What exactly is wrong with EasyCap USB picture quality?

[Pate]

Hi!

My first post, so firstly thanks for creating this forum, it has been very informative for me!

I am in the process of building a proper capture system for transferring VHS tapes to DVD. I have various questions in queue, but actually the biggest thing that bothers me at the moment is that I don't quite understand what exactly is wrong with the capture picture quality that can be achieved with the el cheapo EasyCap USB capture device.

I have been trying to get an understanding of why the EasyCap is rubbish (as mentioned in this thread: http://www.digitalfaq.com/forum/vide...ap-vs-ati.html), and I understand that the build quality is pretty bad, the included drivers are worthless and won't even install, and the audio side is useless (at least with my current setup). But is there something wrong with the actual picture quality (assuming you have a unit that seems to stay in one piece, and have found drivers that work), and and if so, what the problem is?

I currently have two capture devices, some Conexant/Bt878 clone chip in a Nebula Digi-TV PCI card, and the EasyCap USB device. The problem with the Nebula card is that it only has composite input, and as I plan to use S-VHS machine as a source, I would prefer S-Video, which the EasyCap does have.

I have made some comparisons with those two devices using the composite inputs, and it looks like the Nebula card has somewhat darker picture (which probably means that the EasyCap has too bright of a picture), but adjusting brightness and contrast afterwards with Avisynth seems to level this difference out.

So, I would appreciate if someone who has compared EasyCap with a good USB device (like ATI 600) could explain in more detail what makes the EasyCap picture quality rubbish?

Or, if my question makes it apparent that I have no understanding of the things involved in making good quality VHS captures, please be kind to point to some further reading. :-)

Thanks in advance!

Pate

[juhok]

If you want to compare to broadcast capture devices, come on over to Kuokkala with your SVHS deck and we'll compare.
PM me for phone number if interested.

[jmac698]

I happen to agree with you
After calibration, I don't see a big difference. There is a difference that matters more for capturing game systems which have a high quality signal.

[juhok]

jmac698, no big difference between what devices exactly?

[jmac698]

We've discussed this before
http://www.digitalfaq.com/forum/vide...on-vcr2pc.html

I'd like to update this, but if you have time could you compare two capture devices or VCR's on the same material? Both natural images and test screens.

[kpmedia]

The most common problem with the cheapo captures devices is this:

• geometry errors -- example: 4x3 distorted by some % of pixels (BT8x8 was bad about this)
• AGC, IRE and luma errors -- meaning issues with brightness, hots spots, lack of true black/shadows, etc
• color errors -- where colors are shifted, have illegal values, over-downres'd ratio vs luma, etc
• tendency to drop frames outside of normal dropped frames reasons -- commonly a problem with hardware drivers

It's often fairly easy to detect when inferior capture hardware was used. That includes NTSC DV 4:1:1, as well as the cheap chipsets. The values are cooked in a way that gives a unique "amateur" look to the digitally encoded video.

Regarding "brightness" issues -- Most people think that AGC/IRE/luma can simply be corrected by downward "brightness" adjustments. However, quite a bit more signal quality is lost compared against a better device. You're only partially compensating for the problem, not removing it. It's impossible to recover the part of the signal that was lost due to rough overprocessing.

[Pate]

Thanks for the replies!

jmac698: Yes, I had already planned to do some tests/comparisons with various material and with both my VCRs and both my capture devices. I wasn't sure if there was interest in this forum for the results, but I can certainly post my results. I am still waiting for the S-VHS deck to arrive, but as soon as I get that I think I have everything I need for at least preliminary tests.

kpmedia: Thanks, that is the kind of list I was looking for. So, for an amateur like me it looks like there is no reason to ignore the cheap capture devices without first testing them, but I need to be aware of the potential issues. I can then determine after some tests whether the resulting quality is good enough for me, or whether to go for higher-end devices.

juhok: Thanks for the offer, I'll send you a PM when I want to take advantage of it.

Pate

[juhok]

Quote:

Originally Posted by jmac698

We've discussed this before
http://www.digitalfaq.com/forum/vide...on-vcr2pc.html

I'd like to update this, but if you have time could you compare two capture devices or VCR's on the same material? Both natural images and test screens.

The OP is asking about different gear. Do you mean to say based on your past screens that ALL the devices are the same?

I'd like to run some tests screens but for that I want character generator. In the past I've been using DVD-player but that introduces too much artifacts/errors of it's own.

[volksjager]

the ATI600USB in a known excellent capture device and can be had new for under $50 if you shop around.
with good hardware that cheap there is really no reason to skimp.

i have used both the EasyCRAP and the Ion device and they suck.
i wish i saved the captures so i could show you , but i deleted them all when i redid them with the 600usb
even hooked to my AG-1980 via S-video the captures where horrid.

