CAnopus advc-55 to use with jvc S-VHS with TBC/DNR

[MrTemplar]

Hello!

I’m about to capture some old vhs films. I currently have a JVC HR-S7700 S-VHS deck with TBC/DNR and a cheapish plexgear capture card. However, I consider buying the Canopus ADVC-55 and would need some advice on this. How well does it work? Will it be a good complement to my JVC vcr?

Cheers!

Anders

[davexnet]

It records to DV, 13GB per hour. It's a simple device, with reportedly good results.
DV is good for editing, before you convert to your final format.

Here's some reviews:
http://www.amazon.com/Canopus-77010138100-ADVC-55-Converter-External/product-reviews/B...owViewpoints=1

[lordsmurf]

It's just a basic capture box -- overpriced at that. And DV isn't really the best capture format.
There's discussions on this here in years past.
More at http://www.digitalFAQ.com/forum/forumdisplay.php/capture-record-transfer-7.html
Or search this site or that one.
https://www.videohelp.com/search
http://www.digitalFAQ.com/search.htm

Mostly about the 100 and 110 models, but same applies to the 50 model.

You can do better, and for less funds.

[MrTemplar]

Thanks for your input guys!

Refering to your answer Lordsmurf, what would you recommend instead of the Canopus ADVC-55? What do I get for $100, $200 or $300?

What is the ideal capture format?

At best I can get a used ADVC-55 for the equivalent of around $70 without the power unit (My current PC is not equiped with a 6-pin firewire socket so I'll either have to buy a new one or simply wait until I buy/build a desktop PC eventually).

Cheers!

[edDV]

Thanks for your input guys!

Refering to your answer Lordsmurf, what would you recommend instead of the Canopus ADVC-55? What do I get for $100, $200 or $300?

What is the ideal capture format?

At best I can get a used ADVC-55 for the equivalent of around $70 without the power unit (My current PC is not equiped with a 6-pin firewire socket so I'll either have to buy a new one or simply wait until I buy/build a desktop PC eventually).

Cheers!

As for Firewire port, you can add one with a PCI/PCIe card, express card module or new motherboard.

When you look at the capture alternatives, you need to consider your complete workflow to DVD. It comes down to convienience vs. quality of conversion. The quality issues split into pre-capture hardware investment and post capture filtering you intend to do.

The hardware issues have been discussed in detail in other threads. It comes down to investment in the playback deck, addition of proc amp, TBC/Framesync and the actual capture device.

Post filtering is best done from an uncompressed or losslessly compressed capture file. One 2 hr tape can consume 125-250 GB of disk space for the intermediate file. PC hardware must be adequate to handle and filter files of this size and transfer rate.

DV encoding trades off light lossy compression (about 5x DCT compression to 4:2:0 color space for PAL) for intermediate file sizes at more manageable 13 GB/hr CBR. Since the compression/decompression is done in hardware, the CPU is not stressed. DV records frame by frame (no interframe compression) so is better for filtering, effects or compositing vs. MPeg based captures.

At the low end for quality is MPeg capture (hardware or software compression). MPeg is more difficult to edit or filter and suffers more recode loss unless special techniques are employed such as "smart rendering". MPeg capture file sizes reduce to 0.5-4.4GB/hr for SD.

For high definition, uncompressed/lossless capture is economically ruled out at the consumer level. Instead, one must use lossy digital intermediates or attempt "native" MPeg2 or h.264 editing. Any filtering requires a lossy recode.

[MrTemplar]

Unfortunately my laptop PC (Toshiba Satellite series) lacks expansion card slot...

If I capture losslessly and end up with a 250GB file like in your example I suppose I can convert/burn this file into a compressed format so I won't have to buy a giant hard drive for every film I capture.

As I mentioned, I own a JVC S-VHS with TBC/DNR, I guess this will not replace a full-fram TBC. Would I still benefit from aquiring a seperate TBC? and what about proc-amps? Are there any consumer level proc amps available for less than a few hundred dollars and do I really need one?

I haven't checked all my VHS tapes but I've played I few tapes origin from 1984-85 and they were actually in decent shape and all of my tapes contains material either recorded in Sweden or cassettes aquired in Sweden (a few in the UK) and is thus in PAL format.

I guess that I will be doing only minor post capture editing such as cutting.

[edDV]

Unfortunately my laptop PC (Toshiba Satellite series) lacks expansion card slot...

A laptop is not going to work well for uncompressed capture. You really need a power desktop with multiple drives to keep up with uncompressed data flow. DV would work if you had a Firewire port. You are limited to to USB2 for capture so your options are limited.

If I capture losslessly and end up with a 250GB file like in your example I suppose I can convert/burn this file into a compressed format so I won't have to buy a giant hard drive for every film I capture.

