If something in the image is moving, it looks like there are spiky lines sticking out from the sides of the object like the tines of a comb. Hence the effect is usually called combing , though it is also sometimes referred to as feathering. Sometimes, the cadence doesn't stay regular. For various reasons detailed later , the cadence may break from time to time, or perhaps the video was never sourced originally from film. Documentaries, concerts, and made-for-TV material often is shot on video cameras, and then there is no good way to create perfect progressive frames.
Video cameras capture 60 separate fields one field has half the scanning lines of the display per second, and each one is separated in time, so moving objects are in a different position in each field. In any case, if the deinterlacer doesn't see a film cadence, it must switch to video-mode deinterlacing. Here the algorithms get much tougher, and perfect results are just not possible, only different sorts of compromises that look subjectively better or worse. There are two very simple techniques, neither of which is very good, and a host of progressively more complex algorithms.
We'll divide them into five large categories, roughly in order of complexity:.
Secrets Vault
This just involves taking each field and scaling it to a full frame. The missing lines between each of the scan lines in the field are filled in with interpolated data from the lines above and below. Done badly, the screen looks blocky and pixellated. Even done well, the image looks very soft, as image resolution is unavoidably lost.
In Defense of Physical Media: Why You Should Keep Buying Blu-rays and DVDs
These thin lines will fall on just one field of the frame, so they will appear and disappear as the player alternates between the odd fields and even fields. This is the most basic deinterlacing algorithm, and the one that almost every deinterlacer falls back on when nothing else will work. In this technique, each pair of two consecutive fields is merged together to form a frame.
- One of Sony’s best 4K upscaling Blu-ray players is on sale on Amazon for under $100;
- Little foxes (1866).
- .
- One of Sony’s best 4K upscaling Blu-ray players is on sale on Amazon for under $ – BGR.
- Diffusion: Mass Transfer in Fluid Systems (Cambridge Series in Chemical Engineering).
- Staatliche Kulturförderung in Deutschland (German Edition)?
This generally only works well if there is very little or no movement between the two fields, such as is the case with a single frame of film. If there is movement between fields, the image will have combing, which is very distracting. Hence very few deinterlacers use this as a primary algorithm. Because software decoders generally don't have enough horsepower to do motion-adaptive deinterlacing and MPEG-2 decoding at the same time, the players use shortcut techniques to get reasonable-looking results.
Most commonly, they weave together pairs of fields that are stored together as MPEG frames, and soften the image slightly in the vertical direction so any combs that result will look more like double images than combs. This causes loss of vertical resolution and bizarre-looking jitter on movement and pans. Once you notice it, it can become difficult to watch. Static areas of the image also look much softer than they need to, and the overall effect is watchable, but not great.
This is a whole class of algorithms that attempt to switch between different ways of deinterlacing depending on whether an area of the screen appears to be still or moving. If an area is still, the algorithm uses the image data from two fields and weaves them together, but for moving areas, the algorithm just interpolates bobs the current field.
This preserves resolution on the still sections of the screen, where the viewer is most likely to notice it, and reduces combing on the moving sections of the screen, at the expense of resolution. Done well, this looks very good. Most good deinterlacers use some form of motion-adaptive algorithm. A distinction should be made between "per-pixel" motion-adaptive deinterlacing, which makes decisions for every pixel on screen, and "per-field" motion-adaptive deinterlacing, which just changes the algorithm for the whole screen based on the amount of motion in the frame.
The first one is difficult, and when done well looks quite good. The second is not much more difficult than simple vertical filtering, and in practice rarely looks any better. This is something one generally only finds on very, very expensive deinterlacing solutions, and we mention it here for completeness. This involves doing elaborate image analysis to identify the moving areas of the image, and weaving together the same image from two fields, with individual areas shifted to compensate for the movement.
It involves a lot of processing power, and is not found on any DVD player we know of. This is, not to put too fine a point on it, absurd. With the possible exception of some very low-cost progressive players, all progressive players are capable of outputting the entire film frame, without compromise.
Whether the player reads the progressive frame directly off the disc, or recreates it with a deinterlacer in the digital domain, the end result is the same. More importantly, "line doubler" is just a marketing way of saying "deinterlacer. Even worse are "line quadruplers" which are really deinterlacers combined with scalers that double the number of scan lines.
The argument seems to be, "Since a deinterlacer is a 'line doubler' and the scaler then doubles the actual number of scan lines, then the combination must be a 'line quadrupler. The major advantage the progressive player has over the external deinterlacer is that in the player, the deinterlacing can be done to the video in the digital domain, using the digital video directly out of the MPEG decoder, without any intervening analog conversions.
