EX and HVX200

[Steven Thomas]

Not to mention, XDCAM HD has been approved by Discovery HD as a full 100% acquisition format (using 35Mbps HQ mode).
That really says a lot to me.

[Jason Bodnar]

Good info guys, yes I am hoping not to have to buy two different cameras, one for action sports and the EX for everything else. We will have to wait and see actual tests with the EX. I am just going to wait until after the first of the year to make my move I guess. I have seen some pretty good fast action taken with the XDCAM's and they did just fine. In fact I was originally saving up for one until the EX was announced as it is much more reasonable for my applications. I can only hope that the EX will do the same and all the HDV issues with these type of shoots is resolved or mitigated enough to have 90%+ usable footage as long as the Cameraman uses it properly

[Steve Mullen]

Originally Posted by John Bosco Jr.

Yes. As I mentioned in my above comment, the HVX 200 will handle fast motion better than the EX 1 based on the cameras' specs.

Actually there are no motion "specs" for any camera.

There is a marketing claim and while the HVX was possibly better a few years ago, MPEG-2 encoders have gotten much better so the latest HDV cameras have few, if any, motion artifacts. Likewise, the XDCAM HD encoders have gotten better.

(Moreover, if you shoot 24PN, its EFFECTIVE bit-rate is no greater than HDV's 19Mbps. Something most DV100 folks don't understand.)

What hasn't changed is the HVX continues to use "SD" CCD chips to try and fill a 1440x1080 recording frame. (You may remember that 960x540 rez. CCDs were often used by Panasonic on their PRO SD cameras. As I remember, it supported wide-screen SD.)

EVERY objective test has revealed that pixel-shift never increases MOTION resolution more than 15% over the pixels on the chip. The only way to really increase rez. is to use Interpolation as is done by the EIP in Sony's V1. This is a very sophisticated processor that pixel-shift can not be compared to.

But, even interpolation cannot yield the rez. that native 1920x1080 chips do. This is pure science.

[Gabe Strong]

Originally Posted by Steve Mullen

Chris, I know you believe this. But there is a huge difference been a camera that uses pixel-shift to TRY to get enough pixels to fill the recording format -- and one that has has enough pixels to meet the needs of the recording format and uses pixel-shift to gain about 15% more pre-recording samples.

The former is a camera that "under-samples" and the latter is a camera that "over-samples." You'll notice that the company that under-samples now does not specify the resolution of the CCDs they use. Instead, they spec. the size of the CCDs they use. While there are advantages to a large chip -- there is simply no way an SD CCD can yield the same recorded resolution as can a 1920x1080 chip. No amount of pixel-shift can create pixels that are not there.

And, interestingly -- the company that claims pixel-shift works also claims long-GOP doesn't. They are wrong on both counts and it's critical that folks know this.

EXACTLY!!!! You said it much better than I, but that was EXACTLY what I was getting at. Thanks for the clarification.

[Thomas Smet]

Steve I agree with you on the pixel shifting. Pixel shifting is a great way of boosting detail of a native resolution but not so much for filling in missing detail.

In the past pixel shift has been more used to help boost the look of 720x480 chips. It gives the impression of a higher detailed chip that over samples the image and then gives you the 720x480. Starting with a native resolution that is lower then the standard and trying to fill in the missing pixels is a totally different situation. Not even the XL1 went as low as 360x240. It used something like 540x480 pixels and then pixel shifted. In this case it worked fine because it didn't have as much detail to try to fill back in. 540 pixel shifted to 720 wasn't as big of a jump and the vertical full 480 helped keep it in check as well.

When anybody shoot with a SD camera that had 360x240 pixels that were pixel shifted? Nobody would ever want to do such a thing. At the end of the day while pixel shifting can help 960x540 look like it has a lot more detail it is not perfect and is far from the having the resolving power of a 1920x1080 chip. Personally I always felt Panasonic should pixel shift 1280x720 chips so at least they would start with one of the standards and build from there to pull in more detail for 1080p.