For professional uncompressed workflow, the idea is to filter and edit uncompressed (i.e. no recode losses), then encode to DVD MPeg2 or other formats as the final step. During editing, one must cope with very large working files.

As I mentioned, I own a JVC S-VHS with TBC/DNR, I guess this will not replace a full-fram TBC. Would I still benefit from aquiring a seperate TBC? and what about proc-amps? Are there any consumer level proc amps available for less than a few hundred dollars and do I really need one?

I haven't checked all my VHS tapes but I've played I few tapes origin from 1984-85 and they were actually in decent shape and all of my tapes contains material either recorded in Sweden or cassettes aquired in Sweden (a few in the UK) and is thus in PAL format.

I guess that I will be doing only minor post capture editing such as cutting.

This all depends how far you want to push the quality issue. VHS is a worse case source. It is noisy and has large timebase instability compared to camera or tuner source. Unfortunately, time base errors cannot be fixed post capture. The wandering picture elements need to be organized before A/D. Most digital filters and encoders assume stable pixel position.

The proc amp can be bypassed if the levels are approximately correct (e.g. TV tuner source). 8 bit processing leaves little flexibility for large levels corrections after capture. The quantization and rounding errors are too great. Pros digitize to 10bits or more.

So what are your USB2 options?

Normally this would force you to hardware capture directly to MPeg2. If you capture to DVD spec MPeg2 and use a "smart render" editor (e.g. Womble) you will suffer minimal recode loss. This works for cuts editing but any filtering or effects force a lossy recode.

There are some new USB2 capture products that capture directly to a digital intermediate format ideal for filtering/editing. So far, most of these are targeted to Apple's Final Cut Pro. I wish someone would make one that captures direct to Cineform.

[MrTemplar]

Unfortunately my laptop PC (Toshiba Satellite series) lacks expansion card slot...

A laptop is not going to work well for uncompressed capture. You really need a power desktop with multiple drives to keep up with uncompressed data flow. DV would work if you had a Firewire port. You are limited to to USB2 for capture so your options are limited.[/I][/I]

Yes, I'm planning to buy/customize a desktop computer shortly. I've thought about building a PC around the Intel i3 or the i5 processor (I will not play games with this computer) and whether or not to go for a capture device or an onboard video capturing card such as the Black magic Internsity pro, not sure really...

For professional uncompressed workflow, the idea is to filter and edit uncompressed (i.e. no recode losses), then encode to DVD MPeg2 or other formats as the final step. During editing, one must cope with very large working files.

This also requires a lot of RAM memory I suppose

I guess that I will be doing only minor post capture editing such as cutting.

This all depends how far you want to push the quality issue. VHS is a worse case source. It is noisy and has large timebase instability compared to camera or tuner source. Unfortunately, time base errors cannot be fixed post capture. The wandering picture elements need to be organized before A/D. Most digital filters and encoders assume stable pixel position.

I want as good quality as possible, however, my funds are limited and I can't afford more than aprox. $200 or 300. Hopefully my JVC S-VHS can fix some jittering...

The proc amp can be bypassed if the levels are approximately correct (e.g. TV tuner source). 8 bit processing leaves little flexibility for large levels corrections after capture. The quantization and rounding errors are too great. Pros digitize to 10bits or more.

Perhaps as stupid question, but, By "TV Tuner source", do you mean source material recorded from TV? Another stupid question; what do you mean by "8 bit"?

So what are your USB2 options?

Normally this would force you to hardware capture directly to MPeg2. If you capture to DVD spec MPeg2 and use a "smart render" editor (e.g. Womble) you will suffer minimal recode loss. This works for cuts editing but any filtering or effects force a lossy recode.

There are some new USB2 capture products that capture directly to a digital intermediate format ideal for filtering/editing. So far, most of these are targeted to Apple's Final Cut Pro. I wish someone would make one that captures direct to Cineform.[/QUOTE]

Cheers!

[lordsmurf]

4:2:2 uncompressed works fine via USB2, on a laptop. I do it quite a bit, when traveling.
Read this, on the ATI 600/650 cards: http://www.digitalFAQ.com/forum/showthread.php/ati-600-usb-2368.html
There's other posts there, too, on that topic. But I had that one bookmarked.

[edDV]

A desktop machine configured for uncompressed SD 4:2:2 capture would have a small drive for the OS and a fast Raptor or two drive RAID zero (normal PATA/SATA drives) for video capture and playback. The drives are stressed most when searching or scrubbing the edit timeline.

During capture and playback, the CPU/RAM have little to do. During filtering and encoding the CPU carries the load. Two to four GB of RAM are sufficient unless a specific filter requires more.

An alternate configuration captures uncompressed to a "lossless" codec like Huffyuv. The codec uses the CPU to lower the capture bit rate to fit withing the sustained data rate of a single drive. The CPU decodes the video during play, search or scrub. Timeline scrubbing will be limited by the CPU speed.

