Wind against HVX

[wabbit]

wabbit, human kindness is better than robotic netiquette, think about it
Denounce bad action of some people that only want make money (freedom of fraud! It's terrible!) damaging other people, is the first rule of serious people, this oughta be a "forum rule"! Censor amazing thoughts of Benton or me is not democratically good!
Benton SEI UN GRANDE! You're GREAT!!!

You want to discuss politics...no one will censor you...just take it somewhere else on the internet. You want to discuss religion, fine...just not here, it gets ugly and off topic fast.

Pointing out that companies are out to make money is no grand or "amazing" thought. Stating that some companies will behave in corrupt ways to get it is not particularly insightful either. But thanks for clueing us in none the less.

If you got some, even semi-factual, knowledge on why Panasonic should be trusted more or less then any other multi-national corporation, I and others will be happy to hear it. But innuendo that Panasonic should be trusted less because they are out to make a profit while the other camera companies are..."not?" is just trolling and looking to stir the pot.

I hear Sony is going non-profit this year. All profits are going to charity from now on. Also they are ditching their marketing dept since we all know marketing is just 99% propaganda and lies. So if you love your fellow man, buy Sony. :grin:

Sarcasm aside, yes, they all are in it for money but IMHO Panasonic has demonstrated a little better commitment to customer satisfaction (at the professional video level at least) then the others. Other people have had different experiences and are encouraged to share them. Eventually you have to decide what company you mis-trust the least. Or, or course, you could just not buy anything, that works too.

Either way, just keep please keep religion and politics out of it.

Cheers

 

[Barry_Green]

1. Is the camera really an HD camera

Most definitely yes. Seeing the live footage is just stunning.

or is HD achieved by some form of electronic manipulation that does in fact reduce imaging quality? (yes or no)

Big NO to that one.

2. Does one have to really have to spend $50,000 or more US dollars to achieve high quality HD? (yes or no)

No no no. $5995 will do it (well, that plus a P2 card!)

3. Will stations like INHD or Discovery HD accept material shot with the HVX200 if it is properly transferred and meets their individual broadcast standards and delivered on the master format the require ( yes or no)

My understanding on that is yes, but it's up to each network to specify that. I thought Steve Gibby had already clarified this by saying yes...

My really big question is that at the price point the HVX comes in at would one be better off looking at a high quality 2/3" SD camera that shoots 4:2:2 and is 16x9 native like a AJ-SPC700 and is DVCPRO 50?

Well, the thing is, the SPC700 costs more than double what the HVX does. So it's not a direct apples-to-apples comparison. And the SPC700 is standard-def, not high-def. And it'll have better low-light performance, lower noise, shallower DOF, but no variable frame rates, no 720p, no 1080p, no progressive scan (in the US), etc. Very different type of product.

 

[lpcvideo1]

O.K., group hug done.

 

[sanjin]

Sorry, but I've got the message.

 

[brianluce]

If you read the thread you'd know that it is true 4:2:2 in every mode (except DV and DVCPRO25 of cause)

The color subsampling relates to the internal luminance resolution of the codec.

dude, i read the thread and some guy named harddrive had my head spinning with talk of "square pixel rules" and "color matrices". the only matrix I know is the one starring keanu reeves.
but it seemed like he was saying the 4.2.2 in hi def was a misrepresentation of the actual sampling.

 

[harddrive]

( in reply to: "........and especially in a progressive system 4:2:0 evens out the chrominance resolution better than 4:2:2 (or even 3:1.5:1.5.......) - DVCPRO HD may give an unnecessarily high level of chroma resolution vertically, at the expense of the horizontal. ")

Okay, forgive me for being flabbergasted, but are you saying that a system with 1/4 the chroma resolution would be better than one with 1/2 the chroma resolution?

Of course I'm not saying that - read my original quote again: "an unnecessarily high level of chroma resolution vertically, at the expense of the horizontal". In a ideal world no reduction in chroma resolution would obviously be great - sadly that only exists in reality in high end digital cinema.

