After all these times, we are still sitting at 1920x1080 Full HD Recording at 30 frames. How much longer will it be? Is there any upcoming news? Or is it something that is un doable for the next 10 years?

yes,, m too waiting for this..
________________________________
Rachael Ray TV show (http://www.medpro2006.com) muslim marriage sites (http://www.arabloveline.com/muslim-singles.html)

On the HF10/HF100 thread someone mentioned that the top bit rate for AVCHD (24) isn't enough for 60p.

Don't know if HDD has enough bandwidth for that either.

Seggers

Not in this generation. Even Sony's very expensive HDCAM SR VTR (http://pro.sony.com/bbsc/ssr/cat-videorecorders/cat-rechdcamsr/product-SRW5500/) doesn't support 60p at 1920x1080; the max is 30p. Only 1280x720 will give you 60p for now.

AVCHD doesn't include the 1080x60p as part of the spec, so I think you may have to wait for the next generation in format standard. Also taking into account that we've already maxed the bandwidth (24mbps) with the latest Canons on 60i and that 60p will require considerably more bandwidth (may not be twice, but it's going to be more than a few mbps to keep the quality up), it's more or less comfirm that we'll have to wait at another rev of AVCHD or another format.

I would also want 60p over 60i or 30p. But somehow, I don't see that happening for quite a few years. I hope I'm wrong though. :)

Interesting question.

I would also want 60p over 60i or 30p. But somehow, I don't see that happening for quite a few years. I hope I'm wrong though. :)It might be sooner than a few years but I didn't delay my purchase of a HD camcorder in anticipation.

AVCHD (or H.264, or the Microsoft equivalent of VC-1) is a compression technique. These files are large and the volume of data is tremendous but AVCHD can cope with refresh rates of 60p (or 60Mhz) or even 120MHz which is common now in the high end screens.

I'm not certain where the supposed constraint of 25 Mbits/second comes from. Depending on the media, speeds well above that are possible.

From what I have seen, 1920x1080 pixel resolution at 120p is as close to seamless as it gets. From a few feet, it looks real to people with average vision. This will be available at consumer prices within (wild, unsubtantiated prediction ahead) 7-8 years.

For the moment, the Sony and Canon marketing people seem happy to exploit the various combos of 1920x1080 at 60i. Software and PCs have to catch up and the basic camcorder technology has to become mainstream. For example, I would expect to see soon Blu-ray disc camcorders with built-in authoring software.

If you want the best picture that can be displayed by today's top of the line televisions, you would need (as others have stated), 120 frames/second at 1920x1080 and 10 bits per pixel. That would require roughly 80mbps transfer rates for good picture quality. The highest AVCHD supports is 24mbps and is also the fastest transfer rate available right now in the new canon line.

That being said, where there's demand, there will be products. Transfer rates went from 12mbps to 24mbps in a couple of years. If rates double every 2 years, we would hit 80mbps in 3-4 years.

I don't see the avchd limit as much of an issue. manufacturers can introduce a new backwards compatible format that can handle higher rates. I remember people stating we wouldn't have sound better than 5.1 because the standard for hi-def tv used 5.1. I also remember people stating that 480i was the best picture quality you could ever get from a DVD due to it's standard. Within a year progressive scan and upconverting technology blew that out of the water.

My best guess is within 3-4 years we'll have camcoders that can handle the best picture quality available today. Of course, by then, we'll probaly have tv's that can display more colors at a better resolution and refresh rate.

Maybe never, 1080p at 60fps is not in the ATSC spec. Its unlikely that a camcorder maker would make a camera perform beyond what a TV designed to receive. Even the now touted 120Hz refersh rate of TV has nothing to do with the signal that comes in.

I doubt they'd produce cameras that perform at full raster 1080 60Hz unless there was a market of TV's out there that would show it, and since 60Hz isn't a broadcast spec, I doubt they'd make those with the current HD spec.

Given how fast technology has come come in these last few years, I wouldn't be suprised if in 8-10 years ATSC's HD moves out for a newer, better system like the Ultra HD that NHK is experimenting on (or current HD was also once an NHK experiment in the 70's).

