Nero Video Grabber vs Hauppauge WinTV-USBlive2 vs August VGB300

[rgr]

I purchased the following devices for testing:

1. Nero Video Grabber (Recode Stick)
2. Hauppauge WinTV-USBlive2 01341
3. August VGB300

I plan to test them without and with the DMR-ES10 and with AmarecTV. Any advice before I start doing this?

[lollo]

For Haupauge USB-Live 2 use a Y/C time base corrected signal. The same for the others, I suspect.

[rgr]

Round 1 - equipment and installation

1. August VGB300 - quite complete. We get a SCART->S-Video + 3xRCA adapter. There is no extension cable for USB, but there is an extension cable for analog inputs (but only for RCA, not for S-Video). Rating: 4/6.

2. Nero Recode stick - SCART->3xRCA adapter (strange, because without S-Video). So not very useful. There is an extension cable for USB, but it is not long. 3/6.

3. Hauppauge USB-Live2 - it couldn't be less. Just a stick with S-Video + 3xRCA inputs and a short extension cable for USB. No SCART adapter at all. 2/6

So for 2 and 3 you have to buy a SCART->S-Video adapter if your are in PAL region.

I added two standalone devices to the test:
4. AverMedia ER330 (HDMI->SD Card capture device) - complete equipment, but you need to buy an HDCP stripper and of course some SD card from the Endurance series (normal cards, like pendrives, will die sooner or later...). Or an external drive.
5. Sony Digital8 camcorder DCR-TRV210E with S-Video input->Firewire passthrough. Of course, you need an S-Video+RCA cable and Firewire and a Firewire card in your computer. (OK, it's not exactly a standalone device, but if you're recording to tape, it will be )


August VGB300 and Nero Recode do not need any drivers (they install as a generic device), Hauppauge USB-Live2 requires downloading drivers from the manufacturer's website. We also get the Hauppauge Capture program, which is practically useless because it only captures to MPEG2.

Round 2 - capturing 1h tape -- below #11

And here's the problem, because I can't get August VGB300 to work. It communicates with the device, but it doesn't give any image in VDub/Amarec or OBS. Does anyone have it and it works for them?

[lollo]

So for 2 and 3 you have to buy a SCART->S-Video adapter if your are in PAL region.

Or use the S-Video output of the (recommended) devices, which is generally even a better quality connector than the corrispondent pins in the SCART, and which generally provides allows usage of higher quality cables (S-Video versus SCART).

[TimA-C]

On the other hand, how annoyed would you be if you already had the additional scart/s-video/audio cables etc. and you're having to pay for additional ones included with the device? Or maybe you get lumbered with REALLY cheap and nasty cables bundled with the device, and you have to go out and buy a new set anyway? Surely, as long as the item, and everything that comes with it, is accurately described when you're buying it, then that's the best option to have, isn't it?

[lollo]

I have never seen/tested a SCART cable being the same quality as a S-Video cable. In addition, not many good cards are delivered with SCART adaptor, and if they are, the quality is poor for sure, and you need replacement for serious capture work.

[Sharc]

Remember the brightness pulsing tests for the S-Video output of the ES10 (for example) and the recommendation to use the S-wire of the SCART output instead (which is DC-coupled which I personally don't like so much .... )

Edit:
Here Bogilein's test, see the ES10 when it changes from white to black:

[rgr]

Or use the S-Video output of the (recommended) devices

If they have them, and they usually don't.

I have never seen/tested a SCART cable being the same quality as a S-Video cable.

I don't see any reason why they would be of lower quality, provided they are made correctly, of course.

In addition, not many good cards are delivered with SCART adaptor, and if they are, the quality is poor for sure, and you need replacement for serious capture work.

The adapter that comes with August works very well. Is it possible? It is possible.

[Sharc]

Low quality Scart->S-video adaptors and poorly screened cables produce crosstalk between C and Y and may cause crawling ants/checkerboard patterns, similar to Composite dot crawl.
Tests made with a mediocre quality adapter/cable showing the chroma -> luma crosstalk, for example.

