AF101 BBC/EBU test results posted

Status
Not open for further replies.
I can also say that there is a substantial reduction in chroma aliasing when using the Nano at 4:2:2. I wonder if Alan's report used the onboard codec to record (which is how I would expect him to do it, by the way). Because on the shot of the red and green zone plate, if that was shot external to 4:2:2 I can predict that it would look much, much better. The sampling to 4:2:0 creates aliasing rings that simply aren't there on the live monitor output, and therefore may be confusing evaluations of how the sensor is performing.
 
Barry_Green said:
I wonder if Alan's report used the onboard codec to record (which is how I would expect him to do it, by the way).
Click to expand...

Barry, Alan always ignores the onboard recording and measures straight off the HD-SDI output, so I don't think that's an explanation.

Just to be clear, this is an independent report comissioned by the BBC specifically focusing on measuring the sensor & DSP performance in a repeatable, quantitative way.
Those familiar with Alan's excellent reports will know that they are always done the same way, always report the facts - warts and all and avoid the temptation to draw any conclusions regarding suitability of particular cameras to particular jobs - that bit is left up to the reader.
Alan may well have his own opinons, but his reports are always objective and fact-based.

If you're looking for a recommendation on what camera to buy, this is a useful input to your decision, but it won't answer the question for you.

Regards

Dave
 
Barry_Green said:
Well, the sensor isn't diagonal, you can look through the lens port and see that it's square. But maybe the pixels are arranged diagonally on the sensor?
Click to expand...
Yes, that's what I was suggesting, a 16:9 rectangular image sensor with a diagonal pattern of RGBG pixels. Seems to fit Alan Robert's test patterns, but aside from that, I have no idea what advantage that might have.
 
To put things in perspective the 5D didn't even make it past the first test (as in, he didn't even bother testing the camera further) and the Red (old sensor) was pretty much deemed about as good as a typical large body hd camera.

These tests are purely technical in nature, measuring performance on charts. Not to say they aren't useful and interesting to read. Although his statement about DOF compared to an 1/2 sensor appears to be way off, as noted by Barry. That should be painfully obvious to anyone who has used an ex camera and af100 / gh2
 
Last edited:
A simple question but in a lot of cases could it be that the AVCHD codec filters some of the problems out, Alan tests all his cameras direct and not taking into account the recording format.

I think panasonic have made a good choice of using AVCHD for this camera as it is an idea codec for prosumer use, I appreciate that people will also be using nanoflash etc but as said it is a good camera for what it is designed to do which is to offer an alternative to the DSLR shallow DOF shooting but without some of the video nasties associated with DSLR's.
 
I too noticed the softness in my first shoot with it a few weeks ago. I would guess that any sharpness issues could be resolved with firmware updates, given that it uses a multicore processor. Perhaps Panny found that compromises had to be made in what was fed to the codec in order to achieve the best result, perhaps a different, higher-res scaling algorithm could be used when going out through sdi with the increased bandwidth?
 
A lot of that report was extreme pixel peeping, again keep in mind that the same guy said the 5d was unusable, so even a modest report from this guy is actually fairly good when looked at relatively. i do also feel that the downsampling is something that panny can improve on hopefully with firmware updates. we can only hope that they are chatting with the gh2 tech.
 
Yes, I know resolution is not everything.... But when I buy a "HD" camera I expect to be able to use it for shooting high detail wide shots, e.g. landscapes, meadows etc. I maintain therefore my former conclusions voiced in other threads and based on tests with many lenses and comparisons with the GH13: this cam cannot be used for detailed wide shots. The GH13 resolves at a fraction of the price significantly more detail.

Alan Robert's test confirms what all previous tests (e.g. the slashcam.de test) and common sense has already shown: the AF100 does not resolve cleanly 800 lines as advertised by Panasonic. This is the kind of painful truth many AF100 owners like me have to live with. After the initial thrill, a significant discontent among AF100 owners is growing and cannot be dismissed: the camera we have got is "barely" a HD cam.

Panasonic must now come out of the box and comment publicly these findings and tell us what to do: return the camera and pay a full refund or upgrade the cameras to its published specs without additional cost for buyers. What has been advertised is not what has been sold.

Opinions and further experiences are welcome.
 
I can't wait till 4k becomes prevalent. It will make all this nitpicking over 100-200 lines of vertical resolution seem so pointless.
 
danaps point wasn't quite about 4K, it was about a camera resolving the resolution that it captures at.

and i stand firm that anyone that says they need 4k is 90% of the time blowing useless smoke. i hate hypes and buzzwords, makes producers even more intolerable ;)
 
To make my point clear: I only argue that this camera has been shipped with a significantly lower resolution than advertized. The difference between 800 lines (low compared with the EX1) and 610 clean lines is very significant for wide shots. I am also arguing that what I have got is not a "HD" camera. I am not talking about other cams and 4K.
 
I am not surprised. I would be surprised if the BBC would have approved this camera. I've used it for a couple of pro shoots and tried to remain positive but I can't. The electronic noise is horrible and the overall image quality doesn't compare to GH1 and GH2. I dealt with the noise by changing the settings, but after that everybody in my team agreed that image of GH1 is much nicer. Also playback of edited/finished AF101 footage stutters in Quicktime and no that's not a Quicktime problem. And the clearly visible noise in VLC player isn't a VLC player problem either. All the problems are caused by the AF101. My clients do not accept the playback problems caused by the AF101 and they're not interested in explanations or excuses or attempts for cover ups. I could deal with the poor image by using color correction and try to match to the GH2 but the stuttery playback remains forever.

Thanks BBC for confirming all this. Nice to know there are still organizations with proper standards.

My AF101 is for sale, I am now using it as a B-cam and setnoise recorder (it's great for that purpose!), I'm buying a second GH2 and will be happy. For typical run and shoot camcorder work I bought a HMC151. I'll forgive you Panasonic about the AF101 failure and I'll also forgive those at Panasonic who claimed that those who saw noise and other problems "are just stupid". Life goes on, I'll take it as an other lesson.
 
Last edited:
imdjay said:
danaps point wasn't quite about 4K, it was about a camera resolving the resolution that it captures at.
;)
Click to expand...

I wasn't addressing dnaps point.


and i stand firm that anyone that says they need 4k is 90% of the time blowing useless smoke. i hate hypes and buzzwords, makes producers even more intolerable
Click to expand...

Well dobn't threads like this prove that people want and need 4k. There isnt an hd prosumer/consumer that can resolve full hd detail to date. 4k will end that complaint once and for all.
 
sorry astigmatic but you're still missing the point that he was making. even if you get your walmart 4K camera it probably won't resolve 4K, just like the 1080 af100 isn't resolving 1080. so no, 4K will not end anything. And no, threads are not proof that 4K is a "need", that's subjective. There is plenty of data on the side of the argument that 4k isn't needed for the majority, but i have yet to see data which is on your side of the opinion.

In either case moore's law will eventually make 4k for 5K(dollars) at some point in the future, i just don't give a damn, it's a silly pedestal to be pining over
 
Reading that report, it states that the lens adapter was loose and focus was affected by touching the lens, perhaps I'm stating the obvious but I'd like to think the optics were rock steady before doing any sort of performance test
 
Status
Not open for further replies.
Back
Top