Slashdot Mirror

No DV is slightly lossy. It uses a DCT similar to JPEG, but higher quality. http://www.adamwilt.com/DV-FAQ-tech.html for more info if you are curious. It's damn close to lossless, but you can find artificats sometimes if you try hard enough. However, the amount of problems you find are way less than BetaSP.

What I'd like to see is a consumer version of one of the pro HDTV formats out there, DVCPRO-HD would be fine. Tape might need to be larger than MiniDV, that's fine, just something in a more affordiable kind of camera. All the HD cams are too pro oriented and thus too expensive.

I dunno about that. They're a pain in the ass to use (for example, if you set the white balance and then adjust the aperture, you can't change the white balance again without undoing the aperture setting), and they don't really hold up all that well. My school gives them out to film students, and they're in crappy condition after only a couple of semesters of use--people complain about dropouts all the time, even though they're DVCAM and should be more or less immune to that (and that's not to mention all the other problems they've had with them).

If you want a really nice camera in that range, pick up a Canon GL2. Easy to use (put it in automatic mode and you're all set), but highly configurable if you need that capability as well. I've known people who have bought Sony cameras and liked them, but I've also known people who've bought them and regretted them. I've never met anyone (myself included) who has bought a Canon and regretted the purchase.

Whatever camera you get, make sure you put it in a Pelican case or equivalent waterproof case. I see a lot of people buy nice cameras and then put them in soft camera bags--it boggles the mind. If you take decent care of it, and don't let too many people who don't know what they're doing screw around with it, it should last for years even with heavy use. Just clean the heads every now and then, and make sure to watch out for the "conflicting lubricant" issue by either standardizing on one brand of tape or frequently switching brands (more info on that here). Also, get a decent tripod and some sandbags, especially if you're going to leave the camera unattended. It only takes one good fall to destroy your camera completely, and that probably won't be covered under warranty.

Forgive me for stepping on your words:

Basically, 1080i = 540 lines / refresh. 720p has 720 lines per refresh.

but it seems you are implying that 720p is a higher resolution than 1080i. You are correct that more lines are refreshed during each scanning pass in a progressive scan format. However, the refress rate is half as often.

In progressive formats, all of the screens pixels (each frame) are refreshed every ~1/30 second. In interlaced formats, each frame is made up of 2 fields and fields are refreshed alternately every ~1/60 of a second. In the end 1080i has refreshed 1080 lines of resolution in 1/30th of a second, while 720p has refreshed 720 lines of resolution in 1/30th of a second. In the case of 1080i, the gun is spraying electrons much faster than 720p.

The alternation of even and odd lines is virtually invisible to the eye when dealing with slower moving images. Our eye's persistence of vision continues to see the odd lines of the TV image while the even lines are being displayed. The reason films run at 24 fps is because that is that is the point at which the human mind percieves continuous motion from individual stills. (You may have noticed that sometimes when you see silent films that they appear to be sped up. This is because they were often filmed at 18 fps and played back at 24 fps. Film is expensive, especially back then! (Film contains silver.) Shooting a film at 30 fps, instead of 24, would add 25% more cost to film stock and processing. Why raise the budget with very little visual impact?

Also, folks often confuse the way things are acquired and the way they are distributed. Just because something is distributed in 1080i does not mean that it has to be aquired in an interlaced format. For example, the Panasonic AG-DVX100A is a prosumer 24p camera. But it uses the standard MiniDV format to record the image. The image is recorded at NTSC 29.97 interlaced. All of which is explained here by Adam Wilt. By using 3:2 pulldown (RTFA), the 24p images can be encoded and displayed on standard 29.97i NTSC equipment. The same technique is used to produce DVDs.

The difficulty of interlacing is that when motion occurs quickly, the odd and even lines don't seem to match up correctly and can create moire (that weird visual pattern you see when somebody wears a tightly striped shirt on TV) or other visual distractions. However, this effect only applies to images that are acquired interlaced. If the images are acquired progressively and displayed interlaced, you will never notice the problem. 1080i, if care is taken in the acquisition phase, should always looks better than 720p.

Forgive me for stepping on your words:

Basically, 1080i = 540 lines / refresh. 720p has 720 lines per refresh.

but it seems you are implying that 720p is a higher resolution than 1080i. You are correct that more lines are refreshed during each scanning pass in a progressive scan format. However, the refress rate is half as often.

