DV Info Net

DV Info Net (https://www.dvinfo.net/forum/)
-   Open DV Discussion (https://www.dvinfo.net/forum/open-dv-discussion/)
-   -   Live firewire switching - Sony Anycast (https://www.dvinfo.net/forum/open-dv-discussion/79315-live-firewire-switching-sony-anycast.html)

Wade Spencer November 10th, 2006 09:57 AM

Live firewire switching - Sony Anycast
 
Greetings. I'm planning a live shoot that I do every year in December for a local university. It's a choral holiday concert, and we shoot with 4 cameras, live mixed with some post production done later.

Up until this year we've been shooting with Sony UVW-100 Betacam cameras. They have been retired in favor of JVC HD110s. The live mix will be done with the Sony Anycast station with the main mix recorded to external drive via DV Rack, and the independant streams from each camera sent to external hard drive in case I need to correct an error somewhere. And audio signal will just be sent through the anycast so I have something to sync to...the audio is recorded separately with a ProTools system.

In the past I always ran composite BNC cable from the cameras to the mixer, but now with the JVC cameras I have the option of using firewire. I have some questions and would appreciate some help!!

1. I assume that the quality will be better, yes?

It is a 4 camera shoot...camera one will be about 10 feet away from the Anycast, cameras two and three about 45 feet away, and camera four 75 to 80 feet away.

2. I know there is delay with firewire....since camera four will probably have a 100 foot cable between it and the Anycast, do the other 4 cameras need to have the same length of cable to have the same delay? I can delay an incoming audio signal with the Anycast to match delayed video, but I can't delay video inputs independently.

3. Looking at firewire cable through a catalog I have here (I don't think I can mention the name) I see 'extended distance' firewire cables up to 164 feet. I am not sure what is different about them, but I am assuming that if I buy those I won't need to use any sort of repeater?

The HD firmware update is out for the anycast now, and once the HD input cards are available I'll probably upgrading to those and running component. I don't think I'll have them by this concert though.

I appreciate the help everyone!! Thanks!!

Wade Spencer November 11th, 2006 10:53 AM

Anyone at all with any experience with this?

Another forum I could try perhaps?

Thanks!

Vincent Rozenberg November 11th, 2006 12:02 PM

Quote:

Originally Posted by Wade Spencer

1. I assume that the quality will be better, yes?

Yes, a lot better. And that's an understatement.

Quote:

2. I know there is delay with firewire....since camera four will probably have a 100 foot cable between it and the Anycast, do the other 4 cameras need to have the same length of cable to have the same delay? I can delay an incoming audio signal with the Anycast to match delayed video, but I can't delay video inputs independently.
No need, length doesn't matter in this case.

Wade Spencer November 11th, 2006 12:32 PM

So, just to confirm, the delay will be the same even if 1 camera has a 50 foot cable and another camera has a 100 foot cable? If I understand this right, the delay is in the DV codec, so length won't make it any worse or better, correct?

What about a repeater? I won't need one for a 100 foot run with what are called 'extended distance' cables? Or is that just a marketing ploy?

Thanks for the response!

Ervin Farkas November 11th, 2006 05:55 PM

The delay is caused by the digitizing process, cable length doesn't matter... or does it?

The Firewire protocol is built around cables of 25 feet max. But you can use so called "in-line repeaters" to send signal up to 400 feet (one repeater every 50 feet). See http://www.bhphotovideo.com/bnh/cont...=298590&is=REG for a good one. For your application you can buy two 50 feet cables and join them via the repeater - cables longer than that might be good for other applications but not for DV.

Steve Leverich November 11th, 2006 09:07 PM

When connecting digital data streams, one of the main things that affects distance they can be run is cable capacitance - a pulse train needs about 10 times the frequency bandwidth that a sine wave needs in order for the leading/trailing edges of pulses NOT to be rounded off. Cable capacitance limits the length of cable before rounding becomes unacceptable.

Depending on the input circuit the signal is seeing, there is usually a pulse shaper that re-instates nice, square pulses based on crossovers of the input signal - this is also what a repeater does.

There is still a finite delay in any transmission line, but I'm not familiar (yet) with sync techniques used by digital video hardware; it's possible that they are just using the "slowest" video signal and delaying all other inputs to match it. IF so, then cable length should be unimportant beyond causing too much rounding/tilt of individual pulses for the signal to be salvageable.

