F3: F3 problem w Samurai and Ki-Pro

L

Lenilenapi

Carbonite Member
I have been having the most inexplicable problem I can recall in over 30 years.

I shoot a quarterly interview w a CEO against green screen and we've been shooting it with an F3. When they went to green screen we added a Ki Pro which worked fine a few times , then one day the recorder kept stopping and restarting through the run always leaving a few seconds out at each juncture. No steady rhythm- sometimes short sometimes a few minutes before it glitches.

Shooting 1920x1080 29.97 P We are using a Sony paintbox, and though we have a cable loom w BNC and audio cables , we use a separate BNC from the SDI A to the recorder ,

We could not get this to happen again back at the shop no matter what we did - so don't know what the problem was. Tried every combination of settings and triggers.

So we switched to a Samurai - Great for 2 quarters then something similar happened. This time there are little glitches or jumps in the recording where the image jumps to the side for a frame. Occasionally it stops and restarts. Then sometimes it is fine for stretches of a few minutes.
However this time i realized I had erased the drive in the computer rather than reformatting so I figured maybe that was it.

Nevertheless I tested and tested at home - every combination of erasing in computer, triggers, source on HD-SDI, I could think of to no avail. could not repeat the problem. But i figured must be that I didn't Format.

Next quarter it happens again. This time we test beforehand and run for 5 min a few times and it seems fine. But we do have an odd problem in that the Samurai won't start from the Camera trigger. So we hit record on the samurai- and sure enough we see glitches only when our CEO is talking. Thank God for XDCAM internal back-up .

Nowt have just spent a day trying everything I can to repeat this problem and nothing goes wrong - using the same Sandisk Extreme drive 240G SSD and the bad files are still on it - it hasn't been touched.- AC/ battery / with and without triggers on the Samurai or recorder, blah blah blah.

Any ideas. I am at a complete loss. Will send bad file to Atomos and put a small sample up here.
 
Have you tried switching out the BNC cables? They were always the first thing to cause problems for me with external recording from the F3.
 
Yes , both the tech and I are 30 year veterans so we've solved problems like this before. i think we've tried switching everything we can think of. Of course that doesn't mean we didn't miss it in the heat of battle, but we switched cables on the job and thought we h we were OK.
 
Grug said:
Have you tried switching out the BNC cables? They were always the first thing to cause problems for me with external recording from the F3.
Click to expand...

I can double up on this. Bad cables caused me issues. They didn't even have to be "bad". Just micro, micro glitches in the SDI stream that a monitor will cruise through will send a recorder for a loop.

It's either this or the SSD.

Please forgive me, I mean no insensitivity because I know it's work and you need to get it fixed, but it was stuff like this that made me never rely on external recorders again. Just too flaky.
 
Maybe you're correct that its some tiny cable glitch, but then wouldn't changing cables have fixed it? or wouldn't I be able to recreate it? I've tried twisting the cables and connections while recording and it had no effect.
I tried recording to the same SSD over and over without being able to recreate it.

What did your cable problems look like Nate?
 
I can attest that when this has happened to me it's been an SDI cable issue. In some cases, the connection was so fragile that when the operator's hand grazed the cab lightly it would disrupt the connection.

Samurai should start from Sony trigger. Make sure you're on current firmware, then go into the "Timecode" menu and select "Sony" from the "Record Trigger" options. Also, make sure the camera is set to output the SDI record trigger.
 
For me it was just grazing the cable, sometimes any movement whatsoever. It seemed to me at the time like it was more about the fit of the BNC connectors...if it wasn't in FULL contact, (i.e.inner conductor and outer locking barrel with zero wiggle), just the barrel rocking back and forth a 1/100th of an inch would cause enough change in the signal to cause a small glitch. A glitch which, btw, wasn't really visible when not recording. Just when rolling.

Christian's advice I"m sure is important too. I was definitely guilty of not getting cable 100% made for SDI, just getting short BNCs and not checking whether they were suitable for analog composite or SDI. But since I knew it was happening at the connectors, I didn't go farther with things.
 
It's happened with 2 different cameras ( and 2 different recorders) so not the SDI port; firmware is latest on both units and I know how to configure the Sony triggers.

I am listening very carefully to advice above on SDI cables though. Turns out we have been using regular BNC not the more expensive SDI cables so that could be the cause. We did switch cables around a few times when having trouble but perhaps had multiple culprits. Still seems odd that I could not replicate later even with the same cables though.

I'm going to try to replicate again at home using cables I know to be dicey. If I can make it happen that would relieve a great deal of anxiety about this.

Does anyone know if the Samurai is more prone to these problems than say a PIX 240. A cl;ode partner uses the same cables with his PIX and has never had a problem. Uses the same Sandisk Extreme SSD's also.

One thing I will do is get new cables specifically for external recorders though.

Oh here is a long shot left field question:

The one thing I haven't been able to include in my testing as it didn't seem relevant, is that the signal from the Samurai out to our monitor is then looped through 2 more monitors to clients. Could that have any effect on the strength of the signal to the Samurai in the first place????

Thanks for all your help guys. I should have posted this a while ago.
 
Lenilenapi said:
The one thing I haven't been able to include in my testing as it didn't seem relevant, is that the signal from the Samurai out to our monitor is then looped through 2 more monitors to clients. Could that have any effect on the strength of the signal to the Samurai in the first place????
Click to expand...

If the SDI signal is not reconstructed by a chip inside the Samurai (and I doubt it is, on such an inexpensive device), then I think absolutely that *could* be the culprit.

Grounding, impedance, termination issues I think could all wreak havoc with the Samurai if it's a simple loop out through the Samurai (and again, not expressly reconstructed electrically for the loop out)
 
Hmm That's interesting. I'll check with Atomos about that. You'd think it would be designed for that purpose though because its mainly used with gear that has only one SDI or HMI In and the only way to monitor is looped through.

In our case we're going out to a 17" Flanders by the tech then to a 21" Flanders by the client. On occasion I believe we have also looped further into a 5.6" by the operator (Me). However I think I've been pulling my SDI from the monitor-out on the F3 which I assume is a separate draw from the SDI-A out.

Would the stress on the signal be related to the size of the monitor? I ask because I don't have another full size SDI monitor to loop to at home for testing. I could loop out to a TV logic 5.6" however and even to my Alphatron EVF. Would those affect signal strength as much as a 21" monitor?
 
Like you need more advice... I'm going to hazard a guess that it is your SDI cables not being 75 ohm. That right there is more than sufficient cause for it dropping out.
 
Just tried a bunch of tests with terrible BNC cables (actually they were RCA cables with adapters)and putting a 5.6" monitor and an alphatron EVF on the SDI loop through from my 17" monitor . Nevertheless I could not recreate the problem I had when recording in the field. The crappy SDI cable would break up if I moved it and that would halt recording, But it acted the same even without recording (just monitoring) and did not create the kind of video glitches I have been seeing. I really need to be able to replicate this problem to know what actually went wrong.

Sample video compressed to H264 1920x1080 - no sound

https://vimeo.com/108068930

password:

bad samurai

(please no comments on anything but the glitches)
 
Last edited:
Well just heard from Atomos. They did not say whether they've looked at the example, or respond directly to how they produce the SDI=OUT , but i did get this:

"As for your issue, I don't believe it has to do with the cables. Since the cables are digital, the signal should be rendered 100% accurately, or not be present at all. There is a small possibility that the loop out on your first monitor created the glitch, but I don't think this is very likely."

They've asked me to send it in to their shop.

 
You must log in or register to reply here.
Back
Top