Of course detail isn't everything which is why the HVX200 still looks really good. But if anybody is a pixel junky then a native 1920x1080 chip will always have more detail without a doubt. I sometimes wonder if the EX1 will have too much detail. Be carefull for what some of you wish for. High detail can also sometimes be a curse because it will show bad shooting and problems even more. With the EX1 personally I would like to see how it looks with the electronic sharpness turned totally off.

[Steven Thomas]

Bring on the detail!

An image can aways be made soft in post. I agree, it will be interesting on how well the EX1 holds up perceived detail with sharpness off.

I bet it looks great.

I have to disagree with the HVX200. Shutting off detail on the HVX200 had us believing we were out of focus. That's how soft it appears.

[Thomas Smet]

And that is how you tell what the true native resolution of a camera is. With a lot of cameras people are shocked at what a raw tap from the chips really looks like. It is amazing how much detail can be pulled back in by adding in some electronic or digital edge sharpening.

The key to high quality images is to start very high so you never notice the softness. That is where shooting 4k from the Red comes in. Red doesn't use any edge enhancement at all either and it kind of looks soft to some people. So shooting at 4k makes sense because a soft image at 4k still has a massive amount of detail in it. I will be very curious to see how much the EX1 actually resolves and how much is added sharpness.

The DVX100 avoided a lot of electronic sharpness by using pixel shift to create a over sampled image that then resulted in some nice natural detail for SD footage. The HVX200 on the other hand has to use electronic sharpness because it has to use that pixel shift to make the image HD. In order to get the same type of look from the HVX200 that you do from the DVX100 the HVX200 would have had to have 1920x1080 chips pixel shifted to process 4k in the DSP. Then the HVX200 would have looked very clean but detailed like the DVX100.

[Steven Thomas]

Originally Posted by Thomas Smet

And that is how you tell what the true native resolution of a camera is. With a lot of cameras people are shocked at what a raw tap from the chips really looks like.

You're right about this. Even the HD100 with its native 1280x720 CCDs looks a tad soft when the sharpness is set to off. Although, it does appear smooth.

It will be interesting to check the EX1 with the edge enhancement set to off.
When capturing 1920x1080P footage, It might have the same "perceived" detail as the HD100 with it's enhancement set to a middle range. This should give you nice detail and clarity without the nasty hard edges. I never like the affect sharpness has on images with a lot of detail such as grass and leaves. These images really can bring out the nasty effect of edge enhancement.

[Tim Polster]

This is the look we are all after.

Soft edges with overall detail.

Film does this combined with very expensive lenses.

This is tough for a video camera but it is getting better all the time.

[Thomas Smet]

This is why I am excited about the EX1. Even with the sharpness off it should have enough raw resolving power to still look great. Of course we will have to wait and see just what it looks like but I'm willing to bet it will look better then any other 1080 camera in this price range that is out there right now.

[David Heath]

Originally Posted by Thomas Smet

In the past pixel shift has been more used to help boost the look of 720x480 chips.

Yes, very true - effectively it increases the mtf for fine detail without the use of artificial enhancement. As such it's a very useful tool.

Worth noting that it's successful use in most cameras has been in the one dimension only - horizontal. The departure in the case of the HVX200 was to use it 2 dimensionally, and from what I gather, there is no great rush to follow that precedent.

[Chris Hurd]

Sorry, double-axis Pixel Shift in the HVX200 is not a precedent, not even by a long shot. The process has been around for well over a decade, and the most famous examples of H+V axes Pixel Shift in camcorders (the Panasonic AG-EZ1 and Canon) are now ten years old, and it wasn't anything close to a precedent even then.

Also, double-axis Pixel Shift is more common now than you might think; for example it's used in the popular JVC Everio line of consumer HD camcorders. Panasonic gets an unfair rap about Pixel Shift in the HVX200 primarily because those who criticize it the loudest have no idea how effective it really is, or just how commonly it's used in other three-chip camcorders.

Once again, as long as three-chip camera systems have been around and will continue to be around, Pixel Shift in one form or another will continue to be employed (because it works) with very few exceptions (most notably the JVC Pro HD line). Even the new ClearVid sensors from Sony are using a form of spatial offset at the pixel level, which is creates the same benefits of Pixel Shift under a different name.

