HPX300 Preview - In Africa!

-3dB gain, man. I've been waiting for a "No gain" setting on one of my Pana cameras for a while. I'm excited to see it here.

Daylight shooting should look nice on this sucker.
 
Barry_Green said:
It's my understanding that EDIUS is 10-bit internal, but the EDIUS hardware cards only support 8-bit. I could be wrong on that.

My delivered footage was all done in the raw codec, not in an editor, so I used the P2 Portable to export subclips. Kevin edited his stuff in ProRes.

I should test whether EDIUS is limiting the performance to 8-bit on uncompressed output... I don't know the answer to that one offhand.
Click to expand...

From what I understand, Edius can import a 10 bit file, but any effects will be in 8 bit and all output is limited to 8 bit.
 
Bassman2003 said:
From what I understand, Edius can import a 10 bit file, but any effects will be in 8 bit and all output is limited to 8 bit.
Click to expand...

This is not true. Edius is 10 bit processing in the program, the output is 8 bit. So during all of the processing and color/image manipulation it is 10 bit. When it sends a signal to the 8 bit recorders it is 8 bit.

Best,

Jan
 
Thanks for the test Barry and Kevin. The stills sold me already. If I ever shop for an ENG sized cam this would be my prime choice of the moment! I just came back from shooting in India with my HVX and loved the small size. While I was shooting with a self made cine saddle, in the same spot a Belgian team was shooting with an ENG sized Sony and sticks on the road and immediately had 2 cops ask for the official papers......when they walked over to me, they smiled and said "very clever" and left me alone....
 
Yeah, that's the only thing that scares me about the 300 is that it attracts way more attention the my 170 does.
 
I became a fan of negative gain, actually. It cleans up the image very nicely. I've seen 24dB of gain and I don't recommend it for normal use (hah!)
Click to expand...

Barry, have you been able to check on a waveform whether the camera applies gain at the end of the chain or not?

Apparently the EX cameras apply gain after gamma. So for example if you use Cine2 which at 0db gain tops out at around 100ire and then select -3db gain, it then tops out at around 96ire. Obviously not a good thing as this restricts dynamic and tonal range even more.
 
I don't know at which point of the processing the gain is applied.

Negative gain (-3dB) pretty much always robs you of half a stop of dynamic range. That's the tradeoff: you get cleaner images with less overall range, versus noisier images with more range. Works that way with all cameras. Heck, positive gain robs you of dynamic range too, but you typically don't notice it because when you're using positive gain you don't usually have anything bright enough to peak the meters. But if you did, positive gain would take away dynamic range at the top end of the scale, by increasing the brightness and causing formerly resolved detail at the top end to now bloom and clip. Happens on all cameras, by the way, not just a Panasonic thing.

There basically is no such thing as truly zero gain, anyway... zero is a predetermined point at which the engineers have decided offers (usually) the best balance of sensitivity, noise, and dynamic range.

It's pretty much a zero sum game, but the end result is flexible to the point where the engineers have to decide. And even then, it's never truly "over", I mean, look at the DRS function - you can extract significantly different imagery out of a scene, if you're willing to accept the increased noise and potential shading issues that can come from the DRS function.
 
Hello Barry, could you test the DVCproHD on the HPX300? It would be interesting how it performed in comparison to the HVX - I mean especially the resolution.
 
I have a comparison shot where I shot the same image (CamBelles) on the HPX170 and in DVCPRO-HD on the HPX300. It's from a prototype so I don't know how valid the comparison truly is; you're never on fully safe ground when comparing a prototype against a production unit, and usually the prototype will only get better.

I can't conduct any more tests because I no longer have the HPX300 prototype. :(

1234894801.jpg
 
Nice!

Smoother skin tones, smoother transitions from the highlights. The overall sharpness / resolution difference is less apparent in these images until you see the UPC code.

AVC-I should be even better.

Barry's always reliable evaluation has helped me make the decision to buy this camera.
 
Let's not just make this camera look good, though. I have no doubt that's easy to do. I think Barry should put it through one of those torture tests he always talks about. I'm more interested in how bad this camera can look under less than ideal scenarios.
 
markyf said:
Let's not just make this camera look good, though. I have no doubt that's easy to do. I think Barry should put it through one of those torture tests he always talks about. I'm more interested in how bad this camera can look under less than ideal scenarios.
Click to expand...

Trust me, Africa was about as less than ideal as youo can get.

Barry will tell you I was near a nuclear meltdown more than once trying to figure out how we could get even semi-usable footage.

I should say that it has nothing to do with the camera before people start freaking out. :)

You can't get out of the vehicles in Kruger. That little window you see in the pictures of the van was the only shooting angle I had. So we had to move the van to point at the subject.

I arrived a few days earlier than Barry and I tried shooting off of beanbags and anything else I could find to get a stable shot. I was SOL every step of the way.

So finally I put the camera on a rinky dink tripod and a 503 head and wedged the legs inside the van to point the camera out the window.

So, any movement inside the van rocked the camera. With the lens sticking outside the van the wind would catch it and rock the camera.

The tripod and head weren't designed to hold that much weight so that was a cluster in itself.

