So now when specs and footage?

[Damon Botsford]

That tells ya how passionate he is about the subject.
Trust me, he knows what he's talking about... and I wouldn't be too suprised if he's actually seen a private snippet of what this camera is capable of.

 

[Barry_Green]

Well, I type for the whole community, not just one response... so if anyone coming into this thread later finds anything of use, that's what it's there for.

... plus, I type too much no matter what! Brevity is not one of my strong points... :thumbsup:

 

[shaftbond]

Again though, the rest I already knew. You could have just asked me a few simple questions and saved yourself a lot of typing.

He just gave an explanation as to why the exact chip resolution is not as important as the final image in response to your posts about how important chip resolution is to you. If you knew all of that already, why are you putting so much importance on the actual chip resolution?

 

[FatBird19]

That tells ya how passionate he is about the subject.
Trust me, he knows what he's talking about... and I wouldn't be too suprised if he's actually seen a private snippet of what this camera is capable of.

Oh, I trust that he knows what he's talking about. I also believe though that I know what I'm talking about to some extent.

I would only want the HVX to be 1280 pixels wide if they found that it did not affect the image negatively if they added the extra pixels. After all, what good would a 30MP HD camera be if all the pixels were <1micron across? Obviously I would want panasonic to believe that the extra pixels would not degrade the image quality before they added them.
My thoughts though, and I would have to defer to Barry on this to give me the physics, is that a 1/3in chip can support 1280 pixels wide without degrading the overall image. I would guess that Panasonic is just taking the cheaper route. I don't know honestly. Again, I'll have to wait till december to do comparisons, WHICH I WOULD HAVE DONE ANYWAYS.

But in any case, It couldn't of hurt to ask me a little more thouroughly what I was trying to say first.

 

[FatBird19]

He just gave an explanation as to why the exact chip resolution is not as important as the final image in response to your posts about how important chip resolution is to you. If you knew all of that already, why are you putting so much importance on the actual chip resolution?

To restate: I would like more chip resolution ASSUMING THAT IT DIDN"T AFFECT THE OVERALL IMAGE. I assumed that we all accepted that smaller pixels affect image quality; that's why I didn't state any furthur than "I would like more chip resolution." You guys took that as I thought resolution was the only important factor. We're square I guess. Let's move on.

 

[Barry_Green]

that a 1/3in chip can support 1280 pixels wide without degrading the overall image. I would guess that Panasonic is just taking the cheaper route. I don't know honestly.

Well, interestingly enough, there is some evidence to support the notion that it can't be done without having some effect. Witness the split-screen effect on the HD100. The chip is so small, and the pixels so crammed across it, that they have to have two samplers to read out the data -- they can't do it with one or the thing overheats. So they had to go with two -- and if there's any sort of misalignment between the black level of the two processors, you can get the split-screen effect.

Dave Walton at JVC said that the magic number seems to be about 1000 pixels across for the maximum that one sampler could read from a 1/3" chip; he took a guess as to maybe that's why Sony stopped at 960. It raises the question about the Canon though -- if the Canon has 1440 pixels across, will it also be using dual samplers, and will it also then have the potential of the split-screen effect?

And what will that do for latitude and sensitivity on the Canon?

And, let me ask this while I'm at it -- who says the pixels on the canon are true raw resolution pixels anyway? They said 1440 "effective". Could that mean pixel-shift 960 (which would imply that they are, in fact, using the Sony Z1 sensor)? It's something I mean to test for (a simple test will reveal whether it's native-res or using the same 960-wide as the Sony...)

December testing will be an interesting month...

 

[FatBird19]

Well, interestingly enough, there is some evidence to support the notion that it can't be done without having some effect. Witness the split-screen effect on the HD100. The chip is so small, and the pixels so crammed across it, that they have to have two samplers to read out the data -- they can't do it with one or the thing overheats. Dave Walton at JVC said that the magic number seems to be about 1000 pixels across; he took a guess as to maybe that's why Sony stopped at 960. It raises the question about the Canon though -- if the Canon has 1440 pixels across, will it also be using dual samplers, and will it also then have the potential of the split-screen effect?

December testing will be an interesting month...

Would it be any better if they moved over to CMOS?

