Elgato Video Capture vs. Hauppauge 610 USB-Live 2 - Raw Comparison

[CZbwoi]

I set all the Proc Amp settings on both to default. Hauppauge also had the option to control Sharpness in its Proc Amp, Elgato did not.

The first attached video is the Elgato and the second is the Hauppauge. Both captured with Huffyuv through a Sony Hi8 Handycam CCD‑TRV318 via S-Video and mono audio. I'm gonna leave this to you pros, which one is better? Anything that sticks out in favor or not for one?


Some interesting things to note that the Hauppauge does that the Elgato doesn't:

On the Hauppauge's live feed and capture there's a large black border on the right at all times, I don't know why it's capturing so much dead space on the right side. On the Elgato there's an even amount of black on the two sides of a few pixels.



On the Hauppauge's paused feed screen, something weird happens. On the Elgato it just shows the buzz flicker around in the middle and nothing more. On the Hauppauge every couple seconds it goes on a screen like this and it freezes this type of screen covering the whole capture video for a second or two or three. Then it goes back to normal and it does it again. It legitimately looks like how your PC screen turns for a second before getting the Blue Screen, only just in that little capture window and constantly- when the video is paused. Very odd as this doesn't happen on the Elgato, any reason for this?




So yeah, based on the 2 videos and how they play and the 2 odd things I noticed about the Hauppauge, what do you guys think is the better choice? What should I return?

Should I be worried that the USB-Live 2 has those mini-seizures when the videos are paused? And is that extra black bar it adds a hassle in the end since Elgato doesn't do it? Is one capturing more actual film/data than the other (besides the black bar)? Is one a clear-cut winner out of the two, quality/frames/color wise?

I'm sure you guys can notice stuff I can't in the videos so if you do please give your input.

[hech54]

I've been recording (DVD recorders) and capturing (Hauppauge capture CARD and the USB Live|2) for a long time and one of the first things I learned that "pausing" while recording or capturing is a big NO NO.

[CZbwoi]

I don't follow, why? During recording that would make sense because it would mess up your personal recording with a random pause, but when it's simply plugged into the computer and on the Vdub screen you shouldn't pause the tape, FF, RW it? That doesn't make any sense, I've never heard of that. What if you only wanted to capture one part of it and not another anymore and you want to FF to it? You would turn off the camcorder, plug out your USB or S-Video, turn it on and FF, plug it back in to do so? Or you're not recording but it's currently playing and the feed is also going onto your PC, somebody calls for you and you have to pause it, you wouldn't pause it?

I don't understand why you shouldn't pause the camcorder simply because the S-Video or RCA cables are plugged in and transmitting the signal. That is like saying you shouldn't do that when it's plugged into the TV. And I said that it only happens with the Hauppauge, that weird effect every few seconds when paused, it does not happen with the Elgato.

[hech54]

I don't follow, why?

I'm just telling you MY experiences. STOP.....not pause....works for me.

[CZbwoi]

Well when you call something "a big NO NO" that usually means that there are consequences to doing so, as in you shouldn't do so or something terrible could happen. I was just telling a few odd experiences I didn't see with the Elgato and am asking people if things like this should give me cause to pause regarding the Hauppauge.

For comparison's sake since people might ask or are curious, here is what I'm talking about. The first is the Elgato's more normal paused feed, the second is the Hauppauge's. They are .flv files but should be easily played in VLC.

[hech54]

Well when you call something "a big NO NO" that usually means that there are consequences to doing so

Yes....dropped frames, audio sync problems and recording/capturing a screen full of garbled/jumbled video. If I need to capture a 1 hour long VHS tape.....but I only need 3 or 4 music videos scattered throughout that tape.....I run/capture the whole tape, take out what I need, then delete the big capture. That is what has worked best for me. I don't need to bang my head against the wall trying to figure out why pausing caused me problems. I don't care. "This" works.
And if I only need ONE video of the entire tape....I fast forward and get close to the portion that I need....start the capture at least a minute before the portion I need starts.....then again losslessly edit/take the piece I need.....delete the original capture.

just a suggestion.......

