What format is most requested?

[David Heath]

Originally Posted by Glen Vandermolen

I assumed XDCAM HD meant XDCAM HD422.

In context i think it's more likely to refer to the original High definition spec for HD cam - such as used with the 1/2" disc cameras. As such was a family of bitrates and sub-sampled down to 1440x1080.

XDCAM EX was a development which gave full raster 1920x1080 in an mp4 container, but was also 4:2:0 colour sampling.

AFAIK the highest variant has always been known as XDCAM422 or XDCAM 50Mbs.

It does seem that the list is weighted towards older formats - but that may be simply reflecting client demands who are requesting what their workflow revolves around. I'm surprised to see DVCProHD so high on the list, whilst AVC-Intra hardly registers.

Originally Posted by Glen Vandermolen

I did read the review of the JVC HM600, and I'm very impressed. I really like that little camera. Its low light is almost as good as a 2/3" chip cam.

Can I ask what reference you have for the claim about the low light? A 2/3" chip has 4x the surface area of a 1/3" chip, and all else equal a 2/3" camera must have a 2+ stop advantage over the equivalent with 1/3" chips. The HM600 may indeed have things to commend it, but I find it hard to believe it will rival a modern 2/3" camera in low light.

[Glen Vandermolen]

JVC says the HM600 is rated at F11 at 2,000 lux. According to a JVC rep on this site, it's closer to F12.
The 1/2 EX cams are rated around F10. I think the new HPX600 is rated around F12 maybe.

I didn't say they were the same as a 2/3" cam, just that JVC has somehow made them almost as good, probably the best for 1/3" cams. Remember, 1/3" cams usually have very fast lenses.

Here's the JVC link:

JVC Professional Features page

And a good video review:

Exclusive UK Review of the JVC GY-HM600 on Vimeo

[David Heath]

Originally Posted by Glen Vandermolen

JVC says the HM600 is rated at F11 at 2,000 lux. According to a JVC rep on this site, it's closer to F12.
The 1/2 EX cams are rated around F10. I think the new HPX600 is rated around F12 maybe.

I didn't say they were the same as a 2/3" cam, just that JVC has somehow made them almost as good, probably the best for 1/3" cams.

In isolation, figures like "f11 at 2,000lux" don't really tell you very much. To have any meaning, they have to be qualified with noise figures at that setting, it all depends what s/n ratio a manufacturer is prepared to accept.

Hence, a camera quoted as (say) f11 at 2,000lux, might seem far better than one quoted at f8 at 2,000lux? But if measured s/n was (say) -50dB in the first case, -56dB in the second, then intrinsically they have comparable performance. The first camera is effectively performing the same as the second with 6dB of gain in!

And it gets even more difficult when software noise reduction gets used. At first sight this can indeed make a camera seem more sensitive - but it comes at a price, getting rid of the noise can mean other nasty artifacts, and can really cause problems in post work. (Classic example being the problems with the "noise ghosts" of the HPX371. Panasonic were using processing to make the apparent noise level seem lower - but it caused noise trails on moving objects. Solution is to switch the noise processing off - but then it reverts to poorer low light performance. You don't get anything for nothing.)

And in the JVC promo you link to, it's commented on that with 6dB in "some noise reduction is evident", so I suspect a similar story here. None of this makes the HM600 a bad camera - but I am extremely sceptical of claims like " JVC has somehow made them almost as good" (as a 2/3"). But maybe JVC would like to give a formal answer to what the s/n is at 0dB, and whether any artificial noise reduction processing is applied?

I believe the f12 figure you quote for the Panasonic HPX600 also relates to "noise reduction on", and similar circuitry to the HPX371? Turn it (and the "noise ghosts"!) off and expect it to be more like f8-f10, which is pretty well what I'd expect. Compared to a 3 chip 1/2" camera, it will gain about a stop due to the bigger area, but lose a bit more than a stop due to being single chip, not three.

[Glen Vandermolen]

Then don't buy an HM600!
Sheesh...

[David Heath]

Originally Posted by Glen Vandermolen

Then don't buy an HM600!
Sheesh...

That's not the point.

The HM600 has many good and unique features which may make it a good buy for some users. Ability to record 35Mbs MPEG2 to SD card directly (no adaptor) may be one, AVC-HD *AND* the XDCAM codec is also good.

(In the context of this thread, one client may ask for AVC-HD, another may ask for XDCAM 35Mbs - with this camera you can satisfy either with one piece of hardware.)

But if lowlight ability is of high importance to any potential user, don't be fooled into thinking it will be "almost as good" as a 2/3" camera in this respect, and I've tried to give some reasoning above as to why simple quoted figures and claims should be taken with a big pinch of salt. (And that applies to most manufacturers.)

