I have recently sent some vinyl away to be ripped to mp3. I have had a sample output file sent back (320kbps) and a plot of the spectrum looks like the attached image. As can be seen, the upper frequencies roll off around 16k. Is this not to be unexpected with a vinyl rip? Presumably a nefarious activity like high speed dubbing would have much worse results regarding the top end.
Is the cut off above 15 kHz really that important? Most human adults can't hear above 15 kHz anyway. I have difficulties to hear sounds above that threshold and it won't get any better.
Is the cut off above 15 kHz really that important?
Probably not. I should have included that the track sounds fine, at least to me.
Is there a reason that you chose mp3 rather than lossless?
As you may know,
MP3 is lossy and it does throw-away information. But, it tries to throw-away stuff you can't hear anyway (mostly sounds that are drown-out by other sounds) so it's optimized for the
best sound... It's not optimized for good-looking graphs. ;) In most cases, the MP3 will sound identical to the uncompressed original.
There are no 15kHz "notes", only harmonics and overtones (plus noise, and distortion). And although (if you are young) you may be able to hear up to 20kHz with "loud" test tones in a soundproof booth during a hearing test, with music the highest-frequency harmonics and overtones are usually drown-out by the music.
In general, MP3 does a lot of analysis figuring-out what can be thrown-away, but not as much (if any) at the highest frequencies. If you hear a compression artifact (difference between the MP3 and the uncompressed original) it's usually not the loss of high frequencies that you hear.
I have recently sent some vinyl away to be ripped to mp3.
So... MP3 isn't "bad", but some people like to keep a lossless archive. If you do this again, you may want to get WAVs, FLACs, or a CD (or whatever they offer) and then you can make your own MP3s (or other format) and with a lossless "original" you always have options in the future if you want to switch to another lossy or lossless format.
P.S.Noise reduction (if they used it) can sometimes introduce artifacts although it shouldn't cause a loss of high frequencies.
The real weak-link in this whole process is the vinyl record itself, not the MP3 compression.
I agree with DVDdoug. I'm fortunate enough to have a reasonable setup for ripping vinyl, although it is a labour-of-love...
I alway rip and achieve my vinyl (and CDs) in PCM Wave format, then convert to MP3 (CBR 320K) for metadata (tagging) and playing.
Moderation: Shameless plug removed. Contnued off-topic solicitation will not be tolerated and will result in a loss of posting privileges, if not banishment.
I alway rip and achieve my vinyl (and CDs) in PCM Wave format, then convert to MP3 (CBR 320K) for metadata (tagging) and playing.
*achieve should have been archive.
... As can be seen, the upper frequencies roll off around 16k. Is this not to be unexpected with a vinyl rip? ...
To answer your original question, MP3 encoders usually cut off the frequencies above about 15 to 17 KHz. As has been explained, MP3 encoders work by discarding sounds you're unlikely to hear, because they're masked ("drowned out") by other sounds. This almost always includes signals above 15 KHz or so. Cutting them off avoids having to do the processing required to encode them. Also, the space saved in the file can be better allocated to sounds you're more likely to hear. In the case of a vinyl rip, a significant amount of the signal above 15 KHz is likely to be non musical - record surface noise etc - so another good reason to discard it.
Many thanks for the replies. To elaborate a little, I am actually having the vinyls ripped to both FLAC (for exactly the reasons that DVDdoug states) and mp3. Due to the numbers involved they will produce the mp3s for nothing (which will save me some time).
I appreciate the comments about mp3 being lossy and the frequencies at which they cut off, however when I compare the audio spectrum with random mp3s of CD rips that I have done myself the CD rips seem to extend up to 20kHz (using LAME). I have attached some examples.
With the sample mp3 sent through I was surprised to see the upper frequency roll off so early and was curious as to whether this was typical of vinyl in general, or that some vinyls just don't extend beyond 15kHz, or it is a result of the ripping mechanism being employed in this instance.
I suspect the answer may be either one or both of the last two.
If anyone else is ripping their own vinyl to mp3 or to lossless then mp3 it would be interesting to know the frequency responses being achieved.
As originally stated, the mp3 sounds fine. I am not so young so would not be hearing the upper frequency anyway.
I am actually having the vinyls ripped to both FLAC (for exactly the reasons that DVDdoug states) and mp3...
...With the sample mp3 sent through I was surprised to see the upper frequency roll off so early and was curious as to whether this was typical of vinyl in general, or that some vinyls just don't extend beyond 15kHz,
Why not compare the FLAC to the MP3?
From what I've read, it was common to roll-off the extreme high and low frequencies (but there is still some surface noise and possibly some distortion). And to my ear, most records from the 60s & 70s (and earlier) were a bit "dull" sounding. Sometime around the disco era they seemed to get better and toward the end of the vinyl era records seemed to get more "consistent".
however when I compare the audio spectrum with random mp3s of CD rips that I have done myself the CD rips seem to extend up to 20kHz (using LAME). I have attached some examples.
There are quite a few options for LAME so you may not get identical results.
Why not compare the FLAC to the MP3?
When I have the accompanying FLAC file I will indeed do a comparison.
I would get details about the encoding process including the encoder used. I would never pay someone to rip my vinyl to mp3 or any other lossy format. Besides, there have been major improvements to modern lossy formats since the standard for mp3 decoding with all its limitations was created so long ago.
I am also going to call out @Stacker for the continued off-topic solicitation of his software. You have been publicly warned.
There's a distinctive blue cartridge called the "Ortofon DJ" with a reported frequency response that doesn't go above 16 kHz.
It's wouldn't surprise me if that is what has been used here.