GV-USB2 capture filter settings, sharpness issues?

[bakerie]

Hey all,

Just wondering if anyone knows what settings should be used for the capture filter for this card. I wanted it as raw as possible in VDUB, but I feel like these settings are having an effect.

The de-interlacing is set to weave, which seems to be leaving it as interlaced (great), however the Proc amp settings are a bit mystifying to me. Usually the type of thing I leave to default, otherwise I'll end up tweaking until I'm in the grave, but the sharpness being all the way down at 16 is worrying (this was the default!).
I have made a small capture with the sharpness at 128 (middle ground) and instead of being neither sharpening or softening the image, I felt like it added too much sharpness to it.

Has anyone examined these settings to see if there's a optimum to make them as raw as possible?

Image attached.

[lordsmurf]

Preface needed:
GV-USB2 is not a suggested card, for multiple reasons. Yes, it's probably "the best" of the junky cards sold new on Amazon, but that's not saying much. It has multiple issues, from visual quality to signal handling. The card has been meme'd online, mostly by low-knowledge/low-use users (ie, people that don't capture very much, have very low expectations). It also has a few venom-spewing defenders (mostly at VH) -- which is honestly weird.

On to your question...

In theory, sharpness should be at zero. But it can be hard to determine what a true unity setting is. So it does take testing.

For example, on a Panasonic AG-1980P VCR, the middle notch is still too sharp, fake sharp. You have to nudge it right to undo the oversharpening damage, and reduce/remove halo.

The GV-USB2 has had multiple versions over the years; the card model itself is 15+ years old. And multiple drivers. Thus I cannot say with any certainty how each version+driver combo behaves. So again, testing required.

Conclusion...

Or better yet, just get a better card. Rather than try to save ~$50, hit the "easy button" and buy a better card.

(And FYI, that's why I try to keep cards available in the marketplace here. Capture cards are all cheap, under $200, and should be the easiest piece of a workflow to acquire. Save the testing and effort for VCRs and TBCs, and how those interact with your tapes. Crap capture cards just make your life harder. And again, for what? To save the cost of a dinner for two? Eat at home next time, buy a better card with the saved cash. )

I may seem quickly dismissive here, but when it comes to GV-USB2, these sorts of posts are a feature of the card. Feature, not bug. Most of the buyers were duped by fake/low-knowledge reviews, or Youtubes that have no clue. I just hate seeing this, and the meme/trend needs to go away.

[bakerie]

Quote:

In theory, sharpness should be at zero. But it can be hard to determine what a true unity setting is. So it does take testing.

A better question then might be, how do I quantify when the sharpness is "correct" without introducing my own personal bias and also source bias? I just want middle of the road as I'm not looking to do the sort of source by source adjustments you might do on important stuff.

Thinking back to the big backup of tapes I did a few years ago, I'm nearly sure I added a tiny amount of blur (probably some fancy filter) to help remove some of the noise without removing too much detail. That has me wondering if a slightly too high sharpness might be better than too little as you're probably losing detail/definition if it's too low, while too high might just add some extra noise that you can attempt to correct for later.

I have a Sharp VA40HM hooked up related to another thread. Here's some captures from the composite output through a Panasonic DMR-ES10 captured on the GV-USB2 S-Video input.

I made three captures. One with the sharpness at the default 16, another at 64, and one at 128.
Wasn't sure what way to share these, so used H.264. CRF at 15 for the first two, with 16 for the last one as the file size ballooned.

Any input from anyone is appreciated!

[Aya_Rei]

Does the max go to 255 sharpness? I guess 0 in this case would be 128. I know a Pinnacle card has a sharpness of -1 to 2. I leave that to 0. Same with a WinTV HVR-950 card. Default sharpness is 2 out of a max of 15.

Still, this is digital sharpness applied by the capture card which can introduce haloing (or make halos baked into the tape look worse)

It's no different than using a Panasonic AG-1980 with it's adjustable sharpness slider and cranking that all the way to the max, which makes bright distracting halos.

After a while you're just sharpening analog noise..

[bakerie]

Maybe I wasn't clear originally.
It is indeed a 0-255 (8 bit) slider. If the defaults where all 128, then I wouldn't have even taken a second glance at it.
The problem is that the default the stick/driver came with for sharpness was 16, which has me raising a lot of questions.
They quite possibly did this to massively reduce the noise for their shitty capture software that comes with the card.

I cannot find anyone online actually looking at the proc amp settings for this card in dept, so I'm fairly stuck.

It's also not something like a colour issue, where I could just compare with the test card.

-- merged --

Well, I'm a bit further.
I have the datasheet for the IC that does *everything* on this unit.

Like I haven't gone full scale reverse engineer of this or anything, but hardware wise we just have what look like the controller IC, 3 fancy looking crystals (12 and 27 Mhz, which are probably used for deriving clocks for NTSC and PAL?) and then some buck or boost voltage circuitry, protection diodes etc...

