Alan Roberts BBC report on the AF101

[David Heath]

Originally Posted by John Mercer

From what I can gather from his test results for the AF101 there is an issue with the way the camera downconverts its megapixel cmos to 1920 x 1080 that is less than perfect. It doesn't pixel skip as was argued here but its OLPF is either less than perfect or non-existent .......

I said six weeks ago it DID NOT pixel skip, but rather use a binning technique, and all of Alans results match that latter basic theory very well.

At the time, the explanation suggested by me was:

.....resolution is exactly one quarter in each direction of the sensor pixel count. The most likely deduction therefore must be that it's pixel binning on the basis of 4x4 blocks of photosites. Hence 8 green, 4 red and 4 blue photosites in such blocks are having their charges collected (binned) into 3 "bins", one each for R,G,B, before these 3 values are digitised and processed.

Because of a scaling error, I now realise the exact values I suggested were too high (1244x700, and should only have been 1175x661), but the basic theory stands up extremely well. That's in respect of resolution (1/4 each horizontal and vertical of total sensor), symmetry (equality hor and vert), lack of coloured aliasing and relative colour/luminance resolution.

In the report, Alan says (p9):

......there is clearly very little difference between the two colours (blue should always be the same as red anyway), the levels of aliases are identical. This implies that the green signal has been derived from the sensor, at the same resolution as that of red and blue.....

That's exactly what 4x4 pixel binning would give. Think of the 4x4 block as a "super-pixel", with photosites as:

G R G R
B G B G
G R G R
B G B G

R, G, B binning such a block will give equal red, green, blue resolutions, each at a resolution one quarter that of the total no of photosites in each direction, and no differential colour aliasing. Exactly what the test results show.

It's the absence of colour aliasing that Alan uses to discount pixel-skipping - ("It is very obvious that the scaling has not ignored (skipped) sensor pixels, since that would have invoked coloured aliasing, both horizontally and vertically.")

Alans argument ENHANCES the theory in favour of pixel-BINNING, whilst showing there is no pixel-SKIPPING. But I'm open to alternative theories?

[Glen Vandermolen]

Originally Posted by John Mercer

I thought a key element of the XF 305 was the combination with the built in lens, so it was an overall package that created a camera that the BBC are happy with.

That's what I meant Brian by 'surprisingly good quality'. When used by DV Directors and APs, or when used as 2nd multiple cameras, say on something like Top Gear, or less accessible current affairs and news, it produces footage that is considered true HD (unlike the Z1 which is still widely used). But where you shoot drama or studio, or high end documentary then the main camera is and will most of the time be a 2/3" broadcast camera. Something like the XF305 (or any other fixed lens camera) can never be a realistic substitute for these.

Oh, really?

Ever seen the opening segment of the 2010 Emmy Awards TV show? The Emmys was shown in prime time on a major network, watched by over 13 million viewers. The producers could have picked any camera for the opening comedy skit, and they chose - the XF305.

It's been cleared for FULL acquisition for the BBC. The BBC bought 50 XF305s for themselves. Talk about putting your money where your mouth is.

The AF101 was put through the same tests that all the other cameras go through. No biases, no favoritism. It did what it did.

[John Mercer]

It's been cleared for FULL acquisition for the BBC. The BBC bought 50 XF305s for themselves. Talk about putting your money where your mouth is.

For the very uses I have outlined (do you know how many cameras they have in total at the BBC?). Look believe me and I don't mean to burst your bubble Glen (politely I mean) when I say to you no serious cameraman that I know is going to dump his 2/3" camera and only use an XF305 from now on, he'll lose a lot of broadcast work. It does not matter where it was used they will not replace studio and major programme use cameras for the BBC or ITV or SKY in the UK (I can't speak for the US). If you don't believe me take it up with the BBC and ask them what they mean by FULL(or Alan) not me that's not my purpose here, only to comment on the reasons behind approval testing from my own understanding of working in UK broadcast. E.g last month an Independent company producing a well known property series for the BBC asked me here in Spain to shoot with HDCAM for main and yet they had brought 2 XF305s, presumably they or I could of used those, but they were insistent.

