI and BFF for the DV AVI) or you can use Handbrake or even better still, AVISynth (extremely difficult to set up; uses basic code) but the best deinterlacer.
Thanks for the response, your VLC trick worked and there was a difference, I also have run both the DV and the S Video videos through Hybrid with QTGMC deinterlacing and I think the S Video one does look better, though the difference appears relatively minor to be but even I can see that there more saturation and maybe a little bit more sharpness, though I am colour blind so I would imagine the difference is more than my eyes are able to see. I presume S-Video is normally the go to for analog video instead of firewire if maximum quality is the preference?
I presume S-Video is normally the go to for analog video instead of firewire if maximum quality is the preference?
Oooo, you're taking a bit of a risk asking things like that around here!
As with a lot of things in life, it's a tradeoff, but yes, if it's an analogue source eg V8, an analogue process through S-Video will be the preferred workflow for best quality. NTSC DV AVI (4.1.1. colour) is lower quality than PAL DV (4.2.0 colour), so certainly in the NTSC case, analogue is recommended.
if it's an analogue source eg V8, an analogue process through S-Video will be the preferred workflow for best quality. NTSC DV AVI (4.1.1. colour) is lower quality than PAL DV (4.2.0 colour), so certainly in the NTSC case, analogue is recommended.
Traal, that topic is about VHS, a different kettle of fish to a V8 tape played in a Sony TBC-equipped camcorder. FWIW, I've tried PAL DV verses AVI as per the OP's scenario, and I can hardly tell the difference. Certainly not worth getting in a tizz about (not that I am suggesting you are getting in a tizz ).
Oooo, you're taking a bit of a risk asking things like that around here!
I know the comment is in jest, but others may not read the humor between the lines.
- NTSC DV is a thorough butcher job, adding (macro)blocks, blurring, and both altering and reducing the colors. It's only a few shades better than low-end infamous cards like Easycap (aka Easycrap).
- PAL DV is more similar to DVD quality (MPEG-2@4:2:0), but different axis for 4:2:0. It still adds blocks, and is lossy, but doesn't really butcher colors.
DV is 1990s tech, and those ADVC boxes literally had Pentium II minimum specs (or Mac G3)..
So it's not case of "OMG, don't talk about it here!", but rather that we'll let you know how ancient and limited the device is, and that better quality hardware (lossless capturing) has been available for 20+ years.
Quote:
As with a lot of things in life, it's a tradeoff, but yes, if it's an analogue source eg V8, an analogue process through S-Video will be the preferred workflow for best quality.
That's really it. Sort of. In video, tradeoffs compound. So if you're using a crappy VCR, no TBC, and a not-recommended capture card, quality will lousy, often an unviewable mess. Using a DV box makes quality cameras/VCRs and TBCs more important than usual.
Quote:
Originally Posted by Hushpower
Traal, that topic is about VHS, a different kettle of fish to a V8 tape played in a Sony TBC-equipped camcorder.
VHS ~ Video8
S-VHS ~ Hi8
DV ~ Digital8 (digital)
The differences between the 8's are negligible. The same advice applies to both.
But that said, the optics quality of home camcorders sucked. So I can see how PAL users might consider DV 1:1 (or close), when the camera is decent, TBCs in workflow. Hi8 color fidelity does tends to be a wee better that S-VHS, mostly due to the cameras and optics. When you add up all the variables, the net output can be fine. Noting that "adding up the variables" is not something novices can do -- but that why we're here, to help guide the newbies on it. We know what they don't know that they don't know!
Thank you for the responses everyone, much appreciated. I am in the UK and therefore a PAL region. From what I can gather here you’re saying that the difference between the S-Video and FireWire here is more negligible but S Video will still deliver a slight improvement for analog, obviously Digital8 and MiniDV would bee best captured via FW. I was going to post this as another question but I’ll just ask here quickly:
I’m considering getting an ES15 since I’ll be doing lots of family analog tapes including VHS/VHS-C and I already have a DVK-100 in my workflow and I know lordsmurf says they are one of the more viable TBC alternatives. Would I use the ES15 all the time as a pass through or only use it on tapes where the frames are dropping or there are sync issues?
The ES-10/15 is best used as a step up from nothing when:
- Frames are being dropped and inserted to a damaging level.
- Audio/ video sync errors.
- Tearing at the top of the capture and general line wiggles.
The few ES series that perform this wonderful correction also have some negative consequences but search this forum
…others will surely chime in.
Last edited by lordsmurf; 07-27-2024 at 02:11 AM.
Reason: Add bullets, formatting.
Towards the end, it looks at S-Video and DV converted capture of a Hi8 tape from the same D8 camcorder. It shows that the DV color accuracy is definitely worse and there's probably some blown out luma as well through Sony's DV conversion chip. It could be that the ADVC110 has better color accuracy and automatic gain control to not do those things. I can't say that I saw any significant "macroblocking" or blurriness in the DV sample over the S-Video output.
The general color tinting shouldn't have anything to do with DV itself chroma subsampling shouldn't matter when it comes to a large solid colors on screen (as is the case with the Tide container). A solid blue screen should look identical when sampled at 4:4:4, 4:2:2, 4:1:1, or 4:2:0 because all of the sampled colors are the same. Real video most of the time isn't going to have large solid colors, so it can matter on the edges of things or for fine colored patterns in terms of a sense of blurred color on the edges of colored objects, but I really don't find that particularly visually different in the comparison shots.
Tint change are a well known byproduct of the codecs. You are correct, that (mostly) should not happen. But the 1/4th compression can affect colors, and the codecs mangle them. So red sweater turn pink, grandma gets a sunburn, the ocean can tint towards snot green, etc.