In progressive formats, all of the screens pixels (each frame) are refreshed every ~1/30 second. In interlaced formats, each frame is made up of 2 fields and fields are refreshed alternately every ~1/60 of a second. In the end 1080i has refreshed 1080 lines of resolution in 1/30th of a second, while 720p has refreshed 720 lines of resolution in 1/30th of a second. In the case of 1080i, the gun is spraying electrons much faster than 720p.

The alternation of even and odd lines is virtually invisible to the eye when dealing with slower moving images. Our eye's persistence of vision continues to see the odd lines of the TV image while the even lines are being displayed. The reason films run at 24 fps is because that is that is the point at which the human mind percieves continuous motion from individual stills. (You may have noticed that sometimes when you see silent films that they appear to be sped up. This is because they were often filmed at 18 fps and played back at 24 fps. Film is expensive, especially back then! (Film contains silver.) Shooting a film at 30 fps, instead of 24, would add 25% more cost to film stock and processing. Why raise the budget with very little visual impact?

Also, folks often confuse the way things are acquired and the way they are distributed. Just because something is distributed in 1080i does not mean that it has to be aquired in an interlaced format. For example, the Panasonic AG-DVX100A is a prosumer 24p camera. But it uses the standard MiniDV format to record the image. The image is recorded at NTSC 29.97 interlaced. All of which is explained here by Adam Wilt. By using 3:2 pulldown (RTFA), the 24p images can be encoded and displayed on standard 29.97i NTSC equipment. The same technique is used to produce DVDs.

The difficulty of interlacing is that when motion occurs quickly, the odd and even lines don't seem to match up correctly and can create moire (that weird visual pattern you see when somebody wears a tightly striped shirt on TV) or other visual distractions. However, this effect only applies to images that are acquired interlaced. If the images are acquired progressively and displayed interlaced, you will never notice the problem. 1080i, if care is taken in the acquisition phase, should always looks better than 720p.

DV is also lossy. But maybe it doesn't get worse as fast on successive saves?

Because DV video (similar to motion JPEG) uses the same DCT as MPEG-2 and can be made to run off the same ASIC?

Because MPEG-4 royalties are much higher than MPEG-2 royalties?

Because Sony and the other studios are scared of MPEG-4?

I've said it before, and I'll say it again, Episode II is going to suck in terms of picture quality. Episode I looked bad enough (the whole movie looked soft and fuzzy, due to the use of low resolution computer processing on nearly every shot), but Episode II is just going to look plain awful.

The camera they used is built by Sony (with lenses by Panavision), and captures images at 1920x1080 resolution at a 16:9 aspect ratio, the same aspect ratio as HDTV (basically, it is an HDTV camera). In order to form a 2.35:1 widescreen image, this will be cropped to about 1920x800. Compare that with this resolution chart for 35mm film scanning. 35mm film still has scannable information on it at resolutions up to 4096 x 3112.

Of course, the real resolution of Sony's video camera is effectively only about half of the stated 1920x1080, due to the fact that the images are color sampled in a 4:2:2 fashion, which means that every other pixel is just a black-and-white pixel, with the color sample from the adjacent pixel added in. That's not the same thing as true 1920x1080 resolution. And don't forget that the image data is heavily compressed in-camera 9at about a 7:1 compression ratio) using MPEG-2, just to make it fit onto the tape. Can't wait to see all those nice compression artifacts!

Finally, as you pointed out, there are slow-motion shots to consider. Well, guess what? The Sony camera can only shoot at 24fps! Therefore, the Episode II crew had to keep film cameras around for any shots that required slo-mo. Notice that they've been pretty quiet about that. You won't find any information about that on the official Star Wars website! It's a bit of a dirty little secret.

To be fair, shooting with video does have one advantage -- instant feedback. You don't have to wait for dailies to see if you got the shot right. Of course, with video assist systems on most film cameras these days, you can tell a lot already, even though the video assist playback doesn't exactly show you what the final product is going to look like. But when the video is your final prodcut, you know exactly what you're getting right away.

I know I'll probably sound like a Luddite with this post, but I'm really not opposed to video taking over film. It's just that the quality really isn't there yet. Until the resolution and color range of video is increased to match that of film, it just won't look as good. I think George Lucas is getting ahead of himself in deciding to go with video today. However, I think he's so anxious to be known as an "innovator" that he's decided to just do it anyway, quality be damned. Oh well, that's my take on it anyway.