Maybe never, 1080p at 60fps is not in the ATSC spec. Its unlikely that a camcorder maker would make a camera perform beyond what a TV designed to receive. Even the now touted 120Hz refersh rate of TV has nothing to do with the signal that comes in.

I doubt they'd produce cameras that perform at full raster 1080 60Hz unless there was a market of TV's out there that would show it, and since 60Hz isn't a broadcast spec, I doubt they'd make those with the current HD spec.

While we're certainly not going to see a 1080p60 broadcast standard any time soon (decades) as there just isn't the bandwidth to support it, there is already plenty of hardware out there now capable of supporting a consumer camcorder format.

Most recent FullHD TVs will display 1080p60 (even my 2 year old Bravia will) and both the PS3 and Xbox360 can feed them a 1080p60 signal.

All that's missing is an active video format that supports it, and while AVCHD is (currently) limited to 720p60/1080i60 and 24Mbps, this is just a (Sony/Panasonic) branded subset of the full H264 spec which already has 1080p60 support (1080p60 in H264/AVC Level 4.2 and 1080p120 in H264/AVC Level 5.1).

So if Sony (for example) were to release a 1080p60 camcorder along with AVC Level 4.2 playback support on the PS3, then there would be a reasonable market out there for it, and of course they would be able to charge a pretty hefty premium for it too.

After all these times, we are still sitting at 1920x1080 Full HD Recording at 30 frames. How much longer will it be? Is there any upcoming news? Or is it something that is un doable for the next 10 years?

Why do you want something like that?
The amount of light in each frame is what determines the image quality, the more the better. When you are going beyond 24fr/sec you are having dimished return on added benefit of smooth motion. Most people are perfectly happy with the smoothness of Hollywood movies instead of complaining its being jerky. So going from 30fr to 60fr per second you are throwing away 50% of light in exchange of perceived smoother motion. I don't think most people knowledgeable about video would find it worth it. For that reason 60fr/sec cam probably would never be made for mass market.

Most people are perfectly happy with the smoothness of Hollywood movies instead of complaining its being jerky. So going from 30fr to 60fr per second you are throwing away 50% of light in exchange of perceived smoother motion. I don't think most people knowledgeable about video would find it worth it. For that reason 60fr/sec cam probably would never be made for mass market.

When you watch a Hollywood movie, you are watching the product of 100s of professionals who have spent there entire life learning how to make the image look good. They use all kinds of tricks to fool your eyes and divert attention away from the limits of 24fps. In the hands of an amateur 24fps footage usually sucks, as they don't know how to shoot to avoid it's limitations.

To say that 60fps isn't worth it is madness, of course it is. The difference is staggering, especially for point and shooters.

It's also worth pointing out that 99% of consumer video footage is 60fps right now; it's just it's 60i not 60p.

...To say that 60fps isn't worth it is madness, of course it is. The difference is staggering, especially for point and shooters.

It's also worth pointing out that 99% of consumer video footage is 60fps right now; it's just it's 60i not 60p.

So its safe to assume that you would trade 50% of light for 60fps? I think that's madness. 24fps is not at issue here. Its 30fps vs 60fps, as the OP is not happy with 30 full HD fr/sec and searching for 60 fr/sec.
BTW 60i is 30fps and 60p is 60fps. You should first understand the difference before calling other people's assertions "madness".

Interesting question.

It might be sooner than a few years but I didn't delay my purchase of a HD camcorder in anticipation.

AVCHD (or H.264, or the Microsoft equivalent of VC-1) is a compression technique. These files are large and the volume of data is tremendous but AVCHD can cope with refresh rates of 60p (or 60Mhz) or even 120MHz which is common now in the high end screens.

I'm not certain where the supposed constraint of 25 Mbits/second comes from. Depending on the media, speeds well above that are possible.

From what I have seen, 1920x1080 pixel resolution at 120p is as close to seamless as it gets. From a few feet, it looks real to people with average vision. This will be available at consumer prices within (wild, unsubtantiated prediction ahead) 7-8 years.

For the moment, the Sony and Canon marketing people seem happy to exploit the various combos of 1920x1080 at 60i. Software and PCs have to catch up and the basic camcorder technology has to become mainstream. For example, I would expect to see soon Blu-ray disc camcorders with built-in authoring software.

