HVX vs. PMW-EX1

Status
Not open for further replies.
Hi Jan:

I know it's difficult to plan for this but you do realize that there will be a MASSIVE demand for the HPX-170? Do you feel that this will be a repeat of some past product rollouts like the original HVX-200, where a tiny trickle was released in December of '05 but it was difficult to actually buy one until well into the Spring of '06?

I am not counting on an unshipped product but I am wondering if the factory realizes how many people will want an HPX-170 on September 1? I hope that they can produce enough to where we can actually get our hands on one before 2009. I know the HVX was sort of a, "let's hold our breath and see if people want to buy one" because it was so new and radically groundbreaking. The HPX-170 will have a built in audience of probably at least half of all HVX owners will want one, if not more, plus all of the newbies who will see the ads and will want to jump on the bandwagon. I can see B&H and other retailers taking pre-orders soon for hundreds of them or even thousands of them. Multiply that by all of the other retailers and specialty VARs and you will have a massive demand on your hands I am positive.

Also want to thank you so much for being available to speak with us and for working hard to implement all of the features that we request here, it truly is unprecedented in the world of video gear to have a huge manufacturer actually listen, unlike your competitors. Just want you to know that you are well appreciated in the community and we are so happy you are here.

All my best,

Dan
 
The largest appeal the HVX had for me is it's 4:2:2 color space. That is, until I started dont a little more research reading old posts and looking into more detailed definitions of color space.

The EX1 has 1920x1080 sensors. A 4:2:0 color space divides the horizontal AND vertical chroma resolution by 2. So... after the compression, you end up with a 960x540 chroma resolution in 1080p.

The HVX has 960x540 sensors. From what I read, the pixel shifting technique of the HVX only affects the Y' (luma) resolution. So that means the maximum chroma resolution coming from the sensors must be equal to their native resolution, 960x540.

This is the same chroma resolution as the EX1 footage after it has been compressed to a 4:2:0 color space...

Now, I'm not sure about this at all, but is the HVX's chroma resolution then furthar downsampled when it is stored in it's 4:2:2 DVCPro codec? A 4:2:2 color space takes the available chroma information and divides the horizontal resolution in half (as opposed to horizontal AND verticle with a 4:2:0 color space). So does that leave us with a chroma resolution equivalent to somewhere around 480x540 on the HVX when shot at full 1080p?

So, to conclude, it looks like the effective color space on the HVX is equal to if not worse than that of the EX1 (and, again, I am referring to when the EX1 is recorded straight to it's cards in 4:2:0. It would of course be even better recorded through its HD-SDI outs with a 4:2:2 color space).

Correct me if I'm wrong about anything here...
 
Last edited:
Another thought-

The HVX doesn't reach a full 1920x1080 resolution from it's pixel shifting technique, does it? Isn't it something like 1280x1080?

If so, that would meant that the Y' (luma) resolution would only be 1280x1080. Of course, this can be upsampled to your hearts desire, but the effective resolution would remain the same... 1280x1080.

Now, the EX1 shoots in a full 1920x1080, and the Y' (luma) is not compressed in a 4:2:0 color space (the number at the beginning represents luma space and '4' equals full resolution / uncompressed).

So now you're also looking at a 1920x1080 luma resolution with the EX1 versus a 1280x1080 luma resolution with the HVX.

Luma (gamma, brightness) is hardly ever compromized in compression techniques because the eye is much more sensative to brightness changes than it is color changes... thus, compressing chroma may not be visible to the human eye, but it would be much more likely that compressing luma (brightness) would have a visable detriment to the overall image.

I'm saying this because it seems that the difference between the EX1's 1920x1080 luma resolution and the HVX's 1280x1080 luma resolution would be visable... possible more-so than the difference in chroma resolution.
 
