XDCAM-USER.com

Sony recently released version 1.63 of their XDCAM HD firmware. It adds many improvements and a number of small bug fixes as well as making some of the new WiFi options available. My favorite new feature is the ability to display your clip numbers in the viewfinder. Why it’s taken so long to implement this I don’t know, but finally it’s here. This update also allows you to add the Sony WiFi adapter so you can monitor and log material via a smart phone. More details can be found here on the Sony US web site.

Just spotted this tucked away on the Sony website. I’m downloading it now to take a look but it appears to be a plugin that allows you to edit native XDCAM HD MXF’s and proxy files in FCP, as well as native EX files. This could be a really nice tool. Once I’ve tried it out I’ll post back. Click Here to go to the sony site. Sadly it’s not free, but it’s certainly interesting and may open up the proxy workflow to FCP users.

I wrote about this in an earlier post but with the Oscars being given away later tonight it made me think about the biggest financial hit of the year, Avatar. It seems likely that Avavtar may pick up a gong or two at the awards, probably not for the story, but perhaps for the technology used to make the film. Much of Avavtar was created by filming real actors and mapping their actions and facial expressions on to computer generated characters. Cameron even has a system that allows him to shoot a scene with the actors and then play back the animated version almost instantaneously, all very clever stuff. But that’s not the reason for this post. For me the more significant factor was that the live action scenes were shot using HD video cameras. These were not some ultra resolution 4k, made-for-cinema cameras but modified Sony HDC-F950′s. The modifications were made by Vince Pace to make them narrow enough for side-by-side 3D, but essentially the image quality is unchanged from stock F950′s. I doubt there were many cinema goers that realized that they were watching was 1920 x 1080 HD video, and this is I believe significant. What this says to me is that well shot HD is perfectly adequate for mainstream cinema release. I’m not saying that it is the best way to shoot a movie, but that we should not be afraid to use modern high end HD video cameras for movie cinematography where appropriate, possibly due to budget constraints. It also makes me wonder just how many people would be able to tell the difference between 4k, 2K and HD in an average cinema.

Canon has been keeping quiet for some time now and there has been lots of speculation about their next video camera, including many that had hoped for a 35mm sized sensor. Well this morning Canon posted a press release on their website. The main and most exciting point is that the camera under development will be full 1920×1080 and it will be 50Mb/s 4:2:2 Mpeg 2 MXF. Now this looks on paper to be an extremely similar format to XDCAM HD422. It would certainly make edit suite integration a lot simpler if the MXF files are the same as the Sony XDAM MXF’s.

Mockups of a camera seen in recent months have shown a handycam style, fixed lens camera with two slots for some type of memory card, which could possibly be express card slots. The mockup lens looks like it might be an EX1 style lens with manual and auto focus and might be big enough to accommodate 1/2″ sensors. Is this the camera that the press release refers to, or is there another camera in the pipelines? There is a lot of stuff not mentioned in the press release, like recording media. It says “file based”, this could be solid state or it could be something else, maybe optical disc. Might this be a full size XDCAM HD camcorder from Canon? The release gives no sensor or form factor information which I find a little odd. Having shown mockups of an EX1 sized camera why not a more detailed press release with info on the lens, sensors, recording media etc? (apparently there will be another press release on the 8th of Feb).

Canon make some very good video cameras, I had an XL-H1 and it was a great HDV camcorder. I have no doubt that this new camera will be very good and competitively priced. IF it is the EX1 sized camera and it has 1/2″ sensors then this would tick all of the BBC’s boxes for HD. If it’s CCD (which seems likely) it won’t have skew or flash banding. This is a very significant announcement and could push Canon to the front of the Pro Handycam pack. Here is the full press release from the Canon web site.

New Canon MPEG-2 Codec chosen for file-based professional video camcorder promises compatibility with industry-standard editing & processing software

United Kingdom / Republic of Ireland February 2nd 2010 – Canon Inc. today announces the adoption of an MPEG-2 Full HD (4:2:2) file-based recording codec for a new professional video camcorder currently under development. The Canon MPEG-2 codec will enable high-quality imaging and audio performance with up to 50 Mbps data recording and twice the colour data of HDV HDV is a standard for the recording and playback of high definition (1,440 x 1,080 pixels) video and audio on DV-format cassette tapes profile formats. File-based recording helps video operations realise greater efficiencies during post-production processing, making it an ideal format for many industry applications such as news gathering, documentary filmmaking and event videography.

