Lossless compared with lossy data compression

When i first looked into this i was like wtf? too... but this is all to do with the data compression.

Lossless data compression is a class of data compression algorithms that allows the exact original data to be reconstructed from the compression data. This is the opposite to lossy data compression, which allows an approximation of the original data to be reconstructed, in exchange fore better compressions rates.

What's the advantages?

One advantage of the lossy methods over lossless methods is that in some cases the lossy method can be compressed smaller than its counterpart, importantly while still meeting the requirements of the application.

Lossy methods are most often used by data are intended for human interpretation where the mind can fill in the blanks or see past minor errors. Ideally lossy compression is transparent or imperceptible, otherwise you would notice it sounded bad and not use it. If you do notice an anomaly it is called a compression artifacts.

Audio can be compressed at 10:1 with imperceptible loss of quality. The compression rate in lossy compression is 5-6%, where as in lossless compression it is about 50-60% of the actual file.

Cheers Guys,

Matt Price

List of Lossy Formats (links to wikipedia)
AAC
ADPCM
ATRAC
Dolby AC-3
MP2
MP3
Musepack (based on Musicam)
Ogg Vorbis (noted for its lack of patent restrictions)
WMA

List of Lossless formats (links to wikipedia)
Free Lossless Audio Codec – FLAC
Apple Lossless – ALAC (Apple Lossless Audio Codec)
apt-X – Lossless
Adaptive Transform Acoustic Coding – ATRAC
Audio Lossless Coding – also known as MPEG-4 ALS
MPEG-4 SLS – also known as HD-AAC
Direct Stream Transfer – DST
Dolby TrueHD
DTS-HD Master Audio
Meridian Lossless Packing – MLP
Monkey's Audio – Monkey's Audio APE
OptimFROG
RealPlayer – RealAudio Lossless
Shorten – SHN
TTA – True Audio Lossless
WavPack – WavPack lossless
WMA Lossless – Windows Media Lossless

Dither and Quantization Error

Dither is an internationally applied form of noise used to randomize quantization error, preventing large scale patterns such as "banding" in images. 

Example of image "Banding"

Dither is routinely used to processing both digital audio and digital video data and usually one of the last stages in post production for compact discs. 

Quantization Error is the difference in the actual analog value and the quantized digital value. This is due to either rounding of digital views or truncation. The error is sometimes considered as an additional random signal called quantization noise because of its "stochastic" or non-deterministic behavior

Cheers Guys,

Matt Price

5.1 Surround Sound, Dolby or DTS?

This is section explaining what Dolby Digital 5.1 and DTS 5.1 are, along with a bit about why people think one is better than the other etc...

What is 5.1?
Refers to this speaker set up: (left front, left rear, right front, right rear, and center), plus a subwoofer channel (the .1 in 5.1)

What is Dolby Digital Surround Sound?
This is the most common format for surround sound on media such as movies. Its a discrete channel surround sound format because the output has been controlled to come from a variety of speakers, allowing a car to sound like it is moving across the screen etc....

What is DTS?
DTS (Digital Theater Systems) is a digital surround-sound system first introduced in theaters in 1993. DVDs encoded with a DTS soundtrack require a DVD player and stereo receiver equipped with DTS-processing capability. This is partly due to the DTS demands for more data space on a DVD (often sacrificing bonus features), but many believe the audio quality to be superior to that of Dolby Digital 5.1-channel surround sound.

What do people think?
I don't happen to be 'the people' but reading up and around the issue DTS has a higher data rate and so that roughly translates into 'better' sound. Where as many Dolby fans argue that low compression but higher data rate provides 'better sound'....

Ill need another 10 years to provide a better description of 'better sound' but many blogs start with the underdog against the giant, like in all fairy tales...

I should be pointed out that both Dolby Digital and DTS Digital Surround encoding schemes now have even higher sampling rate of 48 kHz at 20-bits per sample, thus yielding an even wider dynamic range between sound level extremes of approximately 120dB.

They both have to be compressed in some form to fit on the disk so that is always going to be an issue, raw data over efficiency is hard to prove when sound is subjective.