It's been decided as sensible that if chrominance res DOES have to be reduced, it's overall better if the reduction is done equally in horizontal and vertical directions. Exactly as happens in a progressive system with 4:2:0. Yes, the vertical resolution of DVCPRO HD is fantastic - but they've robbed Peter to pay Paul - it's achieved by being far greater than in the horizontal direction.

I find it pretty sad the way this discussion has gone whilst it's been night here. It was a good technical debate, and it's a shame to wake up to find it degenerated into the old "my cameras better than yours", and numbers used to prove points rather than in the search for truth. Or worse, people just making cheap jibes.

All my offerings have been referring to CODECS, and I've been deliberately ignoring real world cameras, no of pixels on chip, lens performance etc etc. Hence I wouldn't want any conclusions to be drawn about real life performance. That said, and comparing DVCPRO HD to HDV on a pure CODEC basis, then the point that comes out of all this is that whilst the former is undoubtably superior to the latter in some respects, it is not a one way street. Overall, I believe DVCPRO HD IS the superior codec, though coming at the cost of 4x the data rate. But to pretend it is superior in every respect is absurd, and to gloss over the differences is self defeating.

I've heard that the military strength of countries is going to be defined by reference to an x:y:z system. "X" will represent the total population of the country, and be normalised to "10", and Y and z will represent the ratios of numbers in the army and air force. (My apologies to any Naval personnel reading this.) Apparently, the US has 100 million citizens, 10 million in the army, and 10 million in the air force - it is said to have a 10:1:1 system.

I've further heard Tosh Bilowski has just been made head of an evil dictatorship, "Cameraland", with 10 million loyal subjects. He has recruited 1 million into the army, and 1 million into the air force. In his latest broadcast he reassured the population that his countries military might was completely up to combatting the United States as "..... like them, our armed forces are also at 10:1:1 strength!!" (Cheers from crowd.) A perfectly true statement, but, errr....... rather misleading wouldn't you say?

I'm now writing this from jail, having been thrown there for posting on a political website that Tosh should more properly have said Cameraland has a 1:0.1:0.1 system. If you must make comparisons, make them like with like.

 

[harddrive]

but it seemed like he was saying the 4.2.2 in hi def was a misrepresentation of the actual sampling.

I'm saying Tosh was making a comparison, but not of like with like. Sonys HDCAM subsamples luma in the 1080 mode down from 1920 to 1440 for recording (3/4), so the system starts "3" - 3:1:1. DVCPRO HD in 720 does exactly the same, so there is a lot of justification for anyone to say the same rules should be followed. Since there is one chroma sample for every two of luminance, hence the 1.5. (In HDCAM there's one chroma per three of luma.)

HDV1 uses square pixels, the full 1280 across, so the "4" is fully justified.

 

[Anders Holck]

dude, i read the thread and some guy named harddrive had my head spinning with talk of "square pixel rules" and "color matrices". the only matrix I know is the one starring keanu reeves.
but it seemed like he was saying the 4.2.2 in hi def was a misrepresentation of the actual sampling.

Yes, because he misunderstood the definition of color subsampling.

There is no square pixel rule.

DVCPRO HD, HDCAM and HDV1080 uses a reduced horizontal frame size: 960 instead of the full 1280, and 1280 or 1440 instead of the full the 1920.

AFTER that initial reduction you apply the color sub sampling that we state as 4:2:2, 4:2:0 or 3:1:1.


Hwe also falsly states that the 3 in 3:1:1 of hdcam color subsampling relates to the 1920>1440 difference, it does not.

HDCAM compression first resizes the horisontal resolution to from 1920 to 1440.
This is the luminance resolution.

Then it subsamples the 2 color channels even further. this is where 3:1:1 comes in.
3:1:1 means that the color channels horisontal resolution is a third of the luma channel.
As we just stated the luma was 1440 (not 1920), one third is 480.

So the 3:1:1 of HDCAM really represents 1440:480:480.