[CZbwoi]

Well when you call something "a big NO NO" that usually means that there are consequences to doing so

Yes....dropped frames, audio sync problems and recording/capturing a screen full of garbled/jumbled video. If I need to capture a 1 hour long VHS tape.....but I only need 3 or 4 music videos scattered throughout that tape.....I run/capture the whole tape, take out what I need, then delete the big capture. That is what has worked best for me. I don't need to bang my head against the wall trying to figure out why pausing caused me problems. I don't care. "This" works.
And if I only need ONE video of the entire tape....I fast forward and get close to the portion that I need....start the capture at least a minute before the portion I need starts.....then again losslessly edit/take the piece I need.....delete the original capture.

just a suggestion.......

Oh no, I completely understand lol, when I get to capturing the actual tapes I'm 'recording' everything start to finish. That's what I do too, and most people. I was just giving scenarios of things that would happen when you're currently getting the live feed of the tape but not recording, sometimes you end up pausing it. Especially when you're trying to test out different USB capture devices and how they act, like what I'm doing. I haven't captured a full tape yet because right now I'm tying to see which one of the two is better doing the simple stuff via short clips and tests. This is something that happened with the Hauppauge which I thought was odd on a paused screen while *not* actually capturing but just streaming the live feed to my PC, and for all intents and purposes is, considering nothing this weird happens with my Elgato.

[CZbwoi]

Here are two more comparisons, this time it's a VHS-C tape played on the JVC S-VHS Camcorder, outputs are S-Video and mono audio again.

The large black bar of black dead space is still present on the Hauppauge's capture. I actually now think that what it does is it takes the two black bars that would be on the side and puts it on the right side. Switching between the 2 videos it looks like the Elgato's total amount of black space on the left and right equals the Hauppauge's one side on the right (the Hauppauge also has nothing on the left). It's odd, but that's what's happening. Elgato has the black dead space split on the left and right, Hauppauge has it added up on the right.

Another weird thing I noticed, in this video there's a high pitch noise coming out of the right (it seems) for the Hauppauge capture. No pitch noise from the Elgato. It's almost like a dog whistle, you can barely hear it but it's there. And yes, the plugs were firmly pushed in all the way.

[jagabo]

As expected, the caps aren't too different. The Hauppauge cap is shifted a little left and up but that's not unusual. It's the reason ITU caps usually include a little extra on the left and right (720 vs 704 pixels). Both have captured with too much contrast (darks too dark, brights too bright) but that can probably be adjusted with their video proc amps. The Happuage has crushed blacks at Y=16. You might be able to compensate for that with the video proc amp, or you might not. The Elgato cap has a little more noise in both the luma and chroma but that is probably more faithful to the source.

The audio noise on the Hauppauge cap appears to be the video signal leaking into the audio channel. That's probably because of poor isolation of the analog audio and video signals inside the device. It might go away if you use a Y adapter to send audio to the right channel. Or use a null plug.

Based on only the fist pair of clips (I didn't download the second pair) I'd prefer the Elgato. But you should look at other issues like what happens when you fast forward, rewind, pause, or just capture longer segments. Does either hang up or stop recording? Does one lose more frames than the other? Problems like that can be a big hassle when trying to capture bad tapes.

[LMotlow]

Both Elgato clips have more noise.

[jagabo]

If you plan on doing your own noise reduction you want the capture to have all the noise that's on the video tape. Noise reduction filters in the tape deck or the capture device are inferior to what you can do in software. They may eliminate small, low contrast detail along with the noise. If the capture device is adding noise, that's a problem. You can test for this by capturing a very clean source, like a DVD player playing test patterns.

[LMotlow]

Those gadgets have noise filters? I wasn't aware of that.
Pick your poison, I guess.

[dellsam34]

Here are two more comparisons, this time it's a VHS-C tape played on the JVC S-VHS Camcorder, outputs are S-Video and mono audio again.

The large black bar of black dead space is still present on the Hauppauge's capture. I actually now think that what it does is it takes the two black bars that would be on the side and puts it on the right side. Switching between the 2 videos it looks like the Elgato's total amount of black space on the left and right equals the Hauppauge's one side on the right (the Hauppauge also has nothing on the left). It's odd, but that's what's happening. Elgato has the black dead space split on the left and right, Hauppauge has it added up on the right.

Another weird thing I noticed, in this video there's a high pitch noise coming out of the right (it seems) for the Hauppauge capture. No pitch noise from the Elgato. It's almost like a dog whistle, you can barely hear it but it's there. And yes, the plugs were firmly pushed in all the way.