And I don't know about you, but I have no faith or desire for the Canon. :/

 

[Barry_Green]

CMOS looks like very interesting technology, with its own set of hangups. I haven't spent enough time researching CMOS to know whether it'll be a net gain or not; I'm pretty sure it will, but haven't done enough due diligence to say that absolutely.

 

[FatBird19]

CMOS looks like very interesting technology, with its own set of hangups. I haven't spent enough time researching CMOS to know whether it'll be a net gain or not; I'm pretty sure it will, but haven't done enough due diligence to say that absolutely.

What about the new X3 chips by Foveon ? It'll take a few years to get it right, but I'm looking forward to the idea. :cheesy:

 

[Barry_Green]

Well, Foveon looks very interesting, but as you say, it seems like it'll be a while before they get it adapted to handling video camera speeds.

But we've been talking about Foveon for... heck, how many years has it been... probably four or five years now, how Foveon was gonna be the next big thing. And as far as I know, it's still only being used in a couple of digital cameras, right?

I think the Foveon approach sounds terrific. I just don't know how long it'll be before it's practical for video applications.

 

[FatBird19]

Well, Foveon looks very interesting, but as you say, it seems like it'll be a while before they get it adapted to handling video camera speeds.

But we've been talking about Foveon for... heck, how many years has it been... probably four or five years now, how Foveon was gonna be the next big thing. And as far as I know, it's still only being used in a couple of digital cameras, right?

I think the Foveon approach sounds terrific. I just don't know how long it'll be before it's practical for video applications.

What I do think is promising is the new CCDs by Fuji. I don't remember the title, but it works on the same principle as speakers where one element handles the lower sound levels and the other handles the higher ones. In this CCD array, each pixel is composed of two photosensetive elements: One larger and one smaller. The larger one handles the lower light levels and the smaller one handles the higher ones.

With this technology, I could see how digital video could achieve the same lattitude as film within a few years. :thumbsup:

 

[mmm]

It is statements like this that tell me that people actually believe that to make an HD image it has to have a 720 X 1280 imager. So tell me how are they getting those HD images out of the Adromeda and the SD DVX imager. The imager is an analog device. There is no pixel for pixel relationship from the chip set to the recording so my feeling is that you really should look at the pictures, the resolution as it is recorded, the low light sensitivity, the dynamic range as these are really vastly more useful in determining whether a camera will be an appropriate choice for recording HD.

It is statements like this that make me think the HVX is going to have low res CCDs. My money was on 1280x720, but maybe even 960x720??

And I am not saying it's a bad thing at all!!!! You guys know more than me about getting the best image quality from the camera, and it is certainly in everyones best interests (apart from JVC, Sony and Canon) that the image looks the best possible.:thumbsup:

Go on Jan just tell us... we are expecting it to be low res anyway! Is it people like FatBird stopping you from saying?

 

[Walter_Graff]

I think what Jan says is what I say here every day, HD is a marketing term. There are much more important specs for how a camera makes a picture than resolution numbers.

 

[mmm]

I think what Jan says is what I say here every day, HD is a marketing term. There are much more important specs for how a camera makes a picture than resolution numbers.

Certainly.

I am interested from a purely technical point of view. It will not affect my buying decision at all... I just want to know how they did it.

 

[Jan_Crittenden]

It is statements like this that make me think the HVX is going to have low res CCDs. My money was on 1280x720, but maybe even 960x720??
Go on Jan just tell us... we are expecting it to be low res anyway! Is it people like FatBird stopping you from saying?

Your numbers are wrong as well, and you should not read anything from my statement other than the fact that I feel that people are not camera engineers and as such don't understand the relationship. So when a person posts that the camer has to have a certain pixel map density, or he won't buy it, well what's a girl to think? That there are hundreds more just like him?

You bet, I have been on the boards for ever and I go back to when we introduced the AG-EZ1 and it only had the 270,000 pixel chip set. The Sony VX1000 had 310,000 or some larger number. Everyone said, the Sony was better but truth be known, ours had more resolution on a chart and was vastly quieter. And no one bought into that argument until it was too late for the Panasonic, but they lived with the Canon XL-1 for years, thinking it was the perfect resolution camera.