[lollo]

Yes, in principle the Y/C signals available on S-Video and SCART are the same but:

- inside the machines less care is given to the routing of the signals to SCART connector

- the SCART connectors in the devices are lower quality then the S-Video connectors

- the SCART cables are lower in quality (i.e Monster Cable "Monster Video 3" S-Video versus any of the Monster Cable SCART to S-Video)

- any other reason related to the manufacturers being considering (unfortunately) SCART a low end connector, while in fact it also vehicles RGB and could be the best.

[rgr]

Round 2 - drops/inserts and sync - capturing 1h tape.

Part 1 - TBC off -- Philips VR1100 (JVC HR-S7700 clone), edit/distinct mode, BEST on --> stick (S-Video) --> AmarecTV 3.10

1. August VGB300 - still dead. 0/10

2. Nero Recode

"Amazing" stick
Captures 720x576 25fps... but every other field. Then it scales to 720x576. As a result, we get progressive 25fps 720x576 with every other field. Video-audio sync is also mediocre, although 3 frames (approx. 0,12s) is not tragic.
(I suspect August would work the same, because it reports to the system as an identical generic device.)
3/10


3. Hauppauge USB-Live2

Despite AmarecTV's report of zero dropped frames, USB-Live2 regularly dropped one frame every 8000 frames (or rather 4000 frames, but 8000 fields), which translated into 22 dropped frames during a 1-hour test. A little, a lot? Well, I would say so on average.

Sync - the sound was ahead of the image by 2 frames, which gives a "negative" delay of 0.08s at the beginning and end of a 1-hour tape. A bit better than Nero Recode.


3b. Hauppauge USB-Live2 + VirtualDub capture

After replacing AmarecTV with VirtulDub, the number of dropped frames by Hauppauge USB-Live2 dropped to 10-12 per 1h of recording (two attempts).
Of course, VDub, just like AmarecTV, reported zero dropped frames.
But most importantly, video-audio synchronization was almost perfect one time (<1 frame), and "perfectly perfect" the next (0.00ms delay).

So VirtualDub Capture performed much, much better than AmarecTV.



Rating using AmarecTV: 6/10
Rating using VirtualDub: I don't know? 8/10?

Edit: I also tried setting Clock to OFF in Amarec TV, but the result is similar -- 20 frames dropped, sync: 3-5 frames off.

(Out of curiosity I also checked the included Hauppauge Capture MPEG2 program. The stick only lost 1 frame (2 fields), and the sync was almost perfect. So it can be done well.)

4. Sony Digital8 camcorder DCR-TRV210E -> Firewire --> CaptureFlux

(Since my Vegas Capture wouldn't work with the camera set as passthrough (but would work with the camera set to camera mode), I used CaptureFlux because that's what I had.)
It is clear that hardware solutions can work reliably, and that was the case here. ZERO dropped frames! I checked the file and not a single drop/insert -- every frame was perfectly in place. Sync was also perfect - although at the end of the file audio delay was half a frame (0.02s).

9,5/10



5. Avermedia ER330
6. Panasonic DMR-EH585
7. Panasonic DMR-ES10

I lost the file from Avermedia ER330 somewhere, but since I also decided to test DMR-EH585 and ES10 as passthrough, I will update soon.

[rgr]

5. Panasonic DMR-EH585 --> Hauppauge USB-Live2

a) AmarecTV - 18 frames drops, sync: 1 frame off at the beginning, 4 at the end - 6/10
b) VirtualDub - 6 frames drops, sync: 6 frames - 8/10
Adding the DMR-EH585 doesn't really improve anything.

6. Panasonic DMR-EH585 --> (HDMI) --> Avermedia ER330

Avermedia records a technically difficult MP4 file -- VFR, and the audio has gaps that need to be fixed with EAC3TO. In the end, only 1 insert, but that there are single drops and inserts here and there (but this may also be due to the way LWLibavVideoSource works).

Sync: 1 frame at the beginning, 2.5 frames at the end.

Nevertheless, a good result - 9/10.

[Bogilein]

Can you provide a few details about the source tape?

Have you downloaded the sync test from You Tube and then recorded it onto a VHS-tape (or SVHS-tape)?

Maybe I'll have time and can do the whole thing to see if I get a similar result.

