Best means of analog to digital recording/capture at DVinfo.net

Andrew Hood · October 27th, 2005, 02:12 AM

I have an older camera I recently got (needed to play with something that has manual controls); JVC KY-17E head with interchangable video adapter, and BR411E S-VHS deck (all PAL). The SVHS part works quite well, but inevitably any recording and playback off tape adds to the noise, and degrades signal in an analog system. So I am looking for a practical (and affordable) means of converting to a digital form. Now I can simply record to DV by passing the composite from the SVHS deck, live to my sony TRV18E. But this is composite - even more noise before it goes digital.

This leads me to ways of trying to acquire a better image. There are other connectors on the camera, but I cannot find sufficient specs to make use of them - for fear of damaging connected equipment.
There is a YC443 7-pin connector, which I believe offers S-video, 4 of the pins (y, c, and their grounds), plus timing and power pins. But I cannot confirm this. Does anyone know where I can find, with some certainty, which pins do what?
There is also the RM/VTR 26 pin connection, and I believe that this contains component analog signals, but again I can't find pin assignment specs. Though apparently you can get some pricy converter cables - though I'd rather make my own given I have the coax and connectors - don't know about the impedence though.

Better yet if there's something more useful to be adapted/converted using the 50pin connector straight from the camera head, even better. But I have even less idea what all the pins are for, I'm guessing some may be redundant.
Or if anyone knows a way to connect specific bits/wires to a rigged up adc - might be able to make a more compact unit.

So if it is even possible to get S-video or composite - first hurdle, the next step is to do something with it. S-video I can capture on my camera, or there are capture devices for the computer - just a matter of finding one that's reliable and supports full res and frame rate. Eventually I might get a firewire capture device and have the whole setup 'portable'.

As for component, very few cards/boxes support component video, but the hope is this would give me a better live conversion to DV footage. I think the ADS pyro av link does.

I would really like to be able to get 4:2:2 rather than 4:2:0 capture, but it seems you need to get a component to SDI converter, and then an SDI capture card - very high end and very expensive. And prices rise exponentially the more chroma data you want to preserve.
The real problem is that everything on the low end uses DV (4:2:0 fixed in codec / 4:1:1 NTSC) - be nice if you could hack the codec in the converter, but not likely. Everything above that, is way above and is geared towards uncompressed - at least from what I have seen. Some do mpeg encoding in realtime, but are prohibitive based on price. Is there anything to fill the gap in between?

I realise DV capture, if converted from analog as early in the pipeline will yield much better results than what I'm getting for clarity and reproducability with SVHS tape. But I would really like to preserve some of that colour richness - is that too much to ask? Maybe just too much money.

Also if anybody has any SUBSTANTIATED (eg. not crackpot) ideas for making a camera like this read progressive frames, please make suggestions. I have seen stuff about the Vancecam, but that was a special case 'feature' of some cameras. It may one day be feasible to perform open heart surgey and replace the CCDs with higher def progressive ones, I'll have to see how these other homemade projects go. But for now I just want the recording quality to do justice to the 3 current CCDs and lens of the system.

Andrew

Keith Wakeham · October 27th, 2005, 02:45 PM

Your first stop, if you don't have an oscillascope, is a service manual. Short of finding a service manual you can get into experimentation - not (sometimes though) fun and kind of risky. - I'll try and detail the risky part but please don't blame me if you break your camera, tv, multimeter, etc.

If you can find a multimeter and a ground then just point the camera at a white card while on and make sure most of the video looks white (the closer to being to bright and clipped is best) and measure the dc voltage from the output pins. Find all those that are close to 1.0V. Then put a lens cap on or just make the video all black somehow. You should find one (or two or three) of the 1.0V pins is now 0.3. If you have one then that should be the luma line, if you have three then you have component and you need to get a tv that can test these out. Trying all three one after the other into the tv's luma line (composite may work) and you should be able to determine what is what - hopefully.