Perhaps as stupid question, but, By "TV Tuner source", do you mean source material recorded from TV? Another stupid question; what do you mean by "8 bit"?

"TV Tuner source" would be stable live broadcast vs. tape.

Capture cards typically use ITU Rec.601 specifications. Consumer capture devices usually capture to 8 bit luminance and chroma. Pro devices use 8 or 10 bit capture.

8bit (256 levels) defines level 16 as black and level 235 as white. The black to white range has 224 levels.

10bit (1,024 levels) defines level 64 as black and level 960 as white. The black to white range has 896 levels. The finer level slices reduce digital rounding errors resulting in less blocking artifacts after filtering.

[edDV]

4:2:2 uncompressed works fine via USB2, on a laptop. I do it quite a bit, when traveling.
Read this, on the ATI 600/650 cards: http://www.digitalFAQ.com/forum/showthread.php/ati-600-usb-2368.html
There's other posts there, too, on that topic. But I had that one bookmarked.

SD 4:2:2 capture bitrate is within the USB2 spec but internal issues in a typical laptop make sustained capture risky particularly for long VHS recordings. Many laptop drives won't sustain bit rates over 100 Mb/s* and are subject to frame drops due to OS processes using the drive or disk fragmentation (seek delays). Newer laptops with two drives or a second capture drive connected eSATA would have better safety margin.


* less at the slow end of the drive.

[vhelp]

You can still use (external) hardware for the TBC and Pro-Amp functions. Then connect those to your usb2 capture card (box or device) and then finally capture to a intermediate (lossless) codec in SD 4:2:2 format. Start with the Huffy codec. Lagarith is another lossless codec (smaller filesize) but it require more cpu and on my Acer netbook, it dropped frames while huffy did not.

Just recently, I did several captures using my Atom N450 w/ 1.66 GHz netbook and 1-gig ram over usb2 with the Hauppauge wintv-hvr-950Q card, and it performed quite well on that single core though slow machine. I had a little trouble with vhs but only because of the bloat-ware the machine contained plus the firewall and antivirus that was installed. I sent it back to revert backward the OS, win7se to winxp plus remove the bloat-ware and add 2-gig ram to it. Capturing from a satelite tv receiver was flawless. So, I can't wait to get it back so I can test these scenarios some more. Capturing from this notebooks and netbooks seem to yield better picture quality because they rule out the induced noise you get form the 300, 450, 500, or 600 watt UPS's. So that is another way of getting rid of noise in captured vhs sources. I am testing out different usb2 capture devices.

My next capture equip is the usb2 Pinnalce Dazzle DVD Recorder Plus, then ATI 650/750, which ever one I can figure out is the correct usb2 version to get, probably on amazon since i have two $50 cards of theirs waiting to be used next time one becomes available to them.

Ask lordsmurf for the link to the pro-amp and the exernal TBC, he posted one not to long ago but i can't find that discussion. I was tempted to get it myself but since i no longer have the link.. good luck.

-vhelp 5448

[MrTemplar]

Changing the direction a bit here; Anyone who's got experience using the Black Magic Intensity Pro card?

[2Bdecided]

For professional uncompressed workflow, the idea is to filter and edit uncompressed (i.e. no recode losses), then encode to DVD MPeg2 or other formats as the final step. During editing, one must cope with very large working files.

Standard cut/paste editing of DV is not lossy. Even transitions and titles aren't going to cause any loss that's visible after MPEG encoding for DVD.

With VideoReDo even cut/paste editing of MPEG is lossy.


It's not clear whether the OP intends to do any filtering. If not, capturing to lossless could be a waste of effort (if it requires effort to make it work on his machine).

Assuming the levels are OK, a VCR with TBC and DNR connected directly to a Canopus box is already better than most VHS>DVD transfers. Some people find capturing stops at black or damaged parts of the tape without an external TBC, but I've never seen that (Panasonic VCR, Canopus ADVC110).


A more general point: people always talk about capturing to lossless for subsequent filtering. Undoubtedly this gives better objective quality, but apart from some rare cases where people are rescuing images with more noise than picture (which get trashed by any lossy compression), I've not seen any examples where filtering a lossless capture works, while filtering a DV capture doesn't.

I do more filtering in software than most, and with decent sources it still works OK-ish from MPEG-2(!). There is a slight visible degradation if you look for it, but most normal people would judge it to be pretty minor (assuming 8Mbps CBR for capture and final DVD - lower bitrates can be catastrophic IMO). I can't see any degradation with DV. In any case, the final MPEG-2 encoding for DVD does more harm in complex scenes.

Most of the quality depends on the source - a pristine VHS captured with a cheap VCR + capture card may look better than ropey VHS captured with a decent deck, TBC, proc-amp, capture card, and filtering.

