Panasonic AG-HPX250 and AG-AC160

30 replies [Last post]
Richard Payne
Offline
Joined: Sep 15 2000

I've got a Panasonic AG-HPX250 and an AG-AC160 with me at the moment, so if anyone has any questions on these cameras now would be a good time to ask.
I'm off to IBC early tomorrow morning (Friday) so I might not be able to reply until I'm back on Monday 12th.

noddydog
Offline
Joined: Feb 28 2004

Hi Richard. I'm considering the 250 up against Canon's XF300.. now this is a complete shot in the dark but have you used the XF300 and if so what are your initial impressions of differences between the two?

Other than that have you had time to play with the AG HPX250 to get an idea of how it performs against other contenders such as the EX1?

Richard Payne
Offline
Joined: Sep 15 2000

Don't forget I work for the Panasonic Distributor so I am not an independent!

I've never used a 305, but on paper the 250 has a higher quality codec in AVC Intra 100 and a larger zoom range. It also has Flash band compensation which is a biggy in my opinion. I've always really rated the filmic, naturalistic look of the Panasonic Cameras over the hyper-real and more saturated Canon look, but that is objective.

I believe the 250 will be lighter for hand held use. I'm still waiting for a P2 card to go and shoot some proper footage, might wait for the rain to stop too!

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

You mean that's subjective :)

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Richard Payne
Offline
Joined: Sep 15 2000

;) I know what I mean

infocus
Offline
Joined: Jul 18 2003
Richard Payne wrote:
I've never used a 305, but on paper the 250 has a higher quality codec in AVC Intra 100.......

All the reports I've seen indicate the two codecs to be "comparable", with the EBU going no further than approving both of them for unrestricted broadcast acquisition, and calling them both "quasi-transparent".

On paper, I'd expect AVC-Intra 100 to be better in some respects, XDCAM422 50Mbs better in others. What makes you think otherwise?

Richard Payne wrote:
It also has Flash band compensation which is a biggy in my opinion.

There are two ways of doing flash-band correction - hardware (as with the 250), and software (as Sony do). Both systems work well most of the time - BUT - neither is infallible, and occasionally give a false correction.

With the software approach you can choose to accept the correction or not, so the problem goes away. With the in-camera approach, tough, it gets burnt in at the time of shooting. For that reason, I'd say the software approach is superior.

rogs
Offline
Joined: Jun 16 2006

Now that 'flash banding' is more or less controllable, have Panasonic taken any steps to help with the other rolling shutter issues, now that they are using CMOS sensors in all these new models?

I always thought that their video for the HPX300, where they suggested that the best way to deal with skew was, basically, 'don't pan', seemed a bit lame.

Hopefully, the sensor scans speeds are much faster on these newer models? Or have they included some new 'tricks' to help with rolling shutter issues?

tom hardwick
Offline
Joined: Apr 8 1999

Hopefully it's faster scan speeds because CMOS cameras make prop aircraft and helicopters look just silly. And while we're here, has anyone found the Edius 6.02 (is it?) flash banding filter helps any?

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

There are new tricks, but they're expensive to implement. Don't expect to see major revelatory changes in the near future.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

rogs
Offline
Joined: Jun 16 2006

The message from the last two posts rather suggests that, if you are into aircraft photography -or perhaps any other scenario where high vibration is likely to be an issue? - then grab an HMC 151 while you still can!

Soon (apart from the very top end) it's going to be the option of chosing CMOS or CMOS sensors......

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

That's about right.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Ron Spicer
Offline
Joined: Jul 22 2001
rogs wrote:
The message from the last two posts rather suggests that, if you are into aircraft photography -or perhaps any other scenario where high vibration is likely to be an issue? - then grab an HMC 151 while you still can!

Soon (apart from the very top end) it's going to be the option of chosing CMOS or CMOS sensors......

What's the weight of the HMC 151, Rogs/

rogs
Offline
Joined: Jun 16 2006

I don't own one myself Ron, (sadly!) but I recall the spec. sheet listing the camera as 1.7Kg without battery, and about 2Kg with........so not heavy....

Richard Payne
Offline
Joined: Sep 15 2000

Hi David
"All the reports I've seen indicate the two codecs to be "comparable", with the EBU going no further than approving both of them for unrestricted broadcast acquisition, and calling them both "quasi-transparent".

