Why doesn't the HVX do HDV?

Look Alex - Jan has said Panasonic is not a member of the HDV consortium. It's as simple as that. DVCProHD in the HVX, I'm sure, will live up to what we've come to expect as standard on the DVX = great images! I'm personally more interested in SD at 4.2.2 at the moment and that's what excites me about the HVX. I'd love to have something like the SDX 800 but I've got to think of the future and HD figures in that future. The HVX has got a right now usability that will get me better green screen capabilities and is in the center of the up-coming HD revolution. Here in Japan, HD TV sets are selling, so, it's clear HD is here to stay.
 
ddh said:
Look Alex - Jan has said Panasonic is not a member of the HDV consortium. It's as simple as that.
Click to expand...

Heh! Heh! I know! I'm for DVCPro not against it! Given the choice I would use DVCProHD over HDV! But I have used HDV and it's not as bad as some people maintain. I'm just trying to encourage people not to be prejudice and to investigate things for themselves.

ddh said:
I'm personally more interested in SD at 4.2.2 at the moment and that's what excites me about the HVX.
Click to expand...

I agree. I think it's absolutely fantastic that Panasonic have given us the full swath of DVCPro variants here. The HVX is going to be a fantastic and very versitile little tool.
 
Alex Leith said:
There seems to be a rather a lot of predjuce here against all things that don't have a Panasonic badge.
Click to expand...

well.. this site is called 'dvxuser.com' for a reason...


Judging by some of the responses you would get the impression that HDV looked worse than 3rd generation VHS, and that DVCProHD can see in the dark and gently massages your feet while you film, and that HDCAM - whilst looking good - calls you names and then has a poop in your coffee when your back is turned...
Click to expand...

it is my opinion that HDV does look like poop. The footage I have seen where there is actually something going on, eg: the footage apple uses to show off FCP at the apple store, all the footage of waves crashing on shore look pretty awful, I can plainly see the blocky MPEG artifiacts...

Sure HDV looks great when there is no movement, or it is a simple shot... but start adding in complex imagery, something with high frequency of detail... like grass.. water etc.. and the MPEG-2 compression falls apart... noticing this on a computer monitor is one thing.. but I can imagine it blown up and shown on a 30 foot screen... artifiacts galore.



HDV is not as bad as most people here seem to think. MPEG2 as an interframe compression codec is about twice as effecient as an intraframe codec like DVCProHD - and that's part of the reason why the data rate is so much lower. HDV chroma resolution is lower, but luma resolution is higher. And contrary to popular belief motion artefacts are not as common as some would have you believe - jumping out of the picture, kicking you and then shaking the tripod.
Click to expand...

HDV is just poor... who cares if it is a more efficient compression? That comes at the sacrifice of quality.. it is a trade and always will be. Interframe compression is just plain bad for anything that requires frame by frame work.. DVCPRO-HD is just a plain better format... technically and visually... Intraframe is easier to edit any day.. plus less compression means the image will hold up better over multiple generations of manipulation.

I know what your trying to argue.. I am just making my point here...

I don't see what all the fuss is about anyways.. by the time most of the people actually go into HD production... it will be a 5 years... by that time.. we should have 32 and 64 GB P2 cards....
 
Didn't Jan C. already answer this?

If I remember correctly it was along the lines of something like:

Panasonic wanted to make a distinction from HDV, because if the HVX did do HDV, then what may happen is that HDV's weaknesses and negative qualities would be attributed to the camera. That’s how I remember it anyway.


Anhar
 
high8 said:
I have to say that HD is nice for resolution and color space..but, Digibeta has a "better" color space than mini dv and I would choose to shoot on the DVX over DIGI beta anyday.
if you expose and light properly you can get a GREAT look from these cameras. After all your real audience, the ones you make these videos and films for, will have a very hard time distinguising 16mm from the SD DVX100.
Click to expand...


DV over digi, huh? You want to rethink that for a second?
 
I have shot varicam, dvcpro 50 , dvcam and dv. I have seen alot of hdv...from jvc, and canon. Sorry guys...hdv doesn't look good. Effiecent codec? so what. long GOP is not the way to go.
and at rez fest they were trying to justify why the hdv footage looked bad...and the color space is worse than SD DV. Stay away from hdv, at least for now. It's not even a standard format between manufactures.

dont wast 6-9k on crap ....I'll take good SD over crap HD compressed within an inch of it's life anyday.
 
Most HDV decks in cameras will record to standard DV. Had Panasonic included a deck of that nature in the HVX200 it would completely dominate that market.

Wow think of all the choices!!!

Multiple frame rates, multiple formats, multiple resolutions.

Let's not pretend that anyone purchasing an HVX200 this year wouldn't mind the option of HDV, but alas Panasonic wouldn't sell as many P2 cards.