If you have more noise than picture, then lossless capture + software restoration is the only way to go IMO. Otherwise DV is fine IMO. And some people here think straight to MPEG-2 is fine too. Don't be pressured into spending more money than you mean to. Whatever you do, it'll still look like VHS - but hopefully very good VHS, rather than horrible wobbly soft noisy VHS.

Cheers,
David.

[jagabo]

A more general point: people always talk about capturing to lossless for subsequent filtering. Undoubtedly this gives better objective quality, but apart from some rare cases where people are rescuing images with more noise than picture (which get trashed by any lossy compression), I've not seen any examples where filtering a lossless capture works, while filtering a DV capture doesn't.

Some DV decoders duplicate U and V samples rather than using a smooth algorithm (bilinear, bicubic...) when upscaling chroma to YUY2 or RGB. That can lead to banding in smooth color gradients. You can get around that by using a different DV decoder or a "411" filter afterwards. There's a thread around here somewhere with some examples.

Here's one: https://forum.videohelp.com/threads/239994-DV-CODEC-for-VHS-Capture?p=1421497&viewfull=1#post1421497

And more info later in that thread:

https://forum.videohelp.com/threads/239994-DV-CODEC-for-VHS-Capture?p=2024423&viewfull=1#post2024423

[2Bdecided]

I've seen that illustrated on forums, but not in real life, since I use PAL so don't get the 4:1:1 problem 4:2:0 upscaling doesn't matter so much, since it'll be dropped back to 4:2:0 for DVD anyway.

I use the Cedocida DV decoder, which is great.


This isn't a case where you start with normal VHS with the aim of making a DVD, and lossless capture works while DV capture doesn't.

Cheers,
David.

[MrTemplar]

Just flicking in with a short question here:

Should I go for an external or onboard device when capturing analog video?

Pros and cons?

[lordsmurf]

With VideoReDo even cut/paste editing of MPEG is lossy.

That's not true in Womble MPEG Video Wizard. Cut/paste within timeline does nothing to audio/video, just moves it around. I doubt it's true in VideoReDo either, to be honest.

DV records frame by frame (no interframe compression) so is better for filtering, effects or compositing vs. MPeg based captures.

That's not really accurate -- the MPEG-2 "I frame only" variant is commonly used in the editing and broadcast industry. I deal with it all the time. Some of my most recent MXF master files were SD non-GOP MPEG-2 in the 20-35Mb/s range. Editors like Vegas and Premiere handle MPEG-2 just fine.

I don't really see anybody using DV anymore, and most of the ones that did now can't explain why they did it (excluding marketing, or because "they" said it was best to use). It was always a consumer format.

[edDV]

DV records frame by frame (no interframe compression) so is better for filtering, effects or compositing vs. MPeg based captures.

That's not really accurate -- the MPEG-2 "I frame only" variant is commonly used in the editing and broadcast industry. I deal with it all the time. Some of my most recent MXF master files were SD non-GOP MPEG-2 in the 20-35Mb/s range. Editors like Vegas and Premiere handle MPEG-2 just fine.

You are auguring a special case that is seldom used by "civilians". Consumer MPeg2 is 15 or 12 frame GOP based, interframe compressed.

The OP is in PAL-land. In PAL consumer DV, colorspace is 4:2:0* and intraframe DCT compression is 5x. This is nearly the same thing as 25 Mb/s I frame MPeg2 and both are superior to MJPEG for compression efficiency.

When you bring up 20-35 Mb/s MXF (usually Sony IMX) masters you are in a higher tier of acquisition than consumer MPeg2 capture devices. This is apples vs. oranges.


*DVCPro "PAL" uses 4:1:1 for lower generation loss in typical broadcast news workflow. Consumer DV "PAL" is 4:2:0.

[edDV]

I don't really see anybody using DV anymore, and most of the ones that did now can't explain why they did it (excluding marketing, or because "they" said it was best to use). It was always a consumer format.

For broadcast or industrial SD workflow flow, DV or DVCPro dominate IMX probably 95% to 5%.

If you shoot DV, you are best off editing native DV format.

Again you are arguing a special case. If your camera is DV, there is no advantage to IMX conversion unless you are talking about a special filter situation or importing DV into a 4:2:2 project. Most people doing 4:2:2 would be working uncompressed.

[2Bdecided]

With VideoReDo even cut/paste editing of MPEG is lossy.

That's not true in Womble MPEG Video Wizard. Cut/paste within timeline does nothing to audio/video, just moves it around. I doubt it's true in VideoReDo either, to be honest.

No, it's not, I missed out the word "not", i.e. I meant to type "With VideoReDo even cut/paste editing of MPEG is not lossy."!

Cheers,
David.