How do Live Switchers Work?

[Eric Lagerlof]

Natan, if you keep the FX-1 in HDV it will downrez the video out of the composite and S-video outputs as letterboxed. You get the 'wider' looking picture, but much of the SD 4x3 frame is taken up with black on the top & bottom, i.e. letterboxed. If you shoot in DV, you get the same frame w/out the letterbox. Either way, it's still 720x480. The one advantage of shooting HDV is that in that mode, you do have HDV on tape as backup.

Some of these decisions as to recording format, HD v SD, 16x9 v 4x3 are determined by the specifications of the gear and your decisions around this are going to have to be based on your and your clients needs. In much of what I've done, recording the shows has been secondary to 'i-mag', or image magnification. Putting the video and computer images on big screens live during the event. So video is switched in SD and uprezzed to 1024x768 or higher and a second switcher switches between the video input and computers. Recording is often secondary. All this to say that different clients/situations emphasize different needs.

At your price point, 4x3 SD is probably what your going to both be able to afford and most likely your best distribution format anyway. So make good use of the 4x3 format, embrace SD ;-)

And, if your client(s) want more, i.e. HD, they can pay more...

BTW, NewTeks' VT5 and Tri-Caster, at higher price points, might be worth a look if you get a bigger budget down the line. You can switch, up & down rezz, record and have digital playback all from the same unit. And I believe they will be offering an HD version of at least one of these boxes soon. I checked out a VT4 once and the video looked very good.

[Natan Pakman]

Originally Posted by Shaun Roemich

Not if you're switching composite it's not. PLUS you gain 1 to 2 stops of light by doing so. Recording to tape, you see a huge difference. By the time you get to a composite out downconverted, all things are pretty equal. I know, I do this as part of my business model.

16:9 in SD is strictly a pixel aspect ratio change. If you read my above response, you have all the info you need.

Shaun,

So the bad SD quality that I have seen when recording footage from the HDR-FX1 onto tape does not correspond to the quality it would have if I sent the dv stream via composite or s-video to a switcher and then onto a deck or computer?

Can one send SD footage through component cables? As a corollary, do SD switchers tend to have component, or only composite and S-video?

[Eric Lagerlof]

In your price range, composite and S-video. There is component in SD and the FX-1 can offer it via a menu selection, I believe. But in your price range I doubt you'd find the kind of component switcher you are looking for available.

As I said earlier, use S-video , not composite; composite degrades the signal, especially strong colors, too much.

[Shaun Roemich]

Natan: first, remember you aren't sending a DV stream out the analog outputs, you're sending the signal from the DSP BEFORE it hits tape (and DV compression). You'll need to plug your camera into a monitor and see what the actual output looks like and whether it meets your needs in DV mode.

In response to Eric's post: yes, S-video will provide superior quality HOWEVER the wires used in S-Video cables are typically 28 gauge and there are four sets of braided conductors in there. IF THERE IS ANY POSSIBILITY of your cables being stepped on or having chairs placed over them or being rolled over by carts DON'T EVEN CONSIDER using S-Video. At best you'll end up with a non-functional cable, at worst (this happened to me in a live switched shoot where I couldn't change cable for 5 hours!), the image will short out and go inverse in chroma. BAD BAD BAD!

If you can hide your cables AND you're willing to spend the premium on S-Video cables, go ahead. If not, composite is your friend, just make sure to use good quality 75 ohm RG-59.

[Eric Lagerlof]

Shaun makes a good point about s-cables. If it was me, and I was going to buy and then often use a multicam system, I would be tempted to create special camera cables, i.e. 2 regular bnc cables taped together per camera and a 2-bnc to S-video adapter at each end. The BNC cables are tougher and you get less signal loss on longer runs, i.e. over 100 feet.

You get the best of both worlds. For me, the difference in signal quality would be worth it.

[Natan Pakman]

I still have question regarding aspect ratios.

If I want to switch between two 16:9 (non-letterboxed) sd sources, are there sd switchers that can output the 16:9 non-letterboxed signal?

My biggest issue is getting the non-letterboxed footage.

What switchers could do this?

[Shaun Roemich]

16:9 SD signals can be switched by ANY video switcher AS LONG AS all inputs are 16:9 anamorphic. At the projection/display or recording side (actually editing side) you need to tell the display or editor that the material is anamorphic to get i to stretch back out to widescreen.

Of course, your iso, preview and program monitors will look funny if they aren't switchable as well.

[Natan Pakman]

Shaun,

Thanks for you the quick response. So does a typical, or reasonably logical, workflow consist of the following?

1. Cameras connected directly to switcher (and a computer connected to a scaler, and the scaler to the switcher).
2. Possibly audio equipment connected to the switcher.
3. The switcher connected a a recorder of some kind, possible a hard disk recorder?

Am I missing any necessary elements?

[Shaun Roemich]

My flow is:
- Cameras/scan converters/playback devices/graphic devices attached to iso monitors
- Iso monitors looped through to switcher inputs
- Switcher preview output to preview monitor
- Switcher program output 1 to recorder
- Switcher program output 2 to program monitor
- Recorder output to separate monitor

Audio:
- all audio sources routed to an audio console and mixed "properly"
- audio output of console sent to recorder
- Recorder audio sent to record monitor for verification

The only REAL problem with this et up is that if one starts adding TOO much in the paths, you can begin to introduce signal delay between the audio and video (although in my experience, it's always been tolerable or non-existent)

Of course, this can get far more convoluted when one starts adding in multiple mix-effects busses, several upstream and downstream keyers, graphics units that generate their own alpha channel etc.

I do several events a year using my own setup as above with great results. Barring cameras, my setup is currently in a 10 rack unit case but I'll be breaking everything out into 2 smaller cases shortly.