HPX500 image quality

[bruce edits]

Has anyone compared the image quality of the HPX500 vs. HVX200? I know that the HPX500 has 2/3" chips, but as I recall, they are non-HD chips. The same basic camera, the HPX2000, uses full HD rez chips, but costs twice as much as the HPX500. I was considering the HPX500, but the fact that the chips are from a SD imager is confusing me. Should I just get a HVX200? I already have a Canon Broadcast J16xs8BIRS lens from my Sony DSR-500ws which would fit the HPX500. Any comments would be great.

[Barry_Green]

The notion that they're "not HD chips" is completely misleading. The HPX500 is a high-def camera with a high-def imaging block, and so is the HVX200.

Why not call your local dealer and see if they have an HPX500, and put it to your own test?

[bruce edits]

Barry, I will do that. I guess I need you to explain how the chips in the HPX500 make a HD picture, while the chips in the HPX2000, a $26,000 camera do the same thing? Am I hung up on megapixels? Maybe. My DSR500ws makes beautiful pictures and has 3x 520,000 pixel chips, but is SD. I think the HPX500 has 3x 620,000 pixel chips and is HD. Why?

[kyle.presley]

The 500, like other Panasonic cams uses pixel-shift technology.

[Barry_Green]

We've gone over this a few million times on the board, so you should sit down for a good search session and you'll find lots of discussions. But the simple fact is: CCD pixels do not equal frame pixels, and you cannot look at one CCD by itself and declare that the resolution of that particular CCD is equivalent to the resolution of the entire imaging block.

Megapixels do not equate to better video; I have a 3.2 megapixel Nokia N93, and a half-megapixel HVX200. Want to guess which one's video blows the other one into the weeds?

You can't guess how fast a vehicle is just by looking at the horsepower of the engine (a 500hp truck will lose every time it races against a 50hp motorcycle). You can't guess which one will win a race just by looking at the engine size (a 1.3 liter rotary Mazda RX7 will smoke a 5-liter Crown Victoria). You can even put the same size engine in the same body, but if you don't account for EVERY FACTOR you might miss something extremely important (like the Mini Cooper vs. the Mini Cooper S; if you don't factor in that the Cooper S is turbocharged, you'll miss the whole point). The HPX500 and HVX200 use a form of "turbocharging", the spatial offset system delivers far higher resolution than the pixel counts would indicate (just like turbocharging produces far more power than the simple engine displacement would indicate).

In summary you can't guess how well an HD camera will perform just by counting pixels. There's no simple/easy correlation.

[Erik Olson]

He's a very patient man.

e

[Drew Ott]

We've gone over this a few million times on the board, so you should sit down for a good search session and you'll find lots of discussions. But the simple fact is: CCD pixels do not equal frame pixels, and you cannot look at one CCD by itself and declare that the resolution of that particular CCD is equivalent to the resolution of the entire imaging block.

Megapixels do not equate to better video; I have a 3.2 megapixel Nokia N93, and a half-megapixel HVX200. Want to guess which one's video blows the other one into the weeds?

You can't guess how fast a vehicle is just by looking at the horsepower of the engine (a 500hp truck will lose every time it races against a 50hp motorcycle). You can't guess which one will win a race just by looking at the engine size (a 1.3 liter rotary Mazda RX7 will smoke a 5-liter Crown Victoria). You can even put the same size engine in the same body, but if you don't account for EVERY FACTOR you might miss something extremely important (like the Mini Cooper vs. the Mini Cooper S; if you don't factor in that the Cooper S is turbocharged, you'll miss the whole point). The HPX500 and HVX200 use a form of "turbocharging", the spatial offset system delivers far higher resolution than the pixel counts would indicate (just like turbocharging produces far more power than the simple engine displacement would indicate).

In summary you can't guess how well an HD camera will perform just by counting pixels. There's no simple/easy correlation.

I commend your analogy.

[bruce edits]

You are a patient man. So, what is the advantage of, say, the 3x 2.2 megapixel chips in the new HPX3000 camera vs. the 3x 620,000 pixel chip set in the HPX500?

[TimurCivan]

the HPX200 will probably have SLIGHT advantage in resolution. Though i think the large light gulping, low noise pixels in the HPX500 will produce a far faster camera, with a great clean, sharp image.

