Interesting, Phil -- I had forgotten about line-sync cameras, even though I have one looking at my front porch right now.· I bought it for $8 at a yard sale.· It's been at my front window for eight years.· It doesn't approach the video quality of even the most ordinary consumer vidicams, though.
Jjanes, if you do get a GenLockable camera, it ought to have a name.· I like to give things names.· My pickup is Toulouse (short for Toulouse La Truck); my tiny motor home is The Peregrine Mansion; and so forth.
If you wind up with two cameras, one with GenLock and one without, you could call the syncable one Titanic and the unsyncable one Molly Brown.
Interesting, Phil -- I had forgotten about line-sync cameras, even though I have one looking at my front porch right now.· I bought it for $8 at a yard sale.· It's been at my front window for eight years.· It doesn't approach the video quality of even the most ordinary consumer vidicams, though.
Jjanes, if you do get a GenLockable camera, it ought to have a name.· I like to give things names.· My pickup is Toulouse (short for Toulouse La Truck); my tiny motor home is The Peregrine Mansion; and so forth.
If you wind up with two cameras, one with GenLock and one without, you could call the syncable one Titanic and the unsyncable one Molly Brown.
Cute:· The Unsyncable Molly Brown.· I had a friend who, no matter how serious the situation, would always work a pun into the conversation.
I just bought the Basic Stamp Discovery Kit at Fry's Electronics.· I'm looking forward to working through the labs.
I'm not trying to sync the camcorders because, off the shelf, they are not syncable.· I am just trying to get a reasonably close sync for 3d stereo video so I need to be able to determine what the phase difference is between the two.· I have two Canon HV20 HDV camcorders.
With the circuit and code Phil provided I should be able to get a useable sync.· I don't need 15 minutes:· if I could get 2 or even 1 minute of useable sync reliably, that would be great.· When is the last time you ever saw a shot on television or film last for even 30 seconds?
The next camcorders I get will definitely have genlock.· I'm hoping the new Canon HF s100 will.· Check out the specs if you're interested.· (But I don't see genlock.)
Besides, I have several inventions that I'd like to prototype.· I've been thinking of them for years.· I was wondering when and how I was going to get around to them.
My initial reason for my thread here and for starting the BS curriculum was for measuring sync difference between two prosumer camcorders.
I am on lesson 4 - Servos in the WAM book.· I could probably proceed with Phil's circuit knowing what I know now about the BS but I'd like to finish the WAM book first.
I have since determined a way to measure sync difference visually and fairly accurately.
What do you think:
Getting·two Canon HV20s in (useable) sync for the purpose of making stereoscopic video.
I just heard of a method called, "spinning wheel sync," used to measure the sync difference between two still cameras. I think this method can be used with camcorders, too.
Have a spinning wheel or circle with a line drawn through its center from edge to edge. Have both HV20s showing this spinning wheel in their LCDs.
Adjust the speed of the spinning wheel so the line on the disk remains stationary in one LCD. Note the exact position where the stationary line is: the angle. Do the same thing with the second HV20.
Now compare the difference in the angles between the two stationary lines. This is the sync difference between the two camcorders. You want this difference to be as small as possible to be useable for stereoscopic video depending on your subject matter.
You also want the difference between the two stationary angles to remain as stable as possible for as long as possible but you have no control over this yet.
I wouldn't mess with the spinning disk. There are too many uncertainties, including:
1. Does the camera use frame-at-a-time or rolling shutter exposures? (See this image.)
2. Does the LCD display an interlaced image. (If so you'd see two lines, not one — even more with a rolling shutter.)
By the time you get done building and monkeying with a spinning disk, you could have had the electronics built and working for you several time over.
-Phil
Addendum: After rereading the LM1881 spec sheet, I see that you can (and should) eliminate the 75-ohm termination resistors from my circuit. The '1881 will work as well with a 2V p-p signal as with 1V. -P.
Post Edited (Phil Pilgrim (PhiPi)) : 2/27/2009 7:27:32 PM GMT
Phil Pilgrim (PhiPi) said...
I wouldn't mess with the spinning disk. There are too many uncertainties, including:
1. Does the camera use frame-at-a-time or rolling shutter exposures? (See this image.)
2. Does the LCD display an interlaced image. (If so you'd see two lines, not one — even more with a rolling shutter.)
By the time you get done building and monkeying with a spinning disk, you could have had the electronics built and working for you several time over.
-Phil
Addendum: After rereading the LM1881 spec sheet, I see that you can (and should) eliminate the 75-ohm termination resistors from my circuit. The '1881 will work as well with a 2V p-p signal as with 1V. -P.
(I had to delete my original reply.)
I'll be building the circuit before I do anything else.