[Alister Chapman]

To back up what David says, of all the camera specifications that manufacturers list, sensitivity has always been the most abused. I'm not saying that the camera manufacturers lie, but they will rarely reveal the the gain or noise levels at the stated sensitivity. They do this simply because they can, as most people just read the headline number without understand it's real meaning. It's the same with resolution where once upon a time resolution meant the resolving power of the camera, but now it's become acceptable (or at least common practice) to quote the sensors horizontal pixel count as the resolution, which is very different to the cameras resolving power.

Once upon a time, when cameras were much more expensive, many purchasing decisions were made by engineers who understood the implications of the numbers. At this time the manufacturers had to be much more careful about how they played the numbers game. Today very few production companies have engineers and as a result the camera manufactures and sales companies are able to use bigger numbers to sell cameras, even if those numbers are actually quite meaningless unless you know the full story behind how the numbers were determined.

[Sareesh Sudhakaran]

Originally Posted by Alister Chapman

To back up what David says, of all the camera specifications that manufacturers list, sensitivity has always been the most abused. I'm not saying that the camera manufacturers lie, but they will rarely reveal the the gain or noise levels at the stated sensitivity. They do this simply because they can, as most people just read the headline number without understand it's real meaning. It's the same with resolution where once upon a time resolution meant the resolving power of the camera, but now it's become acceptable (or at least common practice) to quote the sensors horizontal pixel count as the resolution, which is very different to the cameras resolving power.

Once upon a time, when cameras were much more expensive, many purchasing decisions were made by engineers who understood the implications of the numbers. At this time the manufacturers had to be much more careful about how they played the numbers game. Today very few production companies have engineers and as a result the camera manufactures and sales companies are able to use bigger numbers to sell cameras, even if those numbers are actually quite meaningless unless you know the full story behind how the numbers were determined.

+100. That's how this thing works...

[Gary Nattrass]

Sadly a lot of client's haven't a clue about anything technical and will just go with what is the latest flavour of the month, you also get the lo/no brigade who think that they can't shoot on anything less than a RED or Alexa.

I have even had post production supervisors on major TV drama product who have little knowledge about workflows or even shooting formats so I take any such list with a pinch of salt as it may just be what camera is requested as a hire cam or even a self shoot handycam and the EX1 tends to be the most common but more recently the canon XF305.

It's more about right tools for the job and if you are shooting news for ITV you still need a tape based DVcam but if you are shooting a lo/no self funded film project it is also silly to expect a RED or Alexa for free,

sadly most client camera choices these days are based on how cheap can I get it rather than how good will it look and function in the field or more importantly suit a post workflow.

As a P2 user I am surprised that DVCPro HD is even on the list as AVC Intra 100 is a far superior codec for shooting and post and has been around for nearly four years now, or is this list totally out of date???

[David Heath]

Originally Posted by Gary Nattrass

As a P2 user I am surprised that DVCPro HD is even on the list as AVC Intra 100 is a far superior codec for shooting and post and has been around for nearly four years now, or is this list totally out of date???

I think you may have gone a long way towards answering your own question when you say "if you are shooting news for ITV you still need a tape based DVCAM". Point is that an organisation may adopt a format and be with it for a number of years. Bigger the organisation, the higher the likely commitment, the more difficult to make a change. Hence it's very unlikely ITV news would equip themselves around DVCAM if buying now - but there's a lot to be said for making existing assets pay for themselves as long as possible.

And the relevance is that if they hire someone to shoot for them, that person will be expected to deliver in the house format. Hence it's perfectly reasonable to expect such a list to reflect formats that may not now be considered cutting edge.

And you say " AVC Intra 100 is a far superior codec {to DVCProHD} for shooting and post ......". If you're talking about pure quality for a given bitrate, that may be true, but it's far more processing intensive than such as DVCProHD, so in that respect may be seen as inferior. Likewise, DVCProHD can be shot to tape or to solid state. If anyones workflow currently revolves around tape, that puts AVC-Intra out of the equation.

For anyone with a single stand alone edit station, the processing overhead may not be too big an issue. But move up into the world of video servers and it alone may put AVC-Intra out of the question for the time being.

[Gary Nattrass]

points taken David and I am sure are right but in the case of ITV they are dishing out hd handycams to journo's but expecting freelancers to buy 500' s for the same daily rate!

I am working for them on Sat doing the FA cup and we have been told that we have to cable our sony 1500' s as it is impossible to send HD down a triax tie line. well we have been doing that via fibre to triax converter boxes for two years now doing premiership coverage and sony also make the 1550 that can do it without a converter box.

just goes to show that your client may be so far out of date with technology and don't know any better!

[David Heath]

Originally Posted by Gary Nattrass

......... but in the case of ITV they are dishing out hd handycams to journo's but expecting freelancers to buy 500' s for the same daily rate!