HDV consortium? Huh...it ain't the mafia, it ain't the illuminati, it aint the freakin Masons. It's not like it's some club Panasonic couldn't join
 
Huh, I may be wrong but , the Illuminatim and free Masons are all but the same thing?

anyway thats OT, interesting BUT OT, (UNLESS I SHAKE MY HANDS IN THIS MANOR) :wink:

Anhar
 
Barry_Green said:
Specs don't mean crap if the footage doesn't hold up.

Specs? Footage is king. Specs are not. Specs are not what you use in day-to-day use, they're what marketers use to try to sell the product. Footage is what you use to sell your services, not specs.
Click to expand...

Barry, you are a highly respected industry figure and many people, including myself, hang on virtually your every word.

But language like this makes me very nervous. ("Never mind that man behind the curtain.") Too many of your posts devote so much time to split screen and the unfortunate details of the HD100. By your own admission, you get to tag along on behind-the-scenes Panasonic interviews. What I want to hear about are SPECIFICS of the HVX200! That's what we're all here for, aren't we? Respectfully, if I wanted to hear complaints about the HD100, I'd be reading DVinfo.

By your own words, demo cameras are coming to Panasonic's salespeople November 8th. So the specs are now finished!!! Coy wording creates the impression Panasonic is ashamed of the specs. This is NOT what I want to believe!

Where are the HVX technical details and real PICTURES??
 
esperman said:
I have shot varicam, dvcpro 50 , dvcam and dv. I have seen alot of hdv...from jvc, and canon. Sorry guys...hdv doesn't look good. Effiecent codec? so what. long GOP is not the way to go.
and at rez fest they were trying to justify why the hdv footage looked bad...and the color space is worse than SD DV. Stay away from hdv, at least for now. It's not even a standard format between manufactures.

dont wast 6-9k on crap ....I'll take good SD over crap HD compressed within an inch of it's life anyday.
Click to expand...

by the way, ProHD is SHORT-gop HDV... I haven't seen a single motion-compression-artifact in footage from the HD101's HDV...
 
mezelf27 said:
by the way, ProHD is SHORT-gop HDV... I haven't seen a single motion-compression-artifact in footage from the HD101's HDV...
Click to expand...

i stand corrected on that point...

I wanted to like the JVC ProHD...I really did. But the codec is too constrained.
 
Bob Diesso said:
("Never mind that man behind the curtain.")
Click to expand...

Sorry, but I've been fielding the "what's the pixel count" question for about 8 months now. I've answered every possible permutation of pixel counts, I've gone into every circumstance I can imagine, I've pointed out how megapixel CCDs harm video quality in every camera that uses 'em, I've demonstrated mathematically how Panasonic could get a native 1080p picture from as little as 960x720 chips, I've tried to point out a hundred different ways that the actual chip count is nearly irrelevant as far as the final image is concerned...

... and yet people insist on asking this same question. Even though it's obvious we don't have the answer, we haven't been given it.

So I ask again: what would you do with the information, if you knew it? Would it somehow influence your buying decision? Would knowing the pixel count of the chip, in any way/shape/form, change your opinion of the product? If so, then I'd say your priorities are in the wrong place -- you should be making your decisions based on the FOOTAGE, not the pixel count.

And yes, I re-iterate my HD100 experience: on paper it may have sounded good, in practicality it had split-screen issues, chroma noise, poor sensitivity, and my particular unit had a defective slow-shutter system and very poor overall resolution. Should I cling to the specs to comfort me? Or should I look at the footage and say "screw this, this thing's going back to the store?"

I'm not saying we're not curious, I'm sure we'd all like to know the specifications, just like many of us want to know how many horsepower our car has. But we do not have that answer, and won't have it until Panasonic releases it, and may never have it.

Does it matter? Really and truly? Would you rather have a high-pixel-count chip, even if that means a split-screen picture and lousy sensitivity and a noisy image and fixed-pattern noise? Or would you rather have killer footage, regardless of what the pixel count is?

I'd give an answer if I had one. I don't have it. But in the intervening 8 months, I've had plenty of time to think about it, and I've come to the conclusion that the chip count is IRRELEVANT. It really, truly, deep down, doesn't mean jack squat. The only thing that matters is what the footage looks like. If the footage doesn't hold up, the camera's a dud, regardless of what the pixel count is. If the footage sings, the camera's a winner, regardless of what the pixel count is. How would pixel count influence that one way or the other?

Too many of your posts devote so much time to split screen and the unfortunate details of the HD100.
Click to expand...
That's because I bought one and used it and had firsthand experience with it, and knew the answers to the questions people were asking. With any luck, tomorrow I'll receive my RMA# and the HD100 will exit my life and all will be well again. But the point is: the specs sounded good. The footage was not. I decided that I was not comforted by the printed literature that listed the specs.

