Two video outputs from XL H1s
 

I'm hoping you can help to educate me a bit. In summary, I need two video outputs from the XL H1s: one to record; one to monitor.

My application is:
- I use a video camera on an external gimbal (50-100 ft away) to document a scientific experiment occurring a few miles away.
- I've been using a Canon XL2 over the past few years because of it's acceptable price, 1.4 deg FOV and remote control of zoom and focus.
- I've been using a GPS clock and IRIG-B time video insertor to time-tag each (NTSC) video frame.
- I then record the video direct to DVD.
- I can "T" the NTSC video signal through a video amp and then monitor the video while recording. I need to monitor the video in realtime to follow the experiment as it moves around.

I'm considering an upgrade to this system to get finer resolution for the 2-5 mile experiment distances:
- Canon XLH1s
- use SMPTE timecode (from a GPS clock)
- record to Focus FS-CV (via firewire connection)
- I can get firewire extenders to cover the 50-100 ft of cable distance

I'm not sure how to monitor the video and record it at the same time:
- I'll use the firewire output to record to the FS-CV
- is there another video output port than can be active (at the same time) for video monitoring purposes?
- I don't think I can "T" off of the firewire because I think it is a bi-directional data path

Does anyone have any ideas?

Thanks...

Hey there,

The Canon Xl H1s actually has multiple video outputs that are all live.

If you need HD, you can monitor from the HD-SDI or Component outputs.

Their is also various analog video outs, including a 2nd EVF output which will display all of the viewfinder displays on an external monitor.

It really is a fantastic camera.

Hope this helps.

Hey Michael,

Thanks much for the info.

That is amazing! To think that they can simultaneously drive all of those ports! When you consider a single sensor (CCD sensor), that means that the little processors are simultaneously:
- acquiring data from the CCD sensor
- handling SMPTE timecode input, lens control
- reformatting the data for HDV and transmitting out firewire
- formatting and transmitting data out SDI port
- drive the EVF

OK, so now I have to figure out the best way to monitor the video (in the operations lab). I haven't seen any SDI-monitors. So, I guess I'll look for a component/HD monitor. I wonder if I can see the SMPTE timecode in this component/HD output (I'd like to have constant confirmation that I have proper timecode in the video stream)?

Thanks again...

Sure thing Richard!

The camera is actually a 3CCD camera, so it has 3 CCD sensors, one for each primary color.

There are in fact many HD-SDI monitors on the market that will work with the XL H1S. HD-SDI will provide the cleanest signal as it is pure uncompressed HD before the HDV compression. This is something the XL2 didn't have and it provides the ability to record the uncompressed HD-SDI stream, which will give you a clean, very high-end image. This is one reason why the camera costs what it does.

If displaying timecode information is important to you, you have 2 options with this camera. You can use HD-SDI and set the camera to display this information out that port.

You can also use Component out the EVF2 port (using the Canon cable that comes with the camera). This port will show you everything that the EVF of the camera displays (including peaking).

Thanks again, Michael... fantastic help!

I had found the 18 page brochure on the camera. But it doesn't have that level of detail. I'm hoping I can find the full manual in PDF (eventually).

Wonderful news that I can tell the camera to display timecode in the SDI output. And yes, I see... there are SDI monitors. A quick google search shows a JVC 24" for $1700 (expensive for a 24" monitor, but seems reasonable for an SDI-capable monitor).

I'm guessing that the EVF2 output is not HD... is that correct?

I need to decide if I want to record SDI (uncompressed video). It's a lot of data to record and process later. Not sure my end user wants to pay the data handling penalty for the best quality video (video is important but definitely lower priority data). If I record HDV, I don't get quite the quality, but it's better than SD and much lower data volume than uncompressed-HD. I always have the option to record HD-SDI later, if my end user gives me time+funding to implement a solution. So, maybe I'll start with HDV recording.

So, this is really looking pretty good. My approach (at this point) would be:
- XLH1s camera with the 20x HD lens (my existing XL2 lens does not have the resolution required for HD). I believe I can use my existing XL2 1.6x extender.
- SMPTE timecode input to the camera (I think I can just run a 100 ft BNC cable for this).
- Record HDV to Firestore FS-CV. I might need two units to cover a full 12-hour test day (I wish they had removable disks). I can record Quicktime for Mac end users, or .M2T for Windows folks. Then I can copy those data files to the user's primary data disks and they'll have their primary data and video on the same disks. I'm hoping that I can modify my data collection app so that I control the FS-CV programmatically to add experimental metadata (run#, run name, setup conditions) to the FS-CV data files.
- monitor the SDI signal with timecode display enabled. This allows me to verify that I have valid timecode in the video stream (I can also call this timecode out to other experiment players to check their timecode). And, by monitoring the video in HD I can better follow the experiment and help to avoid potential experimental problems.

If anyone has any other ideas or things to consider, feel free to weigh in.

That is correct. The EVF2 output is not HD.

Yes, the EVF2 port outputs a 480 SD signal.

If you want HD, your options will be Component or HD-SDI.

What is your timeframe in purchasing this camera?

Chris, Michael,

Thanks for the replies, and info.

I'm guessing that it would take us 2-4 months to decide if we want to upgrade from XL2 to HD. Once I study more of the options I'll do the economic analysis of costs vs pros and cons.

I think I've passed the feasibility study phase - this seems very possible. I think the next step is to rent/borrow the equipment to try things out. I have some additional operational/physics questions to address:
- given various atmospheric conditions (looking over ocean, on hot/humid day; desert conditions), is it worth it to have HD resolutions (is there too much atmospheric distortion over 2-5 mile distances)?
- If the SMPTE timecode signal is suddenly disconnected from the camera (and I have enabled display of timecode in SDI output), and I have an SDI monitor, will I see an indication of "loss of timecode"?
- Can I really transfer some data out of the FS-CV, while simultaneously recording data?
- Are the FS-CV Quicktime/.M2T files really compatible with Mac/PC?

