4:2:0 kills this camera for me. I'm sure a good key can be pulled, but for CC, it will fall apart if pushed far. That means using an external capture device, which start at another ~$3K. Cineform and a PC it can be less, but then I'm tethered.

The codec kills this camera for me.

I am thinking just the opposite Peter. Does ALL of your work ALL need to be in 4:2:2 or greater color space? Perhaps it does? For me, I shoot a lot of material that doesn't need to be in the highest color space. Stuff that is shown as web clips or only on SD DVD? Heck, one of the best looking shows on TV, Anthony Bourdain: No Reservations is shot on HDV, which is also 4:2:0 and shot on a format that is lower quality than AVCCAM. They color correct the heck out of that show, it looks great, sometimes amazing and won the Emmy last year for best non-fiction series cinematography. Even though I enjoy shooting with my HPX170, the 300s and 2700s, A lot of the time for my clients, it is overkill.

For the few times I shoot green screen, or footage for theatrical display, I could live with renting or buying a Nanoflash or an AG-HPG20. I think that this camera is a very smart move on Panasonic's part. If MSRP is around $6k as has been mentioned, retail might be in the low $5k range. I will use most of my existing Nikon AI and AIS glass, my Canon glass and just hunt down perhaps one micro 4/3 mount fast wide angle lens and life will be good.

I have a feeling that I will be buying one of these cameras. I think it is going to be huge. I wish it had P2/AVCINTRA but this is a good start. I am sure if they sell the heck out of them, they will come out with a higher end version of it eventually. I have to say, I have shot with the RED One twice and worked on several projects this year and I personally hate the RED post workflow, it is cumbersome and time consuming. I much prefer P2/AVCINTRA and DVCPROHD. For me, speed and ease of use trumps ultimate resolution and color space in most cases. I have only shot AVCHD once and it didn't seem any different functionally than the projects I shoot on AVCINTRA, it all ends up as ProRes anyway so why not? If you think about it, Panasonic is damned if they do include P2/AVCINTRA 4:2:2 (it's too expensive, I hate P2 cards, waaah) and damned if they don't (SD cards are flimsy, cheap and unreliable, it only has 4:2:0 color space, waaah). I think that including the AVCCAM and an SDI out is a good compromise.

Cheers,

Dan

I agree with Dan.

In my view, the only way Panasonic could mess this up is to hamper the resolution as to protect the higher end model(s).

If it is full raster 1080p without the DSLR image issues this will be a hit. I know I want one.

I am thinking just the opposite Peter. Does ALL of your work ALL need to be in 4:2:2 or greater color space? Perhaps it does? ...

Dan,

It happens enough (yes I can push it and get fanatical little bit too much) but I don't want to be stuck in a situation where the next camera I buy is not giving me what I need so I have to rent. Also the 4/3rd's frame is great, until you want to deliver 16:9, then there will be a lot of cropping.

It's a very nice camera and will work amazingly well for a lot of jobs. But I also can't help but think it's using a color depth that I want to get away from and it's aspect ratio that (while new for sensors) is kind of retro.

Also the 4/3rd's frame is great, until you want to deliver 16:9, then there will be a lot of cropping.

It's a very nice camera and will work amazingly well for a lot of jobs. But I also can't help but think it's using a color depth that I want to get away from and it's aspect ratio that (while new for sensors) is kind of retro.

Current HDV resolution is 1440X1080, which by itself is a 4:3 image aspect ratio. This is corrected by the use of anamorphic lenses, so I don't see that as being an issue if you're recording 16X9.

Also, if 4:2:0 color space is that big of a problem, you probably wouldn't be considering a camera like this in the first place. Panasonic's new HPX370 records 4:2:2, and several other makes and models offer that feature as well. Keep in mind that every camera out there compromises on something. This is especially so for anything below $10,000.

I'm curious to see what kind of images this thing is capable of recording, and how it will fit into its own little niche.

Shawn, I would hope that there is improvement over HDV. Most of the new cameras are recording full size HD, not 1,440. BTW, HDV records anamorphic by using rectangularly shaped pixels; none of the HDV cameras use an anamorphic lens.

And the new Canon XF is 4:2:2.