"Compression and bit-rate are not the only differences when comparing Dolby vs. DTS formats. For example, the added rear surround channel in Dolby's extended surround format 'Dolby Digital EX', is matrixed over the two left and right surrounds, rather than discrete; instead the DTS counterpart uses a discrete channel. This also explains why DTS ES (Extended Surround) can provide a more precise location for the rear-effects soundstage than the Dolby EX format." - Source here

"Both Dolby Digital and DTS audio are capable of achieving similar end results in delivering surround sound, even though the lower compression/higher bit-rate of DTS Digital Surround should theoretically yields apparent benefits in sound quality.

At the same time, one cannot ignore the fact that these two formats make use of different coding schemes and syntax to perceptually compress audio.

This means that efficiency in terms of data utilization between these two formats is different. Therefore, a Dolby vs. DTS direct comparison based solely on these formats raw bit rates cannot be taken as a measure of sound-quality.

Thus, while it is objectively possible to compare the resultant sound quality for the same audio format encoded at different bit rates, and therefore, to determine whether the same format in a moviehouse application sounds better or worse than in a consumer implementation in home entertainment, it is not so straightforward when dealing with different formats.

Rather, the reality is that for identically sourced audio content, it would be much easier for the listener during a Dolby vs. DTS 'blind' listening test to notice a change in sound quality when changing the playback equipment say between different brands, than when changing from a Dolby Digital to the DTS surround audio track." - Source here

So it is down to choice really and if it you set up your theatre, home cinema etc... then you might notice one suiting your needs better, Sadly this is going to be a never ending debate but at least you can argue both ways forever with your friends.

Thanks

Matt Price

Q+A: Sample rates explained....

Sample rate is the number of samples per unit of time taken from a continuous signal to make a discrete signal, the unit of sample rates are Hertz. We record a discrete signal because we only need lots of little samples of the signal to track how loud it is etc... So the higher the samples the more of these measurements are taken per second. The standard for recording dialogue is 48kHz (48,000 hz) or 44.1kHz and unless your playing on really nice speakers you really wont be able to tell the difference.

For some sound effects recording people record at 192kHz which has advantages for slowing effects down and manipulating them. Also higher sample rates help prevent aliasing, which refers to an effect that causes different signals to become indistinguishable 


also see: Bit Depth

Thanks,

- Matt Price

Q+A: Bit Depth and Bit Rate Explained....

This is a quick guide to bit depth, sample rate and recording volume.


Bit Depth
Bit depth describes the number of bits of information recorded for each sample. This directly corresponds to the resolution of each sample.


a "bit' is an abbreviation for a single binary digit, represented by a 0 or 1. A word in audio is a binary number with more than one digit. A 16-bit binary number for example could be 0111011010110010. The number of bits per word is simply how many digits there are in the corresponding number. Common bit depths are 8, 16 and 24 so the higher the quality is the higher the number. 


The bit depth in recording refers to how many steps the amplitude can be broken down into to be represented digitally. 
16 Bit = 65,536 Steps 
24 Bit = 16,777,216 Steps
You can clearly see that 24 Bit would be considered higher quality as it is able to plot smoother more accurate lines of amplitude.


The Bit Rate is how many of these bits are being transmitted or received per second. This is used for streaming purposes on videos as well. For playback, using the example of a standard CD with a datarate of 48kHz/24, meaning the sample rate was 48,000 times per second with bit depth of 24. The amount of audio data per second is worked out like this: 


48000 x 24 x 2 (for stereo recordings) = 2,304,000 bit/s or 2.3 Mbit/s.


The size of an audio file size can be calculated with a similar equation.The equation below is with the example of recording at 48khz at 24 bit depth, that plays out in stereo (so 2 channels, recording 5 you would x5) for 60 minutes which is 3,600 seconds and divide by 8, you then get this:


48000 x 24 x 2 x 3600 / 8 = 1036,800,000 Bytes or 1.036GB


So bit depth is your resolution for the information for each sample but many would argue there is no massive difference for recording dialogue between 16bit or 24bit but if you have the space why not.


See also: Sample Rate Explained


Thanks,


- Matt Price