You may also want to consider the XHG1 as a lower cost alternative. It also has HD-SDI and HD Component out. It has a fixed lens as the major difference from the XLH1.

They are both fine cameras. And with the money saved with the XHG1 you could get a nanoFlash to gain much higher quality video and audio over the HDV as well as gain file based tapeless work flow advantages.

Actually that should be XH G1S, which has since replaced the discontinued XH G1, but yes this is very good advice. There is no difference in image quality between the XL H and XH series; they both use the exact same image sensors and processors, same menu system, same custom presets etc.

Thanks John,

Maybe you can help me think through this option. I had considered this, but I thought:
- given my 2-5 mile experiment distances I wanted the 1.4 degree FOV that I get with the 20x lens with an added 1.6x extender.
- I have a 1.6x extender already (from my XL2) that will fit the XL H1s lens
- I would need to buy a new 1.6x extender for the different lens on the XHG1s. I think the cost of that is about $3k. So that makes the cost about the same as the XL H1s.
- and, I already have an outdoor enclosure for the XL2. I think the XL H1s will be an exact mechanical match.
- and, I think I could use the XL H1s lens on my XL2. So, my XL2 becomes my back up for the XL H1s, and I gain some "spares".

I like that nanoFlash... I hadn't found that yet... thanks. I did see the Flash XDR, but I like that nanoFlash. I'll keep that in mind if/when we decide to record SDI vs HDV/firewire. If I decide to record SDI then I don't think I can use that same SDI port for an SDI monitor. So, maybe I should look at HD/component monitors so that I leave the SDI port available for a recording option.

A single HD-SDI output out of the camera can "loop thru" to many other devices.

For a practical example:

HD-SDI out of the XLH1s into a nanoFlash HD-SDI input and then out of the nanoFlash HD-SDI output jack into the HD-SDI input of a monitor such as a Panasonic BT-LH1700W.

You could even go on from there if needed.

As to your lens issue, maybe the HLH1s (thank you Chris) would be the better choice afterall.

Thanks John,

I didn't realize that the SDI in/out ports were loop thru type electronics. That is very convenient.

I'm trying to figure out the component ports, to see if a HD-component-monitor would be cheaper and adequate. I'm guessing that the DSub15 (HD15) component connector on the camera would feed a standard monitor DSub15/VGA connector with the proper pinout and electrical signals... is that correct? I understand that a "VGA" monitor would not have the bandwidth to handle the HD signal, so I'd need an HD-compatible monitor ("WSXGA+"). But, I'm guessing that an HD monitor would prefer HD signal on a DVI or HDMI connector. So, maybe the "VGA" port of a HD monitor will not handle the component signal from the camera. Am I getting any of this right? Assuming that I'm wrong, and that an HD monitor with "VGA" connector would handle the XL H1s component signal, my next issue is how to run that signal 100 ft from camera to operations lab. I've been researching cables for VGA signals, but I don't see any that state specifically that they'll handle HD signals. I see loss specs for some cables (dB/100 ft) vs frequency range. And the frequency range is well into the 100s of MHz, but that doesn't mean that the cable is proper for the HD signal (ie. it may have ringing from impedance mismatches; or the loss may be too great for expected signal levels). So, I guess I'm asking:
- has anyone connected the XL H1s component output to a monitor, and gotten acceptable results?
- If so, which monitor did you use?
- what cable did you use (and what was the cable length)?
- has anyone tried running HD/component on 100ft+ cables? If so, what cable did you use?

BTW, I have tried searching the forums for these topics... no hits yet. But, if you know the topic has already been discussed feel free to tell me to use the search feature.

Don't know the answer to your VGA question,
but I can provide info on your component video questions.

Yes, the component out of the XLH1s works fine into a component HD monitor.

But your problem is the 100 ft distance. So if you can, try to stay with the HD-SDI output over the 100 ft cable run as HD-SDI can handle that fine with an HD rated coaxial cable.

But if you are going to try and use a component only monitor, then you could still come out of the camera HD-SDI port thru an HD rated coax cable for the 100 ft run to a Miranda PicoLink model SDM-171p device that takes the long HD-SDI cable run signal from the camera and then converts it to component to feed the component inputs from to the monitor.

Still think you would be happier with a monitor such as the previously noted Panasonic with HD-SDI input.

Thanks John,

I really appreciate your insights and advice.

Very good idea to put the SDI over the 100 ft cable and then convert the SDI to component/DVI/HDMI. I just found an AJA SDI to HDMI. Is that similar to the PicoLink? Then it seems I just need a vanilla HD monitor ($200, 21" monitor). I think I could start out with lower cost monitoring and upgrade later as I gain understanding of pros and cons with my application.

A note about my application vs realtime monitor needs:
- primarily, I need to keep track of where the experiment is (it moves around)
- a typical test day is daylight hours (10-12 hours). We'll collect data in 20-40 minute windows, with maybe a 60% duty cycle (6 hours of collection out of 10 hours).
- the test day is SO busy that there is no time to study the experiment in the video. Only after the entire test week is complete is there time to go back and study the video. So, that processing location is where the very good quality monitors should be used.
- so, in the field, I need realtime monitoring sufficient to:
- track the experiment
- see enough to tell if the experiment is configured properly (not really my responsibility, but I'm another pair of eyes that has at least a little time to monitor, even if I'm 2-5 miles away). I can often see interference coming before the folks in the experiment can see it.
- so, that's why I'm thinking that I could start with a vanilla-quality monitor in the operations lab.