I'm sure this new camera will work wonderfully for a lot of productions; but it doesn't seem like it's for me. Perhaps I've been poisoned by the marketing surrounding the ever delayed Scarlet.

Just to say that Arri D21 uses a 3 x 4 sensor, so for other aspect ratios you need a crop - the advantage being you can use standard film anamorphic lenses. I suppose it comes where in the process the crop takes place and if you are actually losing vertical resolution when shooting 1080p on the Panasonic.

Also the 4/3rd's frame is great, until you want to deliver 16:9, then there will be a lot of cropping.

You seem to be assuming that the 4/3rds in the AF100 sensor is a 4:3 shape. There is nothing in the 4/3rds standard which stipulates that a sensor must use this aspect ratio; the only measurement it has to adhere to is the diagonal of 22.5mm. The sensor used in the GH1 is actually a bit different - it is larger than it needs to be with extra pixels on the sides and top and bottom so that at any given aspect ratio - 3:2, 4:3 or 16:9, it still produces an image circle with a diagonal of 22.5mm.

BTW, the new Alexa EV is 16:9, although they will be coming out with a 4:3 version and an optical viewfinder one as well. That said, I really doubt that Panasonic 4/3" users have anamorphic film lenses sitting around.

You seem to be assuming that the 4/3rds in the AF100 sensor is a 4:3 shape. There is nothing in the 4/3rds standard which stipulates that a sensor must use this aspect ratio; the only measurement it has to adhere to is the diagonal of 22.5mm. The sensor used in the GH1 is actually a bit different - it is larger than it needs to be with extra pixels on the sides and top and bottom so that at any given aspect ratio - 3:2, 4:3 or 16:9, it still produces an image circle with a diagonal of 22.5mm.

John, you're right I am. And I've heard it described as being very close in size to a standard 35mm frame (not S35). But I could very well be wrong.

BTW, the new Alexa EV is 16:9, although they will be coming out with a 4:3 version and an optical viewfinder one as well. That said, I really doubt that Panasonic 4/3" users have anamorphic film lenses sitting around.

Almost everyone will be shooting 16;9, because there isn't a HD 3 x 4 standard, so it's the number of pixels within the 16:9 on the sensor that's important, rather than those outside that area. Could be the sensor may be also being used in a stills camera, so that could be their reasoning for using that size sensor. The important part is that it achieves a full 1080 vertical resolution in 16:9 after debayering. That resolution isn't dictated by the unused areas of the sensor.

But having large unused areas of the sensor makes the 4/3'rds size misleading. 4/3 is twice 2/3, but not if a significant portion is subtracted from 4/3.

I'm not saying the camera will be bad. I'm sure it will be very good. I just think some people are forgetting that 35 is not S35. But like was said above, maybe the sensor will be a 22mm horizontal but not a 4:3 ratio, rather something closer to 16:9.

35mm motion picture film has large unused areas when shooting 1.85 and 16;9 on 4 perf pull down, which is the case on the majority of film cameras,

I believe the many of references earlier were mostly for standard 35mm (width 22mm camera/21mm projection), which is how most film productions are shot.

You're not going to have as shallow depth of field using the same stop, but you can pretty much achieve it by using a wider stop than most DPs actually use when shooting 35mm film (which usually isn't f1.4, more the f2 to f2.8 range or even a f2.8 to f4 split when using a zoom) .

To be fair to Panasonic & Sony, these cameras a huge step into unknown territory.

This camera is an evolution from a $1,000 DSLR which is popular with almost no-budget shooting. They can not stray too far away from that price or they will lose the market that was calling for this camera.

They are making a $6,000 camera that will for all intensive purposes deliver a high percentage of a 35mm camera. What percentage is yet to be seen, but if one takes the Zacuto tests into account, it will most certainly be over 50%.

So to get over 50% of the 35mm workflow for $6,000 is quite a bargain considering 35mm workflows adds zero's to the number.

About the Codec. AVCHD has been shown to be quite nice. I use Edius and I can now edit three streams in realtime. Sure it is not 4:2:2, but there are not many (if any) cameras with 4:2:2 for under $6,000. I think it is a logical choice for this pricepoint. You could always buy or rent a Nano for greenscreen or important work.