Again the 3 figure has absolutely nothing to do with the initial reduction of 1920 into 1440.



Really the odd format here is HDV720 that is using the full horizontal resolution, in 4:2:0 colorspace.

So HDCAM, DVCPRO720, DVCPRO1080 and HDV1080 are all fuzzy math according to him.

Fact is that only HDCAM SR, D7 and HDV720 is sampling the full raster for a square pixel format.


HDV1080 is also reducing the horizontal resolution from 1920 to 1440 and then subsamples the color to 4:2:0. And this is still called a 4:2:0 system.



This is what we call FUD.


We dont know how the HVX200 will be compared to current SHIPPING HDV cameras.

What we have seen is samples from a preproduction unit.

Panasonic has not released info about the CCD resolution, but at the same time the camera also is'nt released yet!


What is true is that on paper HDV720 has 320 more horisontal luma pixels but almost half the color pixels compared to DVCPRO720.

But buttom line is DVCPRO720 and DVCPRO1080 are both righfully called 4:2:2 systems.


But enough is enough.
My point in the beginning of this thread is that in the link in the first post, the authors are making assumptions of a camera that they have not used or seen, and never are going to use for three months. (As they are european)

 

[Anders Holck]

Sonys HDCAM subsamples luma in the 1080 mode down from 1920 to 1440 for recording (3/4), so the system starts "3" - 3:1:1..

And this is where you assume wrong...

HDCAM first subsamples all channels by 1.33 in the horzontal direction, then further applies a 3:1:1 subsampling to the color components.

We start with a 1920x1080 frame YUV.

So Y which goes in at 1920 gets reduced by 1.33 to 1440.

U and V is reduced by 1.33 as well to 1440 then gets further 3:1:1 subsampling which is a third and ends at 480.

Please, this is the specs!

 

[JackFoley]

The FUD tactics of such incompetent sites like this VCM-whatever make me sick.

Still, I can understand that some people will be disappointed when they take a closer look at DVCPRO HD 720 footage when full-rez 720p 4:4:4 is what they really want :happy:

This is a 800x from the native dvcpro HD 720 60p stuff Jarred just posted, right off the FCP timeline. Since the footage needs to be scaled back to 1280 from 960 it looks like there are 3x3 color blocks, not 2x2.

But don't get me wrong, I am far from complaining because this is the best footage one can get in the 10.000 price range.

The only thing that does not look so good in this footage is, again, the edge enhancement visible around bright spots. But Barry already confirmed that this can be turned down, so we should be fine here.

 

[eLeventy]

Here in belgium panasonic isn't a big broadcast player either. And I don't think that will change soon. It doesn't have to for me. I get stunning pictures out of my dvx because it leaves you the ability to tweak it like no other 'handycam' before it.
Over here Sony rules, wich is why they are so unpleasantly arrogant. Allmost like you should be glad they sell you a camera.

Well, 'over here in Belgium', Sony just got a bit smaller and Panasonic a bit bigger. Panasonic just sealed a very large deal with VRT ( belgian's largest broadcaster) to supply the P2 SPX-800 cam as the main ENG camera. This means that at the moment the market is split into 2: the two public broadcasters( VRT and RTBf) going with Panasonic. The 2 main commercial channels are still Sony: RTL-tvi has gone Sony XDcam, and VTM is still using Sony SX

 

[OliverM]

Alé vooruit, nooit gedacht dat uitgerekend de nationale omroep die stap ging zetten. Maar het zat er aan te komen natuurlijk.
Ou êtes-vous francophone?

Sorry for the dutch and the bit of fench everyone else. That's good news. Very good news.

 

[Danilo Del Tufo]

meddesigner, You have said what I want to tell with my message if I could make it into italian language. I'm worried because it's my family that pay this camera, i'm student and I don't work. So you have understood that I want to buy a thing that works, it's for me and my parents a "sacrifice" spend all that money!
Thanks to appreciate my honestly words!
I tried to post yesterday but dvxuser has some problems to open! Sorry to wait my message!

p.s.: I've a problem with dvinfo, I can't post nothing and I dunno why, I've emailed to chris Hurd that often is here in Dvxuser. What I've to do? If someone knows a moderator of Dvinfo that comes here could say to him my problem? I've send him (Chris Hurd) many email but he dosen't reply me!

p.p.s: Cinemek.com, doesn't open! It has some problem! Anyone knows what's happening there?