An external decoder must use the analog signal from the player, re-digitize it to feed into the deinterlacing chip, and then convert it back to analog to feed the display. If it's done well, the image degradation is relatively minor. But all else being equal, the progressive DVD player will always produce a better p image than an interlaced DVD player connected to an external deinterlacer with the same chipset.
When combined with a DVD player that has a compatible digital output, they provide no-compromise deinterlacing and scaling completely equivalent to a good progressive DVD player with the same chipset. These combinations are fairly expensive, though, and not necessary for most consumers who use standard RPTVs.
If the display device requires an external scaler for optimal display quality, because it doesn't display a good image with a p signal, then there's no real advantage to using a progressive DVD player. As long as the deinterlacing chip in the external scaler is of similar quality to the one in the DVD player, the final video quality should be essentially the same whether one feeds the scaler p or i. In addition, many external scalers don't accept p inputs, making the decision moot. We reviewed one add-on product that adds scaling and deinterlacing to an interlaced player, and while there were some issues with that specific product, the idea is sound.
This used to be the most common chip used in progressive DVD players, but has fallen out of favor. The main chip does not, however, have any way of analyzing the video to determine whether it should be in film mode or not, and that's where the gmAFMC comes in. It uses data provided by the gmVLX1A-X to figure out if there is a pulldown sequence coming in, and switches the main chip on the fly between video and film modes. Not all players use both chips. Some players use their own strategy for deciding when to be in film mode and when to be in video mode, so they forgo the gmAFMC. We'll mention that in the individual player reviews.
But most players use the two chips as a set. The video mode on the Genesis is not motion-adaptive as we have defined the term. It uses "vertical-temporal filtering" which appears to us to be a slightly more advanced version of the vertical filtering we mentioned above, with the current field providing more of the input to the finished frame than the next or previous fields. It switches to a simple weave algorithm when the image is still or near-still.
Overall, its video deinterlacing is poor, but watchable. The Genesis chipset offers lots of useful options, including scaling 4x3 input to 16x9 output also called aspect ratio control , and scaling any input signal to a wide range of output resolutions. These features were designed for the scaler market, but some DVD players are starting to make use of them for scaling and zoom features, and doing things like automatically scaling letterboxed movies to make them appear to be anamorphic to the TV. The main Achilles heel of the Genesis chipset is that it doesn't buffer more than one field, so it can't look ahead to see cadence breaks coming in advance.
Accordingly, the Genesis is almost guaranteed to comb on pattern breaks. By the time it knows the cadence has broken, it has already sent out at least one bad frame. In addition, the Genesis is not very good at identifying or handling cadence, which comes up more than you might think. The players that use the Genesis chipset don't all have exactly the same deinterlacing performance.
There are tradeoffs that can be made by adjusting the sensitivity of the gmAFMC chip to cadence mismatches, and that causes some players to do better than others on certain kinds of material. Often, though, on other material the advantage is flip-flopped. There is no perfect setting.
Here is the SiI product brief. These chips have essentially identical deinterlacing performance, but the uses less power and is cheaper to implement. While the bulk of the work is done by the SiI, Silicon Image also recommends an off-the-shelf programmable coprocessor PIC microcontroller be used as well, running their proprietary MC software.
This adds more sophisticated cadence analysis, and improves the handling of complicated material immensely. Almost all DVD players use this extra chip. The Silicon Image chip buffers 4 fields at all times, which it uses for cadence analysis as well as motion analysis in video mode. As a result, it is much better than the Genesis at handling cadence problems and combs very little.
Its video-mode deinterlacing algorithm is motion-adaptive, and is substantially better than the one on the Genesis. It handles cadence with no trouble. Overall, it's one of the best performing deinterlacers on the market, comparable only to much more expensive products from Faroudja, and a few others. One disadvantage of the 4 field buffer is that it adds delay to the video output. If the audio is not delayed the same amount as the video, lip sync problems can result. The chip offers a special digital audio pass-through that will delay the audio the appropriate amount so it synchs up with the video, but very few companies use it.
The video deinterlacing on the Silicon Image is motion-adaptive on a pixel-by-pixel basis, which is very good. The Silicon Image has some edge detection and smoothing algorithms, but they are not as complex as the ones on the Genesis chip. Overall, however, the Silicon Image does an excellent job deinterlacing video material. A feature of the Silicon Image processor is that it has a "film bias" mode, which will raise the likelihood of material being recognized as film. Of course, this also causes it to falsely recognize video as film more often and produces more combing on difficult material.