I really think this will be a big hit as the efforts people have gone through to get this "look" have been monumental and im-practical. This will fill a lot of niches and is at a price that many can afford to add to their lineup.

You seem to be assuming that the 4/3rds in the AF100 sensor is a 4:3 shape. There is nothing in the 4/3rds standard which stipulates that a sensor must use this aspect ratio; the only measurement it has to adhere to is the diagonal of 22.5mm. The sensor used in the GH1 is actually a bit different - it is larger than it needs to be with extra pixels on the sides and top and bottom so that at any given aspect ratio - 3:2, 4:3 or 16:9, it still produces an image circle with a diagonal of 22.5mm.

John, can I ask stupid ?? FWIU, the sensor size refers to the length of the diagonal. So 1/3" sensor is 1/3" across, 1/2" 1/2" across and so on. But if I take 4/3" and multiply it by 25.4 I get 33.87mm. This is not 22.5mm.

So what exactly is 4/3 measuring? Thanks much!

4/3" is measuring the circle that could be drawn around the four points of that rectangle. It's a holdover from the old days before image sensors, when tube cameras prevailed. 4/3" is the diameter of the tube that's big enough to produce the rectangular image area within it. It's an archaic, antiquated convention that should have been done away with many years ago, and yet it still persists.

So 1/3" sensor is 1/3" across, 1/2" 1/2" across and so onNo, that's wrong. Those actual sensor sizes are a bit smaller than 1/3" and 1/2" for the extremely outdated reason I've stated above.

Who needs Wikipedia? Thanks so much Chris.

So it seems to reason that a 4/3rd's sensor does have a diagonal twice as long as a 2/3rd's sensor. It's just that both are smaller than a straight inches to mm conversion would lead you to believe.

I am (was) certain that panasonic is talking about using the 4/3 still camera standard. That would be the mount too.

"The design of the AF100’s micro 4/3-inch sensor affords depth of field and field of view similar to that of 35mm movie cameras in a less expensive camera body. Equipped with an interchangeable lens mount, the AF100 can utilize an array of low-cost, widely-available still camera lenses as well as film-style lenses with fixed focal lengths and primes."

Are we sure 4/3 is an actual measurement? Arri also list 16:9 sensors.

The ARRI chip is close to still APS-C size. That would be Super 35 and designed to take those lenses.

Edit: From wikipedia:

Thus "The Four Thirds refers to both the size of the imager and the aspect ratio of the sensor".[5] Note that actual size of the chip is considerably less than 4/3 of an inch, the length of the diagonal being only 22.5 mm. The 4/3 inch designation for this size of sensor dates back to the 1950s and vidicon tubes, when the external diameter of the video camera tube was measured.

So 4/3 isn't a meaningful size reference.

The 4/3 inch designation for this size of sensor dates back to the 1950s and
vidicon tubes, when the external diameter of the video camera tube was measured.Validates what I was saying above... thanks.

So 4/3 isn't a meaningful size reference.And neither is 1/4, 1/3, 1/2 or 2/3, for the exact same reason.

I calculate the diagonal of a 16:9 crop of the AF100 sensor at 19.8mm, or .75 inches.
(The diagonal of a 16:9 crop of the 5DII is 41mm, for comparison.)

The same article mentions imaging area of 17.3 mm × 13.0 mm, so that seems like a starting point for comparisions.

2/3 is 8.80mm x 6.6mm, 1/3 is 4.8mm x 3.6mm, Standard 16mm 10.26mm by 7.49 mm.

However, don't take the 2/3 sensor sizes as fixed, the 16:9 sensors are different, for one Sony camera it's 9.58 mm x 5.39 mm. Yet another example: the SI 2k is 10.24x5.76mm @ 2k.

I shouldn't have said "AF100" sensor. I took the size off wikipedia, which may not reflect the full 4/3 standard. As pointed out the logical chip may be 16:9 with a 22.5mm diagonal (if the 4/3 standard allows any shape with a 22.5mm image circle)

So it seems 35mmish is "in the ball park" for the AF100, and otherwise we're just guessing.