Hope that is useful and you don't break anything

Andrew Hood · October 29th, 2005, 09:34 PM

Thanks Keith

I've figured out which is the luma and chroma channels from the connector, and I don't think I broke anything. I didn't get much from the voltmeter other than that the amounts fluctuated, but within a narrow range (< 1V). I ended up testing it using my older computer with the capture card - some very sophisticated wires, with reshaped paper clips for the camrera end, and sewing pins for the S-video plug. Though I still don't know which pins on the camera correlate to ground. This could be a problem if I wanted to hook it up to my camera (once I have a proper plug for it) if the grounds aren't wired properly. I'm guessing the capture card just grounds through the computer.

Does anybody know how many capture cards can capture full PAL resolution at 4:2:2? It should have a WDM capability so that something like VirtualDub can real-time compress and store the video - preserving the 4:2:2. Is there detailed information on which codecs can support 4:2:2 with a good compromise between space and quality? If the camera is outputting only analog data, I want to make the most of that data when it's first digitised - especially for any green/blue screen work.

Keith Wakeham · October 30th, 2005, 07:46 AM

Those are some High quality connects - what ever works, glad you found the right lines. I know you didn't have much luck with the multimeter but usually all grounds are not isolated, they all connect, so you should be able to figure out which are gnd by measuring the resistance between the composite ground and the pins o the connector. Something low of <1 ohm would definetly indicate a ground.

can't really help you on the capture card.

Good luck

Daniel Kohl · October 31st, 2005, 02:10 AM

Quote:

Originally Posted by Keith Wakeham

- I'll try and detail the risky part but please don't blame me if you break your camera, tv, multimeter, etc.

Just curious,

What is the risky part?

Andrew Hood · November 1st, 2005, 05:31 AM

OK I think I was mistaken in the last post. Anyway, I now know which pins are the corresponding grounds for the c and y channels. I even had it written down, but just not in an obvious spot. So I just need a way of creating a connector that will hold a 14mm bayonet clip so I can attach to a sacraficed Svideo cable when i get one.

@Daniel - the risky part is when you're dealing with anything electrical and playing with individual wires. Luckily this is 12 VDC going in, and the out voltages are <1. Though voltage is only part of the story. If the current is near you heart's natural current (and the current can change due to the body's resistance I think) it can disrupt or stop it beating. You're in strife when the brain is running low on oxygen from lack of circulation.

I did notice some vertically scrolling fluctuations in the image intensity. I'm hoping this is just noise/interference due to using unshielded cables, and it will be eliminated when I have a better connection. And I will be using the camera adaptor module later on, but I've been testing with the SHVS deck for the moment.

Mark Donnell · December 6th, 2005, 03:06 PM

Andrew - I don't know if this will be helpful to you, but I have used the Canopus ADVC-300 unit to convert all of my old Hi-8 tapes to digital with excellent results using the S-video input (I don't recall if it has component in or not). You can make about ten different adjustments during the conversion including color hue, saturation, etc . The unit also includes line-based TBC and other error-correction routines. I think it sells for about $ 400, but in my opinion it was well worth the money.

Robert Horvath · February 16th, 2006, 11:19 AM

YC443 connector pinout that I found in my I-DEN timebase corrector's manual:

1 - Y(S), Y(U), Y(C)
2 - 1's GND
3 - R-Y
4 - 3's GND
5 - B-Y, C443, C924
6 - 5's GND
7 - NC

As you see, this is a multi-standard connector. It supplies signals for Betacam, U-Matic, S-Video (1-Y, 5-C).

