MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Panasonic SEG <-> LZX sync
MUFF WIGGLER Forum Index -> Video Synthesis  
Author Panasonic SEG <-> LZX sync
smrl
Hey, I've been trying to get a Panasonic WJ-4600C SEG to sync to my LZX with no luck. Not having the appropriate bits to test it otherwise I assumed something was wrong with the unit. When one of the WJ-5600's came up on ebay at a good price, I picked it up. It's exhibiting exactly the same behavior which makes me believe it's not the unit that's messed up but my configuration.

Details:

color camera -> panasonic genlock in (genlock switch illuminates, camera looped thru & terminated at another channel to ensure that the camera and the panasonic are synched)

panasonic pgm out -> LZX video in, sync switch engaged, Red light illuminated. LZX system out -> another channel of the panasonic. Panasonic pgm out to capture device.

I also tried LZX as the master clock source, feeding it into the SEG genlock in the same fashion as the color camera, and I get very much the same behavior: strobing color patterns cascading down the screen...

I've messed with the phase shift of the color subcarrier on the SEG, and can't get the image to stabilize. If I feed some b/w video from the LZX through the SEG the image appears to be stable, though at high frequencies it appears jagged.

I've included pictures showing horizontal/vertical b&w bars at 50% wipe with SEG color bars, then the output adding color. I also included a picture showing the output from the camera is synched and color is displaying.

Also possibly noteworthy is that I notice some artifacts in my color bars as soon as I add the LZX - the magenta and cyan can be seen to have slight fluctuating color patterns...

Any ideas? These units should be AMAZING if I could get them to play nicely with the LZX...


(




lizlarsen
It would be easier to see what's going on if you posted a video, maybe -- but I'm wondering if the Panasonic doesn't like the way the LZX is producing the color subcarrier at the output. In the LZX system, color subcarrier is generated independently in the encoder, rather than distributed as a master reference throughout the system -- this is not an issue at all for display or recording, or mixing thru digital/TBC based systems, but could exhibit some issues when used in configurations like this. Do you have a timebase corrector you can put between the LZX and the Panasonic to test out this theory?
smrl
http://youtu.be/nla7YyAMGoM

Here you go, unfortunately my only TBC is busted, else I would've already tried that...

After I made this video the SEG mysteriously locked onto the color subcarrier (or at least the scrolling color pattern froze) and I shot some footage of messing around with the phase adjustments... Let me know if any of that stuff is helpful. I suppose I could also hook it up to a scope and see if I can get a comparison shot of the two colorbursts - haven't really tried photographing it but I might be able to give some info if there's something I should be looking for...

[edit] i realized the subcarrier froze because I genlocked the SEG to the LZX instead of the other way around...
daverj
The subcarrier out of the LZX encoder is not locked to the genlock input of the LZX. So if you genlock the LZX to the SEG, and feed the LZX into the SEG, the color from the LZX won't be locked to the SEG.

If you genlock the SEG to the LZX, then the LZX encoder should have the right color when going through the SEG. But you won't be able to mix it with the color camera and maintain correct color from both. You would need either a TBC or a genlockable color camera.
smrl
http://youtu.be/fvXBmt2ekbk

Thanks for the input, Dave...

when I genlock the SEG to the LZX output, I get the following - the color subcarrier is frozen but there is a phase shift across the image - in this video clip I'm messing with the phase controls, pretty much at random.
daverj
When you genlock the SEG to the color camera instead of the LZX, do it's colors come out correct?
smrl
Yes, the camera appears to work fine!
lizlarsen
This seems like a weird issue! If there were any sort of color phase weirdness, I wouldn't expect to see that gradient rainbow pattern like that -- I'd just expect to see maybe some jitteriness in the color response -- but I don't know. It definitely looks like a problem with the subcarrier. If you're sending in black & white bars and seeing that saturated color like that, it would seem to me that somethings out of spec with the amplitude of the colorburst coming out of the SEG. Do you see anything different if you send color video out of the LZX into the SEG?

Dave, you have much more experience with interfacing with the older gear than I do, thanks for chiming in on this.

I'm not really sure what to try next, but I'll be thinking about it.
smrl
Yeah, the video i posted had subtitled explanations of what's going on, but the clip is not showing up even though it's been up on youtube for 8 hrs at least... but to clarify, the black and white bars appear just fine, it's when I turn up the color bias or change the LZX video encoder to soemthing other than black/white/grey that this behavior emerges.

The only difference between synching the LZX to the SEG and synching them the other way around is that when the LZX is the sync source, it's possible to stop the 'scrolling' of this color pattern (as in the 2nd video)

i'm going to see if i can figure out why the 1st link isn't working.
lizlarsen
That's very odd! Yeah get that other video working if you can.