The claim that "no amount of pixel-shift can create pixels that are not there" demonstrates a fundamental misunderstanding of what Pixel Shift is really doing. It certainly isn't attempting to create pixels that don't exist. Instead, the idea behind Pixel Shift, or in Sony's case of ClearVid by orienting a sensor diagonally, is to accomplish one thing and one thing only: not to create more photosites, but to create more sampling points per photosite for the benefit of the camera's A/D converter. And these various processes do it very well: they are fully legitimate ways to boost resolution.

High definition or standard definition makes no difference... the gain in resolution is significant and effective no matter how few or how many pixels are on the chip to begin with. If anything, the process is even more critical at SD resolution than it is for HD. Nobody complained about double-axis Pixel Shift in the Canon XL1, which brought its sub-SD chips up to Standard Definition. It was a non-issue then and it's a non-issue now (just look at how little worry there is about on on our JVC Everio forum).

[Craig Seeman]

Maybe this is all better framed for the next Texas shoot out.

Panasonic HVX200 720P60 vs JVC 250 720P60 vs Sony EX1 720P60
To keep things in perspective though, even if JVC and/or Sony "win" it still doesn't mean pixel shift is bad, it's what Panasonic uses to get a bit closer than had they not used the technology. What we may see beyond "win/lose" is how far can pixel shift go against chips in native resolution and up.

We might further add to the test by comparing HVX at 1080P30 vs EX 1080P30

[Steve Mullen]

Originally Posted by Chris Hurd

Even the new ClearVid sensors from Sony are using a form of spatial offset at the pixel level, which is creates the same benefits of Pixel Shift under a different name

That's simply not true.
Sony brings each under-sampled frames into the EIP. The EIP then performs an INTERPOLATION on these pixels to GENERATE A WHOLE NEW FRAME.

This is an ACTIVE process whereas Pixel-shift is a PASSIVE process.

Don't try to claim interpolation offer the "same benefits of Pixel Shift under a different name." The TX Shoot-out showed huge resolution differences between the HVX and the V1. The MEASURED V1 rez about 800x800. (My math model indicates it is about 775x775 -- which is damn close.)

The HVX measured exactly what my math model predicts: 540x540.

Remember it was my math model's backward prediction of the HVX's pixel resolution that finally forced Panasonic to admit it was using SD chips. There was a VERY good reason they tried for months to hide this fact.

Thomas clearly shows why pixel shift doesn't work in SD. There is a word for DV cameras that had 270K chips. They UNDERSAMPLED. It takes sophisticated technology to CHEAPLY build hi-rez chips PLUS the DSP needed to increase their lower sensitivity. Sony has it. And, once again they have leaped over the competition. (JVC also builds exactly the CCDs they need for 720p.) It's time to accept that 1080i requires FullHD chips.

The next TX shoot-off will show the EX1 to deliver 1000x1000 lines. (My math model predicts 1048x1048.) That is 4X the rez of the HVX!

To hold to your belief in the face of both theory and data only makes it look like you are defending Panasonic and/or Canon.

PS1: I always gave Canon negative marks for picture softness in every review I did. I never bought one either.

PS2: I love my tiny HD7. It's great for a $1,200 HD camcorder! But, it's not a V1.

PS3: Actually, the HD7/V1 difference is as others have posted. By bumping the HD7's SHARPNESS 1 tick I get a nice crisp HD pix that I like. The V1 delivers a very smooth pix which is what most folks want. Because I like crisp -- I often found V1 at full wide soft. The EX1 is likely to be the first $7K HD camcorder that can deliver BOTH smooth pix AND great detail at full wide.

[Kyle Self]

Shrugs, I don't care what you call it, how they claim to use it, or what it is supposed to do. Bottom line the HVX 200 is soft and the 500 is not a heck of a lot better. The resolution was a major fault for me. I wanted an HD camera to take HD pictures, not highly monkied with SD ones.

It was very interesting awhile back to see the post in the HVX forum here from people who had been influential users of the camera admitting in the open forum, as one put it "the camera is on a short leash". Why, because the resolution wasn't up to snuff.

K