The engineers at Silicon Image told us that "film bias" mode actually reduces the deinterlacing performance of the chip, because their film detection was about as well tuned as they could get it with the factory settings. Amazingly enough, several customers wanted the feature, so they could turn it on in their product. Perhaps they decided their customers never watched video or other difficult material, only film. Here is the NDV product brief. Since everything is on one chip, it's easy for the deinterlacing logic to take advantage of the flags in the MPEG stream, and in fact the chip does make use of the flags almost exclusively.
This works well when the encoding and flags are fairly standard, but video-sourced material and material with non-standard encoding or incorrect flags are not handled as well. One thing this chipset does incredibly well is scaling. It has a multi-tap scaler that produces smooth results at essentially any zoom ratio. Not all manufacturers make full use of this feature, but when they do the quality is very high. The National's video deinterlacing algorithm is not motion-adaptive, and uses either single-field interpolation or vertical filtering, depending on user selection and possibly the amount of change from field to field.
This is a fairly watchable solution under some circumstances, but if there is a lot of motion in the frame, it starts to look very odd. It also looks stuttery during pans and zooms. If you watch a lot of video material, or material with odd encoding, or really, anything other than major Hollywood releases, this chipset is not recommended. Here is the Ziva 5 product brief. Its deinterlacing is also flag-based, so it suffers from most of the same problems as the Mediamatics chip. The flag reading is more selective than some other chips, in that it looks for specific patterns of flags that it recognizes as film, and treats all others as video.
This is a small help, but no substitute for image-based cadence analysis. When the flags are correct and marked progressive, the picture is good. Otherwise, the picture is not so good. Like the Mediamatics, it has a good scaling engine, though it's not always used by DVD players to best advantage. We were happy that the Ziva fixed the chroma upsampling error on film material that previous Ziva chips had, but were unhappy that it does not fix the error on material.
In addition, the chip goes into video deinterlacing mode on all material, even if the progressive flag is set. Given the amount of film material, especially in Europe, we find this disappointing and puzzling. The chip is completely software upgradeable, so we hope a fix will be forthcoming. In addition, we were shocked to note that the entire chroma channel shifts at least one entire scan line downward when the chip is in interlaced upsampling mode.
This is a fundamental bug in the implementation, and one that we pointed out when the first Ziva5 player came to market a few months back, so we don't see any reason for it to still be there. Again, we really hope a fix is in their future. Here is the Vaddis 5 product brief. Unlike the other two, it has a per-field motion-adaptive deinterlacer, and a cadence analysis algorithm that allows it to do better on oddly flagged material.
Like the other two chips, it has excellent scaling capabilities, though not all players make good use of them. Unfortunately, the chip still has the chroma upsampling bug, and it's just as bad as in previous Zoran chips. Thus we can't really recommend players that use this chip. It may be possible to fix this with a firmware update, but this is just speculation on our part. If it is, we encourage Zoran to put together a fix because in most other ways this chip is a good performer. Here is the CS product brief. This is a new as of December, all-in-one chip like the abovementioned.
LSI, Mediamatics, and Zoran chips. It has all the same problems as the others, and like the others it has an excellent scaling engine. This may just be an implementation detail, as we've only seen this chip in action in a single player.
Making the best quality dvd possible
This chip has no chroma bug on normal or material, though it is tripped up by the common problem with alternating progressive flag, as on Monsters Inc. Here is the FLI product brief. It's a very capable chip, and like the Silicon Image chip does both cadence detection and deinterlacing on a single chip. It doesn't do any scaling, but can be combined with a scaling chip if needed. The Genesis chips go further than other deinterlacers by adding several interesting video enhancement and processing functions. It has a cross-color suppressor that reduces artifacts caused by composite video mastering.
Most good quality DVD movies don't need this function, but titles mastered directly from composite master tapes will look significantly better. It also has a chroma filtering feature that smoothes out the chroma channel and reduces artifacts. This function also has the effect of almost completely hiding the effects of the chroma upsampling bug we talk about later in the report. The filtering of the chroma channel is twofold, and one aspect is good while the other is not so good.
The good part is that the chroma channel is run through a vertical lowpass filter to reduce the chroma resolution to vertical lines. This is fine on DVD, where chroma is never more than lines anyway, and really a vertical lowpass filter is a good way to get rid of 4: The not so good part is that the chroma channel is actually being interpolated from a single field, even when the chip is in film mode. That drops the effective resolution of chroma on DVD to a mere lines. Amazingly enough, it still looks pretty nice.