MPEG-2 Full HD compression and 4:2:2 colour sampling
The adoption of MPEG-2 Full HD (MPEG-2 4:2:2 HP@HL compliant) compression enables the recording of 1,920 x 1,080-pixel full high-definition video. Additionally, compared with the 4:2:0 profile format used in HDV and other standards, 4:2:2 colour sampling offers twice the volume of colour data, providing double the level of colour resolution.

Maximum 50 Mbps data recording
With approximately twice the total data volume of HDV, the codec supports higher resolution and increased colour data to enable the recording of high-quality video.

Industry-standard MXF file format (see note (II))
MXF (Material eXchange Format) is a widely supported open source file format for the recording of video, audio and metadata, developed to suit the latest editing systems used by broadcasters.

Canon partners with major editing and processing software
With the adoption of the MPEG-2 Full HD (4:2:2) file-based recording codec, Canon is working in cooperation with Adobe Systems Incorporated, Apple Inc., Avid Technology, Inc. and Grass Valley to ensure compatibility with major editing and processing software programs widely used within the video imaging industry. Additionally, at future industry events Canon intends to demonstrate the overall video-production workflow, from initial video capture to clip-trimming and final editing. Video clips stored in a file-based recording system and industry-standard software applications will be used.

(II) A format for professional digital video and audio media defined by the Society of Motion Picture and Television Engineers (SMPTE)

Notes to editor

Advantages of File-Based Recording

File-based recording enables video and audio data to be managed and stored by file, much in the same way as computer data. It supports efficiency throughout the production process, from initial video capture to final editing through the entire workflow. Additionally, file-based recording provides users with the flexibility to utilise different editing environments and workflow solutions without the restrictions associated with some other video recording formats, helping to reduce investment costs.

I was reminded of this by Perrone Ford on DVINFO.net. With HD cameras compared to SD cameras the depth of field appears shallower. Why is this and why is it important?

Visually depth of field is the loss of focus as you move away from the object that you have focussed on. If you have two cameras, one HD and one SD and they both have the same lens at the same aperture along with sensors of the same size then the change in focus with distance for both cameras will be exactly the same. However with the HD camera, because the image is sharper to start with, any small changes in focus will be more apparent than with the softer picture from the SD camera. So visually the HD camera will have a shallower depth of field. Now if you take that HD image and convert it to SD then the depth of field appears to increase again. This can be calculated and measured and is defined by the “circle of confusion”

So why is this important? Well lets look at what happens when you shoot an interview or face. The human brain is very good at looking at faces, we “read” faces day in and day out, taking in expressions, skin tone and subtle changes. We use these tiny visual cues to gauge emotion and see how someone is responding to the things that we do. Because of this any imperfection in the look of a face in a video tends to stand out (thats also why you normally expose for faces). With HD it’s quite possible to have a shot of a face where the tip of the persons nose or their ears are in sharp focus while the eyes are slightly soft. With an SD image we would be unlikely to notice this because of the greater depth of field, but HD with it’s visually shallower DoF can show up this small difference in focus and our brain flags it up. Very often you see the HD face and it looks OK, but something in your brain tells you it’s not quite right as the eyes are not quite as sharp as the nose or ears. So this apparently shallower DoF means that you can’t just focus on a face with HD but you must focus on the eyes, as that’s where we normally look when engaged in a conversation with someone.

Ever since the release of the XDCAM EX cameras users have been having problems getting good looking SD pictures out of downconverted HD.
Why is this and what can be done about it? This is an issue that effects all high resolution HD cameras and is not unique to the EX’s. There are two key issues. The first is the way basic software converters handle fields in interlace material and the second is the amount of information in an HD image that must in effect be discarded to get a SD image.
At first glance you would think that starting off with lots of picture detail would be a good thing, but in this case it’s not. Let’s see if I can explain.
Imagine that you have something in you HD picture that over 4 pixels goes from light to dark, in Hd you get a gradual transition from light to dark and all looks good. Now what happens when you take those 4 pixels and convert them to SD. The 4 pixels become just 2 and instead of a stepped change from light to dark the picture now goes instantly from a light pixel to a dark pixel. If these pixels were the edge of a moving object, as it moved the pixels would be switching instantly from on to off and unless the object moved at exactly one pixel per frame you will get a flickering effect. Clearly our nice gradual transition from light to dark has been lost and if there is any motion we may now be seeing flickering edges. Niether of these look good.