Like I said talented professionals will use all these cameras at one time maybe and we'll see, I'm sure some great things. I don't really want to get into any arguments here. :)

[Glen Vandermolen]

My bubble isn't burst. As a matter of fact, I did sell my 2/3" cam for the XF305. And I am a professional cameraman (kinda serious). And I still get work. You're gonna have to trust me on that. :-)

Sure, I would rather have the new HPX3100. But for the money, the XF305 is a great little cam.
I did look at the AF100 before I bought the XF. I really liked the AF, it had a lot of nice features, and fitted with a PL mount, it was sweet! However, for my needs, the XF fit better. I don't know why some of you are dissin' on the XF305, but it is a really nice cam.

[Brian Drysdale]

I think there are going to be a few cameras in this price bracket later in the year, some of which will meet full broadcast HD spec, while others won't. The choice will range from 1/3", 1/2", 2/3" to S35, with standard video type or RAW workflows depending on the manufacturer. The right camera being the one that works best on the type of productions you're shooting.

That's quite a range of choices and not everyone actaully needs a camera that meets the broadcasters' full HD requirements.

[John Mercer]

Originally Posted by Glen Vandermolen

And I still get work. You're gonna have to trust me on that...

... I don't know why some of you are dissin' on the XF305, but it is a really nice cam.

Glen why should I not trust you, I don't know you :) I am only talking of the situation as I've experienced it in the UK market up to now, as Brian says it is changing. But I still hire out my DSR 570 (here in Spain for broadcast and Sky news occasionally), a HPX 500 (I know, I know the XF305 has a better picture but..so does my HD101 but they never believe it) regularly am asked to get Sony/Panny 2/3" HD for other jobs. I know I could hire out a XF305 too and I am seriously thinking of getting one for the quality and the Handycam form factor, it is top of the list for me. But I would lose a lot of work, at the moment, if I just had an XF305 here because of many other factors than pure picture quality, that is all I am talking about, obviously YMMV and I respect that.

I am not dissing the XF305 for sure. It is the best camera in that form factor bar none and the picture/lens/codec quality seems to be excellent even compared to much more expensive cameras, though I have to be honest I have not personally used one. Let me say categorically as well infact I truly hope that the XF305 becomes widely accepted because I'm finding it finacially painful to keep up. But as you say you would rather have the HPX3100 and I am sure you would concede it is a much better camera, recognized as such by major broadcasters. I don't write any rules either I just have to follow them sometimes.

[David Heath]

Originally Posted by Brian Drysdale

..........not everyone actaully needs a camera that meets the broadcasters' full HD requirements.

Indeed no. But given the choice of a fully approved broadcast camera, and one which doesn't meet minimum HD standards, then if roughly the same price, and assuming similar basic features and facilities, then there's nothing to lose by going for the approved version, but potentially a lot to gain.

And the real point is that the 101 is up against other cameras of comparable price, which do meet the spec. (Albeit if needing a nanoFlash in some cases.)

Even if you don't actually need the approval, it's still a good marketing point - "I use equipment which is fully approved to latest broadcast spec".

[Brian Drysdale]

Although, for some people it could come down to a shallow DOF driving the purchase, rather than other factors.

[Steve Rosen]

I've had an AF100 on order since December, but have been putting off delivery. Every day there is another nail in the AF coffin.

It seems that Mr Roberts was truly disappointed with the outcome of his tests.. I think, from the tone of his comments, that he really wanted it to perform better than it did.

In my case, I do make documentaries intended for broadcast. My trusty HPX500 is now considered obsolete (although I still like the picture personally).

As an independent documentary filmmaker, I refuse to spend another 20+ grand on a 3100. So, I've invested in having Leica-R lenses Cine-Moded, bought Cineroid finder, and even acquired a Zuiko lens - all in anticipation of getting the AF100 for my next project, which starts in April. If these tests were better, I'd even consider a Nano.

But, all in all, the Canon looks like a better choice. I'd be happy if someone could convince me otherwise.

[David Parks]

I'm pretty sure Panasonic has been very clear about where this camera falls in the marketplace. It has better ergonomics and fewer issues for moire and such than DSLRs shooting HD.If you don;t like shooting with a DSLr then the AF100 is a very good option for you. Personally I think it is a little stupid for everyone to consider the BBC to be the judge and jury on what camera one should use for documentary production since very few of us will actually have something air on the BBC. If you are in fact shooting for the BBC
then shoot 2/3 broadcast.