I scanned the picture - too bad cannot post attachments :(

Cheers!
Rob

October 27th, 2005, 02:12 AM	#1
Andrew Hood Major Player Join Date: Oct 2005 Location: Latrobe Valley, Vic Posts: 320	Best means of analog to digital recording/capture I have an older camera I recently got (needed to play with something that has manual controls); JVC KY-17E head with interchangable video adapter, and BR411E S-VHS deck (all PAL). The SVHS part works quite well, but inevitably any recording and playback off tape adds to the noise, and degrades signal in an analog system. So I am looking for a practical (and affordable) means of converting to a digital form. Now I can simply record to DV by passing the composite from the SVHS deck, live to my sony TRV18E. But this is composite - even more noise before it goes digital. This leads me to ways of trying to acquire a better image. There are other connectors on the camera, but I cannot find sufficient specs to make use of them - for fear of damaging connected equipment. There is a YC443 7-pin connector, which I believe offers S-video, 4 of the pins (y, c, and their grounds), plus timing and power pins. But I cannot confirm this. Does anyone know where I can find, with some certainty, which pins do what? There is also the RM/VTR 26 pin connection, and I believe that this contains component analog signals, but again I can't find pin assignment specs. Though apparently you can get some pricy converter cables - though I'd rather make my own given I have the coax and connectors - don't know about the impedence though. Better yet if there's something more useful to be adapted/converted using the 50pin connector straight from the camera head, even better. But I have even less idea what all the pins are for, I'm guessing some may be redundant. Or if anyone knows a way to connect specific bits/wires to a rigged up adc - might be able to make a more compact unit. So if it is even possible to get S-video or composite - first hurdle, the next step is to do something with it. S-video I can capture on my camera, or there are capture devices for the computer - just a matter of finding one that's reliable and supports full res and frame rate. Eventually I might get a firewire capture device and have the whole setup 'portable'. As for component, very few cards/boxes support component video, but the hope is this would give me a better live conversion to DV footage. I think the ADS pyro av link does. I would really like to be able to get 4:2:2 rather than 4:2:0 capture, but it seems you need to get a component to SDI converter, and then an SDI capture card - very high end and very expensive. And prices rise exponentially the more chroma data you want to preserve. The real problem is that everything on the low end uses DV (4:2:0 fixed in codec / 4:1:1 NTSC) - be nice if you could hack the codec in the converter, but not likely. Everything above that, is way above and is geared towards uncompressed - at least from what I have seen. Some do mpeg encoding in realtime, but are prohibitive based on price. Is there anything to fill the gap in between? I realise DV capture, if converted from analog as early in the pipeline will yield much better results than what I'm getting for clarity and reproducability with SVHS tape. But I would really like to preserve some of that colour richness - is that too much to ask? Maybe just too much money. Also if anybody has any SUBSTANTIATED (eg. not crackpot) ideas for making a camera like this read progressive frames, please make suggestions. I have seen stuff about the Vancecam, but that was a special case 'feature' of some cameras. It may one day be feasible to perform open heart surgey and replace the CCDs with higher def progressive ones, I'll have to see how these other homemade projects go. But for now I just want the recording quality to do justice to the 3 current CCDs and lens of the system. Andrew

October 27th, 2005, 02:45 PM	#2
Keith Wakeham Major Player Join Date: Jan 2005 Location: St. John's, NL, Canada Posts: 416	Your first stop, if you don't have an oscillascope, is a service manual. Short of finding a service manual you can get into experimentation - not (sometimes though) fun and kind of risky. - I'll try and detail the risky part but please don't blame me if you break your camera, tv, multimeter, etc. If you can find a multimeter and a ground then just point the camera at a white card while on and make sure most of the video looks white (the closer to being to bright and clipped is best) and measure the dc voltage from the output pins. Find all those that are close to 1.0V. Then put a lens cap on or just make the video all black somehow. You should find one (or two or three) of the 1.0V pins is now 0.3. If you have one then that should be the luma line, if you have three then you have component and you need to get a tv that can test these out. Trying all three one after the other into the tv's luma line (composite may work) and you should be able to determine what is what - hopefully. Hope that is useful and you don't break anything __________________ www.engr.mun.ca/~wakeham/index.htm