On paper, I'd expect AVC-Intra 100 to be better in some respects, XDCAM422 50Mbs better in others. What makes you think otherwise?"

AVCINTRA at 100MBS might be very close at the first generation, although I too would expect a more modern variable block size iframe compression would certainly look better in certain situations, but as you progress through the production chain the iframe nature of the codec would start to reveal more and more advantages over XDCAM at 50MBS.
This is based on a conversation with a Panasonic Codec engineer at IBC.

infocus
Offline
Joined: Jul 18 2003
Richard Payne wrote:
..........as you progress through the production chain the iframe nature of the codec would start to reveal more and more advantages over XDCAM at 50MBS.
This is based on a conversation with a Panasonic Codec engineer at IBC.

There is a massive flaw in that engineers argument as it stands, which is that it assumes any degradation in post is due to de/recompressing the GOP structure alone. In practice, most such work will also involve de/recompression of the I frames themselves. Hence, relative codec performance will depend on type of post work and the nature of the material. It's just as likely that AVC-Intra will degrade faster than XDCAM422, dependent on what's being done.

It's the same principle as opening a still image in Photoshop, doing work and recompressing - do it repeatedly and there will inevitably be degradation. And a still image is the ultimate "I-frame only" system..... :)

Richard Payne
Offline
Joined: Sep 15 2000

Is the compression used in JPEG really the 'Ultimate' compression David? Surely things have moved on a bit since then? Just as they have moved on from the old MPEG2 based codecs to the shiny new MPEG4. :-)

infocus
Offline
Joined: Jul 18 2003
Richard Payne wrote:
Is the compression used in JPEG really the 'Ultimate' compression

No, it's not, and I've never said it was. JPEG, MPEG2 and AVC-Intra are all based on DCT, so no, I don't think MPEG2 or AVC-Intra are the "ultimate" by any means. Wavelet has been talked about a lot for the future.... we'll see.

Quote:
Surely things have moved on a bit since then? Just as they have moved on from the old MPEG2 based codecs to the shiny new MPEG4. :-)

Well, applying that logic further, then I'm sure you'll agree that things have moved on and P2 (based on the old Cardbus technology) has now been left behind by SxS (based on the shiny new ExpressCard technology)? :)

At heart, DCT is at the heart of all these codecs. New "tricks" have been developed to get a better quality at the same bitrate (or comparable quality for lower bitrate), that's quite true, but they may be better seen as more like mathematical formulae than technology. They don't go out of date in the same way - just as Pythagorus Theorem is just as valid (and useful) today as it was when first formulated. Equally, the newest codecs still need to use many of the old "tricks" - they just use newer ones AS WELL. The down side of using all these "tricks" is that they need more computer power to cope.

Regarding the XDCAM422/AVC-Intra comparison, then it's true that AVC-Intra uses "tricks" that MPEG2 doesn't - but inter-frame compression is one of the most effective "tricks" yet thought of, and that's used by XDCAM but not AVC-Intra. That's why a direct comparison of the two codecs is so difficult - on some types of footage one "trick" will work incredibly well, on a different scene far less so. That's before we even start to consider not just different types of footage, but different post processes.........

And that's why all the EBU have said after very extensive trials is that BOTH are highly acceptable codecs. I've not seen any credible reports that give EITHER of them an advantage OVERALL, and that's borne out by the paper specs. Please don't think I'm knocking AVC-Intra - it's good, very good - but there just is no evidence to say it's better than XDCAM422 overall.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Don't confuse the message with the medium. P2 and SxS are card formats, not compression algorithms.

In general, an I-frame compressor needs about double the bitrate of an inter-frame compressor for the same level of visible artefacts. AVC-I has the advantage over MPEG2-I (or JPEG or DV or Digibeta) in that the block size is not fixed, so compression can be better distributed. The next result is typically 80% of the bitrate for the same level of degradation.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003
Alan Roberts wrote:
Don't confuse the message with the medium. P2 and SxS are card formats, not compression algorithms.

Errr, no confusion - the earlier comment was meant tongue-in-cheek (hence the :) ). If "newest is best" should be the mantra, then surely it must mean that the newer SxS is better than P2?

Alan Roberts wrote:
In general, an I-frame compressor needs about double the bitrate of an inter-frame compressor for the same level of visible artefacts. AVC-I has the advantage over MPEG2-I (or JPEG or DV or Digibeta) in that the block size is not fixed, so compression can be better distributed. The next result is typically 80% of the bitrate for the same level of degradation.