Stopping the "scrolling" of the gradient when LZX Color Encoder is the sync source makes sense. The gradient being there at all doesn't.
smrl
now it's working
lizlarsen
Thanks! Yeah, I see the issue. It looks like sync-lock is happening just fine (synchronization of Vertical & Horizontal) but it's having trouble with the color subcarrier. That's expected when you're not using the LZX as the master sync source, but it's the weird gradient that happens that is confusing me!

I'll alert my engineer Ed to see if he has ideas. Maybe Dave will have some ideas on what's happening, too.

At this point I'd suggest you try any other devices (cameras, VCRs, etc.) you have on hand to see if anything else recreates the same issue you're seeing with the LZX input.

What happens if you run the camera thru the LZX and use that as a video source instead of bars? Doesn't quite make sense that it would make a difference, but maybe the SEG doesn't like the hard-edged waveforms.
daverj
Strange that the color camera looks green. If you genlock to the color camera, it should look like normal colors. I suspect the subcarrier lock or phase adjustment of the SEG genlock is not set correctly.

When genlocking the LZX to the SEG it makes sense that the color image from the LZX through the SEG has flickering and rolling colors, since it's subcarrier isn't locked to the SEG. But even free-running subcarrier from the LZX should be closer in frequency to the frequency of subcarrier in the SEG and camera.

My only guess is that the subcarrier frequency of the LZX is off by a fair amount and that monitors can handle that difference, but the genlock in the SEG can't. On the other hand, the fact that when locking to the color camera it is green makes me think that the subcarrier lock of the SEG is not correct.

When you are genlocking the SEG to the LZX, you are not also trying to lock the LZX to the SEG at the same time, are you? That would cause a loop that would screw things up. If genlocking the SEG to the LZX, try not putting anything into the LZX genlock input.
lizlarsen
Another point to note here, is that there is a tuning trim-capacitor for the genlock frequency on the back of the color encoder module itself. It is a little green trimmer on the top board. You can try adjusting that while viewing the output to see if it makes a difference for you.
smrl
YES! It was the frequency trimmer on the back of the video encoder! Works perfectly now!

Thanks to both of you!
lizlarsen
Awesome!! I don't know why I didn't think of that sooner.
I should probably have something like a properly calibrated SEG around here so that I can trim that module to within a tighter tolerance range.
smrl
It's so odd, I would've thought that the SEG would have a wider tolerance than other devices. Isn't the subcarrier usually locked with a PLL?

Anyway, here's a screen grab of my success. There's still some wobbling in the LZX video and I notice some 'ringing' or 'ghosting' kind of artifacts on the SEG colorbars - I don't really have the terminology here, it looks like slightly wavy lines of higher luminance in the color bars, near the edge of a few of the colors. I will spend a bit more time tweaking it. I'm also eager to see if the LZX will now magically work with my WJ-4600...



LZX is on the bottom. Also, I noticed with the bias controls I can get the video signal out of SEG compliance - so bright the image starts getting corrupted & losing sync. It's only at the extreme top range, though. My SEG probably could use calibrating. Now to find the service manual...
lizlarsen
Awesome! Yeah, due to the asynchronous subcarrier, there may be some color weirdness with a setup like this. Ed suggested that we could possibly modify a CVE with an expander panel to accept an external subcarrier input (from the SEG) if we wanted a more proper genlock option for the LZX system -- it's probably something I'll try to investigate on my own at some point. The simplest thing, though, is probably having a TBC for the LZX in a setup like this. Our upcoming Color Video Decoder & Frame Synchronizer module, as an output conditioner for the CVE before the composite video leaves the LZX system, would do the trick (and you get all sorts of extra features for your final output then, too.)
daverj
If you are talking about fine zig-zag pixels moving up at the edge of where two colors meet, that's called "chroma crawl" and that is the nature of composite video. It's the point where sudden brightness and color changes cause the subcarrier to be picked up as luminance information.

It happens whenever two high saturation colors of different hue are right next to each other. It happens more with certain color combinations.
smrl
re: chroma crawl. I think that might be what I'm talking about, but is that an exclusively static phenomenon? I'm noticing these kind of wispy patterns slowly moving up the left-hand side of the cyan.
lizlarsen
can you take a 720x480 screen cap?
smrl
Wish i could. Having trouble getting my capture card (bt787-based hauppage) to capture at a higher resolution than that.
daverj
Chroma crawl (aka dot crawl) is easier to notice on static images like color bars or colored graphic shapes, and colored text. It happens in encoded images. It doesn't happen in component video.

You seldom see it in camera images because you don't get those high saturation colors right next to each other in most real life images.
MUFF WIGGLER Forum Index -> Video Synthesis  
Page 1 of 1
Powered by phpBB © phpBB Group