AVCHD is not a compression technique. It's a file format. H.264 (or AVC) is the video compression used in AVCHD, but they're definitely not the same.

The spec of the AVCHD format is available here (official AVCHD format site):
http://www.avchd-info.org/format/index.html

Note that:
1. 1080x60p isn't listed as a supported format.
2. 24Mbps is the max bitrate listed.

For AVCHD to go over those limits, it would require an revision of the format (AVCHDv1.1 or v2) so that there would be a distinction between the current limits and the new ones.

After a format is public like AVCHD, a company like Sony can't go and revise the max capabilities without reving up the format, as it has repercussions not only on the camcorders, but also on the potential players that would need to support those new limits. Imagine for example that a player with the AVCHD branding wouldn't be able to play those theorical new 1080x60p 30Mbps files. Users would scream murder that their "AVCHD" capable player doesn't support their new files.

Don't get me wrong. I'm all for 1080x60p and I would love to get out of that deinterlacing business without sacrificing framerate or resolution. But bottom line, as the spec is layed out, I don't see how 1080x60p will be possible with AVCHD.

If it happens, it'll probably be JVC, Samsung or Casio that will offer it, as they decided to offer their HD camcorders/hybrids using other formats than AVCHD which could theorically support 1080x60p.

While we're certainly not going to see a 1080p60 broadcast standard any time soon (decades) as there just isn't the bandwidth to support it, there is already plenty of hardware out there now capable of supporting a consumer camcorder format.

Most recent FullHD TVs will display 1080p60 (even my 2 year old Bravia will) and both the PS3 and Xbox360 can feed them a 1080p60 signal.

All that's missing is an active video format that supports it, and while AVCHD is (currently) limited to 720p60/1080i60 and 24Mbps, this is just a (Sony/Panasonic) branded subset of the full H264 spec which already has 1080p60 support (1080p60 in H264/AVC Level 4.2 and 1080p120 in H264/AVC Level 5.1).

So if Sony (for example) were to release a 1080p60 camcorder along with AVC Level 4.2 playback support on the PS3, then there would be a reasonable market out there for it, and of course they would be able to charge a pretty hefty premium for it too.

Agreed, but that theoretical camcorder won't be able to be branded AVCHDv1 because of the spec limitations I've mentioned above.

Also, from what I understand, the only reason why we see those new 120fps TVs is because 120 is divisible by 24, 30 and 60, and thus making the playback of all those framerate possible via frame repeats without changing the fps mode. It wasn't possible for a 60fps TV to support 24 natively otherwise.

I'm confident that we're quite a long way before we see consumer recording solutions that will break the 60fps barrier (in realtime). The next gen recording format (which is IMAX quality) will probably the 4x resolution of 1080p, and it's still using 24p and 60p. And don't expect this format to be used for broadcasting anytime soon.

So its safe to assume that you would trade 50% of light for 60fps?

Not sure why this is a concern though. Most consumer cam default at 60i and with minimum 1/60 s shutter speed (some would drop to 1/30s in low light condition). So this would not change with 60p. And those cam with 60p capability could have switchable 30p or 24p option.

Not sure why this is a concern though. Most consumer cam default at 60i and with minimum 1/60 s shutter speed (some would drop to 1/30s in low light condition). So this would not change with 60p. And those cam with 60p capability could have switchable 30p or 24p option.

Agreed... I'm all for user choices. :) Getting 60p, 60i, 30p and 24p would be the best.

BTW 60i is 30fps and 60p is 60fps. You should first understand the difference before calling other people's assertions "madness".

60i and 60p both have the same temperal resolution. 60i samples 60 unique points in time per second, just the same as 60p does. The difference is that with 60i the samples are of (alternating) halves of a full frame.

Played back correctly on a modern progressive TV, 60i will end up being displayed as 60 de-interlaced images every second. When motion is high these frames will simply be the fields displayed doubled up (bob), whilst during static scenes they will be combined from two fields (weave), with various possible motion compensation algorithms for points in between these extremes.