Okay. Have fun debating while I go to work tomorrow and shoot more footage on our AF 100. And don't forget to clean off the smudge marks from your computer screens from your noses being too close. :)

[Brian Drysdale]

Perhaps it's best to wait for the BBC to decide, which probably be a marker for Nat Geo, Discovery Gold HD etc. If your market isn't one of those high end HD broadcasters, the AF100 could be a possible camera for you.

The EX series into Nanoflash and the Canon are non 2/3" options if you do.

[Simon Wyndham]

I'm pretty sure Panasonic has been very clear about where this camera falls in the marketplace.

I don't agree. They are marketing it as a 1080p camera, when it clearly isn't capable of even doing 720p well. Say what you will about the BBC tests, but they do show, in an official measured form, the true performance of the camera. Some here were not that happy at the chart I posted previously. Jan claimed on that thread that the results shown on that chart were nothing like the ones that she and Barry Green had measured. She claimed that the camera showed 800 lines on her chart. We still haven't seen this chart, and now we have this BBC test that confirms the low resolution of the camera.

What I'd like to see is some transparency. If Panasonic were open about what the camera is doing instead of going about it with cloak and daggers there wouldn't be a problem. But they do, and they try to sell it as a 1080p camera that solves all the issues of aliasing etc that the DSLRs suffer from, when in reality it does absolutely nothing of the sort.

[Andrew Stone]

Simon without knowing the circumstances around your findings, I am willing to give Jan Crittenden the benefit of the doubt until she responds.

I think you have to put this in the context of the whole development of the AF-100/101. Jan had put herself out there unlike any product manager to date in the development of the camera. The user input on the camera was unparalleled. She stated the criteria for the camera was to give the best camera they could for the target price which turned out to be 6 grand USD. They responded with changes to the camera from the user input on forums like this and others, the camera was released to great fanfare and the momentum for the camera took off and the momentum was largely user/forum based. The expectation for the camera was huge. Everyone thought they were getting away with the farm. Now people are finding out the camera is now not like the $20,000 that they had hoped but a $6000 camera.

I feel a lot of sympathy for the AF owners right now. It is like they have been kicked in the gut and it's not pretty but it has to be put in perspective. Granted if Jan has made a claim about a chart on a public forum she should in the light of this report put up the chart and quell the anger.

She's banked a lot of karma, I would give her some time. There is probably a lot of corporate fallout from this and she is answerable to more than just herself.

[Simon Wyndham]

Simon without knowing the circumstances around your findings, I am willing to give Jan Crittenden the benefit of the doubt until she responds.

Andrew, I think you are missing the point that my chart matched almost exactly with that of the BBC research document. Where mine came from and the circumstances are now an irrelevant question as you can see the BBC conclusions.

Now people are finding out the camera is now not like the $20,000 that they had hoped but a $6000 camera.

Which could buy a GH2 with a load of accessories and seriously good glass for the same performance. Okay, not the ergonomics, but for that price difference are ergonomics alone worth that amount of money?

There is probably a lot of corporate fallout from this and she is answerable to more than just herself.

I agree. But this seems to be a continuing theme. I only ask for one simple thing. For Panasonic to open up and stop being so coy about their camera specs and performance. As you can see from the people who contribute to these forums, we aren't stupid, and can quickly find out if there is an issue, or of the camera under performs compared to how it is marketed. The solution is to square with us in the first place.

[David Heath]

I commented on this at the time - as far back as early Feb. ( http://www.dvinfo.net/forum/panasoni...ml#post1614658 - also post #58 in that thread)

I now feel even more certain of what happened in the case of a lot of people. Cameras were pointed at charts with only horizontal and vertical lines, a response was seen at the 800 line blocks and the conclusion drawn "oh good - it's resolving 800 lines".

A reasonable assumption to make - but unfortunately completely wrong, as we now know. What they were seeing was aliasing. That's why I went on to say that these sort of assessments need to be done with a zone plate (as Alan used) or at very least a circular res chart with bands (as Simons examples). With such, it's impossible to mistake real detail and aliases - the alias circles have a false centre.

Unfortunately, the high levels of aliasing in this camera made the mistake even more difficult to spot. I suspect Panasonic may be putting in an order for zone plates to replace their old charts to prevent a repeat embarrassment!