Take a look at these images:

The original frame with a box showing the area of interest.

Original HD image

Image down converted to SD then resized same size as original

As you can see the down converted SD is very blocky and there is some strange patterning (aliasing) going on amongst the bricks of the houses in the background. This does not look good and if there was motion the brickwork would shimmer and flicker.

So what can be done?

Well the best way to improve the SD down conversion is to soften the HD image before it is down converted to prevent this single pixel light to dark switch from happening. You need to end up with an SD image where you go from full light to full dark over at least 3 pixels to prevent flicker (Twitter).

How much you will need to soften you HD by will depend on how sharp it is to start with. Simply turning down the cameras detail settings can be a big help, but even then the best results are often obtained by applying some kind of blur filter in post production. In FCP I find the flicker filter works quite well. As you can see from the frame grab below the difference in the quality of the downconvert is quite striking.

SD downconvert with blur added to HD before conversion

I have also found that another problem is that the detail settings on an HD camera are not optimised for SD. The detail correction edges created in HD are very thin and when these are down converted to SD they all but disappear and can cause further aliasing. The solution is to make the detail correction edges thicker (on an EX turn detail frequency down to -60 to -99) but this then looks ugly in HD. The bottom line is that a camera optimised for HD works best in HD and SD will be a compromise.

Update: PDW-700 now on BBC’s official approved list.

This is getting very confusing! Last week I was told by several different sources working on productions for Sky HD and the BBC that the BBC & Sky were going to agree to a common minimum standard of 50Mb/s with 1/2? sensors for HD production. Today I hear from a production company that is making HD programmes for the BBC that they were told at the end of last week that 35Mb/s from the XDCAM EX1, EX3 and probably the new PMW-350 is acceptable. Right now there is a lot of confusion over what is and what is not acceptable. I wish the BBC or Sky would publish, in public their HD production guidelines, I was told this would happen last Thursday but so far no-one has been able to find anything. The confusion over what is acceptable makes it impossible to plan productions or make any equipment purchases and this can only harm the industry as everyone holds off waiting to find out what they can or can not use. Come on BBC and Sky… lets us know whether 35Mb/s is acceptable. My guess, reading between the lines is that if you shoot at 50Mb/s 4:2:2 either with a PDW-700/F800 or with an EX with a NanoFlash you will be fine, whatever the BBC or Sky’s decision. I also think that 35Mb/s from the EX’s will be allowed under dispensation or with restrictions to usage and workflow. One thing that appears to be lost in this whole debate is that the EX at 35Mb/s 4:2:0 has the same compression ratio as the PDW-700 with it’s 50Mb/s 4:2:2. The extra bits used by the 700 are taken up by the extra colour information. So in terms of blockyness and robustness through the production chain there should really be little difference between the two. My feeling is that much of the confusion is caused by the older PDW-F350 type XDCAM HD cameras which recorder 1440×1080 at 35Mb/s. This “thin” raster signal causes a lot more problems in post than the full raster of the EX cameras. The re-sizing from 1440 to 1920 (and back in the case of Sky who broadcast 1440×1080) leads to soft pictures and nasty compression artefacts, so I think that these cameras will NOT be allowed while the EX might.

There’s lots of rumors flying around of a pact between Sky HD and the BBC over the minimum acceptable standards for HD acquisition. It’s been known for some time that the BBC wants a minimum bit rate of 50Mb/s and camera sensors that are 1/2 inch or bigger. At the same time there have been many Sky productions produced using XDCAM EX cameras (Road Wars for example) which are shot using XDCAM EX and are 35Mb/s. The BBC have also been using XDCAM EX within many HD productions. Now however it looks like Sky is going to follow the BBC and insist on 50Mb/s for all HD production. 50Mb/s is not a surprise, but what’s being said now is that there will no longer be any exceptions to this rule. So RawCut, the producers of Road Wars are now having to work out ways of shooting at 50Mb/s while still using their existing EX camcorders. The NanoFlash from Convergent Design is an obvious solution.