Hi,

There is no argument that when you start out with a stationary object and the cameras are at 1080, the EX1 looks sharper. Move the object and things start to change. Both of these quotes by dj are indicative of watching too many of the Sony presentations. The words are very much what they say. They make it sound as though the number of color samples is what is important to cut the green screen. Let me put it this way, 4:2:2 is a ratio. It doesn't matter what the numbers are, it is a ratio of color to luminance to its own signal. That's it. You can compare how many this to that all day long and it doesn't change the necessary numbers in the ratio.

And when the signal comes off of the CCDs, remember it is an analog signal not a digital one like the CMOS, it is initially sampled in the 1080P mode and all processing happens there prior to the recording.

Thanks,

Jan
 
Last edited:
Na, I haven't watched any Sony presentations. I just found out that the EX1 / EX3 exists about a week ago. I've been working a lot and have been out of the 'camera tech' loop for a bit. I came back on to do a little research because I'm going to be buying a new camera.

Are the things I said the kind of things that Sony says in their presentations?

I'm saying all this very openly... as in, I'm not positive about it myself, but putting out there what it looks like based on my own research.

The number comparisons are important to me because I do quite a bit of greenscreen work. Plus we all know that talking about them boosts our egos as we feel inteligent in a collegiate sort of way :).

It makes sense to me that more raw pixels would yield better greenscreen results... there is just more detail available to draw a finer line between the subject and the background. And, as you said, color space is a ratio to luminance. Luminance is usually not compressed and is equal to the full res signal. Thus, the higher resolution the signal, the higher resolution the chroma should be, right?
 
Last edited:
I ran tests last week with green screens I shot on the HVX-200 and the EX-1. IMHO, they key equally well. The compositor said there was no difference on his end in pulling clean keys with footage from either camera.

Was kind of surprising to me but I was glad that I can shoot with either without worrying about how the key will look.

Dan
 
There is a lot of talk about EX1 having problem with flashing light and fast movement. I have now done a couple of tests with my stroboscope run in different speeds and I can't find any problems with the shutter method on the EX1.

Is there a specific light situation that trigger the effect?

Regards / Jimmy
 
Hi,

I have seen some Green screen done with the EX1 and it isn't bad, but I have yet to see singular blond hairs that are blowing in the wind get cut on a 4:2:0 camera. This I have seen on the HVX200. On the EX1 I see a sort of a rigid line not as bad as the HDV stuff but I see a rigidity that I don't see on the HVX/DVCPRO HD material. Some things are easier to key than others, but one of the worst is blond hair blowing in the wind.

And yes DJ, that is exactly what the Sony presentation says, that because the EX1 has as many color samples as the HVX, it will key just like the HVX. And that just isn't how it works. 4:2:2 is a ratio, that is the reason for the colons, 4 samples of luma, 2 of red, 2 of blue.

Jimmy, The inexpensive CMOS Chips, read EX1 through the Red, have a rolling shutter, and you should know how and why this will come into play while you are shooting or you will come back from a shoot where it has happened thinking you have it in the can and have one of the worst experiences while you are watching your footage. as perhaps some will be really unuseable. Go to Google, type in EX1 and Rolling Shutter. See what others have experienced and then compare their experiences to what you do and see if you can recreate it. Knowledge will help you.

Best
 
Right, 4:2:2 is a ratio... that's the basis of what I have been elaborating on. This ratio is dependent on initial resolution... And, although pixel shifting boosts the luma of the HVX, the chroma resolution starts at 960x540, half that of the EX1.

Now, in proportion to the initial, uncompressed resolution, there is a 25% chroma resolution difference between 4:2:0 and 4:2:2. The 50% difference in initial resolution outweighs the 25% color space difference between the two compressions.

In theory, the HVX would need to record in 4:4:4 to match the chroma resolution of the EX1 recorded in 4:2:0.

So, I'm not surpised that the keys from the EX1 are being reported to be as good as the HVX. I'm actually suprised the HVX keys as well as the EX1...



Now, this is in theory... like a lab test. There are other factors like, as you mentioned, the supposed resolution drop with fast movement on the EX1... although if it only drops a couple hundred lines, it wouldn't make much difference to my point.