[rgr]

Can you provide a few details about the source tape?

Have you downloaded the sync test from You Tube and then recorded it onto a VHS-tape (or SVHS-tape)?

Maybe I'll have time and can do the whole thing to see if I get a similar result.

Recorded on this VCR from a DVD. My test DVD with a video clip, on which I numbered each frame and added a few shots for testing. This test may be from YT (I don't remember), I corrected it in Vegas, because we know what quality YT files are.

[rgr]

7. Panasonic DMR-ES10 --> Hauppauge USB-Live2

AmarecTV -- two inserts at the beginning, but then 22 drops -- overall 20 drops. Sync: audio is 2-3 frames ahead of video. AmarecTV sucks as always. 6/10

VirtualDub -- 6 frames dropped. Desync: 3 frames. Still about 8/10. (By the way, I discovered that it is better to disable the video display on the screen (even in Overlay mode) during capture. When enabled, any load on the graphics card causes drops.)

[Bogilein]

Could you upload your test DVD?
Then I could also work with it and the results could be compared.

[Alwyn]

audio is 2-3 frames ahead of video

What process are you using to determine the audio sync?

[rgr]

Could you upload your test DVD?
Then I could also work with it and the results could be compared.

Priv.

audio is 2-3 frames ahead of video

What process are you using to determine the audio sync?

VirtualDub --> View --> Audio display

[Alwyn]

VirtualDub --> View --> Audio display

Anybody can view a waveform in any number of editors. What process are you using? What's your source material? How was it constructed?

My experience with AmarecTV is totally different to you. I've captured dozens of dodgy home VHS video tapes and I've never had any more than 2-3 dropped/inserted frames over 3 hours thru an ES-15.

Where are you getting the Amarec dropped frames count from?

I therefore have my doubts about the validity your other tests.

[rgr]

No process -- I simply loaded a file generated by AmarecTV or Vdub into VDub (+deinterlace if necessary).
Maybe AmarecTV doesn't report drops (it doesn't report me -- 0 drops), but it drops a lot of frames, most in the test.
The test procedure is described above, you can also see the screenshots -- these are 180000 numbered frames recorded on a VHS tape. It was enough to set the dump to a frame with a counter of 0 and move forward 180,000 frames (I checked every few to a dozen thousand). Then you can see that AmarecvTV's reports have nothing to do with reality.

[Alwyn]

...

[Alwyn]

How are you accounting for a VCR that is not playing at precisely 25fps? The digitiser/software will be capturing at precisely 25fps, but if the VCR is running slow or fast, then the frame count display from the tape will bear no resemblance to the frame count on the captured file, would it not?

[Alwyn]

Here's a test I did:

[Attachment 85362 - Click to enlarge]
AmarecTV and VDub reported dropped frames at 39898, 111533, 141008.

I see zero evidence of multiple unreported dropped frames.

[Sharc]

As I understand it (I might be wrong though) a slight deviation (within allowed tolerance) of the actual framerate of the VCR and the nominal ("exact") framerate should not matter. The "exact" (=nominal) framerate of the digitized capture (like 25.000fps) is re-timed when the captured fields/interlaced (woven) frames are read out at playback time - and can be altered in avisynth with assumefps() for example. The capture device follows the sync signal of the video and captures field by field, and only if the "next" sync pulse or vertical sync pattern is not received within a certain time window (nominal+/-tolerance) a frame is dropped/inserted. So the main reason for dropped frames are poor quality (distorted) sync signals and the tracking performance of the capture device (tracking range of a PLL circuits, or simple timer gates, a technical compromise) rather than slight framerate deviations from nominal values (within tolerances). Details are probably with the chip manufacturers.

[rgr]

How are you accounting for a VCR that is not playing at precisely 25fps? The digitiser/software will be capturing at precisely 25fps, but if the VCR is running slow or fast, then the frame count display from the tape will bear no resemblance to the frame count on the captured file, would it not?

Not necessarily. VirtualDub can capture irregularly arriving frames and adjust the audio (resample) so as not to lose synchronization (see UGREEN capture device test below). It is possible that Amarec cannot do this, because there are no options there and then such frames are dropped.