This was also an accurate comment on that video:
"Some flaws in transfer methods stand out more on bad VHS tapes than real good sources."
... and realizing most homemade VCR/camcorder tapes are "bad" in this context.
The ES-10/15 is best used as a step up from nothing when:
- Frames are being dropped and inserted to a damaging level.
- Audio/ video sync errors.
- Tearing at the top of the capture and general line wiggles.
The few ES series that perform this wonderful correction also have some negative consequences but search this forum
…others will surely chime in.
Quote:
Originally Posted by lordsmurf
This is why ES10/15 is not suggested to be used at all times. It's not a TBC replacement.
However, if you have no actual TBCs, you have no choice.
- bad = nothing
- "doesn't entirely suck" (not a compliment) = ES10/15 + low end gear
- passable = ES10/15 + better non-TBC S-VHS VCR, better ATI/Pinnacle type capture card
- quality = actual frame TBC or modded frame TBC(ish), S-VHS with line TBC, best capture cards
So ES10/15 only if you must, but it's never suggested, unless it's least-worst option.
Thanks for the detailed information.
Let's say my S-VHS VCR playback doesn't show any tearing at the top of the capture, no general line wiggles, and I don't have any audio/video sync issues. Given this, I'm wondering if I still need to use an ES10/ES15 TBC, especially considering the potential negative consequences mentioned.
Assuming I have an ATI AIW capture card. Based on the workflow options you've outlined, I'm considering these possibilities:
- Use my S-VHS VCR + ATI AIW capture card without any additional TBC
- Use S-VHS VCR + ES10/15 + ATI AIW capture card
- Look for a higher-end S-VHS VCR (Like Philips VR 1x00) with a built-in line TBC + ATI AIW capture card
Which of these options would you recommend for the best quality capture, assuming my source material doesn't have obvious sync or tearing issues? Is it still worth using the ES10/15 in my case, or would I be better off without it?
You still lack frame TBC (or the puny ES10/15 non-TBC frame sync), to prevent signal issues at capture. The invisible errors, which often present as dropped frames, duped frames, audio sync errors, etc.
Best quality will be
JVC S-VHS with line TBC >
> some form of frame TBC, either actual TBC, or modded TBC(ish) >
> quality capture card, including ATI AIW
Sync is not obvious at first. Even tearing can be non-obvious. You only see/hear the issues later, on a detailed scrub, or the actual content watch. The worse the gear, the more gear is missing, the more time you must spent to verify the captures.
You still lack frame TBC (or the puny ES10/15 non-TBC frame sync), to prevent signal issues at capture. The invisible errors, which often present as dropped frames, duped frames, audio sync errors, etc.
It got me thinking about some of the stuff you've mentioned before. The video agrees that line TBC is super important (no surprise there), but then it talks about some workarounds for frame timing issues without using a frame TBC.
He mentioned tweaking VirtualDub settings to deal with sync problems. In one of his tests, he only got 92 duplicated frames in a 4.5-hour capture without a frame TBC.
I know you're all about getting the best quality possible, but I'm kinda curious - what do you think about this approach? Is it ever a viable option, or is frame TBC still a must-have in your book?
Also, given that frame TBCs can cost $1000 or more, do you think preventing those 92 frames is worth the investment for hobbyists or folks on a budget?
Your insights on this would be awesome. As the expert here I'd love to hear your take on whether this kind of approach is more like investing an extra 80% (in money and effort) for that last 20% of quality?
Depends a lot on the condition of the tapes, If you have mediocre recordings you're going to have a miserable experience even with a proper frame TBC. Capturing analog tapes is not a "plug and play" task like recording digital channels with a modern DVR. It's a process that requires a lot of know how.
That video is flawed. He's still dropping video frames, but has instead told VirtualDub to ignore the drops by speeding up and slowing down audio randomly. That creates issues where grandma might sound like Barry White, or grandpa sounds like Speed Gonzales on crack. Messing with audio speed is not the answer to dropped/inserted frames.
It's not viable whatsoever, unless you truly do not care about quality.
Video is overall a cheap hobby, DIY task, or business expense. "OMG $1000!!!" is ridiculous, as these are just tools required for video work. Washing machines, cameras, etc, are also "OMG $1000!" tools. The same people generally spend $k on their tech toys: phones, gamer computers, fancy NAS, etc. So it's really about priorities, and the "OMG $1000!" is the excuse. Just say "I don't want quality" and "I don't want to buy tools", without any pretext. I don't want to fix my own car, so I don't buy tools for it. I don't want 90%+ ground beef, I want cheaper 80/20 or even 73/27. I don't hide this by saying "OMG, too expensive!!!" This vastly differs from a person saying that they cannot afford the tools. However, I rarely come across a person that both (1) wants to convert video, (2) truly does not have money. In general, those do not coexist, as not having money creates a more urgent priority than sitting on your butt staring a screen as it ingest videotapes. Mostly students, but still noting that most students have money, and in fact more disposable income that older folks with responsibilities. Sometimes people get pissed at me for saying all this, but it's because I have correctly shown them to be nothing more than cheapskate, and not some destitute person in need of converting VHS tapes.
BTW, DVK in that video is not $1000, and it's not even a full true TBC. So there's also that.
latreche34 said it well:
Quote:
If you have mediocre recordings you're going to have a miserable experience even with a proper frame TBC. Capturing analog tapes is not a "plug and play" task like recording digital channels with a modern DVR. It's a process that requires a lot of know how.
).
