Does the AF-100 Line skip?

[Mark David Williams]

Brian

Yes looking forward to Olof and others testing the camera.

Happy Xmas everyone!

[James Campbell]

Mark,
I'm also interested in making sure that the AF100 is all that Panasonic is saying it is before I purchase one also, but I wanted to clarify one of your statements:

Jan Crittenden Livingston clearly states above: "there is no line skipping being used."

You responded after an initial post: "Thanks, Jan. All questions answered. Sounds real good now!"

David Heath then provides some great input and possible alternatives that Panasonic may be using in lieu of line skipping, and your next statement is: "so it's clear now that it does line skip." In David's statement, he never ended with the perspective that the AF100 was line skipping. David actually states in a later comment in this post, "a camera can pixel skip, but not line skip." How did you come away with the position that the AF100 is definitely line skipping from his statement (on 12-22-10 @ 5:38 am)?I'd be interested to see if it can be clarified if the AF100 is pixel skipping instead.

[David Heath]

Originally Posted by Nigel Barker

No disrespect to David but I think to prefer the speculations of some contributor on an Internet forum over an assurance from a representative of the manufacturer is an odd choice.

I think quite a few people really need to CAREFULLY reread EXACTLY what has previously been written, and then they may see there is not necessarily any conflict between what I've put and Jans statement. In particular:

Originally Posted by David Heath

As far as "line skipping" itself goes, line skipping is a special case of pixel skipping - a camera can pixel skip, but not line skip.

So I'm not directly contradicting anything Jan wrote - she said it doesn't line skip, NOT that it doesn't pixel skip. They are NOT the same thing. Though the terms are sometimes (wrongly) used interchangeably.

The other statement she made was that the use of an OLPF means it doesn't suffer the aliasing problems that a DSLR does. There is no question that is anything other than absolutely true.

Maybe, to avoid any ambiguity whatsoever, the original question posed should be reworded to something like "can it be categorically stated that the AF101 reads out all 12.4 million photosites each frame, forms the image from all of them, then downresolves the result to form the final 1080 image"?

As it is, the confusion centres more around nomenclature than what is happening technically.

Brian - yes, tests will eventually independently confirm exactly what is going on - but the process shouldn't be underestimated. It needs proper test cards, not just evaluation of real world footage. As far as "real world" results, then again I repeat what was previously said:

Originally Posted by David Heath

There's an (incorrect) widely held belief that pixel skipping necessarily means "nasties" such as moire, aliasing etc. The belief therefore goes that if no moire/aliasing - therefore it proves no line (pixel) skipping. That's completely untrue.

I think some people need to write that out 100 times. Pixel skipping doesn't necessarily mean aliasing/moire. It's only in the absence of a suitable OLPF that the two really are linked.

[Brian Drysdale]

Yes, testing with charts would be a part of the process.

[Mark David Williams]

I apologise for my mixing up of line skipping with Pixel skipping.

So the real question is HOW does it do its magic OR another word for possibly something along the lines of does the camera pixel skip use blocks of four etc as clearly that statement and Jans statement regarding the use of the OLPF intertwined with her statement that it doesn't pixel skip leaves more questions.

I can see pixel skipping as being possibly equated with some form of compression.

One thing that seems clear to me is that the magic that is happening is removing something from the original image and replacing it with a new clever algorythmn of some kind. Something I think that a professional camera like the Sony EX1 doesn't do.

If I get any of the above wrong I am not an expert Just a seeker of the truth to enable me to decide whether or not to purchase the AF101.

[Brian Drysdale]

With single sensor cameras you need some algorithms, it just comes down to how invisible they are in the final image.

[Mark David Williams]

Originally Posted by Brian Drysdale

With single sensor cameras you need some algorithms, it just comes down to how invisible they are in the final image.

Yes I managed to work that out. It did initially confuse me as I thought the answer that there was no line skipping meant no other type of algorthmn either like pixel skipping because I took for granted and would have expected that to have been helpfully offered as part of an answer. It's what your not told and wording that you have to be careful of I can see that now.

[Robert Lane]

Mark,

You're overly concerned about something you'll never have to deal with in real-world usage. "Line skipping"? Trust me, that's something only measurebators concern themselves with when looking at a technical specs between camera models. You will never, ever see any aspects of line-skipping from the AF100. And how can I say this with such certainty?

Going back to the days far before AVCHD was even a speck of electrons in a engineers' brain, Panasonic has always been an engineering company, first and foremost. What that means is that when they research a new product they work the *science* behind the product first, then they work out the feature details and human interfacing. This is quite different from Sony or Canon who's methodology is always to address features and aesthetics first, then bring in the technology behind it. Which is exactly why both Sony and Canon products have always had a slightly more clean-aesthetic appearance where Panny products often look more... industrial in comparison.