IIf you think about it, Panasonic is damned if they do include P2/AVCINTRA 4:2:2 (it's too expensive, I hate P2 cards, waaah) and damned if they don't (SD cards are flimsy, cheap and unreliable, it only has 4:2:0 color space, waaah).
Damned if you do - damned if you don't. A tough one. An innovative manufacturer may therefore try to think of a third way as an alternative to either of those two. One which will be praised by it's customers, is technically easily implemented, and which is cost effective.

So what about AVC-Intra 100 to Compact Flash cards? You simply don't need P2 to record 100Mbs video these days. The nanoFlash is living proof that a 100Mbs video bitstream can be reliably recorded to fairly basic spec Compact Flash. The new Canon has gone for Compact Flash - why does the Panasonic choice have to be either (expensive) P2, or SDHC - why can't the Panasonic camera record AVC-Intra 100 to CF?

Or would that have their marketing people giving the loudest "waaah!"?

Are we sure 4/3 is an actual measurement? Arri also list 16:9 sensors

4/3rds is an actual measurement, but it is not a ASPECT RATIO. 4/3 is not the same as 4:3. Nothing in the standard dictates that it has to use any particular aspect ratio.


I calculate the diagonal of a 16:9 crop of the AF100 sensor at 19.8mm, or .75 inches.

But there is no crop - at least not one that reduces the size of the diagonal. Early reports from Jan Crittenden said the sensor was based on that from the GH1. While the GH1 sensor uses cropping in every aspect ratio, this is because the sensor is larger than it needs to be so that it can accommodate a 4:3, 16:9 and 3:2 image all with a diagonal of 22.5mm. So it is exactly the same as though you are pulling the 16:9 image from a 16:9 sensor.

a draw to illustrate the size

http://www.dvinfo.net/forum/attachment.php?attachmentid=17770&d=1272467528
bye

So what about AVC-Intra 100 to Compact Flash cards? You simply don't need P2 to record 100Mbs video these days. The nanoFlash is living proof that a 100Mbs video bitstream can be reliably recorded to fairly basic spec Compact Flash. The new Canon has gone for Compact Flash - why does the Panasonic choice have to be either (expensive) P2, or SDHC - why can't the Panasonic camera record AVC-Intra 100 to CF?

I think by the time this camera actually gets released SDXC will be the new standard or at least the wave of the future. CF currently enjoys the highest transfer rate now, but not for much longer.

I think by the time this camera actually gets released SDXC will be the new standard or at least the wave of the future. CF currently enjoys the highest transfer rate now, but not for much longer.
Quite possible - the point is that making the native codec AVC-Intra 100 *does not* technically neccessitate the use and expense of P2. At the moment, CF would be the strongest candidate but yes, it's conceivable that in a years time with SDXC it would be possible to reliably use those cards.

What's technically possible, and what is marketing strategy, are two wholly different things and it wouldn't surprise me if for the latter reason Panasonic were keen to keep AVC-Intra linked to P2 and AVCCAM linked to SDHC. It just needs to be absolutely clear that technically a 100Mbs stream like AVC-Intra can be reliably recorded to far cheaper media than P2.

For their higher range products, P2 will no doubt still make more sense than SDXC would. For a sub-$10,000 camera like this, AVC-Intra to CF or SDXC would be far and away the best compromise.

Here are some size comparisons compiled for Red
Some of these may be wrong, but it gives a rough comparison to 4/3

2/3" - 3072 x 1620 - 3K
10.1x 5.35mm MX
native 16:9 aspect ratio, approx 11.5mm diagonal,
3.3 micron pixels, approx: 5 mpx, 3.75 FOV crop factor.

R1 - 4520 x 2540 - 4K
24.4 x 13.7mm Mysterium
native 16:9 aspect ratio, approx 28mm diagonal,
5.4 micron pixels, approx: 12mpx, 1.5 FOV crop factor (APS-C).

S35 - 5120 x 2700 - 5K
30x15mm MX
native 16:9 aspect ratio, approx 33.5mm diagonal,
5.4 micron pixels, approx: 14mpx, 1.25 FOV crop factor (APS-H).

FF35 - 6000 x 4000 - 6k
36x24mm Monstro
native 3:2 aspect ratio, approx 43mm diagonal,
6 micron pixels, approx: 24mpx, 0 crop factor (FF DSLR).