The eye is just not terrifically sensitive to chroma. Still, we would love to see newer versions of the chip keep as much of the chroma detail as possible. The vertical lowpass filter is only turned on when the CCS cross-color-suppression filter is turned off. Cross color artifacts are specific to composite video, so this filter was more appropriate in the laserdisc era. Most DVDs are not composite video, so we favor turning the CCS filter off , to get the benefits of the low-pass filter.
If the low-pass filter is not on because the CCS is on , you still get some chroma bug hiding because of the single-field interpolation of chroma. Good, it wasn't clear to us at first. Faroudja apparently has some secret technical reasons for making the CCS and low-pass filter mutually exclusive. The reason for the single-field interpolation appears to be cost savings. In Faroudja's defense, it's been there for years, since the very first Faroudja processors, so it's taken a long, long time for someone to notice it.
This really helps hide the most obvious artifacts when deinterlacing moving video material. It makes it difficult to tell when you have dropped out of film into video mode. No matter how good a deinterlacer is, sometimes it must scale up all or part of a single field to a full frame, which produces ugly stairsteps on diagonal lines.
It is important to note that it only helps with video mode deinterlacing, which is used on video sourced material, and during those times when the deinterlacer can't get a good lock on the cadence on film-sourced material. When a deinterlacer scales a line field to a line frame, another word for that process is upsampling, because it uses lines worth of input samples or pixels to create lines of output samples.
Hence the number of samples is going up. To do this, each pixel of the line output is created by applying a weighted average of several of the input pixels. Under normal circumstances, those input pixels will be the ones just above and below the output pixel's location. In other words, the sampling angle is completely vertical or 90 degrees. When creating an output pixel, the algorithm looks at a small local patch of input pixels, and looks for a strong diagonal contour. If there is one, then the sampling direction is set to be perpendicular to the local contour. For example, if the algorithm determines that there is a degree diagonal line running through the pixel in question, then the input samples will be gathered along a diagonal line that crosses the line in the image at a right angle or 0.
When there is no easily identifiable contour, the algorithm falls back on the standard angle of 90 0. The result of all this math is a much smoother image, with fewer annoying jagged edges. It doesn't necessarily look exactly like the "true" image that you'd see if the source were higher resolution, because the algorithm can't magically recreate details that aren't there in the source, but it does represent a better interpolation of the image, more like what a human might do if asked to smooth out the image by hand. It's also possible to see artifacts at times where the algorithm looks worse than the simpler strategy for example the resolution wedge on the WHQL disc , but those are few and far between.
But wait, there's more! It also has frame rate conversion and noise reduction. OK, maybe not the food processors. We have, as of this writing, not seen any released shipped units that use this chip, but what we've seen in prototypes has been impressive. And you thought we would never have anything nice to write about Genesis. This is a brand-new series of chips from Trident, who is best known as a manufacturer of PC video cards.
They've come out of the gate with a set of chips that all have similar basic functionality, but are customized for different applications. First off, the deinterlacing was superb. The chip sailed through our deinterlacing tests and passed every single one. We dragged out a few alternate torture tests that have tripped up the Silicon Image in the past, and it passed those as well.
This is the first chip that has challenged the Silicon Image and Faroudja chips and come away unscathed.
Secondly, the video decoder was also superb, at least in terms of capturing the full resolution of the signal. We're used to ADCs in external deinterlacers losing at least TVL of the signal, but this chip had no obvious visible effects on the resolution pattern in Avia. That's amazingly good, and unexpected. The internal scaling was excellent as well, among the best we've seen. There was no aliasing on the resolution wedges when the signal was scaled to higher resolutions like x or p. Clearly they are using a good quality multi-tap FIR filter or something similar.
We hope to see more of this chip in players , and better implementations, because the scaler we looked at had some other serious video problems hot voltage levels and chroma delay some of which may have been caused by the chip, or may have been just a poor implementation. Still, what we saw was very exciting, and clearly Silicon Image and Genesis need to watch their backs. From a consumer standpoint, a new high-quality chipset in the market should produce more and better inexpensive progressive DVD players, and better internal deinterlacing on display devices, both of which are good things.
These two chipsets are clearly the best available right now for progressive DVD applications, and the question often comes up of which is better. Unfortunately, it's not a simple answer. For video-sourced material, the Genesis has the better picture with fewer obvious jagged artifacts. It's not a panacea, and it can't make video look like film, but it does improve things.