I don't disagree with any of that, but the more you go into it, the more and more complex it becomes. One very big difference between AVC-Intra and XDCAM422 that has not so far been mentioned is that AVC-Intra is 10bit - XDCAM422 is 8bit. At first sight that would seem to be a plus towards AVC-Intra - in practice, few but the top end cameras are likely to have noise figures good enough for it to make any practical difference. (Certainly not such as the XF300, EX1/3, and the HPX250, HPX371 etc.)

For the purpose of this discussion, all it means is that 20% of the bitrate of AVC-Intra effectively gets used coding noise, and the most significant 8 bits of the video get coded with 80Mbs - not 100. This is compared to 50Mbs for XDCAM422.

Now using Alans figures, the interframe nature of XDCAM422 means this translates to (very roughly) about 100Mbs of an equivalent I-frame only codec, the variable block size nature of AVC-Intra means the 80Mbs translates to 100Mbs equivalent for a similar (fixed block size) codec. In other words - on paper, there's an equivalence.

Before anyone else says it, these figures really need to be treated with a pinch of salt since the actual performance will vary with all sorts of factors. But they do give some idea why on paper neither codec is really likely to have an overall quality advantage over the other, and to even see any difference at all is likely to mean quite a lot of pixel peeping.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

I was deliberately trying to avoid it getting too complicated, because we could go on for about a year discussing the differences between coders. So we stick with the generics. If you really want to know the full details, then you need to be in a research lab, reporting to an EU committee.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

MAGLINK
Offline
Joined: Mar 8 2007

Well said Alan and as said before both codecs are acceptable for broadcast so any further discussion just goes around the same circles again.

I do find though that some of the RED and Alexa graded footage on screen looks very flat and lifeless to me so at the end of the day it is down to what you do with it rather than the pure pixel spotting aspects!

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Quite right. Footage shot with the large-format film-style cameras is generally shot look like it came from film negative, so it looks flat from the camera. It's up to the colourist to impose personality on it, and with such footage it's easy to do, but some seem to want to leave it looking flat. The best extreme example I've seen of that is the French series 'Engrenages' ('Spiral' in English) on BBC 4 now (1st series being repeated).

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

rone01
Offline
Joined: Feb 9 2009

interesting you should mention this Gary - I'm finding now I really like the lack of Punch that a flat Red One gives. It's less immediate than a high contrast image but you gain so much more in subtlety.

You can add punch of course - but I usually don't add so much as built into processed images from our other Panasonic cameras.

Oh and AVC-I makes a great codec for storage and distribution of locked R3D footage! I'd imagine the 10bits of being put to full use there.

Red EPIC / Scarlet - Panasonic AC 90 -  besq.co.uk

Ron Spicer
Offline
Joined: Jul 22 2001
rone01 wrote:
interesting you should mention this Gary - I'm finding now I really like the lack of Punch that a flat Red One gives. It's less immediate than a high contrast image but you gain so much more in subtlety.

You can add punch of course - but I usually don't add so much as built into processed images from our other Panasonic cameras.

Oh and AVC-I makes a great codec for storage and distribution of locked R3D footage! I'd imagine the 10bits of being put to full use there.

For some particular purposes, I agree. A bride looks more beautiful if the features are not so clearly HD cut. I remember Judy Garland being treated that way in her time in films. Amazing results!

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Vaseline on the lens was the usual treatment.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

noddydog
Offline
Joined: Feb 28 2004
Alan Roberts wrote:
Vaseline on the lens was the usual treatment.

Erm, just to be clear for noobs, that's vaseline on a cheap UV filter, not directly on your lovely multicoated lens:-)

MAGLINK
Offline
Joined: Mar 8 2007

Or for the posh set-ups a promist!

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

But vaseline was being used pre-war, before Promists :)

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

noddydog
Offline
Joined: Feb 28 2004
Looking good

When you view at 1080: http://www.youtube.com/watch?v=IH2S-Y8up2U

.... but it's always hard to know what compression occured in the process.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

I had a quick look at the 250 on Thursday. Not long enough to form a full opinion, but it looks good enough to be worthy of testing for broadcast in that it's a parallel camera to the Canon XF305.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

MAGLINK
Offline
Joined: Mar 8 2007

HPX250 review here: http://vimeo.com/29509312