Getting the Most out of Your Audio Gadget Storage - Let's Encode
(Page 4 of 5 )
Let’s Encode!
Now the time has come for us to encode. Finally! First, start up your chosen encoding application that supports HE-AAC v2 encoding and hopefully is powered by NeroAAC. Throughout this section, for the sake of the example, I am going to use dBpoweramp. The custom configuration specifics are quite similar with any other tool.
Note of warning: Keep in mind that HE-AAC is a relatively new audio compression scheme and it is not guaranteed that every gadget/cell phone has the capability to play back these formats. It’s not as popular as MP3s yet. If you face issues (funny noises included) while trying to play back the files, then chances are your gadget does not support the format. Try LC-AAC.
Check out the attached screen shot to notice the configuration I personally am satisfied with. Of course, this shouldn’t be considered to be anything other than my subjective opinion; it only serves the purpose of providing a simple guideline.
I’ve opted for 32kbps ABR (Average Bit-rate) with HE-AAC v2, therefore I’ve picked the “Force HE v2” option from that drop-box. Don’t forget to set your output location, as well as select “m4a Nero (AAC)” at the converting scheme. Eventually add your preferred DSP effects; in my example below the “Volume Normalize” is active.
You should try out several configuration schemes — for example, 32kbps ABR, 32kbps VBR, 48kbps ABR, 48kbps VBR, and eventually CBR for both bit-rates. You may likewise try going lower than 32kbps; the quality should become acceptable, but it depends on your equipment (gadget, headphone) as well as your “ear tolerance.”
All in all, even if you are a passionate audiophile with pretty high quality standards, I bet that you are going to be satisfied with the 48kbps bit-rate transcoding. Don’t forget that going higher than 48kbps while staying on HE-AAC v2 isn’t proven to yield better quality. The sweet ideal point is 48kbps for HE v2. But you can experiment!
The attached screen shot below is a real-world illustration of the gains of HE-AAC encoding. The original source file was approximately 3.84MB on MP3 128kbps format, 3 minutes and 40 seconds, 44100Hz Stereo. I’ve encoded this aforementioned file with HE-AAC v2 on ABR 32kbps and VBR circa 32kbps estimation. Check out the results!
We cannot leave without clarification of those bit-rate prefixes: VBR, ABR, CBR— variable, average, and constant bit-rates, respectively. CBR maintains a constant bit rate and therefore it is not optimal in case of low bit rates. It degrades quality because it “neglects” complex sections, which would require a higher bit rate. CBR is acceptable on really high bit rates like 256-320kbps because in these cases the ceiling is higher.
ABR works on the basis of arithmetic mean: it can contain high bit rate parts but also low bit rate ones, as long as they are equal to the required average. Therefore, this is preferred over CBR on lower bit rates because it is more efficient. Once run on your desired bit rate, the encoder tries to maintain that average, leaving a “ceiling edge” for particular sections that require high bit rates, but goes lower when it is possible.
The only drawback of accurate ABR transcoding is that the encoding process is slightly slower because its algorithm is based on a two-pass encoding. This is pretty much required; the encoder cannot know which parts call for a higher bit rate and which do not. First run “observes” and “analyzes” while the second run “does all the work.”
VBR stands for variable bit rate, therefore it also “varies” during a track. In fact, ABR is a derivative of VBR; they work on the same principle. To understand them best I am going to point out the differences. With ABR we specify the average bit rate and thus we can almost exactly predict the output file size.
With VBR we specify the required quality; encoders give us a rough estimation of the expected bit rate depending on the specified quality. In fact, the encoder is able to choose the optimum distribution of bits without being limited to maintaining an average. The algorithm is based on acoustic models and quality levels to determine where and how to place the bits. We cannot know precisely the output file size.
The rule of thumb generally goes like this: VBR > ABR > CBR. But don’t take this for granted because it is highly relative and dependent on other factors like bit rate. Some people like to use CBR because it preserves the most data, but still neither AAC nor MP3 are lossless encoding solutions. But the point of this article is becoming frugal and that’s why we care about file sizes and want the best quality-per-size.
Next: Final Words >>
More Hardware Guides Articles
More By Barzan "Tony" Antal