Believe me, I would never wish that situation on anyone in a million years.

I want to go back and shoot some more but not until I figure out a better way to shoot from the vehicle.

Yet, Barry is still happy with the footage I shot. So go figure.

So yeah, I shot the camera in way less than idea situations and it still performed.
 
That's fine but I think a lot of people would want to SEE the bad footage. An honest evaluation of this camera shows both the good and the bad. These motion box files have almost no movement in them. They're more like photography then film. Sure, they look gorgeous. But if these sorts of reviews are going to be more useful, we need to see under what conditions the camera falls apart. We need someone on these forums that's going to be brutal but honest.
 
The "torture tests" were always about the places where something could fail - the long-GOP codec, and the rolling shutter. This camera doesn't have a long-GOP codec, so there's nothing to test there - instead, it has the best codec you can get in a camcorder. And the rolling shutter is exactly the same as it is on the Sonys and everything else, and the situations it will fail in are the exact same situations that I've been trumpeting for YEARS: flashes, off-sync fluorescent/HMI/ballasted light sources, and telephoto skew/wobble. You want brutal honesty? I've been crucified enough for pointing out the emperor's new clothes (CMOS) and people still seem to accept CMOS, so -- CMOS is coming and that's just the way it is. Should I go and repeat everything I've been saying? Just read this article, it'll tell you everything you need to know about potential shortcomings in the HPX300 (and also the EX1, the EX3, the V7U, the S270, the Red, the Scarlet, the Epic, the Infinity, the V1U, the FX7, the HV20/HV30/HV40, the Canon AVC-HD lineup, and every other rolling-shutter CMOS camera on the market).

Point is, if people find the EX1/EX3 acceptable, they should love the HPX300 -- it gives them the same CMOS issues but removes the long-GOP issues. And if people find the EX1/EX3 unacceptable, they'll find the HPX300 unacceptable too.
 
But there's different CMOS technology, is there not? The HV20 gets more flack then the EX1. There not all exactly the same. We don't know how limiting these chips are on the HPX300. Panasonic and P2 people want to know about the CMOS shortcomings on the HPX300 because P2 technology is what they're invested in. Saying it's equal to the Sony isn't really providing value in terms of a review, especially since many Panasonic people won't really know the shortcomings of an EX1, other than hearsay.
 
Thanks for the great review!
A very worthwhile read, I'll probably read through it a few more times. Thanks for takign the time to write this up!
 
markyf said:
But there's different CMOS technology, is there not?
Click to expand...
No there is not. It's all the same rolling shutter technology. The difference is that some have optimized the read/reset speed, but that only changes the matter of degree, not the kind. It's the same kind of thing and subject to the same kind of issues, just perhaps less so.

Some day there might be commercially available global shutter technology, and when that happens then yes, that's a horse of an entirely different color and will deserve to be treated very differently. Until then, it's all rolling shutter and it all performs similarly.

The HV20 gets more flack then the EX1. There not all exactly the same.
Click to expand...
The HV20 has a very slow read/reset time and makes itself more prone to exaggerated wobble. But the EX1 still wobbles, and they are fairly equal when it comes to partial exposure/flash/rolling bars, in that they will all show the same basic effect.

We don't know how limiting these chips are on the HPX300. Panasonic and P2 people want to know about the CMOS shortcomings on the HPX300 because P2 technology is what they're invested in. Saying it's equal to the Sony isn't really providing value in terms of a review, especially since many Panasonic people won't really know the shortcomings of an EX1, other than hearsay.
Click to expand...
There's amateur cameras like the HV20, which have a slower read/reset than the more pro cameras like the EX1/Red/HPX300. But (and this is the big, huge, massive "but" that I end up fighting left and right) somehow people think that the faster read/reset is "fixing" the rolling shutter issues, and it most emphatically does not. It might make the wobble less, but there's still wobble, etc.

An illustrative test might be a wobble-off, where we get an EX1, Red, and HPX300 all together and show how they perform under various rolling shutter circumstances. Expose them all to flash and strobelights, set them all to the same field of view and pan simultaneously, put 'em under magnetic HMI lighting, etc., and show how they all perform. I don't have access to any of the three at this time, but that's something I'll put on my wish list for when production HPX300's are available.

Thing is, they're all going to have issues. I've tested a Red (Build 17) under magnetic HMI, and an EX1, and an HPX300. They all have scrolling bars. Whether one's is slightly less than another is irrelevant, the shot is ruined under all of them, know what I mean? They all catch partial exposure under flash circumstances. Whether one obscures 50% of the frame and another obscures 60% of the frame, they're both subject to the issue and it's going to cause a problem in any footage shot. The only area where there might could be a difference would be in a wobble-off, and there we'd put them under the same field of view and try them. If there's a significant difference from one to the next, well, that'd be valuable information to know.
 
Last edited:
This camera looks very nice.
I currently shoot with the EX1 and look forward to checking this out and maybe getting back with Panasonic. But;
I keep reading broad statements about AVC-intra being a superior format.
What makes it so special over the XDcam format?
 
You must log in or register to reply here.
Back
Top