Unfortunately high quality motion compensated deinterlacing from 60i to 60p is very rare on the PC, and most software takes the lazy route and just deinterlaces to 30 progressive frames per second. Some let you choose to deinterlace to 60fps (bob deinterlace) at the expense of loosing vertical resolution, but few seem to do it properly. This may have distorted your view of what 60i is capable of, and if that is so then I recommend you view some 60i camcorder footage directly on a good quality TV.

I recommend you view some 60i camcorder footage directly on a good quality TV.

I agree. If I view my Panny SD9 footage on my 1080i LCD pj, the component video is clear, motion is smooth, and is identified as 1129i .

Not sure why this is a concern though. Most consumer cam default at 60i and with minimum 1/60 s shutter speed (some would drop to 1/30s in low light condition). So this would not change with 60p. And those cam with 60p capability could have switchable 30p or 24p option.

I think the OP's question is why not 60p being an industry standard, and my response is that 60p is going to look worse than 60i in most cases. If there is no benefit what is the point in making 60p an option. Its just like saying why not making 10 speed shift a standard in all cars, after all people would still find 5 speed shift available in those cars.

60i and 60p both have the same temperal resolution. 60i samples 60 unique points in time per second, just the same as 60p does. The difference is that with 60i the samples are of (alternating) halves of a full frame.


I think there is a misunderstanding here about how 60i is acquired on CMOS/CCD type of HD cams. When an HD cam acquires data for a field in 1/60 sec it reads the whole 1080 lines off the CMOS chip and then generate 540 lines for the field. By alternating 540 lines on 1080 matrix it creates 60i video. As you can calculate yourself, in each frame (not field) each pixel gets 1/30 sec of light in the case 60i video (in max amount). On the other hand in each frame in 60p's case, each pixel gets 1/60 sec of light max. Therefore 60i mode is a great compromise for both fast shutter speed and maximizing the amount of light for each pixel.
That's why for 60p you are likely to find 720p offerings, because each pixel on 720 is 2.3 times as big as 1080, which is more than enough to compensate for the loss of light, but you do lose resolution going from 1080i to 720p.

..
Played back correctly on a modern progressive TV, 60i will end up being displayed as 60 de-interlaced images every second. When motion is high these frames will simply be the fields displayed doubled up (bob), whilst during static scenes they will be combined from two fields (weave), with various possible motion compensation algorithms for points in between these extremes.

Unfortunately high quality motion compensated deinterlacing from 60i to 60p is very rare on the PC, and most software takes the lazy route and just deinterlaces to 30 progressive frames per second. Some let you choose to deinterlace to 60fps (bob deinterlace) at the expense of loosing vertical resolution, but few seem to do it properly. This may have distorted your view of what 60i is capable of, and if that is so then I recommend you view some 60i camcorder footage directly on a good quality TV.
I am not sure I am following what you are trying to say here. For most new HDTVs deinterlacing 60i shouldn't be an issue.

I am not sure I am following what you are trying to say here.

The process of capture and display using a consumer camcorder can be split into 3 broad sections.

1. Acquisition of image on the sensor.
2. Processing and encoding of the image into a Video stream on the camera.
3. Decoding and display of the Video stream on a display device.

For (1) 60i and 60p are identical. Assuming the shutter speed is 1/60th of a second, every 1/60th of a second a new image will be captured by the sensor. At this point there is no difference in the processes between 60i and 60p.

For (2) let’s assume for the sake of simplicity that the sensor has exactly 1920x1080 pixels. This is the point at which 60i and 60p diverge. For 60p, every 1/60th of a second the entire 1080 lines are read from the sensor and encoded as a full frame in the video stream. For 60i the even lines are taken from the image and encoded as the first field of a frame, and then 1/60th of a second later the odd lines of the next (new) image on the sensor are taken and encoded as the second field in the frame. The important point to note here though is that the even and odd lines are from different images taken at different points in time (1/60th of a second apart).