645 - 9334 x 7000 - 9k
56 x 42mm Monstro
native 4:3 aspect ratio, approx 70mm diagonal,
approx: 65 mpx, 0.6 crop factor (MF).

617 - 28000 x 9334 - 28k
168 x 56mm Monstro
native 3:1 aspect ratio, approx 177mm diagonal,
approx: 261 mpx, 0.25 crop factor (MF).


Crop Factor refers to Field of View (FOV) crop factor, or Focal Length Multiplier (FLM), referenced to Full-Frame 35mm.

APS-C (Canon1.6, Nikon1.5): Canon Betacam, 550D, Nikon D300s, D90.
APS-H (Canon 1.3 crop): Canon 1D MkIV.
FF DSLR: Canon 1Ds MkIII and 5D. Nikon D3s/x and D700.
MF (Medium Format): various Hasselblad, Leaf, Mamiya, Pentax, Phase One.

I think by the time this camera actually gets released SDXC will be the new standard or at least the wave of the future. CF currently enjoys the highest transfer rate now, but not for much longer.

Anything but proprietary storage. There's little benefit to the small SD size in a camera like this.

Quite possible - the point is that making the native codec AVC-Intra 100 *does not* technically neccessitate the use and expense of P2. At the moment, CF would be the strongest candidate but yes, it's conceivable that in a years time with SDXC it would be possible to reliably use those cards.
AVC Intra 100's data rate is 100 megabits -- or 12.5 megabytes -- per second. Panasonic's E-series P2 cards can reach 1.2 Gbps, or 150 MBps. SDXC cards following the SD 3.0 specification were scheduled to reach 104 MBps (UHS104) last year, and SD 4.0 specifies a maximum transfer speed of 300 MBps. So, on speed alone, SDXC should easily handle AVC Intra 100, and probably AVC Ultra (200 Mbps). Reliability is, of course, another matter.

UDMA CF is necessary with Canon CF to reliably do 35mb/s.

I've been looking at GH1 movies and I have to say the color is nice. Seems very well controlled, especially skin tones. Canon can get a little pink or go too dull at times(with Caucasians). That chip with a proper video anti-aliasing filter should be very nice in the AF100. Especially if the chip is read faster.

I wouldn't want to buy 4/3 lenses, but it doesn't sound like that will be necessary.

UDMA CF is necessary with Canon CF to reliably do 35mb/s.
I think you are confusing MegaBITS with MegaBYTES. The data rate of the EX is 35Mbs(bits), the Canon camera will be 50Mbs (MegaBITS/s). UDMA CF is rated up to 45 MBs (MegaBYTES/s) or no less than 360Mbs - 10x the 35Mbs data rate you refer to.

Something like Sandisk Extreme III (much cheaper) is rated to 30MBs - or 240Mbs. That's still nearly 7x the speed of the 35Mbs codec, and about 5x the speed of the 50Mbs codec (as used by the new Canon).

Now I take the figures with somewhat of a pinch of salt, and ones which need a big safety margin to be allowed. But in real world usage that speed of card is approved with the nanoFlash for the 100Mbs datarate - so about 2-2.5x seems to be an adequate safety margin.

And if the card can record 100Mbs reliably in the nanoFlash, there is no reason why it shouldn't do the same for AVC-Intra 100 in a Panasonic camera. As far as reliability goes, Compact Flash is relied upon by the vast majority of the professional digital stills photo industry. The only failures I've heard of first hand was someone who bought a very cheap, unbranded card - needless to say, he sticks to decent brands now, with no further trouble.

The point I was trying to make was that the speed rating alone didn't matter for avoiding buffer overrun. The card needed to be UDMA.
I do wonder if more than 50 mb/s is necessary. I am curious if current CF cards can really do sustained 100 mb/s.

I am curious if current CF cards can really do sustained 100 mb/s.
Don, I don't think there's any doubt about it, and that's relatively cheap cards at that. Higher spec ones will reliably record much higher bit rates still. As far as the 35Mbs rate goes, many people are reliably recording that to pretty low spec SDHC via an adaptor in an EX.