What I want to hear about are SPECIFICS of the HVX200!
Click to expand...
You've heard all that we have. When more is known, it will be discussed. It cannot be discussed before it's known.

By your own words, demo cameras are coming to Panasonic's salespeople November 8th. So the specs are now finished!!! Coy wording creates the impression Panasonic is ashamed of the specs. This is NOT what I want to believe!
Click to expand...
I doubt they're "ashamed" -- they all seem pretty darn proud of the product. But I have been told repeatedly that they will let the footage do the talking.

Here's two scenarios for you. Let's presume that in both scenarios, the footage is absolutely drop-dead gorgeous, okay? Let's also presume that the chip pixel spec is startlingly low -- let's say 400 x 300 or something ridiculous like that, and that through some sort of camcorder engineering wizardry, they make it deliver a full native resolution image. And, let's say that in side-by-side demonstrations, it's shown to be clearly sharper, cleaner, more detailed, with better sensitivity and better latitude than all the competition (XL-H1, Z1, HD100, etc).

In scenario 1, they announce the chip spec before footage is available.
In scenario 2, they announce the chip spec after footage is available.

What would happen in scenario 1? I'll tell you what would happen, people on this board and others would shriek holy hell and raise a hue and cry and stink about it and complain, they'd pronounce the product dead on arrival, they'd suggest that anyone still considering it would have to be an idiot, etc. Now, mind you, these are people who are NOT camera engineers, do NOT understand imaging technology, and would have never seen a frame of HVX footage. But oh how they would complain. And yes, that would taint the product and put a stain on its introduction, and many people may make up their minds to look elsewhere.

Now, in scenario 2, the footage gets released first. People see it and they go ga-ga. They go wild, they love it, and they start chanting "show us the specs! show us the specs!" And Panasonic eventually says the number. What's people's reactions at that point? More than likely they'd say "are you kidding me? Amazing! How did they do that? This is incredible, I didn't think you could do that, but the footage doesn't lie." Buzz continues to build, and instead of being excoriated, the engineers are praised for their amazing technological accomplishment.

It's not unlike filmmaking. Make a movie on a $500 budget and most of the jaded distributors will say "uh, no thanks." Let them see the movie first and get them excited to where they want to buy it, have them actually offer you a million bucks for it, and THEN tell them you made it for $500, and they'll be blown away. Same movie. Same budget. The difference is all in how you manage the info.

So -- again: I don't know the specs. I'm not worried about it, I've seen the footage. I look forward to finding out what the specs are, and I look forward to testing the daylights out of the camera to see what it's capable of. But I am not worried about what the specs are. I've seen the best specs, and I've seen what comes of that footage. Specs are overrated.
 
Barry_Green said:
So I ask again: what would you do with the information, if you knew it? Would it somehow influence your buying decision? Would knowing the pixel count of the chip, in any way/shape/form, change your opinion of the product? If so, then I'd say your priorities are in the wrong place -- you should be making your decisions based on the FOOTAGE, not the pixel count.
Click to expand...

Obviously the footage rules, but the footage is a result of DSP-ed light that was captured onto a sensor. Quality of footage and quality of CCD (in this case, number of pixels) is obviously very much connected.

Footage of a 400x300 camcorder might look good or fine, just as long as you know: it looks good for a 400x300 sensor-camcorder. Problem is, if sharpness and definition is what you're after, there is NO substitute for pixelcount. yes, you can interpolate missing pixels to enhance the resolution. You might do it linear, cubic, bicubic or even with spline-interpolation. The last one is very, very good. It's just no native resolution and it never will yield the same true definition in your footage. Pixels-shift is another technique, arrange the 3 CCD, so that the short in resolution is covered by the fact that colors that are given priority by our eyes are represented best. Better than interpolation? Better than linear interpolation. Probably no better then spline interpolation. DEFENATELY no replacement for native resolution CCD's.

If you can't have full res, pixelshift is the best. No doubt. But if you can have full res: even better. a whole lot better.

I understand the sentiment you're trying to make, but I also understand that someone is anxious to know the real specs, because they matter too. It isn't possible to mathematically prove that 960x720 is enough for 1080p, because it isn't. It means that the highest detail you can get in your camcorder from the outside world is your frame surface devided by (960 x 720). You can uprez that, but higher order corrections (like interpolation) are just 'mathematical guessing techniques' estimation that nothing weird happens in between scan lines. Obviously you lost that info at the CCD and you won't get it back.

Truly dosens and dosens of techniques exist ('nearest neighbours', ....), but they all have in common that they estimate what would be in between pixels, by information out of the rest of the image. No real world solution there, only continuous mathematical models. Nice solutions, no substitute for the real world.