Also, as you mentioned, the CCDs output an analogue signal versus a digital signal with the CMOS... although I don't know what difference this would make if the signal is captured in a definite pixel resolution anyways. I mean, in the audio world, playing vinyl has advantages because the source (the record) it's self is analogue, so there is no definite limit to the bandwidth... But it makes more sense to carry a signal of digital origin through a digital signal...

Having everything processed in 1080p after the signal is captured sounds nice... but you couldn't boost the resolution of what was originally captured in the first place.

Jan, don't get me wrong, I'm not trying to stir up an argument nor am I sitting in any specific 'camp' so to speak. I just need to type it as it seems clear to me... This way others can be informed and/or I can be corrected.
 
Last edited:
Also... What is the big deal about rolling shutter? I saw an example of this in slow motion when a camera flash was recorded... I almost laughed because, based on the footage I watched, this might be the biggest 'non-issue' I've ever seen so much fuss about.

I suppose if you are recording a wedding or something with a ton of camera flashes, and you want to play the footage back at 3 fps, then it might be a concern. For me, though, I can't see myself shooting around camera flashes enough to worry about it. Based on the footage I watched, I'm really not even worried about the times I will be around camera flashes.

Perhaps I'm misinformed on this one though.
 
Last edited:
Do a search on the internet/google about rolling shutter and the EX1.

It's hardly an "non issue" for what is reported there, and at a bunch of other boards.
 
Hi DJ,

See the problem is that everyone looks at the pixel count on a CCD and assumes that there is a one to one relationship to what is recorded. That is why I am very clear to bring up that only in a CMOS is there the potential for this as it is a digital signal. An analog signal is presented to the digital domain at 1080P. So referencing the chips as the "count factor" isn't rally relavant.

Compairing a camera that is recording at a 4:2:0 color at a long GOP to a camera that is doing 4:2:2 in I-frame only at 100mbs, is really apples and oranges. I have seen some good examples over on the EX1 form but nothing like the good examples I have seen from the HVX. Singular blond hair blowing in the wind. That needs chroma resolution. Oh and BTW, you should read the white paper on how the HVX gets more resolution out of its CCDs that its count, because yes we can add to the resolution:

Tech Paper - AG-HVX200's Advanced Progressive CCDs under the support and resources tab of the HVX200 http://catalog2.panasonic.com/webap...roupId=14569&surfModel=AG-HVX200&displayTab=R


And yes the rolling shutter would nail your footage to the firing line if you were 1. shooting a red carpet event, 2. shooting a press conference, 3.a wedding, 4. wante vertical buildings in your panning shots. Do a little more research as David suggests. Keep in mind that between the rolling shutter and the long GOP, you may not alway "know" what you footage is going to look like and the place to find out that it is unuseable is in the edit suite. Not good. This is not to say that some footage that has been shot with the EX1 camera is bad, some of it is quite lovely, but so far i haven't seen too much that would qualify for much more the credit rolls. Just talking about the pace of it here.

Best regards,

Jan
 
I have been taken by this comparison for the last 2 months.
My first impression of the EX-1 couldn't be better. I really like the LCD the VF the sharpness, the menu after a few days seemed very smart and complete. The finals results were substantially better than the ones I was getting with my HVX200.
Then I did a whole day of handheld. No big vertical buildings, no flash photography, no fast pans, just plain verite documentary style. And... There're problems.
The producer got disappointed I was very concerned. At the end we were able to save most of the footage and the issue was minimized. The part that was "clean" looked amazing, better than anything I have shot with the HVX, but the stress was to big.
Just go ahead and rent both camera and see which one works for you better.
 
I think that's a good idea... renting both cameras and trying them, that is.

Does pixel shift technology increase chroma resolution? I was assuming it didn't based on only one opinion, so I'm not sure actually.

Maybe it's in those white papers...
 