My take-away on this industry is that many read the specs and thinks that is the performance, yet it really doesn't work that way on some things. S/N yeah, right on, the higher the number the better, with chips, the density should tell you that the denser it is , esp. on a 1/3", the more fixed pattern noise, the less the sensitivity, and less latitude. And there may have to be some additional filtering to rid itself of "too much" information, that reults in ringing and aliazing.

Look what the guys from Purdue are doing with the DVX, a HD picture from the DVX SD chip set. My hat is off to them, they probably understand exactly what I am saying. It is not about the numbers it is about the pictures.

When the picture on the camera is finished and we have done all the resolution tests and it shows that the HVX200 wins hands down, we may disclose, but in the meantime, just know that we are determined that this little camera will be the best in class.

Hope this finishes the topic.

Thanks,

Jan

 

[Jan_Crittenden]

Certainly.
I am interested from a purely technical point of view. It will not affect my buying decision at all... I just want to know how they did it.

Well when we figure out how to get it out of "engineering speak" we will tell you how we did it, probably before we tell you the numbers. From what I understand though it is pretty cool and the initial pictures I am seeing in 1080i look great. I can hardly wait to see the Progressive images.

Best,

Jan

 

[mmm]

... we are determined that this little camera will be the best in class.

Hope this finishes the topic.

Sorry, not quite.

Surely you can't beat Canon resolution wise though? I mean, due to the DVCPROHD codec isn't that impossible? I would guess you would have them pipped in latitude etc though.

Oh, BTW, is "PAL" DVCPROHD 1080 1280x1080 or 1440x1080.

Many thanks.

 

[stephenvv]

Sorry, not quite.

Surely you can't beat Canon resolution wise though? I mean, due to the DVCPROHD codec isn't that impossible? I would guess you would have them pipped in latitude etc though.

Oh, BTW, is "PAL" DVCPROHD 1080 1280x1080 or 1440x1080.

Many thanks.

Codec does not equal resolution anymore than CCD equals resolution. Otherwise, we would not be able to tell 35mm film apart from DV when watching a DVD (NTSC MPEG2 codec).

DVCProHD has all the necessary codec space to render extremely sharp resolution images but the camera is much more important than the codec. Given it's lower compression and especially great color space, it delivers images in a different class than HDV or DV.

Sure, the XL H1 has HD-SDI out which is very intriguing on paper but until a portable and/or low cost method of capturing images, it falls into the Andromeda category. Interesting but ususable for 99% of owners.

 

[mmm]

Codec does not equal resolution anymore than CCD equals resolution. Otherwise, we would not be able to tell 35mm film apart from DV when watching a DVD (NTSC MPEG2 codec).

DVCProHD has all the necessary codec space to render extremely sharp resolution images but the camera is much more important than the codec. Given it's lower compression and especially great color space, it delivers images in a different class than HDV or DV.

Sure, the XL H1 has HD-SDI out which is very intriguing on paper but until a portable and/or low cost method of capturing images, it falls into the Andromeda category. Interesting but ususable for 99% of owners.

I agree Stephen, I am just talking about pure recorded resolution. I assume (could easily be wrong) that the Canon will be able to resolve the highest res (aslong as 24f is as they claim) simply because it has 1440x1080 chipsets and records at 1440x1080 luma res. The HVX on the other hand is limited to the 1280x1080 of DVCPROHD, and is very likely to have even smaller pixel count on the CCDs. I would also GUESS that the Canon lens is better, but it is $3K more expensive.

Personally the HVX sounds like it will produce a "better" picture because they have probably made wiser compromises for this price bracket.

 

[stephenvv]

I agree Stephen, I am just talking about pure recorded resolution. I assume (could easily be wrong) that the Canon will be able to resolve the highest res (aslong as 24f is as they claim) simply because it has 1440x1080 chipsets and records at 1440x1080 luma res

I'm not sure this amounts to much of anything since the Sony HDV are capable of the but many other issues (cited here) have a much bigger effect on image rez than CCD and codec specs. If you could feed the codec "pure resolution", maybe, but the Canon lens, CCD block, scanning method, 24f DSP and most importatly, the 4.1.1 HDV codec will be much bigger factor is what the recorded resolution measures.

The Sony HDV have resolved no where near 1080 lines on res charts tests I've seen.