Image scaling

Scaling is also known as Resize. Sometimes, Resample is even called scaling, which is not entirely unreasonable. Image scaling refers to the computer graphics process which increase or decrease the size of a digital image. As a matter of fact, an image can be easily scaled by an image viewer or editing software. In addition, an image can also be scaled automatically by a program, which can greatly help the image to fit into an area of different size without efforts. People can make use of a lot of methods in order to reduce an image. However, the most popular way adopted by people is a type of sampling called undersampling, which can help to maintain the original quality. It is more complicated to enlarge an image as there is larger area to be filled with more pixels. 

Generally speaking, scaling is a non-trivial process which involves a trade-off between efficiency, smoothness and sharpness. With bitmap graphics, the pixels forming the image will become more and more visible when the size of the image is being reduced or enlarged, which can make the image seem “soft” if the pixels are averaged or jagged. 

In effect, scaling will not change the image pixels in any way. To be honest, scaling will only change the single number of dpi (ppi), a number simply stored separately in the image file in an arbitrary way. The only equipment makes use of it is the printer. What’s more, it only changes the size the image will print on paper. The images on the computer screen will not be influence by the number at all. Actually, the camera can never know how people want to print the image, and thus just makes up a number. As a result, people are supposed to fix the number before printing anything.

About Black Bars

Generally speaking, there are some common problems associated with aspect ratio. For example, the video seems stretched horizontally or vertically when you play it on your DVD player. Fortunately, such problem can be easily handled by tweaking hardware settings. Users just need to configure the DVD player or TV set to the correct aspect ratio in order to make the video as normal. 

In addition, there is one problem related with aspect ratio. People will encounter with it from time to time when playing video. How to remove the black bars from the edges of the video when you are playing the video? 

As a matter of fact, black bars often happen when a widescreen video (16:9) is converted to 4:3 if the user is making use of the Letter Box resize method. To be honest, the proportions of the image have nothing uncommon, but there are black bars existing on the top and bottom of the video. If you are watching video on a widescreen TV, then this problem will lead you to feel a little bit uncomfortable and distract you to a large extent. The TV adds its own bars to the sides of the video to display it on the screen. 

On the other hand, black bars will appear on the left and right sides of the video image when a 4:3 video is converted to 16:9 using Letter Box resize method. In this case, users can make use of the Crop resize method in order to get rid of the unwanted black bars and return the video to 4:3 aspect ratio eventually. Moreover, people can also remove the unwanted black bars by adopting some professional software designed to perform some very powerful functions. In the end, the video will be displayed normally and people can enjoy watching it without being distracted by the black bars.

Hardware Compression vs. Software Compression

Software compression is more well-known among people when compared with hardware compression. The reason is that the majority of people do not have need to apply hardware compression in their daily life. In contrast, software compression does meet their requirements in many aspects. Generally speaking, software compression is cheaper and more easily accessible solution compared with hardware compression. On the other hand, hardware compression demands specialized equipment which is designed to deal with specific workload. Even though hardware compression costs users more, it does have its own advantages over software compression. First of all, the specialized hardware enables hardware compression to be faster than software compression to a great degree. Software compression just requires a general purpose processor in order to perform its job. Secondly, hardware compression will not cause any extra burden to the host processor because its calculations take place within its own hardware. Software compression cannot make it. Under most circumstances, software compression is likely to degrade the performance of the host during heavy use and other operations. If you are compressing a large amount of data while using your computer at the same time, then this may pose a great threat.

There is no doubt that software compression has advantages as well. First, it costs less. Second, software compression offers users a lot of options to choose from. Users are allowed to control the process of how the data is archived, compressed and formatted. In comparison, users are given very few, or no options at all with hardware compression. Users have no say in the process of how the data is compressed before being stored into the media. Everything has been pre-programmed into the hardware by the manufacturer.

In conclusion, software compression is better if you are going to store compressed data for a long time. Hardware compression is usually device specific, which may cause great problems when your device fails without anything can replace it.

Hardware Compression

Generally speaking, the hardware compression is performed on the data path level. As a matter of fact, the hardware compression is only available for the data path which directs data to tape libraries. Under this circumstance, the uncompressed data will be sent from the client computer to the media through the data path. Therefore, the data will be compressed by the tape drive hardware before being written to the media.

There is no doubt that hardware compression is faster than software compression most of the time. The reason is that hardware compression is operated by dedicated circuitry. As a result, the hardware compression is ideal for direct-connect configurations in particular where the subclient and MediaAgent are hosted by the same physical computer. In such configurations, the drives are able to compress the data at the same rate as it is sent by the subclient as there are no network bottlenecks throttling the data transmission to the media drives. On the other hand, the hardware compression can boost the virtual capacity of the tape but also the performance of the data protection thanks to the tape storing more data per unit with higher operation speed. 