Still not convinced? Just take a look at the relevant Convergent Design website page - Media | nanoFlash | Video Recorders and Converters (http://www.convergent-design.com/Products/nanoFlash/Media/tabid/1653/Default.aspx) . You can't just take any cheap unbranded card and hope for the best, but that reference gives a range of cards qualified by CV for given bitrates - all the way up to 280Mbs!

The nanoFlash has been in use for quite a while now - if CF usage for recording sustained 100Mbs was a problem, we'd have heard a lot about it by now.

I hate to be literally the only one on here who seems to bring this up but... what about SKEW!

You are all talking about size of chip, but not one word about actual function. I have a horrible feeling this is just going to be a lovely tease just like the 300 was. It will be plagued with skew. They really need to be on their game and eliminate the skew issues down to something much closer to a global shutter feel, for this camera to be taken seriously.

Why am I the only one who actually cares about the MOTION in motion pictures? How can anyway stand watching any of this footage from these DSLR's without it screaming SKEW...?

I just don't understand. And for everyone complaining about 4:2:0 color space... it has HD-SDI out... the camera is only 6k I think you can buy one of the few capture boxes out there and get your 10bit 422. But please for the love of the camera gods, push Panasonic to eliminate SKEW!

I will only buy this camera IF there are no skew issues, otherwise its in the same category to me as a DSLR, and thats just sad.

My guess is that it is down to pure processing power to read each pixel at the same time for every frame.

I hate skew but I think it would be addressed by now if it was an easy fix.

I agree that it seems like as much attention is paid to limiting a camera to fit in a range rather than just whooping the competition. I know this is their busines model and they need to make money, but it is frustrating as a consumer.

What bugs me about skew is that this situation conforms shooting with these cameras to a cinema type approach, ie limited camera movement. Which actually is closer to the role of the AF-100. But for the smaller chipped cameras it just does not fit. These are video cameras and video like it or not, often involves moving the camera.

I sure it wil be fixed over time.

As I understand it, skew comes down to sensor read-reset times (the time it takes to scan a line of the sensor and move on to the next) and processing horsepower. I know Panasonic has mitigated skew somewhat in their higher end cams through processing, but possibly it's been too expensive to implement in their smaller cams to this point. As most things, if you see it in the more expensive line, generally it will trickle down to the less expensive lineups eventually.

My guess is that Panny will up the read-reset times with whatever newer chip they stick in this thing. Who knows if it'll be feasible to cram the processing in there by release time.

We live in AMAZING times in filmmaking today. Cameras released this year will be able to do, for less than $10k, things that we never would have dreamed of just a decade ago.

The camera will most assuredly appeal to the "Pro World." even DSLR cameras out right now do.

Panasonic doesn't need to "beat RED." RED is a very small company that doesn't pose even a minute threat to Panasonics business. The companies are in totally different classes. If you want something like a red, BUY A RED.

"One perfect camera"? Give me a break. the thing doesn't exist. Has never existed. and never will exist. And the idea that it would make panasonic MORE money is laughable. Thats like saying that the Ford Motor company would make more money if it made one perfect car that could run 10 million miles. Planned Obsolescence has ALWAYS been a part of consumerism, and always will. If it wasn't people would be out of jobs.

If you don't like the camera, don't buy it. There's an OCEAN of options out there for you. That's how amazing the times are right now.

As I understand it, skew comes down to sensor read-reset times (the time it takes to scan a line of the sensor and move on to the next) and processing horsepower. I know Panasonic has mitigated skew somewhat in their higher end cams through processing, but possibly it's been too expensive to implement in their smaller cams to this point. As most things, if you see it in the more expensive line, generally it will trickle down to the less expensive lineups eventually.

My guess is that Panny will up the read-reset times with whatever newer chip they stick in this thing. Who knows if it'll be feasible to cram the processing in there by release time.

In addition there's the number of channels on the sensor. It appears there are chips with about the same number of photosensors as finished resolution. So about a 2K chip. But there are also the DSLR that do some sort of aggregation to produce a smaller final image. It's possible with CMOS that this aggregation could be done on-chip.
There's going to be different technical approaches to solve these problems. It appears most videocams will switch to CMOS. With that change happening I'm sure they have ideas about limiting skew.