I just looked at more EX1 'rolling shutter' footage examples... It still seems like a pretty small issue to me. I also downloaded some footage of Red exibiting the same problem. It's like when you are using a film camera and you're using lights that are out of sync with the shutter. Film cameras show some diagonal artifacts on pans sometimes too...

People were saying you can almost completely, if not completely eliminate the rolling shutter anyways by setting the shutter to "0", or turn it off, which lets the camera default the shutter speed to whatever the frame rate is. I haven't used the camera yet though, so I'm not exactly sure what they are talking about with this yet...

I know in my own shooting, I typically am never around flashing lights of weddings, red carpet events, police cars, strobe lights, planes, UFOs, etc. I also almost never shoot handheld... I usually set up more controlled shots on a tripod / track. So, sure, I could see how someone who typically shoots flashing things like this in most of their work would benefit from an HVX which won't have these 'rolling shutter' issues.

Most of my shots will greatly benefit from the higher resolution, less noisy image that the EX1 can render. The benefit of this far outweighs the seldom detraction of the possible rolling shutter and skew issues.

But, ya, that is of course an assessment based completely on my own shooting style / preferences.
 
Last edited:
alexdias said:
My first impression of the EX-1 couldn't be better. Then I did a whole day of handheld. No big vertical buildings, no flash photography, no fast pans, just plain verite documentary style. And... There're problems.
The producer got disappointed I was very concerned. At the end we were able to save most of the footage and the issue was minimized. The part that was "clean" looked amazing, better than anything I have shot with the HVX, but the stress was to big.
Click to expand...

Thanks for the balanced evaluation. Too many people only voice good or bad as if there is no grey. Both are decent cameras. So try them out, know what you need and get what works for you.
 
Why I'm not buying an EX1

Why I'm not buying an EX1

I've been shooting for 20 years, mostly with Sony cameras and other 2/3" cameras.
My work has been on every major TV network and many cable networks. I had a Panasonic DVX100A and also a Sony Z1. When the EX1 came out I was excited about it primarily because of the 1/2" chips. I wanted to like it. I seriously thought I was ready to buy it. Then I got an EX1 to use for a week and a half. I borrowed a friends HVX200 and went and shot the same things with both cameras. The EX1 clearly has a cleaner, sharper picture than the HVX 200 and more latitude. Maybe the HVX 200A reduces the EX1's advantage here. The EX1 lens is excellent. But in my opinion the camera body is pretty horrible. Very hard to hand hold. Tiny buttons in awkward places. No way to use a remote focus or iris. I could almost live with that except that is shoots MPEG2 and even though it is at a higher data rate than regular HDV, I can clearly see the compression artifacts in water and when panning. I have a Z1 and am pretty tired of HDV compression. Some shots with light to dark gradients are almost unusable for me because of the compression and that's before any processing. The 50% higher data rate of the EX1 is not enough to solve all the compression problems. I could also see CMOS rolling shutter artifacts on tall vertical objects. I tried to match the colors that the HVX makes by using the EX1's extensive menu controls. I came close but couldn't get it "just right". I've been tweaking cameras for a long time. Some things the Sony EX1 allows you to do in menus is very, very good. But the bottom line was the picture. I shot a shiny black truck with both the EX1 and the HVX. The EX1 delivered three different shades of black all mixed with magenta. When the driver opened the door, the light changed and the door was really off. The HVX delivered very nice deep rich blacks. No weird colors. Everything I shot with both cameras was far better to my eye with the HVX than the EX1. I could never completely dial out the "Sony video look" out of the EX1. The HVX is more film like and the colors really pop by comparison. If you have the opportunity to look at both cameras shooting the same thing, I think you'll see the difference. Resolution isn't everything. As far as I'm concerned, it's no contest. The EX1 isn't in the same league as the HVX as far as picture is concerned. For me the HVX makes a much better picture and that's what's it's all about.
 
Status
Not open for further replies.
Back
Top