However, the problem is that hardware compression is not supported by disk library. It is only applicable for tape libraries.

If the data secured by data protection operations must compete with other data for network bandwidth, then hardware compression may be not that useful. When the network is congested, the tape drives will be starved for data for the data cannot be supplied quickly enough. Under such circumstance, the drives can compress as well, but the drives are likely to stop and restart the media in order to wait for more available data. Therefore, the compression performance may not be very ideal, which may lead to some potential problems.

Software Compression

Client Compression

The client compression is specified on the subclient level for most agents. Generally speaking, it is available for all storage media. If people adopt this way, then the data on the client computer will be compressed using the compression software. Next, the compressed data will be sent to the MediaAgent that directs it to the storage media in turn. When the client and MediaAgent reside on separate computers and the client has no choice but to send the data through a network, then client compression will be of great use and convenience for it can reduce the network load to a large extent.

Replication Compression

Replicated data can be compressed between the source and the destination computer. If the compression is enabled, then the data will be compressed on the source computer, replicated across the network to the destination computer and uncompressed on the destination computer, which can reduce the workload of the network to a rather large degree. As a matter fact, replication compression is specified on the Replication Set level and applies to all of its Replication Pairs. Therefore, people have the capability to enable or disable client compression between the source and destination machines for a given Replication.

MediaAgent Compression

The MediaAgent compression is specified on the subclient level for most clients. If the data path does not have hardware compression enabled, then people have the capability to enable or disable MediaAgent compression for a given subclient or instance as appropriate. 

Actually, MediaAgent compression is available for all storage media. The data will eb compressed on the MediaAgent using compression software in the MediaAgent. Afterwards, the compressed data will be sent from the MediaAgent to the storage media. When the MediaAgent software resides on a computer more powerful than the client computer, then the MediaAgent Compression can be of great use and convenience.

Data Compression

Data compression options are provided for data which is secured by data protection operations. Generally speaking, compression is a very useful way to reduce the quantity of data sent to storage, which will double the effective capacity of the media in return (depending on the nature of the data). Moreover, the system will automatically decompress the data and restore it to the original state when the data is later restored or recovered.

The following data compression options are provided: software compression and hardware compression. Software compression offers users to compress data in the Client and MediaAgent while hardware compression for libraries with tape media at the individual data path. In addition, as compressed data often increases in size if it is again subjected to compression, the system will apply one type of compression for a given data protection operation. Therefore, users are able to redefine the compression type at any time without damaging the ability to restore/recover the data.

If the hardware compression is available and applied, then it will have some kind of priority over the other compression selections. Whenever hardware compression has been enabled for a data path, all data conducted through the data path will be compressed in the way of hardware compression automatically. Otherwise, the data will be dealt with in accordance with the software compression selection of each subclient which backs up to the data path. Under such circumstances, people are able to choose from the following options: Client compression, MediaAgent compression, or no compression.

Last but not least, bear in mind that hardware compression is not applicable for disk libraries. As a result, the software compression selection for subclient is adopted by people for data paths which are associated with disk libraries. It is advisable to have a good understanding of such knowledge before compressing data.

Different Effects for Different Parts of an Image

It seems that brightness, contrast, saturation and sharpness are the four simplest controls of image. On the surface, they are four mutually exclusive controls. However, as a matter of fact, they are related to each other and intertwined in a way to such a degree that the change of any one of them will lead to rather complicated effects in the image in terms of the other three controls. Only when users have mastered a rather good knowledge about how these four controls are related and how to make use of them in a harmonious way can users achieve the desired image effects in the end. It is wise for users to think twice about what they really want to accomplish before taking actions to change the brightness, contrast, saturation and sharpness, whether to increase or to reduce them.

Generally speaking, the overall effect of brightness, contrast, saturation and sharpness varies according to different contents in the photo. Take the increase of contrast as an example. With the increase of contrast, the shadow will be darker while the highlight will be brighter. However, if most details in the photo are very bright, for instance, an overexposed sunset, then we will end up with less contrast with the increase of its contrast. The reason is that there is no shadow in the photo at all, which means that the separation of shadows and highlights in an image containing only highlights will just compress the highlights. As a result, the image will get less contrasty. We can safely come to the conclusion that it is vital to have a good understanding about how these four simplest controls affect each other and how they work in a harmonious way. It is a bit of art to use brightness, contrast, saturation and sharpness to achieve a balance.