Help - Search - Members - Calendar
Full Version: EZCD-DA best lame SQ settings?
Hydrogenaudio Forums > Lossy Audio Compression > MP3 > MP3 - General
icsterm
Hi,


I ask for an advise for setting my Easy CD-DA software for the best lame parameters in terms of SQ. My portable player supports FLAC, but it drains the battery much faster than mp3.
Here is a screenshot of my current settings.


I am not sure if leaving highpass and lowpass automatic or disabled. Any advise?
I chosed VBR0 because it's best, especially for portable players, and no real advantage over 320kbps.
pdq
Yes, leave lowpass and highpass automatic.

Most people here would agree that V0 is overkill for a portable player, but you should decide that for yourself after testing other values.
icsterm
QUOTE (pdq @ Oct 16 2009, 16:14) *
Yes, leave lowpass and highpass automatic.

Most people here would agree that V0 is overkill for a portable player, but you should decide that for yourself after testing other values.

I was concerned about mp3 codec cutting off the higher frequencies, making the output quality a lot worst than using flac. I use high-end IEM's on-the-go. I also played a couple of mp3's with Foobar, and i saw on the spectogram that the higher parts of the tracks have information erased (cut-offs). Disabling the filtering under settings can make the mp3 use full frequency levels? Sorry for my poor english.


Also i have saw OGG format don't cut-off frequencies at same bitrate as mp3. Can't this frequency cut-off filter be disabled to make the output perceived quality better?
pdq
You can raise the cutoff frequency, but this will not improve the perceived quality, and may actually make it worse. You should NOT use spectrograms to judge quality, only your own ears.
DVDdoug
QUOTE
I chosed VBR0 because it's best, especially for portable players, and no real advantage over 320kbps.
The main advantage of V0 is smaller file size. It will use high bitrates only when necessary, whereas 320kbps CBR will use 320 to encode silent parts or simple "easy to compress" passages, as well as complex passages where you might really need 320.

Apparently file size is not an issue for you, since you've been using FLAC. Go ahead and use V0 or 320kbps even if it is "overkill".

QUOTE
i saw on the spectogram that the higher parts of the tracks have information erased (cut-offs). Disabling the filtering under settings can make the mp3 use full frequency levels?

...Also i have saw OGG format don't cut-off frequencies at same bitrate as mp3. Can't this frequency cut-off filter be disabled to make the output perceived quality better?
These are all lossy compression schemes and they are all going to throw away data. All of these formats also rely on psychoacoustic algorithms to determine what data can be thrown-away without affecting sound quality. It's usually best to let the algorithm do the work instead of manually forcing it to do something.

AAC is supposed to be "more advanced" than MP3. (Probably OGG too, but I don't know much about OGG.) However, at high-enough bitrates all of these formats should sound identical to the original uncompressed file in a blind ABX test. So... you don't need a more advanced format unless you're trying to squeeze the best quality into the smallest file size.
kornchild2002
QUOTE (icsterm @ Oct 16 2009, 08:37) *
I use high-end IEM's on-the-go.


It doesn't matter. People often have the misconstruction that high quality headphones and/or high quality car stereo systems will allow them to discern between lossy and lossless more easily. The key here is that these are all used in portable environments. Granted, not everyone is going to stand next to a jackhammer while listening to their music. However, portable environments are less than optimal when it comes to discerning the performance of lossy encoders. Even with high quality equipment, you will likely have great difficulty in distinguishing between -V 4/3 Lame and the source lossless material.

Do yourself a favor and conduct blind ABX tests using the headphones that you would normally turn to when it comes to portable listening. Additionally, you need to answer this question: do you look at music or do you listen to it? Never, ever, ever, ever judge lossy performance by looking at frequency plots. I can use a decade old Xing mp3 encoder at 64kbps to produce a full spectrum plot that looks great. However, the sound quality would be horrendous especially when compared to something like -q0.35 (~130kbps VBR) with Nero AAC (that produces strange looking plots) or -V 5 Lame mp3. That is why we never judge lossy encoder performance by looking at spectrograms. They have their place in the audio community but it isn't for judging lossy audio quality.
icsterm
Thanks for helping with some clear answers. I guess the spectrogram won't show any real flaws the ear can hear, only some frightening frequency cut-offs that won't reflect the distinquished quality of the human brain. My Sansa player boosts almost double battery life with mp3 encoded files, while FLAC/ogg use the most cpu power from the device. Also, because mp3 files are smaller, i just tried to clear out what are the best quality settings produced by this type of encoder(Lame 3.98). V0 should be totally OK, because i really can't do critical listening on-the-go, while the perceived is at it's best. While on the desktop side where i keep my home equipment, i always keep lossless material for transcoding.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2009 Invision Power Services, Inc.