Is there a raison to why you are capturing the camera display info, Did you know that you can turn that off from the camera menu?

[CZbwoi]

Does one lose more frames than the other?

Do frames usually get lost for capturing? For USB capturing? Like is it a normal thing to occasionally happen or should I be expecting nothing lost...if there are frames dropped it equals trouble? I'm going to be relying on Vdub's side panel for all of that information if that's okay, unless there's something more reliable.

If you plan on doing your own noise reduction you want the capture to have all the noise that's on the video tape. Noise reduction filters in the tape deck or the capture device are inferior to what you can do in software. They may eliminate small, low contrast detail along with the noise. If the capture device is adding noise, that's a problem. You can test for this by capturing a very clean source, like a DVD player playing test patterns.

Any popular test patterns to play on a DVD player you would recommend to DL or rip from YT?

Those gadgets have noise filters? I wasn't aware of that.
Pick your poison, I guess.

I don't think the Elgato has it inside of it, but in it's software if you choose to use it. I'm using Vdub so that doesn't matter to me. Why poison lol, what else would you suggest I buy?

Is there a raison to why you are capturing the camera display info, Did you know that you can turn that off from the camera menu?

For the VHS-C camera, just because I want to see where I am when I'm capturing test samples, the exact seconds so I can compare. When I get to actually capturing full tapes from it it'll be turned off. On the Hi8 camcorder the display info stays on the camcorder screen but it doesn't get sent out to the capture's feed, so that's neat and helpful for that device.


From what you've all seen so far would you take the Elgato over the Hauppauge?

[CZbwoi]

And here is the data from capturing longer segments, Elgato about a half hour and Hauppauge an hour. I let the second one run twice as long because I left and couldn't go back to stop it, but the data from the Elgato should be enough to compare with Haup's hour.

Anything bad or noteworthy you see from either results?

[jagabo]

Do frames usually get lost for capturing?

People often have problems with dropped frames, especially from tape sources.

should I be expecting nothing lost

From video tape you will probably get a few lost frame here and there. From cleaner sources you should be able to cap with no dropped or inserted frames.

if there are frames dropped it equals trouble?

Yes, you'll be missing parts of the video and audio will get out of sync.

I'm going to be relying on Vdub's side panel for all of that information if that's okay, unless there's something more reliable.

I think it's generally reliable. Certainly a place to start.

Any popular test patterns to play on a DVD player you would recommend to DL or rip from YT?

Many THX certified DVDs include test patterns. You can create a colorbars test pattern with AviSynth's Colorbars() filter. These forms are full of test pattern images and videos. Even some DVD ISO images.

And here is the data from capturing longer segments, Elgato about a half hour and Hauppauge an hour.

Both look ok. The elgato cap has one inserted frame and the hauppauge 2, but the hauppauge captured for twice as long. So the number of inserted frames per unit time looks to be about the same. And one or two dropped/inserted frames an hour isn't a problem.

[CZbwoi]

Thank you, I think I'll stick with the Elgato due to the odd pause effect Hauppauge gives and the way it decides to add a huge right border-frame instead of making it even like the Elgato. It's just weird how it does things like that and I'd rather leave it before anything worse potentially comes up. Good thing I didn't return this Elgato before testing them both at the same time.

[jagabo]

But you've only looked at one (a few?) tape. You should look at a variety of tapes, especially difficult one like very old ones, ones recorded at very low speeds, tapes that were recorded over many times, etc.

[CZbwoi]

Alright, I'll check out some VHS-C tapes and some 8mm tapes that simply look older. I found one or two with scratched out titles so those were recorded over.

ones recorded at very low speeds

How do you determine this? Just older looking tapes?

[jagabo]

ones recorded at very low speeds

How do you determine this? Just older looking tapes?

I'm not real familiar with 8mm and Hi8 decks but VHS players can usually tell you want speed they are playing at, SP, LP, SLP.

[CZbwoi]

ones recorded at very low speeds

How do you determine this? Just older looking tapes?

I'm not real familiar with 8mm and Hi8 decks but VHS players can usually tell you want speed they are playing at, SP, LP, SLP.

Ah, now I see what you mean. I'm not at the VHS part of my conversion process as I don't have a good enough deck yet, and I'm almost certain from what I know that this doesn't apply to 8mm and Hi8, maybe VHS-C but I don't know how to check on those.