[Barry_Green]

I'd have to see it to see what real-world benefits it actually delivers.

The spatial offset technique employed in the HVX200 is good enough for about 1400 x 810 and the HPX500 has the potential to resolve even more; either is basically enough to take advantage of the DVCPRO-HD format's maximum recording capability. The vertical res leaves a little on the table but the horiz is maxxed out.

The HPX3000 comes standard with AVC-Intra, which is capable of recording a full-raster 1920x1080, so it may benefit from having a chip block that's denser. But that's speculation; it'd be much more useful and informative to just set 'em up side by side and see the differences (if any differences are apparent).

I don't go in much for "bragging rights" on pixel counts, I know some do but I think that's a complete waste of breath. What matters is the end result. A 960x540 HVX matches every other HD camcorder in its general price bracket for sharpness and resolved detail. The HPX500 outdoes it by a small margin. I would expect the HPX3000 will excel in as many ways as possible; it is, after all, Panasonic's top model. But even if it does show notable resolution gains, you wouldn't necessarily be justified in attributing all that to a simple swap of the chip block; the recording format and the DSP and everything else in there is likely to be of much better quality and higher tolerances etc.

I don't expect the HPX3000 to be as light-sensitive or to have the dynamic range of the HPX500 though, specifically because its pixels are 1/4 the size.

[GaryinCalifornia]

Barry...

I was wondering... I believe that you said in another topic using the M2 or any other type of adapter... if you could or can do a slow mo... you can only shoot it at the most at 1/60th...

I was wondering if you have any knowledge if the same applies if you would use the Pro 35 with the 500... or can you shoot different frame rates with the Pro 35 and not see the ground glass...

Thanks in advance...

[Barry_Green]

I've only used the Pro35 once, and that was with an F900 and no variable frame rate stuff, so... I don't know for sure, but I would bet that the same issue would apply.

The MovieTube is a static ground glass (i.e., non-moving) and that makes it immune to any issues related to slow-mo. But any moving-ground-glass adapter is potentially subject to the issue; like I said before, I encountered it with the $11,000 Mini35 so I expect that the Pro35 would also have the issue. But without trying it we don't know for sure.

[GaryinCalifornia]

Thanks Barry...

One more... I can't seem to find out anything about the Angenieux HD Converter... I have a pdf of it... its I guess basically the same thing as the Pro 35... you can use 35mm pl mount lens on a 2/3 camera...

Been wondering if it has the same dof as the Pro 35... but since I can't remember where I got the pdf from and knowone seems to know anything about it...

Even if you do a google search and their company names come up...

Update... I found out SMS in C town used to have it... but since no one asked about it.. they don't have it any longer...

He said it was a static converter... costs about $15 to $16k...

He did say that it has the same dof as the Pro 35... like a chump didn't think about asking about light loss till now... another hunt for that information...

Here's a link to their page that has some information on it...

I'll see if I can still track down where I got the pdf from...

http://www.smsprod.com/products/lenses/angenieux4.html

[Barry_Green]

Never used it, but if it does what it says it does, it'd be better than the Pro35 because it should give the same preserved 35mm angle of view (which should mean the same depth of field) but no reprojection/ground glass in the way. Don't know offhand how they could do that, but that's what it sounds like.

You could try calling SMS, looks like they have it in their rental catalog.

[GaryinCalifornia]

Thanks Barry that's who I called... they dropped it because no one asked about it...

I put in an email to a dp friend in LA... see if she can find out anything about it...

[bruce edits]

I have used the Pro35 P+S Technic converter many times. I must say, the picture is wonderfully film-like. This was used on both an F900 and Panasonic Varicam. I had a set of 35mm T1.3 Zeiss Super Speeds and the image was beautiful. There was, as I recall, a two-stop light loss. The Mini35 was upgraded a year or so ago and no longer has the spinning groundglass, but now has the oscillating goundglass similar to the Pro35. Just my two cents!

[GaryinCalifornia]

An update...

I found out about the Angenieux HD converter... all it does is like keep a 28mm lens a 28mm lens on a 2/3's camera... it does not have the save dof like the Pro 35mm...

Also by the time you put the lens on the camera it sticks out 14"...

So that's why it didn't take off.. was told this today by the lens guy at Birns and Sawyer...