[jmac698]

These are my thoughts though - for a particular chip (there's only a few in current use), you're given a reference design, it's only a matter of what size/shape you want to make the device, and how many inputs. Pretty simple. The chips themselves are pretty consistent and do most of the work.

The only way to screw this up is a mistake in the first revision, bad quality control, stupid driver defaults, or something I'm not aware of. I'd like to know exactly why it's bad. Something like levels can be from simply a wrong resistor somewhere.

How were they horrible? How would you describe the problem?

And when I'm generalizing to easycap I'm thinking it's the same chip, same design, different brands but probably the same manufacturer. But yes I haven't tested an actual easycap, perhaps I'll order one just for fun.

I'm also curious about how people form opinions. I can't remember it all off the top of my head but people have simple instinctive rules. Nothing wrong with that, but it's interesting - if someone is aware of how it works, they could create a better brand reputation.

[Pate]

Okay, I got my S-VHS deck and took some captures from the menu screens using my Nebula capture card, and the EasyCap USB capture device. The short version? The EasyCap device is rubbish.

The longer version: Composite input seems to be totally useless, but it is interesting that there is such a difference between a low-quality S-Video cable (the one I used came with a satellite receiver, I think) and a higher-quality S-Video cable. It looks to me like there is a horrible amount of crosstalk between the luma and chroma signals, so that all the chroma info is actually mixed to the luma signal before it even gets to the actual capture chip.

I am thinking of soldering a short (1 meter or so) higher-quality S-Video cable directly to the EasyCap component board and seeing if that has an effect on the quality. But in any case it seems like the ATI 600 is the way to go, so if I break my EasyCap device completely during my soldering it is all for the best. :-)

Here attached are a couple of images showing the differences in the captures. In actual video the differences are not as pronounced. Of course if you have an idea as to what might be wrong in my test setup, let me know.

Pate

[lordsmurf]

Blue screens are known to set off certain devices more than others, because they're not always 100% to-specs PAL or NTSC. However, there are some other issues that are visible, aside from flickering chroma/luma, which I doubt are in any way connected to format spec issues. That very obvious herringbone pattern, for example. Those are common on poorly-shielded devices. I would bet the cheapo EasyCap leaks like crazy.

EasyCrap is a well deserved nickname.

[juhok]

Looks better than my 1200€ Brighteye with the JVC's character generator (BE goes total batsh*t with the menus). It's crappy signal. Test with some real life material instead.

[lordsmurf]

Animation makes for good testing -- much easier to see the flaws.
I've been known to use a lot of Cartoon Network stuff from my recording days, including episodes of The Smurfs.

[Pate]

Okay, here is a quick comparison between EasyCap USB (top, using S-Video input via the 20 meters long cable) and Nebula (bottom, using composite input), from the 20th Century Fox logo in the beginning of Star Wars Episode I (recorded from a satellite channel onto a S-VHS tape).

I'll have to hunt for some animation material, I don't think I have much of that recorded, sorry. I'll see if I can find some.

I believe the logo shows some of the same flaws as the blue-screen image, but the flaws are less noticeable. Replacing the S-Video cable (and connector) and possibly adding some extra shielding would probably be useful in any case.

Pate

[juhok]

20 meters?! Oh my. I've got some spare (short) HQ s-video cables if you want to buy one asap.

Are the images downsized or do the cappers digitize at weird size?

[Pate]

Thanks for the offer juhok, but I think that would only help with a part of the problem. The EasyCap device itself has a crappy 15cm extension S-Video cable (which looks to be practically unshielded), which probably already causes much of the leakage. I need to practice my soldering skills, been a while since I did any so this is a good practice project.

Pate

[Pate]

Okay, one more nail to the EasyCap's coffin.

I decided to check what the histogram view looks like, and compare that to the histogram from the Nebula card. Here attached is an image showing the differences.

I tried various settings using the STK1160 Video Proc-Amp settings page, but all the settings do is just move those stripes around.

I guess an ATI card is really the only good option for me. Sadly I have not been able to locate any outside of USA, and ordering from there it looks like the postage will be more than the cost of the actual device! Would be nice to find one in Europe, at least.

Pate

[lordsmurf]

Quote:

Originally Posted by Pate

looks like the postage will be more than the cost of the actual device!

I wouldn't worry about that. If I refused to buy things that carried shipping costs as much as (or more than) the item I was buying, I'd almost never be able to buy anything from overseas. But I don't worry, so I do buy. It just costs more to send something on airplanes and boats than it does over land locally. Set a max budget, and forget about the internal number breakdowns.

The ATI 600 was sold exclusively in North America. So you probably won't find one in Europe. Just import it.

Also, for future reference, there's a nifty before/after feature on the forum: http://www.digitalfaq.com/forum/news...er-images.html
You may find that useful for showing comparisons between devices.