When I was a direct consultant to Panny during the P2 HD Tour of '07 I got to see their methodology put into action firsthand whilst they were developing more P2 products (HPX500/2000). The science behind DVCPRO and it's HD variant were technically amazing when compared to the other codecs available at the time, not to mention the industry-leading technology built into the P2 cards themselves. (Note I said built "into", not the physical cards).

The AF100 is in point of fact lock-step with Panny's tradition of "science first" design; they've taken both CMOS and AVCHD technology a few steps beyond the competition and provided a visibly superior output. And at the price-point it's being offered there's not even a close competitor both in output and feature-set.

If you want to be concerned about any aspect of using the AF100 - or any other CMOS-based, AVCHD-type camera for that matter - which really does impact image quality/usability, it's rolling shutter. And unfortunately about 90% of all "video" cameras made today use a rolling shutter. That's the real headache today, not the sensor, not the codec and certainly not the price!

Happy Holidays to All!

[Mark David Williams]

Thanks Robert.

Taken on board what you have said!

Happy hols to you too.

Mark

[David Heath]

Originally Posted by Robert Lane

You're overly concerned about something you'll never have to deal with in real-world usage. "Line skipping"? Trust me, that's something only measurebators concern themselves with when looking at a technical specs between camera models. You will never, ever see any aspects of line-skipping from the AF100.

Well, in one respect, I have to fully agree with that. I posted a long time ago in this thread (and later restated) "There's an (incorrect) widely held belief that pixel skipping necessarily means "nasties" such as moire, aliasing etc." So no, I fully agree, pixel-skipping does not necessarily mean bad pictures - not as long as a good OLPF is used. (And I'm sure in the AF100 it is.)

But in another respect pixel-skipping can be of concern to all users. It's not "just for measurebators". The theory doesn't just predict the possibility of moire etc (if no OLPF is used), but also much lower than expected sensitivity. And to turn things around, it's the (relatively) poorer than expected measured sensitivity that is one of the main reasons for currently supposing that this camera DOES pixel shift!

You'd expect it to be about 2 stops better than something like an EX1 with this size of chip if some sort of pixel skipping wasn't happening. (Three stops better due to 8x chip area, then lose about a stop as one chip versus three.) In practice, the stories I'm hearing are about equivalent to an EX1.

I posed the open question before Christmas: "can it be categorically stated that the AF101 reads out all 12.4 million photosites each frame, forms the image from all of them, then downresolves the result to form the final 1080 image"? If Panasonic were to give a categoric "YES" to that question, it still leaves the question about sensitivity wide open. As it is, pixel skipping ties a lot of (observed) technical points up very neatly.

Does it matter though? The EX1 is seen as pretty good in low light anyway, isn't it?

If the AF100 is used with fast primes, that argument holds pretty true most of the time. But use solely with primes can be very limiting, and it's using a camera like this with zooms when the sensitivity issue starts to matter. If the sensitivity was 2 stops up on the EX, it becomes possible to use apertures of about f3.5 and maintain usage parity in low light with an EX. And f3.5 zooms for such large format sensors are likely to be reasonably affordable.

But use an f3.5 lens on an AF101, and it will be trounced in low light by the f1.8 of the EX1.

The AF100 is in point of fact lock-step with Panny's tradition of "science first" design; they've taken both CMOS and AVCHD technology a few steps beyond the competition and provided a visibly superior output. And at the price-point it's being offered there's not even a close competitor both in output and feature-set.

I'm sure that's exactly the Panasonic party line, but in truth, if their policy really was "science first", they'd have developed a purpose designed large sensor of more like 4 megapixel, not adapted what is first and foremost a still imager. I fully accept their approach has led to a product which is far cheaper than it otherwise would have been, and (by virtue of the OLPF) should give moire/alias free video. But it shouldn't be seen as equal to a "designed for" sensor.

At the moment, you're probably right that it is in a space of it's own. The F3 has many features the same, is unquestionably better in other respects - but is a lot more expensive.

But surely it's the future, and the NXCAM little brother of the F3 that Panasonic must be really worrying about? Price reckoned to be about the same as the AF100, same native codec, many other features similar - but having the same purpose designed sensor as the F3?

If I wanted a large sensor dedicated video camera NOW, I'd get the F3 if I could afford it, and the AF100 if I couldn't. But if there was no pressing need, surely it must be better to at least wait until NAB and see exactly what the F3 little brother will consist of? Latest rumours are of far better ergonomics than the AF100 (or the F3, for that matter!), let alone the far better sensor.

