The SSE Skinny

[Jim Giberti]

There's a lot of specualtion from, "it's been fixed in recent firmware versions" to "it only happens at 18db in low light".

Given our experience in the past month with the HD100 and my recent interaction with the people who should be most knowledable regarding this at JVC, I'd have to say that neither of those speculations are true.

I mentioned recently that it didn't make sense to say that SSE was eliminated in new cameras simply because JVC has infomed everyone of it's "potential" under certain circumstances and how to work around it.
In my production company's change over to HD I accepted this potential as the way that JVC had implemented true 24p/720p. The camera makes beautiful pictures and is certainly among the best of the cameras out there right now.

JVC never fudged the issue nor accepted it as a flaw...it is simply something that could happen given the unique way that their R&D people found to deliver true 24p/720p.

So, to the point, here is my interpretation (I'm not going to quote people verbatim or claim things categorically) but this is from my converations with the informed people there.

It doesn't matter what your serial number is.
There has been no new firmware update that was meant to "fix" SSE.
There have been two firmware upgrades - one to address an improper reading of Anton Bauer systems and another that I can't recall specifically.

The 24p system is based on the two DSP model and they need to be set to very demanding tolerances to insure that they display the exact image on both sides.
This calibration is done on a bench with wavefrom and vectroscope and the settings are burned into the EPROM.
There is also, as you probably know from your literature, a separate circuit that monitors the two signals to keep them in parity.
It's not a firmware issue...it's a calibration issue.

Now I did receive one email that was a bit "cryptic" in that it used this expression - "in light of the current "implementation" plans at JVC...."
The word "implementation" was actually set off in quotes in the email.
Again, this is only my interpretation, but that leads me to think that if anything, perhaps JVC has refined the way they calibrate the DSPs.

Most importantly though, if you haven't seen SSE (and it hasn't been a problem here except for the one time), then I think JVC has made it's point.
They put out a great HD camera at a great price and delivered true 24p in a way that has a possible issue under certain potential circumstances.

The point that was made that "if youre getting SSE at 0db then send your camera into JVC makes sense.

But..."If you're seeing it" is different from "if you have seen it".
One would suggest that it's a common occurence, the other could well fall under the "it could happen" that JVC stated before you bought the camera.

Was the camera not "warmed up" enough when those scenes were shot?
That's possible.
In fact it's very cold here in the mountains right now and I'm going to take the HD100 out tomorrow and try and recreate the issue by intentionally shooting it cold under similar conditions and backgrounds.

Again, JVC has been exceptional in their response to us, and as we've continued to tweak this camera, I've gotten more and more comfortable moving forward with a second and third HD100 purchase.

The newest settings that are a hybrid of our tests, Tim Dashwood's and Paolo's DSC color adjustments, are really great. I've got an All-purpose scene file that is, IMO, better in color and latitude than my previous experience with the Canon and Panasonic.

It's pretty much all good.

[Stephen L. Noe]

Originally Posted by Jim Giberti

....Again, this is only my interpretation, but that leads me to think that if anything, perhaps JVC has refined the way they calibrate the DSPs....

You are correct in your assumption. There is a machine now that does the calibration that replaces the 'by hand' process. I understand there is only one of them (USA) and it's in California. This was brought up at the ProHD users group meeting here in Chicago.

[Marc Colemont]

I also experienced, if I detect the SSE, reboting the camera 2x seems to trigger the calibration between the 2 DSP's enough to get them right.
I thinks it's just a matter of time when they figure out a better algorithm to auto-calibrate in the firmware. If they would make a short-cut in the firmware were you need to hold two buttons on the camera for a few seconds, to force a recalibration. That would help a lot. Changing the gain settings does trigger the re-calibration also sometimes to get it right again.

[John Vincent]

Originally Posted by Marc Colemont

I also experienced, if I detect the SSE, reboting the camera 2x seems to trigger the calibration between the 2 DSP's enough to get them right.
I thinks it's just a matter of time when they figure out a better algorithm to auto-calibrate in the firmware. If they would make a short-cut in the firmware were you need to hold two buttons on the camera for a few seconds, to force a recalibration. That would help a lot. Changing the gain settings does trigger the re-calibration also sometimes to get it right again.

That's a neat tip Marc - I had also heard that re-white balanceing often will make the SSE go away (but, luckily, never had the SSE yet).

John

[Marc Colemont]

Yes, the white balance works too sometimes.
But I found it quicker to just turn off and on the camera. Without screwing around with my settings.

[Jim Giberti]

Originally Posted by Stephen L. Noe

You are correct in your assumption. There is a machine now that does the calibration that replaces the 'by hand' process. I understand there is only one of them (USA) and it's in California. This was brought up at the ProHD users group meeting here in Chicago.

Well it's not a new machine but a new, bettter implementation that's being done by QC in California.