Music File Formats

By jsalk

29
Oct
2015

PCM (Pulse Code Modulation)

With the advent of CD’s came a digital music standard called “RedBook.” RedBook is a PCM (pulse code modulation) standard that features a word length of 16 bits and a sampling rate of 44,100 samples per second. It was calculated that this combination could adequately reproduce frequencies up to 22,050 cycles per second with a dynamic range (the difference between the softest and loudest passages) of about 90dB (which was better than either audio tape or vinyl at the time). Since humans can only hear to about 20,000 cycles per second and less as they age, this was felt to be plenty sufficient for the task at hand. For many years, RedBook remained the go-to standard for digital recordings.

Starting in the late 90’s, however, recording studios began recording at higher resolutions, primarily 24/96. These recordings would then be “down-sampled” to RedBook in order to master CDs. It was felt the the better the original recording, the better the final product, even though the resolution had to be decreased in the process. In fact, many commercial recording studios continue to produce 24/94 recordings today, since many recording engineers believe that anything above that is not all that audible. That said, the march to higher and higher resolutions continues unabated. Today, it is not at all uncommon to see PCM (pulse code modulation) file resolutions from 24/192 up to 48/384. Can you hear the difference? Probably up to a point, but where that point is, is the subject of some debate.

PCM files can be stored in a variety of formats, such as uncompressed .wav files, or compressed file formats with names like .flac, .aiff, .mp3, etc. Other than file resolution (which impacts sound quality), is makes relatively little difference which you use. Player software can generally deal with all of these file types and produce a PCM bit stream to send to the DAC.

Up until the last few years, most all consumer DACs were strictly PCM-based. But to complicate the situation, the consumer audio industry has lately been heavily promoting a completely different format – DSD.

DSD (Direct Stream Digital)

In 1999, Philips and Sony jointly introduced an entirely new digital concept for recording and reproducing sound. They called it SACD (Super Audio CD). While the format never really caught on with consumers, the file format, DSD (Direct Stream Digital), did take hold with some in the recording industry and is now being promoted by the consumer audio industry. Rather than using a multi-bit sample like PCM, DSD is a single bit format utilizing extremely high sampling rates. The dynamic range (the difference between the softest and loudest signals the format is capable of resolving) is much higher than that of a CD (PCM). DSD, in use, has a dynamic range of about 105dB or greater from 20 – 20kHz compared to about 90dB for CDs. Human hearing is capable of resolving about 120 dB of dynamic range.

Can You Hear It?

When music is stored and played back on streaming devices, you are no longer limited to 16/44 file resolutions. As stated above, PCM files can have resolutions as high as 48/384. But can your hear the differences?

A report on double blind listening tests conducted in 2004 found that test subjects could not discern any difference in audio quality between DSD and 24-bit, 176.4 kHz PCM music files. Even so, today the audio industry is aggressively promoting DSD.

I had a conversation with a DAC manufacturer a while back. He told me he had no interest in producing a DSD DAC because studies have shown that listeners cannot hear the differences DSD brings to the table. My point to him is that while he may indeed be correct, his position was missing the point. If the industry is promoting it, consumers will come to the conclusion that DSD is audibly superior to PCM and will eliminate non-DSD DACs from consideration. It matters little if your DAC is a superior product in every other respect. Due to the industry’s promotional efforts, if it does not include DSD, a significant number of consumers will pass it by.

If there is any negative influence on the adoption of DSD by consumers, it is that there is currently a limited supply of original program material recorded in that format. Many that are available tend to fall into the classical and folk genre since these can easily be recorded with a single stereo microphone. When you purchase these recordings, you are essentially purchasing a bit-for-bit copy of the studio master tape. No intermediate processing is required.

But what about the vast library of existing music that was not originally recorded in DSD? A case can be made that DSD files created from original analog master tapes can at least deliver the full audio quality of those recordings. And since mastering techniques have improved over the years, it can be argued these files can actually sound superior to the originals because they are not subject to the limitations involved in the transfer to vinyl or audio tape (primarily compression related to the dynamic range capabilities of those media).

But most commercial recordings released after the late 1980’s were recorded originally in a digital format and limited to 16/44 or, at best, 24/96. Even so, you often see high resolution up-sampled re-releases of this material on sites offering high resolution downloads. While mastering techniques have improved over the years and a re-mastering may indeed increase audio quality, one could legitimately argue that the higher resolutions offer little in the way of increased audio quality. After all, you can’t capture additional information from a music file that does not contain it. All you can do is use computer algorithms to “synthesize” additional information based on the content of the original file. That being the case, it would certainly be legitimate to ask yourself if these new high-resolution versions of older digital music are really worth the extra cost, or are they just a marketing gimmick used to sell older material at an inflated price? You be the judge.

Conclusion

Music that is well recorded and expertly mastered at RedBook (CD) resolutions can be thoroughly wonderful in every regard. If you have a large selection of CD’s and simply want a more convenient and versatile way of playing them, then your streaming system does not need to be able to handle file resolutions that are any higher. Just streaming them to a more capable and polished DAC will improve the sound quality of your library.

If you are interested in higher resolutions, there may be no reason to have to deal with formats higher than 24/96 or 24/176, since studies have seemed to indicate that nothing above these resolutions is significantly audible.

But if you like to be on the cutting edge and want to future-proof your system, DSD is probably something you should include in your system capabilities since it appears that the audio industry will continue to aggressively promote the format.

The nice thing is, no matter which path you decide to take, you will most likely experience an increase in audio quality on your journey.

NEXT: Compression...

Return to Blog Topics