That said, the Silicon Image video deinterlacing is very good as well. But we'd have to give Genesis the edge here. In film deinterlacing, the two are also very good, though they each have their strengths and weaknesses. The Silicon Image is better at staying in film mode when there are hiccups in the cadence, which is good because is maintains the full resolution as much as possible. The Genesis chip's tends to go into video mode more often and stay there longer.
Welcome Secrets of Home Theater and High Fidelity
Because the Genesis chip goes into video mode more often, it is slightly less susceptible to combing, because at the first indication of combing, it immediately goes into video mode, and stays there until the cadence is clean again. On the other hand, during the time it's in video mode, the viewer is losing much of the fine detail in the scene. We found quite a few films that tripped up the Genesis, even though the cadence was fine.
On most of these, the Silicon Image would stay in film mode the whole time, but the Genesis would drop in and out of film mode. We've done extensive testing on the pulldown performance of the two chips, and it's clear that the Silicon Image is much better at detecting pulldown and staying in film mode with material. On the flip side, though, the Silicon Image combs more often on video material, because it more often incorrectly identifies video as progressive. This just serves to highlight how difficult it is to always 'do the right thing' in deinterlacing.
Still, much of this is splitting hairs. Both of these chips are excellent, and in our view represent what you should expect from a progressive DVD player. Either one of them will deinterlace the vast majority of discs with no noticeable artifacts. If you watch a lot of video sourced material, or you want the chroma bug hiding or cross-color hiding features, the Genesis is probably a notch better. But for film sources, it's a tossup, with perhaps a slight edge to the Silicon Image. Before we get into specific problems, be aware that progressive DVD players can be divided into two basic types: Flag-reading players use the flags embedded in the MPEG stream to make decisions about what deinterlacing algorithm to use.
Cadence-reading players ignore the flags, and analyze the content of the frames as they come out of the MPEG decoder to figure out which algorithms to use. Flag-reading players have two major difficulties: This is watchable, but the main reason to get a progressive DVD player is to get film-mode deinterlacing. You can get video-mode deinterlacing from the deinterlacer in the TV.
I am having the same experience. So explain to me how Hollywood continually puts their very high quality film products onto a regular DVD, they are more often than not OVER 2 hours in length, and they look fantastic? How does that work? I want to make that same kind of product.
I have a feeling it's a special disk, and not a DL either. So, many factors are involved. Making the best quality dvd possible COW Forums: Making the best quality dvd possible by Vasja Mihelcic on Jun 22, at 6: Return to posts index. Making the best quality dvd possible by Vasja Mihelcic on Jun 22, at 7: Making the best quality dvd possible by Filip Kubis on Jun 22, at 8: Making the best quality dvd possible by Vasja Mihelcic on Jun 22, at Making the best quality dvd possible by Michael Slowe on Jun 22, at Making the best quality dvd possible by eric pautsch on Jun 22, at Making the best quality dvd possible by Phil Bates on Jun 29, at Making the best quality dvd possible by Michael Slowe on Jun 30, at 3: Making the best quality dvd possible by Phil Bates on Jun 30, at 3: Making the best quality dvd possible by Dave Haynie on Jun 24, at 8: Making the best quality dvd possible by James Reeve on Jul 28, at Making the best quality dvd possible by Dave Haynie on Jul 29, at 2: Making the best quality dvd possible by Rob Robertson on Sep 30, at 4: Making the best quality dvd possible by Morten Kristiansen on May 14, at Making the best quality dvd possible by Steve Curtis on May 18, at 2: Making the best quality dvd possible by Jeff Pulera on May 26, at 2: When laid out like it is in the graphic above, it becomes obvious that you can easily stack the pixels supplied by old standard definition content multiple times in the same display space provided by HDTV sets.
You can, albeit with mixed results. Video looks best when the source video is displayed on a screen that shares its native resolution. With the advent of HDTV sets, however, it became necessary to make old content fit newer screens. This is where video scaling comes in. Nobody wants to watch standard definition reruns of the X-Files, for example, where the screen looks like this. Instead what happens is that a video scaling algorithm is employed and that p video is expanded to fill the screen.
All modern HDTV sets perform video scaling, in this application known as upscaling as they are taking a small image and scaling it up to a large one. Whatever video input they get, they scale up to their native resolution 1, x in the case of resolutions sets. Most sets have lackluster scaling algorithms and do a poor job enlarging the image.