@Paul
 

Hi Paul,
Thank you for the question. I work intensively with RED and it is a blesing! The DIT on my last show knows all the big guys at RED and he has all kind of information and charts regarding RED and other cameras. Hope that helps.
Cheers

Thanks Luben great info.

The job is done!
 

Well, this thing ended up bigger than Ben Hur. I digested the advice of many on this and other fora and ended up going completely insane and spent over $18,000.00 on stuff that was recommended to me.

So, next week I should take delivery of the PMW-EX1R, a couple of BPU-60 batteries, an extra 32GB SxS card, RODE shotgun mic, Sennheiser wireless lav mic,nanoFlash unit and two 32 GB CF cards to go with it, Miller 1852 tripod system, a Zoom H1 recorder, custom cases for the gear, cords and cables.

I have moved from "it doesn't matter if I get work to pay for the camera" to "Jesus Christ! I hope I can get some work to pay for all of this!"

Still, the wife hasn't threatened me with divorce and I do my first wedding in February - even if it will be a "love job". At least I'll have something for my portfolio.

Thanks again, everyone, for the considered responses to my ponderings. Be prepared for a barrage of questions as I try to figure out how to make it all work.

Cheers

Russ

But there's H264 and H264, the quality varying widely with bitrate and other factors.

And the REAL problem with aliasing is not that it necessarily looks so bad immediately, or even through the edit, but that it can cause problems at the final stage, when the material gets compressed more heavily.

So, an aliased image will require a higher bitrate than the same image without aliasing, for the same quality and using H264 in each case.

It may not sound obvious, as it's easy to think that if you can't see any problem on the original material, so what? The theory is that aliasing on video will move in the opposite direction to a moving object that is it's source. It may just seem like a bit of "twinkling" to the eye on the original - but will cause the motion estimation part of a codec disproportionate problems, cause it to waste data rate, or drop quality.

David hits on key points about H.264 and compression.

The people that I've spoken to at Canon (not higher level people) have said they have no intention of changing the form factor of their DSLRs (which does not preclude coming out with a video form factor camera if they were so motivated). I've often thought the next step in their evolution would be a camera or firmware upgrade which would allow the complete (albeit already pixel row skipped) signal out of HDMI. That would allow an Apple ProRes recording rather than H.264.

Some people have assumed that the given that the DSLRs H.264 is "better" then that found in AVCHD cameras because it has a higher data rate. Others have pointed out that the DSLR recording is actually only I and P frames (I do not know this directly) rather than IBP GOP. That would be an example in which the higher data rate may possibly be lower quality given the inefficiency of the codec in DSLRs. That would could also mean another "evolution" would be DSLRs that can record 24mbps AVCHD with greater efficiency.

Of course one would argue that the above two things are addressed in the Panasonic AF100 but it remains to be seen what real world results are once we get based the marketing hype stage and real world workflows are engaged.

It may be there's another generation to go before large chip sensor cameras work through the "kinks" in the price range of the EX1

Russell -- that sounds terrific. One more item you might consider if you aren't aware of it: instructional DVD set for the EX1 and EX1R by Doug Jensen (Vortex Media), who is one of the regulars here on DVInfo. It was included when I got my camera last month. Good way to learn the basics.

Sorry Craig, that won't really help. Even if you recorded uncompressed, let alone ProRes, all that would mean is that the motion aliasing gets faithfully preserved through the main production chain - ready to hit the final coder and mess it up. The only answer is to not let the front end of the camera produce the aliasing in the first place.

Unfortunately, the only easy way of doing that with DSLR type chips severely reduces the still capability of the camera. So not much use in a DSLR...... ;-)
The two things that the AF100 really addresses are usability etc, and in quality terms by putting an optical low-pass filter on the chip, which is still fundamentally a 12 megapixel still camera chip. It seems that the AF100 chip will still pixel-skip - but the OLPF will mean that the aliasing level will be substantially reduced. That's a big step forward - but don't then expect the AF100 to have good resolution as a still camera!

It seems that the F3 will have a dedicated video sensor, so no need to pixel-skip at all. The implication is it will give better performance still, and I'd expect substantially in terms of sensitivity. That's likely to be reflected in the extremely low noise figures being quoted, and that's what means they can sensibly start to consider 10 bit recording and S-log. There's no point even thinking about that in cameras with higher basic noise levels.

Hi Keith,

Already done! Winging its way over from the Yooonited States as I type this. I actually ordered it before I ordered the camera.

Thanks

Russ