Have you seen this thread

http://dvinfo.net/conf/showthread.php?t=49398

Doesn't sound like your path is gonna be glitch-free... Steve

Pat Miller November 11th, 2006 10:11 PM

firewire cable length is built around the standard of 15 feet. If you're looking at the cables from Markertek you won't need repeaters and the delay you'll actually have is in the beginning of triggering the recording process in your Anystudio. Once you've begun recording on all cameras (and I would suggest still using tape as backup) you'll be recording each cameras video and sound output in real time. I do it alot with 2-4 cameras all the time and the longer firewires' have worked fine.

Greg Boston November 11th, 2006 11:14 PM

Everything Steve said is true. In addition to capacitance causing the pulses to become misshapen over long distances, the internal resistance of the cable wiring will attenuate the voltage levels over a long distance and a repeater should also restore the signal voltage level.

The long distance cables will have less internal capacitance and resistance per foot, but a repeater might still be a good idea.

-gb-

Wade Spencer November 13th, 2006 09:52 AM

Thanks everyone. I think at this point it sounds like I'd be best to use a repeater with two 50' cables.

For the cameras that are under 50 feet away though...will I need to still use the repeater with two 50' cables? Will my delay be different?

Thanks!

Steve Leverich November 13th, 2006 03:11 PM

As I've already stated, the delay WILL be different; the question is, does the Anycast compensate for this?

I followed the link I posted in my last post, and it does NOT lead to a pdf; so I looked around a bit, and only found a product bruchure - this doesn't say one way or another whether the DV inputs (or any inputs) are automagically delayed to match each other.

Since we all know the meaning of Ass-U-Me, I'd recommend you call Sony tech support and ASK about this - if that's not possible for some reason, I would then "Ass-U-Me" that the unit does NOT automagically correct, and have enough cables/repeaters on hand to make all your firewire lines electrically the SAME.

The one thing that's kept my backside from being warm and wet (as in, yer butt's in hot water now) over the years is pessimism; I try never to assume something will be OK if there's ANY reasonable way to MAKE SURE it will be.

If it's cost of extra cables/repeaters that's keeping you from doing this, here's one possible option - if said cables/repeaters come in sealed bags, buy enough for identical 100' runs, set up with just the lengths you need, and DON'T OPEN the extras til you see if it works without them; then you may be able to return the extras for credit after the gig.

If you talk to Sony tech support (and can understand them thru whatever swahili accent) make sure you ask them HOW the inputs are delay compensated - if they can't answer that, I'd STILL get the extra cables/repeaters (how's THAT for pessimism :=)

That's my take on this; just call me pessimistically prepared... Steve

Wade Spencer November 13th, 2006 03:26 PM

Quote:

Originally Posted by Steve Leverich
As I've already stated, the delay WILL be different; the question is, does the Anycast compensate for this?

I followed the link I posted in my last post, and it does NOT lead to a pdf; so I looked around a bit, and only found a product bruchure - this doesn't say one way or another whether the DV inputs (or any inputs) are automagically delayed to match each other.

Since we all know the meaning of Ass-U-Me, I'd recommend you call Sony tech support and ASK about this - if that's not possible for some reason, I would then "Ass-U-Me" that the unit does NOT automagically correct, and have enough cables/repeaters on hand to make all your firewire lines electrically the SAME.

The one thing that's kept my backside from being warm and wet (as in, yer butt's in hot water now) over the years is pessimism; I try never to assume something will be OK if there's ANY reasonable way to MAKE SURE it will be.

If it's cost of extra cables/repeaters that's keeping you from doing this, here's one possible option - if said cables/repeaters come in sealed bags, buy enough for identical 100' runs, set up with just the lengths you need, and DON'T OPEN the extras til you see if it works without them; then you may be able to return the extras for credit after the gig.

If you talk to Sony tech support (and can understand them thru whatever swahili accent) make sure you ask them HOW the inputs are delay compensated - if they can't answer that, I'd STILL get the extra cables/repeaters (how's THAT for pessimism :=)

That's my take on this; just call me pessimistically prepared... Steve


Thanks Steve. I cannot delay video inputs independantly on the Anycast, and I wouldn't bank on the unit doing it automatically, but I will call to Sony tech support to confirm that.

I think my safest bet is the scenario that you mentioned....buying enough cables and repeaters for four identical 100' runs. Cost is not prohibitive.

Thanks for the response!

Sharyn Ferrick November 13th, 2006 05:38 PM

Something to keep in mind
When we talk about the difference in quality of composite vs firewire transfered dv, we are usually talking about recorded difference, NOT an un recorded signal.