This new, stricter ruling is not good news as it makes it harder to produce HD programmes with small budgets. If HD is to grow and become the norm for TV production, excessively ridged regulations like this will not help. Surely it would be better if exceptions were allowed. Perhaps following an evaluation or consultation process to allow the use of cameras that fall outside of the 50Mbps, 1/2″ minimum criteria for shows such as Road Wars where filming inside police cars with full size cameras simply is safe or practical.

I’ve seen Road Wars on Sky HD and it always looked fine to me. What will Sky and the BBC’s view be on all the Discovery and National Geographic produced material which doesn’t meet the Sky/BBC specs?

This reminds me of when DVCAM was launched. The BBC and others all announced that DV was unsuitable for broadcast. Well look at SD production now. DVCAM is the workhorse format used by just about everyone for low budget work. The BBC owns thousands of DVCAM camcorders and even has it’s own special “DV Solutions” department. I suspect that once the BBC and Sky realise that they will be closing the door on many interesting HD projects this ruling will gradually disappear into the mists of time.

OK, here’s my take on the situation.

If money is no problem then the safest bet is to purchase a good quality HD lens, expect to spend at least £8k.

If you budget is restricted then the situation is much less clear. There are now several low cost 2/3″ HD lenses designed for cameras such as the Panasonic HPX500. In my opinion these lenses are just not worth the money. They might be cheap (£4k ish) but the one’s I’ve played with have been pretty grim, suffering from lots of CA and soft corners.

If your on a tight budget the best thing you can do is take your camera to a good dealer and go through their second hand lenses, trying them on the camera. Check for resolution (use a chart), corner softness, CA and contrast. I did this and ended up with a Canon 16x8x2 IF lens. I found that lenses with lower zoom ratios tended to be better than those with higher ratios. I’m really pleased with my lens and when compared to the latest HD equivalents I can not tell the difference in real world use. It certainly outperforms all the budget HD lenses I’ve tried.

One interesting thing that I have discovered in my research into this subject is that Contrast is what makes the biggest difference in lens performance, not simply resolution as one might expect. Visually the next thing you notice is CA. This is a tough one as when you increase the resolution or sharpness of a lens you also tend to increase the CA.

Until lens manufacturers start to release MTF curves for their lenses the only thing we have as buyers to go on is the advertising blurb. It’s easy for a manufacturer to claim improved performance or new glass or other technology, but without accurate MTF curves it’s all pretty meaningless. You would only need the tiniest resolution improvement to be able to claim that you new HD lens range is sharper than your SD range, it could just be a fraction of a percent difference.

I was recently given the opportunity to use Sony’s new PDW-F800 on an airshow shoot. We had the camera for 2 days along with a beautiful Zeisss 6-24mm T1.9 Digizoom lens.
The F800 is very similar to the Sony PDW-700, it shares the same body and much of the same electronics including the sensors and laser deck. It does everything the already well respected PDW-700 does plus a few extras. The “F” in the F800 denotes that this camera is from the Sony Cinealta range.