Now for (3) things start to get complicated. 60p is obviously easy to display, as we just display each full frame every 1/60th of a second. For 60i we have many options; the simplest of which is to just decode and display each field as a unique frame every 1/60th of a second so we effectively end up with 540p60. Alternatively we can blend two fields and display them every 1/30th of a second, which will approach (assuming little motion) 1080p30. The ‘proper’ method of displaying 60i though is using motion compensation systems to get the best of both worlds, displaying 60 images per second that are intelligently created from interpolation of data in fields behind and ahead of the current one, sacrificing resolution during periods of high motion, and getting near to 1080 lines when motion is low.

With regards to your comments about light sensitivity, there is no difference between 60i and 60p (assuming the same shutter speeds). Light sensitivity is entirely dependent on part (1) of this process, the acquisition of the image on the sensor pixels. In this respect 60i and 60p are identical, as every 1/60th of a second (assuming that is the shutter speed) a new image will be exposed onto the sensor pixels. The process of sampling these pixels into the scan lines that are encoded into either a 60p or 60i video stream has no effect on the light that falls on the sensor pixels.

... For 60i the even lines are taken from the image and encoded as the first field of a frame, and then 1/60th of a second later the odd lines of the next (new) image on the sensor are taken and encoded as the second field in the frame. The important point to note here though is that the even and odd lines are from different images taken at different points in time (1/60th of a second apart).

..

I think you are confusing how the digital sensor acquired the data with how old fashioned CRT does it. In digital sensors there is no scanning guns therefore no need to skip the lines. Lets say there are 1080 lines in the sensor, for the first 60i field, the processor takes rows 1&2, 3&4, 5&6 and so on to form one set of 540 lines, and then for the next interlaced field it takes the data from rows 2&3, 4&5, 6&7 and so on to form another set of 540 lines. So no data collected on 1920x1080 is wasted in the process.
I think if you know that then the difference between 60i and 60p becomes pretty clear.

I think you are confusing how the digital sensor acquired the data with how old fashioned CRT does it. In digital sensors there is no scanning guns therefore no need to skip the lines. Lets say there are 1080 lines in the sensor, for the first 60i field, the processor takes rows 1&2, 3&4, 5&6 and so on to form one set of 540 lines, and then for the next interlaced field it takes the data from rows 2&3, 4&5, 6&7 and so on to form another set of 540 lines. So no data collected on 1920x1080 is wasted in the process.

The details of how the image processor samples the sensor pixels to produce the scanlines is highly camera dependent, which is why I left out these details in my last post, to keep things simple. In recent cameras the sensor pixel count is likely to be higher than the video stream resolution (say 3-5 MP) and so each scanline is being produced by the processing of a number of sensor pixels using some kind of sampling kernel.

From your posts though, it seems that the crucial point that you're not taking into account is that the sampling of the even and odd fields occurs from different images. To use your example sampling technique.. "rows 1&2, 3&4, 5&6 and so on to form one set of 540 lines"...these rows are from one image on the sensor and then..."rows 2&3, 4&5, 6&7 and so on to form another set of 540 lines"... are rows of a new different image that is captured 1/60th of a second later.

.. In recent cameras the sensor pixel count is likely to be higher than the video stream resolution (say 3-5 MP) and so each scanline is being produced by the processing of a number of sensor pixels using some kind of sampling kernel.

well if you understand that then you can reach the same conclusion that in each of 30 frames in 60i each pixel has 1/30 sec of light while in each of the 60p's frame each pixel has 1/60 sec of light. That's my point-- 60i has twice as much light as 60p in each frame. The tradeoff for 60p is twice the frame rate for half the light in each frame.

...From your posts though, it seems that the crucial point that you're not taking into account is that the sampling of the even and odd fields occurs from different images...
What does that have to do with 60i having twice as much light as 60p.

well if you understand that then you can reach the same conclusion that in each of 30 frames in 60i each pixel has 1/30 sec of light while in each of the 60p's frame each pixel has 1/60 sec of light. That's my point-- 60i has twice as much light as 60p in each frame. The tradeoff for 60p is twice the frame rate for half the light in each frame.

Sorry GodobeHD, but this is completely incorrect. There is no difference in light gathering between different video resolutions or frame rates, interlaced or progressive, for a given shutter speed.

I don't know where you are getting your information from, but you have been misled.