DSR500s, or PMW500s?

[Gary Nattrass]

Originally Posted by David Heath

DSR500s, or PMW500s?

PMW500 !

As a side note ITV have probably never seen a sony 1500 as they are generally ten years out of date!

Dsr 450 or 570 have been Dvcam std for that time.

[Alister Chapman]

Originally Posted by Gary Nattrass

As a side note ITV have probably never seen a sony 1500 as they are generally ten years out of date!

It's an interesting situation, only a few years back broadcasters would not generally adopt a new format, standard or particular camera until after it's been available for a year or so. This was to allow the bugs, issues, pros and cons to show themselves before making a commitment to a large investment. Then once the format was decided it would take a few months to get the budgeting approved, and a few more months for tenders and supply contracts to be sorted. So more often than not the cameras didn't arrive in the hands of the end users until a couple of years after the launch. These days with cameras coming and going in 18 months to two years how will the broadcasters and larger organisations keep up. I know that most manufacturers will give larger organisations sneak peaks at what will be launched around 6 to 9 months in advance, but even so. I guess many of the larger organisations with wheels and cogs that turn slowly are at the grass roots level at least destined to always be acquiring new tech just as it get replaced by newer and supposedly better tech. Then they have to teach everyone how to get the best from it, fine tune the workflow and get the whole infrastructure up to speed. Meanwhile the rest of the outside world has already moved on. I'm sure this in part explains why so many, what I would consider legacy codecs and formats are still very much in demand with rental houses. It may be that as many operators these days own their own kit and like to have the latest, current kit, that when requests come in from the larger organisations that are still struggling to recoup the investment made on 5 year old technology that the operators can only supply the "out of date" tech by renting it.

It's quite scary to think that HDV and DVCPro HD are codecs that were very much De-facto HD codecs only 5 years ago but by many are now considered archaic and completely out of date. For broadcasters used to getting 15 years out of Betacam etc that's a hard pill to swallow. It's also in part why Mpeg 2 in the form of XDCAM HD422 continues to hang on as a broadcast codec because the workflow etc is now so well sorted and ingrained in so many organisations that even though it's "only" 8 bit and a somewhat dated codec by todays standards, it looks like XDCAM will still be around in another 5 years, even though by then it really will be old technology.

[John Richard]

only" 8 bit and a somewhat dated codec by todays standards, it looks like XDCAMHD422 will still be around in another 5 years, even though by then it really will be old technology...

Maybe from an economic standpoint in very tough economic times we are hovering around the point of diminishing return considering the size and quality of the display venues (generally mis-calibrated flat panels in the 40" to 50" range) and the clamped pipeline delivery schemes of satellite and cable. Maybe HDCAM422 is already far beyond the distribution and display portions of the chain for broadcast. I would imagine if I were in the shoes of the "suits" responsible for turning a profit, it would be tough to make new investment that doesn't translate in a noticeable way down to their customers.

We are so lucky these days with the equipment and price points. It's exponential.
Ghosting tube based black and white behemoths to Betacam. How long and costly.
Betacam to Mini-DV/DVCam ........ HDV.......P2 ......External Recorders XDR/nanoFlash HDCam422 ..DSLR video..10bit codecs..C300.. F3.. BMCinema Cam in DNGRaw..................faster and faster.. better and better.. cheaper and cheaper ---- Big Boys - Red, Alexa, F65

[Alister Chapman]

I'm not really knocking XDCAM as a codec, it does an excellent job and I use it day in, day out and get great results. But, within the same 50Mb/s bandwidth you could have a 10 bit codec that would deliver higher quality images. Yes, we are in an area of diminishing returns and yes delivery is often far from ideal. We are lucky here in the UK to have a few very high quality HD channels and you can see the difference between high quality shows and lower quality ones. Conventional delivery over the air will continue for many years to come. But we are now seeing an ever greater number of viewers watching content on computers and personal viewing devices with delivery via the internet. As a lucky owner of both a Retina iPad and Laptop, I can tell you that good quality content looks simply stunning. The screens might be small, but when your only 18" from the screen you really can see the quality. This is very important and it's something the accountants are starting to wake up to, because it's much easier to control and monetize content delivered over the web than via conventional broadcast methods. It's also easier to profitably cater for smaller audiences as you don't need a traditional broadcast infrastructure. Where once it was simply not viable to produce a programme that might only attract an audience of 10,000 in a single country, with the web your audience can be global. Then 10,000 viewers in 20 countries becomes 200,000 viewers and if you make a dollar from each viewer that's a pretty decent budget for a documentary or short film. As the internet gets faster, as 3G and 4G becomes common place and personal viewing devices become less of a luxury item and more of a commodity item, this audience will grow.