What I would suggest is that you get access to a professional monitor, and carefully compare the analog output from the camera to the dv output converted to analog and see if there is the difference you expect.

If you are looking to record the output from each camera to dv, i think there is an option for that on the anycast but i am not sure if it restricts the number of inputs. but i think it will not allow you to do that AND send the final signal via dv to the DV rack, I think it uses at least ONE of those connections. (sony hasthis method of on each firewire output connecting a drive and recording two synced files of dv) If on the other hand you are looking at recording the dv output from each camera, then I think you might have problems recording the stream and also feeding the stream to the anycast.

IMO what I have seen is that DV as a recording format vs analog is of course vastly superior, BUT a analog output from the camera, compared to a dv converted to analog output is IMO identical, and of course has none of the DV conversion delay. We tend to forget that most of the quality lost is during the recording not the transmission. IN a pro setup, you would have the option of Composite Component or SDI for the transmission. With the camera you have you pretty much have Composite, Y/C and with the method youare using DV converted.

Videonics and Focus Enhancements had a DV switcher that was less than a major success. IMO a lot of the problem was the issues with long firewire cables, the data errors that were introduced, and that as a INPUT DV was lacking since it needed to be converted back and forth. Personally I think the future is more along the lines of the BLACKMAGIC intensity card where via HDMI you can input UNCOMPRESSED HD video, and the the solutions that take compressed dv and un compress them are less attractive.

One other problem that i have found with the firewire connectors etc is that they are extremely Fragile and unreliable in a live event situation. If someone could come up with a quality locking connector it would be a great improvement. It is clear IMO that the connector was designed to be used for transfer to a computer etc, and not in a live environment. I use the DV Rack and it is a fantastic product, but attached to the live camera the connections are problematic. Attached to the mixer as you are proposing is fine.

Anyway just some thoughts to consider. I remember questioning at an NAB years ago several of the pro camera companies when they were demoing their cameras for resolution and quality etc, and noticed that most of them were simply being connected via composite bnc cable to the monitor as to why they chose that method. They then were the ones talking about the differences in transmitting vs recording.
Sharyn

Wade Spencer November 13th, 2006 06:03 PM

Thanks for the reply Sharyn. I have been wondering myself about the firewire connections in a live event, but the JVC cameras do have 6 pin connections, which will be a little better. The cameras will all be mounted on tripods, so not a lot of moving around will be happening where the cable could get pulled loose.

I'll only be using 4 of the inputs on the anycast...I'll be using an extra video input card for the third card and can use the DV connections on that to output a main mix firewire signal. The two extra firewire connections on the two in-use cards will be used to record the full streams from each input to external hard disk. I did that last year at this same even and it worked great.

I'm not sure if this will be worth the extra effort...I would just like to get some better quality if I can. SDI would be preferable, but I don't have four HD250s, or the SDI cards for the anycast at this time. If I can I'll do some tests and post up some screenshots.

Sharyn Ferrick November 14th, 2006 12:48 PM

It will be interesting to see how the tests go for you
We did a test using the videonics/focus enhancments dv switcher and there was no advantage over y/c in input quality. I have found that long cables or multiple repeaters increase the errors in the data, and my experience has shown that these errors based on firewire from camcorders can only be fixed via error correction vs re transmission (firewire for networking or hard drives is a different implimentation as far as errors than firewire from camcorders) SO again in our experience, we simply had more errors.

I found that for the best quality outside of component or sdi was to use dual rg6 coax, with a y/c to dual bnc connections. The 18awg core of the rg6 and the fact that the cable was identical in length avoided any problems with running long yc runs. The output in analog was then only converted once.

Anyway should be a fun situation

we do a lot of live multi camera stuff, mainly using for portable set up an upgraded mx50. We have a full Utah Scientific pro mixer but that is a royal pain needing genlocking or syncing.

I am looking at the new hd camcorders with hdmi or component out and the Blackmagic intensity card. Since it only has two inputs, I have been trying to get more information about using a hdmi switcher on the front end, and also some of the gefen vga component to hdmi converters. It still will be pretty rube goldberg but the potential for a significant step up in quality makes it interesting. I was talking to Blackmagic about supporting more than two cards in a pc but the problem seems to be more slots that will work
Sharyn


All times are GMT -6. The time now is 01:57 PM.

DV Info Net -- Real Names, Real People, Real Info!
1998-2024 The Digital Video Information Network