A 1/30th second shutter will gather 2x the light of a 1/60th second shutter, but with the same 1/60th second shutter, 720p60, 1080p30, 1080i60 and 1080p60 will all have exactly the same light sensitivity as they are simply different ways of sampling the same image on the camera sensor.

willc111, I don't think I am going to convince about the difference between 60i vs 60p in terms of light sensitivity.
But as you yourself had mentioned earlier, each data point in 60i's field comes from an area on the sensor twice the size of that in 60p's frame. So even though the shutter speed is the same, each data point (which is a pixel in 1920x1080 matrix) in 60i has twice the light input, which yields better image quality in each frame.

willc111, I don't think I am going to convince about the difference between 60i vs 60p in terms of light sensitivity.
But as you yourself had mentioned earlier, each data point in 60i's field comes from an area on the sensor twice the size of that in 60p's frame. So even though the shutter speed is the same, each data point (which is a pixel in 1920x1080 matrix) in 60i has twice the light input, which yields better image quality in each frame.

So that’s the problem. Your assumption that an interlaced pixel will always be sampled from twice as many sensor pixels as a progressive pixel is false. Yes, this may be so in your simplistic from-a-text-book example used to explain to people how interlacing works, but we live in the real world.

Modern consumer camcorders have CMOS sensors with arrays of single-color sensitive pixels. The actual output pixels used to encode in the video stream are constructed from multiple sensor pixels using a bayer filter and demosaicing algorithm. In this case the number of sensor pixels used to sample a progressive output pixel, and the number used to sample a half-output-pixel-offset interlaced pixel will be the same.

But you're right, we're not going to agree on this. So let's agree to differ.
Apologies to the OP for the thread derailing.

Again what does Bayer filter (or Clearvid if its Sony for the matter) have to do with the size of the sensor area we are talking about here. Filters just subdivide the sensor for color info, they are located before the sensor and they don't restrict the area of data collection.
I think the image quality from the light sensitivity is precisely the reason for the lack of 1080p60fr that OP was wondering about.

Sorry GodobeHD, but this is completely incorrect. There is no difference in light gathering between different video resolutions or frame rates, interlaced or progressive, for a given shutter speed.

I don't know where you are getting your information from, but you have been misled.

A 1/30th second shutter will gather 2x the light of a 1/60th second shutter, but with the same 1/60th second shutter, 720p60, 1080p30, 1080i60 and 1080p60 will all have exactly the same light sensitivity as they are simply different ways of sampling the same image on the camera sensor.

I have to agree with willc111.
A image recording data path can be roughly viewed as 1. image sensor, 2. ISP, 3. CODEC engine.
Those video formats (720p60, 1080p30, 1080i60) are for CODEC engine, which encodes buffered images into the specified formats (resolution, raster/interlace).
While the amount of light per pixel will receive is set on the front end---image sensor. That is to say, the shutter speed control has nothing to do with how encoder works.

So, regardless of encoded image quality, the hardware basically can do whatever combination of shutter speed and video format.

@1/30 1280x720 into 720p60, no problem.
@1/120 1440x1280 into 1080i30, why not?

If shutter speed and encoding image format are tied together, can you see what will happen when you want to record 1080p60 under lowlight (say required shutter speed is 1/30)?

Again what does Bayer filter (or Clearvid if its Sony for the matter) have to do with the size of the sensor area we are talking about here. Filters just subdivide the sensor for color info, they are located before the sensor and they don't restrict the area of data collection.
I think the image quality from the light sensitivity is precisely the reason for the lack of 1080p60fr that OP was wondering about.

FIY, there are camcorders that can do 720 60p.

FIY, there are camcorders that can do 720 60p.

and most of those do it poorly

Alright, here's a $1M question:

If 60 frame/sec is not necessary, then why would all the brand name HD camcorder give us:

http://en.wikipedia.org/wiki/720p

Some broadcasters use 720p50/60 as their primary high-definition format;

720 p at 60 frames/sec?

I mean, if we don't need anything higher than 30p, then how come at resolution of 720, they come w/ 60 fps?

I went to chat w/ a Canon rep. at a trade show, he said even at national geographic, they use only 30p, I don't think that's true, but again, why would industry give us 720 at 60fps?

so they can make money on all those 720p TVs that they pushed on the public.