The standard 3.5mm jack has at the tip and going up:· sound left, video, ground, sound right.
How do you recommend I take the output from the camcorder?
I was thinking about taking a 3.5mm jack and cutting the cable and just attaching the jack's video wire to the input of the circuit.
Let's say, as you suggest,·I eliminate the 75 ohm resistor at the input making this branch open.
In either case, why do I need to connect the camcorder ground to the circuit ground as shown in your schematic?
(Keep in mind that I am rusty on my electronics.)
Is there any way I could fry something in a camcorder following the above procedure?
That's a four-conductor plug, so your plan to use a pre-terminated cable is a good one. I've had to do this myself and recommend this part, which you can cut in half — one half for each camera. You definitely need to connect both video and ground to the sync detector. All circuits require a complete current path to function. (That's why they're called "circuits"! ) If you're only going to be using the circuit single-ended to the sync detector instead of daisy-chained to a PC capture card, say, you might as well leave the 75-ohm resistors in. Otherwise, it's best to omit them or make them jumper-selectable. As to frying the camcorder, just be careful not to short anything in your cable setup. Even then, its outputs are likely to be protected from this sort of thing.
It would be very interesting to use the delay info to synchronize both videostreams.
I would like to encode a field-sequential 3D videostream out of two cheap camera modules without sync capabilities. As a newbie I am
asking, if one or two Propellers could do following job:
No, there's not enough RAM in the Propeller to store an entire frame of video. Even if there were, you'd get false 3D artifacts from the cameras not being in sync with each other, arising from motion across the field of view.
Phil Pilgrim (PhiPi) said...
That's a four-conductor plug, so your plan to use a pre-terminated cable is a good one. I've had to do this myself and recommend this part, which you can cut in half — one half for each camera. You definitely need to connect both video and ground to the sync detector. All circuits require a complete current path to function. (That's why they're called "circuits"! ) If you're only going to be using the circuit single-ended to the sync detector instead of daisy-chained to a PC capture card, say, you might as well leave the 75-ohm resistors in. Otherwise, it's best to omit them or make them jumper-selectable. As to frying the camcorder, just be careful not to short anything in your cable setup. Even then, its outputs are likely to be protected from this sort of thing.
-Phil
I do plan on leaving the resistors in.
I agree that all circuits must have a complete path from·- to + for electron flow.· And your circuit with the two LM 1881s does.· But....
But I don't have a problem with connecting the camcorder output ground to the circuit ground.· I was just thinking it was unnecessary in this case.· (Maybe) I'll try both ways and see what happens.· Do you think this would be alright?
jjanes said...
I was just thinking it was unnecessary in this case. (Maybe) I'll try both ways and see what happens. Do you think this would be alright?
No. It is necessary to connect the grounds. I'm a bit perplexed why that would be a problem.
jjanes said...
I was just thinking it was unnecessary in this case. (Maybe) I'll try both ways and see what happens. Do you think this would be alright?
No. It is necessary to connect the grounds. I'm a bit perplexed why that would be a problem.
-Phil
Certainly, it is not a problem for me.
It goes to the heart of my understanding of electronics and electricity.
There was a gap here.
I think I have resolved it.
I'm kind of smart but I'm kind of stupid, too.·
Just to go back to line syncable cameras, most major CCTV manufacturers do PAL and NTSC colour cameras that are line sync and exceed 480TVL resolution. They produce *excellent* pictures. While the 480TVL models are quite expensive the 380TVL models are quite affordable and offer a wide selection of lenses. Panasonic offer a range that is also completely genlockable for not much more $$$.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Cardinal Fang! Fetch the comfy chair.
Comments
Jjanes, if you do get a GenLockable camera, it ought to have a name.· I like to give things names.· My pickup is Toulouse (short for Toulouse La Truck); my tiny motor home is The Peregrine Mansion; and so forth.
If you wind up with two cameras, one with GenLock and one without, you could call the syncable one Titanic and the unsyncable one Molly Brown.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net
I just bought the Basic Stamp Discovery Kit at Fry's Electronics.· I'm looking forward to working through the labs.
I'm not trying to sync the camcorders because, off the shelf, they are not syncable.· I am just trying to get a reasonably close sync for 3d stereo video so I need to be able to determine what the phase difference is between the two.· I have two Canon HV20 HDV camcorders.
With the circuit and code Phil provided I should be able to get a useable sync.· I don't need 15 minutes:· if I could get 2 or even 1 minute of useable sync reliably, that would be great.· When is the last time you ever saw a shot on television or film last for even 30 seconds?