(Note: Sorry if this comment shows up twice, slash seems to have eaten it the first time, yet it's still complaining about it being a duplicate.)

I've said it before, and I'll say it again, Episode II is going to suck in terms of picture quality. Episode I looked bad enough (the whole movie looked soft and fuzzy, due to the use of low resolution computer processing on nearly every shot), but Episode II is just going to look plain awful.

The camera they used is built by Sony (with lenses by Panavision), and captures images at 1920x1080 resolution at a 16:9 aspect ratio, the same aspect ratio as HDTV (basically, it is an HDTV camera). In order to form a 2.35:1 widescreen image, this will be cropped to about 1920x800. Compare that with this resolution chart for 35mm film scanning. 35mm film still has scannable information on it at resolutions up to 4096 x 3112.

Of course, the real resolution of Sony's video camera is effectively only about half of the stated 1920x1080, due to the fact that the images are color sampled in a 4:2:2 fashion, which means that every other pixel is just a black-and-white pixel, with the color sample from the adjacent pixel added in. That's not the same thing as true 1920x1080 resolution. And don't forget that the image data is heavily compressed in-camera 9at about a 7:1 compression ratio) using MPEG-2, just to make it fit onto the tape. Can't wait to see all those nice compression artifacts!

Finally, as you pointed out, there are slow-motion shots to consider. Well, guess what? The Sony camera can only shoot at 24fps! Therefore, the Episode II crew had to keep film cameras around for any shots that required slo-mo. Notice that they've been pretty quiet about that. You won't find any information about that on the official Star Wars website! It's a bit of a dirty little secret.

To be fair, shooting with video does have one advantage -- instant feedback. You don't have to wait for dailies to see if you got the shot right. Of course, with video assist systems on most film cameras these days, you can tell a lot already, even though the video assist playback doesn't exactly show you what the final product is going to look like. But when the video is your final prodcut, you know exactly what you're getting right away.

I know I probably sound like a Luddite with this post, but I'm really not opposed to video taking over film. It's just that the quality really isn't there yet. Until the resolution and color range of video is increased to match that of film, it just won't look as good. I think George Lucas is getting ahead of himself in deciding to go with video today. However, I think he's so anxious to be known as an "innovator" that he's decided to just do it anyway, quality be damned. Oh well, that's my take on it anyway.

Existing digital camcorders compress their signals right from the word go, to manage their consumption of tape and help keep the transfer bandwidth over IEEE1394 reasonable. However, to keep the data stream easily editable, only intra-frame compression is used; DV uses a fixed 5:1 DCT-based compression, some "pro" formats use a fixed 3.3:1 compression ratio. The MPEG-based HiDef cameras do the same.

The 5:1 compression used in DV is just visible(see here for the SMTP analysis) but for most people the only really noticeable effect of the chroma subsampling used in DCT-based compression is that it makes chroma-keying (e.g. bluescreening) much more difficult because of the averaging - the trick there is to use green as your key colour, because DV camcorders typically retain more information on green than other colours.

Remember that one of the things we're getting back from using Digital is less generation loss along the path to the finished product - it's often staggering how many generations a film print can be from the original negative, and none of that loss applies to digital. The end-stage compression is a non-issue.

720-line, 60fps Hi-Def video is breathtaking. I shoot a lot of sports footage on DV, and my Canon camcorder can shoot either 25fps progressive-scan (no interlacing) or regular interlaced. The "look" of the two is completely different when viewed on a TV - the motion strobing of the lower frame rate isn't really noticeable by itself, but flicking between modes the interlaced footage is noticeably smoother. It's subtle.

One of the very interesting perceptual phenomena that regularly appear on the DV-L mailing list is that progressive-scan video shot on camcorders like the Canon XL1 actually looks, according to viewers, like it originated on film. Why isn't 100% clear; perhaps it's just that the lower frame rate of film is something we're trained to notice. See here for more on the film-look phenomenon. The one aspect of film look that digital video cameras generally can't provide is shallow depth of field; that takes a large imaging area - compare 35mm film versus a 1/3" CCD, and you'll see why film wins.

In fact, 24fps isn't going away despite the move to higher frame rates for HDTV systems. See here for Sony's blurb on 24fps progressive.