So if I can check the registers on the IC directly, I can probably see what the driver is setting for different sharpness values.

https://www.bdtic.com/DataSheet/Inte...POM35SpYDv.pdf

EDIT: There's some fun stuff on this! For example, it has a register you can check to see if it detects a Macrovision signal! Which is relevant to another thread I posted in recently.

There is further components on the rear, which I imagine probably contain an interface IC, but I can't remove the PCB to check what's there without possibly damaging the board...

-- merged --

I did make some images from the videos I'm going to add here.

Original digital Video clip:
Original.png

VHS captured at 16 Sharpness

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

VHS captured at 64 Sharpness
16.png

VHS captured at 128 Sharpness

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

[lordsmurf]

I saw your dupe/clone thread at VH.

The outcome was that you're now seeing what I've said all along: the card has flaws. Multiple issues. Not the worst card out there, but also far from best. All cards have at least 1-2 flaws (nothing is flawless), but the point is that GV-USB2 has a much higher concentration of issues (3-5 issues minimum).

VH is also where the meme was started (then parroted on Reddit), solely due to overly vocal users. Furthermore, I've never understood why lollo and Alwyn shill so hard for this card. You'd think they were getting kickbacks (or reach-arounds?) for the amount of aggressive defending they do of GV-USB2 (and by extension, AmaRecTV, the only software that really works with it, and an inferior software at that). Both of them were members of this site (lollo2, Hushpower), but left due to nasty disagreements over GV-USB2 and AmaRecTV. Essentially, I would not allow their false claims to go unchecked, I detest misinformation. I just do not understand their aggressive attitudes, nor their ignoring and down-playing serious issues. Note that neither of them do capturing at any volume (ie, not enough times to see/notice problems), nor with any professional usage. Most GV-USB2 lovers are just low-end home users with "beer goggles" to quality.

If you want a GV-USB2, fine, but do it for the right reasons: you have determined it's best for your needs.

But if you see flaws, and cannot fix those, then trust your instincts. Verify. And I've/we've verified it for you. GV-USB2 sharpness is crap. Your not wrong, you see what you see. The solution is (a) better card, or (b) live with the flaws.

Sharc is awesome. But he, lollo and Alwyn all speak from PAL experience. That card has less flaws on PAL sources. But for NTSC, you can essentially ignore anything they say, as the card far more mishandles NTSC footage.

Seriously, honestly, capture cards should not be this hard. GV-USB2 has too much friction. Even the PITA install process of AIW is not that bad.

[matt314159]

Quote:

Originally Posted by lordsmurf

I saw your dupe/clone thread at VH.

The outcome was that you're now seeing what I've said all along: the card has flaws. Multiple issues. Not the worst card out there, but also far from best. All cards have at least 1-2 flaws (nothing is flawless), but the point is that GV-USB2 has a much higher concentration of issues (3-5 issues minimum).

VH is also where the meme was started (then parroted on Reddit), solely due to overly vocal users. Furthermore, I've never understood why lollo and Alwyn shill so hard for this card. You'd think they were getting kickbacks (or reach-arounds?) for the amount of aggressive defending they do of GV-USB2 (and by extension, AmaRecTV, the only software that really works with it, and an inferior software at that). Both of them were members of this site (lollo2, Hushpower), but left due to nasty disagreements over GV-USB2 and AmaRecTV. Essentially, I would not allow their false claims to go unchecked, I detest misinformation. I just do not understand their aggressive attitudes, nor their ignoring and down-playing serious issues. Note that neither of them do capturing at any volume (ie, not enough times to see/notice problems), nor with any professional usage. Most GV-USB2 lovers are just low-end home users with "beer goggles" to quality.

If you want a GV-USB2, fine, but do it for the right reasons: you have determined it's best for your needs.

But if you see flaws, and cannot fix those, then trust your instincts. Verify. And I've/we've verified it for you. GV-USB2 sharpness is crap. Your not wrong, you see what you see. The solution is (a) better card, or (b) live with the flaws.

Sharc is awesome. But he, lollo and Alwyn all speak from PAL experience. That card has less flaws on PAL sources. But for NTSC, you can essentially ignore anything they say, as the card far more mishandles NTSC footage.

Seriously, honestly, capture cards should not be this hard. GV-USB2 has too much friction. Even the PITA install process of AIW is not that bad.

I'm sure I've got to be missing it but is there a thread with a guide to help navigate the world of capture cards? I see a few sticky threads in this forum but they're for particular capture cards. I fell for the GV-USB2 trap as well but I'm not sure what to replace it with.

[RayNotes]

I'm still researching this capture card setup at well. You basically need two things done cleanly: Convert analog to digital, stream digital video into a PC. I'm guessing the Blackmagic stuff is likely the best split point for price/performance but I haven't tested that hypothesis yet. May be worth some research though.

Right now I'm using two devices (RetroTINK 2X Pro to Magewell Gen2 [which is not recommended but I like the results]) to do that but maybe there's a device that does both. Even with my modest setup, I don't evoke any capture filters. I take what I can get as close to what the signal gives me. That I do adjustments in post with software when necessary.