The next camcorders I get will definitely have genlock.· I'm hoping the new Canon HF s100 will.· Check out the specs if you're interested.· (But I don't see genlock.)
http://www.usa.canon.com/consumer/controller?act=ModelInfoAct&fcategoryid=177&modelid=17992
Besides, I have several inventions that I'd like to prototype.· I've been thinking of them for years.· I was wondering when and how I was going to get around to them.
So here I am.· And here I come, Basic Stamp!
jj
I am on lesson 4 - Servos in the WAM book.· I could probably proceed with Phil's circuit knowing what I know now about the BS but I'd like to finish the WAM book first.
I have since determined a way to measure sync difference visually and fairly accurately.
What do you think:
Getting·two Canon HV20s in (useable) sync for the purpose of making stereoscopic video.
I just heard of a method called, "spinning wheel sync," used to measure the sync difference between two still cameras. I think this method can be used with camcorders, too.
Have a spinning wheel or circle with a line drawn through its center from edge to edge. Have both HV20s showing this spinning wheel in their LCDs.
Adjust the speed of the spinning wheel so the line on the disk remains stationary in one LCD. Note the exact position where the stationary line is: the angle. Do the same thing with the second HV20.
Now compare the difference in the angles between the two stationary lines. This is the sync difference between the two camcorders. You want this difference to be as small as possible to be useable for stereoscopic video depending on your subject matter.
You also want the difference between the two stationary angles to remain as stable as possible for as long as possible but you have no control over this yet.
jj
····
1. Does the camera use frame-at-a-time or rolling shutter exposures? (See this image.)
2. Does the LCD display an interlaced image. (If so you'd see two lines, not one — even more with a rolling shutter.)
By the time you get done building and monkeying with a spinning disk, you could have had the electronics built and working for you several time over.
-Phil
Addendum: After rereading the LM1881 spec sheet, I see that you can (and should) eliminate the 75-ohm termination resistors from my circuit. The '1881 will work as well with a 2V p-p signal as with 1V. -P.
Post Edited (Phil Pilgrim (PhiPi)) : 2/27/2009 7:27:32 PM GMT
I wouldn't mess with the spinning disk. There are too many uncertainties, including:
1. Does the camera use frame-at-a-time or rolling shutter exposures? (See this image.)
2. Does the LCD display an interlaced image. (If so you'd see two lines, not one — even more with a rolling shutter.)
By the time you get done building and monkeying with a spinning disk, you could have had the electronics built and working for you several time over.
-Phil
Addendum: After rereading the LM1881 spec sheet, I see that you can (and should) eliminate the 75-ohm termination resistors from my circuit. The '1881 will work as well with a 2V p-p signal as with 1V. -P.
(I had to delete my original reply.)
I'll be building the circuit before I do anything else.
The standard 3.5mm jack has at the tip and going up:· sound left, video, ground, sound right.
How do you recommend I take the output from the camcorder?
I was thinking about taking a 3.5mm jack and cutting the cable and just attaching the jack's video wire to the input of the circuit.
Let's say, as you suggest,·I eliminate the 75 ohm resistor at the input making this branch open.
In either case, why do I need to connect the camcorder ground to the circuit ground as shown in your schematic?
(Keep in mind that I am rusty on my electronics.)
Is there any way I could fry something in a camcorder following the above procedure?
jj
-Phil
It would be very interesting to use the delay info to synchronize both videostreams.
I would like to encode a field-sequential 3D videostream out of two cheap camera modules without sync capabilities. As a newbie I am
asking, if one or two Propellers could do following job:
The AD9280 ADC is 8-Bit, 32 MSPS www.analog.com/static/imported-files/Data_Sheets/AD9280.pdf or
as DAC using the ADV7170 www.analog.com/static/imported-files/data_sheets/ADV7170_7171.pdf and
as ADC using the ADV7180 www.analog.com/static/imported-files/data_sheets/ADV7180.pdf
having own Sync in/outputs so no LM1881 is needed anymore.
- Valerio
No, there's not enough RAM in the Propeller to store an entire frame of video. Even if there were, you'd get false 3D artifacts from the cameras not being in sync with each other, arising from motion across the field of view.
-Phil
Yes, correct. Did you already see such artifacts? I am wondering if they would be noticeable in 3d video-glosses with a resolution
of 640x480 ...
tnx a lot
Valerio
I agree that all circuits must have a complete path from·- to + for electron flow.· And your circuit with the two LM 1881s does.· But....
But I don't have a problem with connecting the camcorder output ground to the circuit ground.· I was just thinking it was unnecessary in this case.· (Maybe) I'll try both ways and see what happens.· Do you think this would be alright?
jj
-Phil
It goes to the heart of my understanding of electronics and electricity.
There was a gap here.
I think I have resolved it.
I'm kind of smart but I'm kind of stupid, too.·
Thank you.
jj
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Cardinal Fang! Fetch the comfy chair.