Hi,

Like many people, I want to convert my Cd's collection into MP3 files, using the
Lame combining with the Razorlame GUI. But, before beginning, I ask myself
some questions; one of this is the following : has this a sense, when converting
Wav files previously extracted from a CD, to resampling the extracted files in
48 Hz in place of conserving them in their original 44,1 Khz?
Logically no, we cannot get more precision that this giving originnaly from the
CD; but, i try, and I obtain a file that has not the same frequencies response that
this obtained with conserving the original 44,1Khz sampling rate of the CD;
the difference is basically in middle and high frequencies though being soft.
Who can me explain that fact??

Thanks,

Deuterium

no, it doesn't make sense. most lossy codecs (that try to achieve transparency) are tuned to 44.1 anyways.


later

ABX test and get back to us.

Don't use razorlame either. Use the binary and rip with EAC or use LameDrop. That way you can use the presets.

I use razorlame all the time for the pretty output - just specify --preset standard under custom settings

Yeah, razorlame has no problem using the presets.

I don't like these "preset" self-made algorithms, because each of them always
lose a part of information, especially in the high frequencies.
I personnally work with 320 kb/sec, in full bandwith, no low/high pass filters
allowed and no problems at all; sounds exactly like the original CD's. but my
previous question was : can a higher sample rate (saying 48khz) still sounds
better??

Deuterium

I think you've joined the wrong forum Deuterium.

Hydrogen audio is very strong on facts, science, and very carefully carried out listening tests.

Resampling to 48kHz loses a little data due to rounding errors, and doesn't add anything real to the signal. However, some sound cards resample everything to 48kHz badly internally, so doing it well yourself can improve the sound quality.

xmixahlx is right that lossy codecs like mp3 are tuned at 44.1kHz - the performance at 48kHz is not nearly as well optimised or tested.

If you read the FAQ, you'll find out why a lowpass filter before a lossy codec is a good thing. Even at 320kbps, you're thowing away 3/4 of the signal - it's daft to force the codec to keep parts that you can't hear in preference to parts that you can!

I can't argue against your desire to use 320kbps for the best possible quality, but you might want to check you can hear the difference between that and a preset which typically gives a much lower bitrate, and has been tested extensively and shown to sound identical to the CD in 99.999% of cases.

You know that even 320kbps mp3 is going to sound different from the CD in some very rare cases, don't you?

You are, of course, welcome to ask any questions you like - but you might make yourself look a little silly if you don't read at least part of the FAQ and the Terms Of Service first. The mp3 section of the FAQ (see the button, FAQ, top right of the page?) is a great place to start.

Hope this helps.

Cheers,
David.

well, you are going from a 44.1 source, so... i'd say the answer to "can a higher sample rate (saying 48khz) still sounds better??" is NO.

if you are wanting 320cbr, then your answer is "use --preset insane".


later

Please God, make it stop.

Suck your God and other all supports for crimes against humanity, like
all the affected people who seem to be the lot of this ...petrogenaudio site!
You won, I go away!!

Thank you.

Shakespeare hath spoken

Wow. Now we need to create a "petrogenaudio" webpage. Full of wonderful new ideas, like ABXing is the tool of the devil and real men only use 320kbps and hand tuned settings and that high end expensive oxygenated speaker wires are the only way to listen to music, and so forth...

We have it.. it's called the Recycle Bin.

Which is where this thread should probably head.

Who are we to say that resampling would make it sound worse? I think Deuterium is right, we should stop using all these arrogant presets. I suggest he start with my favorite command line for LAME: -b 32 --cbr -m s -q 9 -k

-b 32 : SACD is only 1 bit, and it sounds far better than 16 bit CDs, thus you clearly want to use the lowest bit rate.

--cbr : No annoying fluctuations in the bitrate that will otherwise comprimise the sound quality.

-m s : Mixing the left and right channels together? That can't possibly be good, so force stereo.

-q 9 : q for Quality, which we want lots of, thus the highest number.

-k : I can hear clearly up to 22 kHz, don't you dare filter out any of that music!

I find this greatly enhances the musicality in all cases.

Thanks for your support; sure, on this forum stay some guys which have problems
with their ego; maybe was I speaking with the ego, not with the man...
But you're not fully right : VBR will always be better than CBR or ABR and...
the best quality is q=0, not q=9, the lowest the best with this parameter!
For the other guys, I'm not an audiophile with golden wires and so on, but, for
respect for the artists, I will take out the best possible quality, and all that
presets always cut data's in the stream; I've still have a look to predifined
presets furnished with the Lame package : if you look to the "studio" VBR
preset, no passfilters at all, full bandwith...why to buy a Porsche for driving like
a turtle?

If you're limited to 320kbps, forcing it to keep in low and high frequencies will also force it to cut out other data, which will make more of an audible difference to the sound.
edit: and the post above yours was a joke

I'll make the pain go away now...
If you expect people to give you the answer you already decided on why ask?