 

[redindian]

I tried to post yesterday but dvxuser has some problems to open! Sorry to wait my message!

p.s.: I've a problem with dvinfo, I can't post nothing and I dunno why
I've send him (Chris Hurd) many email but he dosen't reply me!

p.p.s: Cinemek.com, doesn't open! It has some problem! Anyone knows what's happening there?

you are being systematically disintegrated from all the US based forums for your fiery posts - your cell phone is being tapped, and dorm is being watched... BOO

so now - do you still maintain that Panasonic is evil? hmm..wait till they switch off your electricity!

 

[eLeventy]

Alé vooruit, nooit gedacht dat uitgerekend de nationale omroep die stap ging zetten. Maar het zat er aan te komen natuurlijk.
Ou êtes-vous francophone?

Sorry for the dutch and the bit of fench everyone else. That's good news. Very good news.

Vlaams :thumbsup:

I think large corporations like VRT are less and less falling into the PR trap from big company sales reps, and more and more listening to the intented users. Finally!

On the other hand, Panasonic Europe can take a lesson or two from Sony Public-Relations wise.

 

[Barry_Green]

Of course I'm not saying that - read my original quote again: "an unnecessarily high level of chroma resolution vertically, at the expense of the horizontal".

But -- that still makes no sense to me. Let's take the example of DV50 in PAL: chroma sampling of 360 x 576. Would you classify that as "unnecessarily high vertical resolution?" Would you instead argue that 4:2:0 (assuming progressive encoding) would result in a better, more-balanced picture? Because it would be 360 x 288? Not a chance. Not a chance at all; DV50 spanks DV 4:2:0 from here to pittsburgh, as it should.

Likewise it is with chroma resolution in DVCPRO-HD. There is simply not a solid ounce of ground to stand on to argue that 720x540 is going to look better than 640x1080. It just doesn't work that way.

It's been decided as sensible that if chrominance res DOES have to be reduced, it's overall better if the reduction is done equally in horizontal and vertical directions.

By who? Last I checked 4:2:2 was universally, without a doubt, unquestionably preferred in every shooting circumstance over 4:2:0. 4:2:0 reduces chroma equally in the horizontal and the vertical, as you state. 4:2:2 reduces the chroma horizontally, but not at all vertically. Yet 4:2:2 is obviously, easily, and vastly preferable over 4:2:0. I really don't follow your statements at all here.

Is this all an attempt to justify HDV1, by trying to exploit the lower horizontal subsampling of 720p DVCPRO-HD? Because if you want to make a specific case like that, stick to the specifics. But to say that it's been somehow universally decided that cutting res in both directions is preferable to NOT cutting res, I think that's a statment that is completely indefensible.

That said, and comparing DVCPRO HD to HDV on a pure CODEC basis, then the point that comes out of all this is that whilst the former is undoubtably superior to the latter in some respects, it is not a one way street.

It pretty much is though. There is no practical circumstance where HDV (either format) is going to deliver a demonstrably superior picture versus DVCPRO-HD, unless you want to concoct a specific motionless lab test that focuses only on luminance and restricts all motion. In that one case, which is a grossly unrealistic scenario, then yes, 720p HDV will show more resolved luminance detail than 720p DVCPRO-HD. But once things start moving, and once you add color to the mix, 720p DVCPRO-HD will take a decided advantage. And once you start talking about 1080i, DVCPRO-HD is undoubtedly substantially superior. It is pretty much a one-way street. Or, more accurately, it's an entire city full of streets that only go one way, but there's one little rarely-used back alley that runs the other way.