Besides Barry, if that would be so - I was out of a job. No need then for us (mathematicians and physics masters) to refine new techniques...
 
I think Panasonic are just jealous that HDV can record higher resolution at 1/4 the bitrate of DVCPROHD...

:evil:


In all honesty, HDV can look pretty darn good at first generation. At present I would prefer the compromises of ProHD to DVCPROHD on P2, but in a year or two I am sure I will feel different. From what I have seen, 1080i HDV at only 25Mbps is pushing it. That does mean that I won't buy an HVX though. The footage will determine that.

If I am doing higher end work, I use more than a 1/3" camera. IMHO 720p HDV is a good match to this level of camera. I wouldn't go recording a $100M feature on it though!
 
Barry Green - Hail 2 the king

Barry Green - Hail 2 the king

Barry_Green said:
...And yes, I re-iterate my HD100 experience: on paper it may have sounded good, in practicality it had split-screen issues, chroma noise, poor sensitivity, and my particular unit had a defective slow-shutter system and very poor overall resolution. Should I cling to the specs to comfort me? Or should I look at the footage and say "screw this, this thing's going back to the store?"
Click to expand...

Barry - I, for one, have deeply appreciated your view/review(s) of the JVC 100. I run a small independant film company, and for us, the next camera we purchase will be the only one we use for the next 2 to 3 years, along w/ whatever else we need to make it sing (which editing system, etc).

Having never used the DVX I had no real loyalty to Panasonic and thought that, on paper, the JVC was the obvious way to go.... But then all those rumors about the SSE and other problems began to surface.

Out of all the reviews, on this site and others, I felt that yours was the most thought-out, least emotional, and most informative of those I had read. With your most recent postings, plus the others I had read, I now feel that I made the right desicion to wait for the HVX....

And that is what makes boards like DVXuser such a power tool. As a small shop w/limited time and money, the decision on which equipment to buy is invaluably aided by postings by yourself + others who take the time to help the rest of us who are not experts in cameras, et al.

Specs are interesting - but the picture is King.

Hail to the King baby - and keep up the great work on this board. Thanks....
John
 
mezelf27 said:
But if you can have full res: even better. a whole lot better.
Click to expand...
This is where the whole concept just doesn't work though.

Is native res better for resolution? Maybe -- keeping in mind that there is not a 1:1 relationship between CCD pixels and frame pixels.

But is native res better for IMAGE QUALITY? Not necessarily! And when it comes to 1/3" sensors, the answer is almost definitely NO. For 2/3", I'll agree with you. For 1/3", I don't.

The pixels get too small, and too much surface area is wasted with the dividing lines between the CCD registers. And the pixels get smaller than the resolving capability of the lens. So you end up with unnecessarily small pixels, which results in wasted image quality.

There is a see-saw effect, with competing, arguing factors in play. On one side you have resolution. On the other side you have everything else: latitude, sensitivity, noise, signal-to-noise ratio, everything. They are in direct competition. If you go for native-pixel res, you will be tilting the scales in favor of resolution, but necessarily and unavoidably harming sensitivity and latitude and SNR. If you go for larger pixels, you put more weight towards sensitivity and latitude and SNR, at the expense of maximum potential resolution.

There's a critical mass of pixel density and pixel size. Once you get above the "critical mass" you can expect great performance with native-res pixels; 2/3" cameras accomplish critical mass, where the sensor pixels are large enough that making them larger wouldn't add much power to the sensor, but would harm resolution. However, 1/3" sensors are way, way below that critical mass.

Look at it this way: what effect does adding another computer to your rendering farm have on the overall farm's performance?

The answer is, of course, that it depends on how large the farm was to begin with. If you have 100 computers in your rendering farm, adding one will have a negligible impact on performance. If you only had one computer, adding a second one will have a tremendous impact.

That's the way it is with these CCD pixels -- a megapixel 1/3" CCD has pixels that are so unbelivably tiny that we run into limitations as to how they can resolve an image. The ratio of dead area (between pixels) on the sensor vs. sensitive area grows dramatically as the pixel size shrinks too. Add a little size to that pixel and you have a massive impact on its performance.

With a 2/3" chip, it doesn't matter -- each pixel is so big that adding a little size to it would have minimal impact -- so at 2/3" size, native res is the goal.

But with 1/3", native res does more harm than good. The pixels are just too small to begin with; making them smaller than necessary just hurts video performance.

I'd love to be proven wrong, of course -- I'd love to have maximum resolution AND maximum latitude AND maximum sensitivity AND maximum SNR. But I don't think it's possible.
 
You must log in or register to reply here.
Back
Top