How can anyway stand watching any of this footage from these DSLR's without it screaming SKEW...?

that may be a little bit of an exaggeration!

personally the only time ive noticed skew in footage from my 7d is in a shot when ive accidentally kicked the tripod or whipped the camera around to reframe

what about SKEW!


what about it, every CMOS camera has it,
it is in nature of rolling shutter, I don't think any manufacturers any time son will be able to make CMOS imager fast enough to eliminate skew and at the same time keep the camera cost affordable;
if skew problem would be so important to me as it is to you Giuseppe, for the next couple years for sure, I'd stay with CCDs,
but i don't like crazy camera moves, and I love my DSLRs :)

I agree that skew is not a big issue. Chances are if your footage is unusable because of skew or jello, then there is most likely an underlying problem with the footage which would make it unusable anyway (eg panning too fast, vibrations from strong winds or a moving vehicle) Personally if I ever see skew in my GH1 viewfinder I take it as a warning sign that I'm doing something wrong and need to slow down my movements or stabilise the camera.

Exactly my point. What if you are not filming a "movie" but shooting video? Is there then a problem with your filming if you have to pan or catch a fast moving object?

I know CMOS is a way to get things to market, but it sort of forces a cinema approach to all shooting unless you can live with diagonal verticals in certain instances.

As a side note, I was shooting in a testing machine shop with some large air compressors with an EX-1. I had a fair amount of jello on locked-down shots from the slight floor vibrations. It is not always operator error! This stuff is real and I could not use those shots.

what about SKEWFewer swish-pans.

How is that not a good thing?

As a fan of classic cinema, I am beginning to look at CMOS/rolling shutter skew as possibly a good thing. It forces people who have no concept of how to move a camera cinematically to do so. Cameras are not meant to be waved around like a hose, it makes the audience sick.

MTV camera movement was a reaction to visual boredom and because operators often did not have access to dollies, jibs and Steadicams. Now that you can buy a slider for $99.00 and a jib or Steadicam device for under $1,000.00 IMHO, there are few places or times when lousy camera movement is warranted.

Even the worse skewing camera can make beautiful images if the operator knows how to utilize the movement to good effect instead of just randomly shaking and moving the camera around to give a shot or composition energy. Of course, skew when the camera is not moving is not desirable but if it makes you feel better, 98% of the audience never notices it and when they do, it is still no big deal. It bothers filmmakers and videographers MUCH more than it bothers audiences. These days, it is becoming accepted, to a point, as part of the look that people use.

Dan

Where did anybody say they were moving their camera like a hose? This was the mantra of the HPX-300 defenders...until they actually fixed the skew a bit.

Not everybody shoots movies with their video cameras. If you shoot any kind of sports, movement or live events this is part of your world. And you do not always have the luxury and cinematic camera movement. Problem is that in a few years, CCD will be gone. So I hope somebody who is making the cameras thinks skew is bad so they eliminate it in the future.

I just don't get the "nobody will notice it" attitude that so many have with today's cameras. If somebody made a light that flickered once every 3 minutes, would you use it? Or be happy when folks told you nobody would notice it or care? Just seems like a degredation of the professional toolset to me.

Although I do expect the AF-100 to have skew, this camera is more of a cinema camera, so it should be less of an issue for use.

Well i've never once saw a video posted on vimeo or youtube and thought, "wow that skew ruins the look of this". Unless they were trying to show examples of skew of course.

I have seen many videos that had that horrible CCD bright line that goes up and down the screen when a light or the sun makes it into the frame. I noticed that even before I saw people online complaining.

Skew seems to be an invisible problem to the laymen like me.

If you dont like skew, you just have to wait for 48p, 50p, 60p. But i like 25p.

i think its fair to want some improvements in the rolling shutter.. of course nobody would complain about that! but i think the main point being made by users of dslrs based on experience so far (with the canons at least - the nikons skew is certainly more of a challenge) is that if your movements are exteme enough to give you noticeable skew then perhaps the viewers are going to be more put off by the extreme camera movements before they would by the wavy verticals..

for me at least i think it would be nice if it was fixed or improved but it wouldnt be a dealbreaker of any kind as its never been an issue for me, there are of course many styles of shooting though so just speaking for myself!