 

[harddrive]

Likewise it is with chroma resolution in DVCPRO-HD. There is simply not a solid ounce of ground to stand on to argue that 720x540 is going to look better than 640x1080. It just doesn't work that way..

Regarding 4:2:2 and 4:2:0. Firstly I'm glad to see you add the rider "assuming progressive encoding" above, as I did, my references were specifically to 720p following on from the original query. For PAL, or rather any interlace system, there is more point to using such as 4:2:2 to 4:2:0 as the interlace puts subsequent samples two lines apart anyway - using 4:2:0 would put them 4 lines apart.

These issues don't exist with progressive. In an ideal world 4:4:4 would be the way to go, realistically that's only practical for digital cinema at the top end. From the early days colour television has exploited the eyes reduced colour resolution relative to luminance, and in analogue systems chroma bandwidth has been well under half that of luminance. If the no of chrominance samples is fixed at 1/4 the no of luminance, surely it's better to have half as many vertically and horizontally, rather than one quarter horizontally, and full vertical sampling!?! A balanced system?

The numbers aren't quite like that for DVCPRO HD though, but just ask yourself whether half the number of vertical samples may not be better put to improving the horizontal luminance and chrominance resolutions, or lowering the anount of intra frame compression? What's the point of compromising those factors, only to use samples to increase vertical chrominance resolution beyond what the eye can resolve!?!

Is this all an attempt to justify HDV1, by trying to exploit the lower horizontal subsampling of 720p DVCPRO-HD? Because if you want to make a specific case like that, stick to the specifics. But to say that it's been somehow universally decided that cutting res in both directions is preferable to NOT cutting res, I think that's a statment that is completely indefensible..

I'm not trying to "justify" HDV1, nor deride DVCPRO HD. It's just that I don't see the world in the black and white terms you do, and I don't see why Tosh (or you) should selectively quote to try to make that seem true. Comments such as "DVCPRO HD is better than HDV overall" I wouldn't argue with, neither would I if the former was portrayed as light grey, the latter dark grey. But white and black? That just shows bias.

There is no practical circumstance where HDV (either format) is going to deliver a demonstrably superior picture versus DVCPRO-HD, unless you want to concoct a specific motionless lab test that focuses only on luminance and restricts all motion. In that one case, which is a grossly unrealistic scenario, then yes, 720p HDV will show more resolved luminance detail than 720p DVCPRO-HD. But once things start moving, and once you add color to the mix, 720p DVCPRO-HD will take a decided advantage. And once you start talking about 1080i, DVCPRO-HD is undoubtedly substantially superior. It is pretty much a one-way street. Or, more accurately, it's an entire city full of streets that only go one way, but there's one little rarely-used back alley that runs the other way.

All of the above really illustrates the "black-grey-white" point. HDV will be superior to DVCPRO HD (all else equal) for a motionless scene as you say - it will have better horizontal resolution, and the I frames will be compressed less than in DVCPRO HD (all I frames). Add a certain amount of motion and it will cope fine - still look better than DVCPRO HD. Increase the motion beyond a certain point and the data allocated to the B,P frames will not be enough to cope transparently, increase more and the quality will fall below that of DVCPRO HD.

I really don't want to be too much of an apologist for HDV - I really do think DVCPRO HD is overall better. (It damn well should be at 4x the data rate!) But it really isn't as black and white as you make out, and trying to do so weakens your case overall.

Likewise, if the chips and front end of the HVX are superior to such as those of the Z1, all this becomes less relevant anyway. That may well outweigh any advantages of HDV for low motion scenes. Time will tell. Let's try and keep open minds, eh?

 

[Danilo Del Tufo]

you are being systematically disintegrated from all the US based forums for your fiery posts - your cell phone is being tapped, and dorm is being watched... BOO

so now - do you still maintain that Panasonic is evil? hmm..wait till they switch off your electricity!

I'm still can't post anything, but I'm ever reiceving "Hello Danilo is great to see you again!". I've take no banning info, it's some error I think. But nobody wants help me! Chris Hurd where are you? There I can't post!