[rgr]

7a. Panasonic DMR-EH585 --> (HDMI) --> Hauppauge HD PVR Pro 60 HD --> SD Card (standalone recorder)
7b. Panasonic DMR-EH585 --> (HDMI) --> Hauppauge HD PVR Pro 60 HD --> PC (as USB device)

The most expensive and crappy device. In 576p50 Duplicates even frames and drops odd ones (like Nero Recorder), and in 576i mode it simply deinterlaces to 576p50 while also removing every other frame. (Like Nero Recode). Complete misunderstanding. Disqualification.


8. Panasonic DMR-EH585 --> (HDMI) --> Elgato Game Capture HD60 X --> PC

HDMI->USB device. But a failure - no matter what signal it received (576i, 576p), it only sent the "incorrect resolution" screen to the computer. I suppose it only works with 60fps, but I didn't do any further tests. DSQ.

[rgr]

9. Panasonic DMR-EH585 --> (HDMI) --> MS2109 (cheap chinese device) --> PC

66 inserts (no drops). At beginning desync is 1 frame, after 1h it increases to 6 frames. What to expect from a $5 device... And instead of two audio tracks we get one at 96kHz. Still better than Hauppauge HD PVR Pro 60 HD or Elgato 3.5/10.

10. Panasonic DMR-EH585 --> (HDMI) --> UGREEN Full HD 1080P 60FPS Capture Card HDMI to USB --> PC
https://www.amazon.com/UGREEN-Capture-Streaming-Recording-Compatible/dp/B0CFQ2BMPZ

Interesting device. Something is wrong with the data transfer from the device to the PC (bad timestamps? irregular transfer?), so in VirtualDub options I had to enable "Ignore video timestamps from capture driver"). Otherwise I had drops or the video had jumping frames (it became VFR).

(As an aside -- the bitrate of the MJPEG stream is the same as the bitrate of the UTVideo YUV422 compressed video - about 50MBit. Very high.)

Results: 5 drops, but video and audio synchronization is perfect. Rating: 8.5/10.

[aramkolt]

I was thinking about doing a similar audio sync/frame drop test myself across several devices myself, ironically using that same YouTube video as the source haha.

What I ultimately decided is that it would be better to use LTC (linear time code) audio on one of the audio channels and a timecode generator inline when making the recording test tape that burns in the timecode that is identical to the LTC audio. The idea is that each frame gets its own number visually burned in and the LTC audio decryption in a linear editor can be compared to the "realtime" time code.

You could also then see after the fact if frames were duplicated or dropped because they'd each have a unique timecode burnt into the frame.

I can't tell if that is what you are doing here with your burnt in timecodes or not, but perhaps it is. My guess is you used the linear editor to visually burn-in a unique timecode to each frame when making the test tape, but you don't have the LTC audio to go with it. Most editors won't let you make LTC audio for free is my understanding.

Just something to consider.

Now that I've acquired all the hardware imaginable I'd want to test (with the exception being a green AVT 8710), I really need to get around to doing more systematic testing like you've done!

[Alwyn]

In addition to post #23 which disproves the claim that AmarecTV does not record drops/inserts, I have just done a bunch of tests with a very bad tape. My conclusion is that AmarecTV and Vdub both report Inserts accurately, verified by going through the capture files in VDub with Shift ].

All my captures were stabilised with either a Pioneer 645 or a ES10. I therefore conclude that the claim that both do not report dropped/inserts is invalid.

In any case, every time my tape was stabilised and captured with ATV, there were no drops/inserts.

@RGR, if you are getting drops/inserts capturing with ATV, you have a serious problem with your setup. Perhaps the 585 is U/S.

FWIW, VDub 1911 and VDub2 generated a lot of inserts, even through the ES10.

[rgr]

The idea is that each frame gets its own number visually burned in and the LTC audio decryption in a linear editor can be compared to the "realtime" time code.

Yes, I have over 180,000 frames, each numbered. The video-audio sync test is at the beginning and at the end (after 1h). In the middle I have a couple of test clips to compare video quality, but that's later -- for now I'm just gathering material.