I just want to put this one out. It bothers me to see how careless many of you are when you are doing conversions and DRM removals. In yoour posts on this forum. I just tested removing a DRM off a WMA file and it did not sound as good. After taking a closer look I have to slap myself in the forehead. By pressing buttons and not paying attention I screwed up that tune. It occurred to me I might not be the only one just pushing buttons. I wanted to make anyone aware of conversion pitfalls. My suggestions are do not convert unless you really need to. Very few conversions are ‘clean’ you will most likely lose some quality or gain file size. Be careful about making sure you do not lose quality by mistake or add extra useless data by mistake. Try to maintain the same format during DRM removal. I warn you I am not an expert but I figure I know more than many of you about audio compression. If you have anything to add feel free to put in your two cents. There are 3 compression techniques I know of. Lossless – no information is lost but the file can be less than half that of a wave. Primitive – I could find little on this. It is a lossy compression. CDs were created to be able to approximate an analog signal. The format is a fixed bit rate of 12,000. There is room enough to store ultrasonic sounds. Analog stores them so the CD stores them. 60 minutes of silence is the same size as 60 minutes of complex music. Humans lose the ability to hear high notes due to damage and age. Hearing loss starts in our teens. By as early as 25 most persons have suffered a loss of at least 2,000 Hz. The tones at the limit are only faintly heard even though they are loud. Most of the data in a symphonic music CD is to reproduce tones no one can hear. It makes sense to ignore data hogs no one can hear if you want to conserve space. There are hypersonic effect studies that may. counterdict what I just stated. I contend the study was far from conclusive. See the bottom if you want to read dribble on the stupid experiment. I would not double the size of my files over the barely significant results of those experiments. Psychoacoustic compression – This is very advanced compression. This is also a lossy compression. Believe it or not, the brain does not distinctly identify every tone made by the over 300 strings in an orchestra. Studies were made as to what we actually hear when exposed to music. We don’t hear those strings at all during the boom of a base drum boom. We hear loud base noises over less loud high pitched tones. Loud base tone require very little data to reproduce verses all that string data that is washed out from the boom. Psychoacoustic compression removes tones we will not hear. Below is a detail description. http://en.wikipedia.org/wiki/Psychoacoustics Wav and some other lossless formats have no compression and are huge. Flac and Apple lossless use only lossless compression. CBR MP3 uses lossless and primitive compression to produce a fixed amount of data per a unit of time. It is the only lossy format in this overview that doesn’t use psychoacoustic compression. Advanced Systems Format (ASF) WMA, M4a and many others. These use all the compression techniques to produce a fixed amount of data per a unit of time. Many persons wrongly equate these with MP3 (CBRs) of the same bit rate. ASF can keep more high tones if psychoacoustic compression was more successful. At the lower bit rates, like 128 these will sound better to a trained ear. At the higher bit rates you probably can’t hear what was lost. I see a problem with these files. You have no clear idea of the music quality. You only know that is higher than an equivalent mp3 CBR file. Mp3 VBR files. These are called variable bit rate mp3s. They also use all the compressions. They differ from the ASF files in that the quality is constant but the bit rate fluctuates. I prefer this format because it delivers the most data compression but the quality is known and is constant. The drawback for these is the time sequence on the player will not be accurate. To know the quality you need a tag reader that is able to display the quality setting. Using the bit rate your guess will be at least as good as with the ASF files. The power of this compression other than you set the quality is when you have silence or a big boom you save data. With ASF you save more high tones that you can’t hear because there are none (silence) or they are under the boom. Much of the savings from psychoacoustic compression are not realized. What does this mean when you convert files or remover DRMs? Be conscious of converting advanced audio formats with primitive mp3s. Try to keep an ASF as an ASF. It is best to keep the same format when removing DRMs. If you must convert to an mp3 because of playability, use VBRs. Try to approximate the bit rate. For example, you are trying to convert a wma 128 to an mp3 vbr. You move the side watching the approximate BR until you find something close. You can get away with a bit less BR because the vbr is more compressed for the same quality. I use the next lower br. The hypersonic study: I contend the study was far from conclusive. The proof was barely significant and is very likely to be chance. When results are that close you always re-test. You publish the re-test to make your first test believable since they did not publish a re-test I can only assume the first test was due to chance. The experiment looked fair on the surface but muddled important factors which you only do if you don’t have a case. There was no published attempt to correlate hearing range with the increased alpha waves. That can only mean there was no correlation to anything meaningful. I have done enough experiments to know a ‘cooked experiment’ when I see one.
One point I would like to make about the hypersonic study. Not mapping the subject’s hearing range to alpha waves was very likely a cover up. The wide age range of the subjects used in the study would suggest a wide range of hearing range from 15-20 KHz. Since subjects do not hear the high end of their audible range unless the tones are loud. In actual music this are soft resonances of stringed instruments. It would have been VERY interesting to compare someone with an effective 13 KHz range was affected to 24 KHz verses someone with an 18 KHz range. It could be the same effect as persons claiming they can hear tones beyond the range of their ear buds. If you are going to listen to lossless on your mp3 player at least get ear buds that will create the ultrasonic sounds you can’t live without. Very few do since people can’t hear those tones anyway. Some buds do produce 22 KHz but I have not seen any that produce 24 KHz tones. At least the study used speakers that produced ultrasonic tones.