Alister Chapman filming with PDW-F800

That means that it will shoot at 24P, well 23.98 to be exact, out of the box without having to buy any options. On the PDW-700 you can get 24P by getting an optional upgrade kit fitted, which includes a Cinealta sticker! The other key feature of the F800 is it’s ability to overcrank and undercrank, it can in fact shoot at any frame rate from 1 frame per second to 60 frames per second, something that the PDW-700 can’t do and won’t ever do. Picking up the camera for the first time I was reminded of how well balanced the 700/F800 cameras are. They sit very nicely on the shoulder with all the controls where you would expect them to be. Switch the power switch to on and within a couple of seconds you are able to shoot. As the PDW camcorders have a memory cache that can buffer video before writing to the disc, the camera can actually record images before the optical disc drive has spun up to speed making it very responsive. Even when the camcorder is in save mode there is no delay between pressing the record button and pictures being recorded, unlike most tape camcorders where it takes a small amount of time for the tape to start rolling.
Talking of cache record, this is one of the greatest features of the Sony XDCAM HD camcorders. On the PDW-700 and F800 you have upto 30 seconds of memory cache to play with. There is even a rather neat “Disc Exchange Cache” that allows you to swap discs without actually interrupting the saved recording, great for shooting stage shows on conferences. At the airshow I made good use of the picture cache to shoot aircraft engines being started. You never quite know when the pilot is going to press the start button. With a tape camera this meant pointing the camera at the plane and rolling lots of tape waiting for that moment when the engine starts to turn over. With cache record all you have to do is point the camera at the plane and as soon as you see the engine turning over press record, knowing that the previous 10 seconds (or more) are stored in the cache and will be recorded onto the XDCAM Professional Disc.
The Professional Discs come in 2 sizes, 23Gb and 50Gb. The 23Gb discs will record around 45 minutes of top quality 1920×1080 XDCAM HD 4:2:2 footage, currently they can be purchased for around £12 each which is less than HDCAM tapes, in fact they are not much more expensive than pro grade HDV or DVCAM tapes. They are incredibly robust and not affected by water and moisture in the same way that tapes are. You can re use them many thousands of times or simply keep them as you would with tape. However the format is still a file based format bringing with it a whole host of benefits such as faster than real time workflow, lossless copying, proxy files and metadata.
Anyway back to the shoot and the F800. One thing I wanted to play with was the S&Q mode. In this mode you can select the frame rate that the camera shoots at. I chose to shoot some crowd scenes, overcranking the camera and shooting at 60 frames per second. When you shoot above 30 frames per second the vertical resolution gets halved and I wanted to see how this looked. I am pleased to report that the overcranked material looks really good when played back. The motion is very smooth with no jerkyness and the drop in resolution is not all that noticeable. It’s certainly a reasonable price to pay for silky smooth slo-mo. One of the nice things about the F800 is that you can play back the slo-mo clips in camera, something that can’t be done with some of the competitions cameras! This means that if you were shooting for news and feeding a satellite truck you could playback your slo-mo straight from the camera without needing a computer to conform the material first.

Frame Grab from PDW-F800 with Zeiss 6-24mm DigiZoom

So what else can the PDW-F800 do that the 700 can’t? Well it has the ability to invert the image, this allows you to use 35mm lens adapters like the Pro35 much more easily and without needing picture degrading flip prisms. It can add markers such as safe area to HDSDi output No.2, this makes it easier to use an external monitor as a viewfinder. It has 2 filter wheels (the 700 only has one), one for colour correction and the other for ND filters. To be honest I like the single wheel system and then electronic control as on the 700, but having dual wheels does give a bit more flexibility and the option of adding extra diffusion filters or similar behind the lens.
Another one of the F800′s stand out features is the ability to superimpose a bar graph in the viewfinder that tells you how sharp the center of the image is. This acts as a focusing indicator and is very accurate. Simply adjust your focus until the graph peaks and you know you are as sharp as you can be. This worked really well when following fast moving aircraft as you could see the graph starting to drop off before you could actually see the slight loss of focus in the viewfinder. Very nice, wish I had that on my 700.
The PDW-F800 had the latest version of Sony’s XDCAM firmware, version 1.5. This has some nice new features including the ability to write the proxy files from the disc to a USB memory device such as a pen drive, memory stick or even hard drive. Another new feature is ALAC or automatic lens aberration correction. This does however need a lens with the appropriate digital hand shaking between it and the camera so I was unable to test it as none of the lenses we had would talk to the camera. However the Zeiss Digi Zoom didn’t really need it.
The pictures from the Digizoom were beautiful! Almost no CA and really rich. Its a seriously heavy lump of glass and not perhaps ideal for shooting an airshow, but it was a delight to use and the pictures are impressive. I’ve been doing a lot of research into lenses and lens choice recently. One of the things that I have discovered is that resolution is only a very small part of the whole HD lens equation. Of more importance than resolution is contrast, a lens that can provide very high contrast will show more detail than one that has less contrast. This Zeiss was one of the most contrasty lenses I have ever used, giving very rich deep blacks and solid highlights with very clean transitions from one to the other. Comparing the Zeiss with a Canon HJ11 the pictures with from the Zeiss just had a richness that wasn’t quite there with the Canon.

Frame Grab form PDW-F800 with Zeiss Lens

So would I buy a PDW-F800? Well probably not, but that’s just because I don’t normally shoot 24P and for occasional overcrank shots I can use my trusty EX1. But if you shoot drama or wildlife then the PDW-F800 would be a very good choice of camera. The picture quality is superb, the workflow is robust and reliable and the feature set of the camera is incredible.