October 30th, 2005, 07:46 AM	#4
Keith Wakeham Major Player Join Date: Jan 2005 Location: St. John's, NL, Canada Posts: 416	Those are some High quality connects - what ever works, glad you found the right lines. I know you didn't have much luck with the multimeter but usually all grounds are not isolated, they all connect, so you should be able to figure out which are gnd by measuring the resistance between the composite ground and the pins o the connector. Something low of <1 ohm would definetly indicate a ground. can't really help you on the capture card. Good luck __________________ www.engr.mun.ca/~wakeham/index.htm

December 6th, 2005, 03:06 PM	#7
Mark Donnell Major Player Join Date: Nov 2005 Location: Silver City, NM Posts: 385	Analog to DV Andrew - I don't know if this will be helpful to you, but I have used the Canopus ADVC-300 unit to convert all of my old Hi-8 tapes to digital with excellent results using the S-video input (I don't recall if it has component in or not). You can make about ten different adjustments during the conversion including color hue, saturation, etc . The unit also includes line-based TBC and other error-correction routines. I think it sells for about $ 400, but in my opinion it was well worth the money.

February 16th, 2006, 11:19 AM	#8
Robert Horvath Tourist Join Date: Feb 2006 Location: Budapest Posts: 2	Yc443 YC443 connector pinout that I found in my I-DEN timebase corrector's manual: 1 - Y(S), Y(U), Y(C) 2 - 1's GND 3 - R-Y 4 - 3's GND 5 - B-Y, C443, C924 6 - 5's GND 7 - NC As you see, this is a multi-standard connector. It supplies signals for Betacam, U-Matic, S-Video (1-Y, 5-C). I scanned the picture - too bad cannot post attachments :( Cheers! Rob Last edited by Robert Horvath; February 16th, 2006 at 12:19 PM.

October 29th, 2005, 09:34 PM	#3
Andrew Hood Major Player Join Date: Oct 2005 Location: Latrobe Valley, Vic Posts: 320	Thanks Keith I've figured out which is the luma and chroma channels from the connector, and I don't think I broke anything. I didn't get much from the voltmeter other than that the amounts fluctuated, but within a narrow range (< 1V). I ended up testing it using my older computer with the capture card - some very sophisticated wires, with reshaped paper clips for the camrera end, and sewing pins for the S-video plug. Though I still don't know which pins on the camera correlate to ground. This could be a problem if I wanted to hook it up to my camera (once I have a proper plug for it) if the grounds aren't wired properly. I'm guessing the capture card just grounds through the computer. Does anybody know how many capture cards can capture full PAL resolution at 4:2:2? It should have a WDM capability so that something like VirtualDub can real-time compress and store the video - preserving the 4:2:2. Is there detailed information on which codecs can support 4:2:2 with a good compromise between space and quality? If the camera is outputting only analog data, I want to make the most of that data when it's first digitised - especially for any green/blue screen work.

November 1st, 2005, 05:31 AM	#6
Andrew Hood Major Player Join Date: Oct 2005 Location: Latrobe Valley, Vic Posts: 320	OK I think I was mistaken in the last post. Anyway, I now know which pins are the corresponding grounds for the c and y channels. I even had it written down, but just not in an obvious spot. So I just need a way of creating a connector that will hold a 14mm bayonet clip so I can attach to a sacraficed Svideo cable when i get one. @Daniel - the risky part is when you're dealing with anything electrical and playing with individual wires. Luckily this is 12 VDC going in, and the out voltages are <1. Though voltage is only part of the story. If the current is near you heart's natural current (and the current can change due to the body's resistance I think) it can disrupt or stop it beating. You're in strife when the brain is running low on oxygen from lack of circulation. I did notice some vertically scrolling fluctuations in the image intensity. I'm hoping this is just noise/interference due to using unshielded cables, and it will be eliminated when I have a better connection. And I will be using the camera adaptor module later on, but I've been testing with the SHVS deck for the moment.