[James Campbell]

David,
Can the concept of pixel skipping be viewed similar to how AVCHD compression was viewed a few years back? I recall reading posts across many forums complaining about how there was no future in using AVCHD... now its use is widespread. Would it be accurate to say that many next generation cameras -- perhaps just prosumer at first -- will likely be using some means of assimilating MOST but not all of the pixels that the camera sensor is reading? Could there be better implementations of pixel skipping just as there are now better implementations of AVCHD compression than there were a year or two ago?

It seems like many people are currently wondering if the AF100 is something to jump at or does one of the major camera companies have what you're referring to just around the corner. It certainly always comes back to needing to make the decision to "pull the trigger" at some point and just purchase the best technology that is available at the time. I just off-loaded all my HDV equipment, and my plan is now to use the GH2 I just purchased to work as a pair with the AF100. But if the AF100 is using a stopgap measure that will be addressed in 6 months by a different camera, I'll wait. It'll be interesting to see if the AF100 (or the Sony you're referring to, or perhaps the next Canon??) has the answer to what you're asking:

"Can it be categorically stated that the AF101 reads out all 12.4 million photosites each frame, forms the image from all of them, then downresolves the result to form the final 1080 image"?

I guess a follow-up question is the one you also ask above: Does it matter though... if you have a fast lens and if the video you're capturing is great footage? Would the issue be that there would be minimal professional use for the footage due to the limitations you're mentioning?

[David Heath]

James - what I think is important to realise is that pixel-skipping has got a bad name because it's associated with the moire/aliasing in DSLRs, not because it's inherently a bad thing. But that's due to their lack of an OLPF - not a problem in a made-for-video camera such as the AF100.

In picture quality terms it's something of an irrelevance - as long as the right OLPF is used. That's why nobody can look at a camera output, see no moire/aliasing etc, and draw the conclusion of no pixel skipping.

The *GOOD* thing about pixel skipping is that it enables a made-for-still sensor to be fairly easily adapted for video use at quite lost cost and low power consumption/heat generation. The *BAD* thing is regarding sensitivity issues - which can manifest as either/or signal-noise or ISO rating.

In terms of the future, I'd expect the way forward to be more in terms of made-for-video sensors with an appropriate number of photosites for the output resolution (more like 3-4 megapixel than 12), rather than improving the way what are fundamentally still camera sensors are adapted.

[David Heath]

Originally Posted by David Heath

I posed the open question before Christmas: "can it be categorically stated that the AF101 reads out all 12.4 million photosites each frame, forms the image from all of them, then downresolves the result to form the final 1080 image"? If Panasonic were to give a categoric "YES" to that question, it still leaves the question about sensitivity wide open. As it is, pixel skipping ties a lot of (observed) technical points up very neatly.

A month on, I think it's now possible to be a lot more certain about what the AF100 does and does not do in a technical sense.

Firstly, I believe it is correct to say it is NOT line skipping. I now believe that what it IS doing is "pixel binning", as opposed to "pixel skipping", That should give better sensitivity results, but worse resolution. That means it is reading the entire sensor, but not forming a high res image and downscaling - "pixel binning" processing is far simpler. (That's why the cost, power consumption, etc are what they are.)

It also seems to not have 12.4 million photosites making up the active frame, but rather 13,932,800 in 16:9 framing, so nearer 14 megapixel. That works out to be 4976 x 2800, and this figure can be easily obtained from the GH2 specs. (The 12.4 million would be the figure if a GH1 chip was used.)

Put it on a test chart, and the measured resolution comes in far less than I originally expected, roughly equivalent to a 720 sensor. This is in line with various posted images which seem to show it looking somewhat soft compared to cameras with 3 1920x1080 sensors. Looking at the null points of the aliasing, it seems to give way to aliasing at 1244 pixels (622 line pairs) horizontally, and 700 pixels (350 lp) vertically. The chart is symmetrical - so no line skipping.

The eagle eyed will have spotted that that 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.

Hence, the image seems effectively formed from 1244x700 "super-pixels" each with an R,G,B value, and each the averaged result of 16 sensor photosites. It's therefore equivalent to a 3 chip design with 3 chips each of 1244x720 or about 0.87 megapixel each.

[Jan Crittenden Livingston]

Gentlemen,

The camera does not do pixel binning or line skipping. It uses an optical low pass filter to resolve the moire and aliasing.

Thanks,

Jan

[Simon Wyndham]

Thanks Jan, however this does not explain many of the results being obtained in tests. For example, if the camera is not pixel binning, this doesn't explain why the camera is only really producing around 650-700 TVL of good resolution, rather than 900-1000 odd TVL that one would expect from a true 1080p camera.

Would it be possible for you to tell us or find out from the engineers how the data is being read out and interpreted?