I guess my testing process for now will consist of older 8mm and VHS-C tapes.

[CZbwoi]

Okay I'm back with another round of results.

Here's what I got from 2 hours and 3 min of recording time with the two from a 8mm tape that was recorded over a few times.

Elgato:


Hauppauge:


NOTE: The first time I ran the 2hr long Haup-capture I got 8 inserted frames. I decided to do it again after the Elgato and got 3 this time. The Elgato also always inserts a frame at the start of the capture if I'm at the stop screen and press play while capturing (hence the 9 it has) but barring that one they would be equal the first go around with 8 each. But I don't know what that really means as this second time the Hauppauge had only 3 frames inserted.

So looking at these with what I wrote in mind, what's the consensus? Why is the Elgato's video size 9 gigs bigger; good thing, bad thing? And what's up with the video compression ratio being different between the two?


Here are video examples as well, proc amp set to default, and this time I turned the DNR on the camcorder off.

Which looks better?

I'll run a long test with an old VHS-C capture as well.


Running the same tape test again with the Elgato, this time there's 5 frames inserted and it's a video compression ratio of 2:3:1...should I be worried of all these different compression ratios, especially the 2 different ones from Elgato itself?

Here's the picture, same tape, Elgato capturing the same thing once again but the stats are different. File size, video size and video compression ratio (audio slightly different too):

[jagabo]

The first time I ran the 2hr long Haup-capture I got 8 inserted frames.[/B] I decided to do it again after the Elgato and got 3 this time. The Elgato also always inserts a frame at the start of the capture if I'm at the stop screen and press play while capturing (hence the 9 it has) but barring that one they would be equal the first go around with 8 each. But I don't know what that really means as this second time the Hauppauge had only 3 frames inserted.

Obviously, 3 inserted frames is better than 9. But out of 200,000+ frames neither is a problem.

Why is the Elgato's video size 9 gigs bigger; good thing, bad thing?

The ElGato capture has more noise. Noise is hard to compress. Hence the larger file.

And what's up with the video compression ratio being different between the two?

The compression ratio is just the size of the raw uncomrpessed frames divided by the size of the compressed frames. So the difference in frame size follows naturally from the difference in file size due to noise.

Here are video examples as well, proc amp set to default, and this time I turned the DNR on the camcorder off.

Which looks better?

The Elgato is able to capture Y<16, the Hauppauge looks like it's crushing everything below Y=16. Again, adjusting the proc amp might take care of that. The Elgato has a little more noise but may also have a little more fine detail. I don't see any strong reason to prefer one over the other. Maybe the Elgato has a tiny lead.

Running the same tape test again with the Elgato, this time there's 5 frames inserted and it's a video compression ratio of 2:3:1...should I be worried of all these different compression ratios, especially the 2 different ones from Elgato itself?

No.

[CZbwoi]

Thanks for the information, looks like the Elgato it will be.

The Elgato is able to capture Y<16, the Hauppauge looks like it's crushing everything below Y=16. Again, adjusting the proc amp might take care of that.

What should I be looking to do when setting the proc amp settings? Is there a guide somewhere or a general thing to do regarding the 4 settings? One being higher than the other, one you should barely alter, etc.?

[jagabo]

The best way to check your levels is to use a waveform monitor:

https://forum.videohelp.com/threads/340804-colorspace-conversation-elaboration?p=212156...=1#post2121568

You have to adjust brightness and contrast controls to keep your video from clipping at dark and bright levels. The correct settings to use will vary depending on the capture device and the video source.

[CZbwoi]

Okay, would you mess with the saturation in the proc amp at all?

[jagabo]

The hue and saturation in your latest samples looked about right. As long as they're not too far off you can make adjustments when you filter later.

[Alyosha2001]

Just recently bought and tested the Hauppauge USB Live 2 card. Has anyone noticed that in all the above examples the usb live 2 has the video delayed with exactly 8 frames?

[Alyosha2001]

I'm having the same issue with my card on longer captures and can't find a solution (the desync is gradual)

[lordsmurf]

Most capture cards, and that card especially, cannot capture well (or at all) without proper TBCs in the workflow. You cannot connect a random VCR directly to a capture card, and expect quality results, or any results.

So ... what is the full workflow hardware in use here?