Hi there

I want to convert my old tapes into MP3's. Whats the best quality to convert them to in order to keep the original data?

I thought theres no point in doing them as 192Kbps coz Im assuming they wont be that good in the first place. I thought 128 is the best?

I do realise that the quality will differ depending on the quality of the cassette as it loses quality over time but whats the best I can get from a brand new tape?

Thanks.

As no cassette, not even metal or chrome, will produce anything relevant above 16 kHz, I'd say -V6 should generally suffice. So I'd suggest -V6 -q1 or -V6 -h.

But as always: just try out a couple of -V presets (I'd start out at -V3 (which is between --alt-preset medium and --alt-preset standard) and go down to, say, -V7) and listen if you can hear any difference between them.

Erm, what does the above translate to in English? Are you talking about using certain types of software? Whats the best thing you have in mind?

Now that I've noticed that this is only your second post here, and you obviously don't know what -V or --alt presets are (no offence, everybody has to start somewhere), I suggest you look into the stickies at the top of both the General and the Tech MP3 forums here, along with LAME's own command-line switches reference. Use the board's search function and simply read through the numerous relevant topics that've passed along. That should prevent you from having to ask any more basic usage knowledge questions.

The only way to preserve the original quality would be a conversion to a lossless format, such .wav, .flac, or .ape. An mp3 at any bitrate will degrade the sound quality, although it may not be noticable to you at the higher bitrates.

Well, in that case, he's degraded the quality by digitizing it. There will be frequencies above the Nyquist limit that are not kept when going digital, and there will be amplitudes that are not captured by any bit depth. The thing to remember is it that doesn't matter. If he can't hear a difference, no quality has been lost.

You should definitely do a listening test yourself to find what settings are acceptable. Take Polar's advice and do encodes starting from -V 3 and going up. When one setting sounds unacceptable, use the previous instead. (Of course, it's good to know what -V settings are and how to use them. Most people here assume you know how to use LAME.)

[edit:] Hey! This was my 100th post! Only 5605 before I overtake Roberto!

Badass is assuming that the physical quality of the cassette can be matched to a mp3 bitrate. I'm just trying to point out that a poor sounding original will sound even worse if an inferior bitrate is used.

Rather than choosing a specific bitrate for your tapes, I would personally error on the side of caution and let --preset standard (e.g., -V2 in LAME 3.95 or higher) do the deciding for you. You should end up with a pretty decent transfer if my experiences hold true. Make sure to use Dolby NR and the correct tape bias if those switches were used when recording the tape to begin with.

Now that I'm done with all my CD's, I decided to pull out all my old cassettes which have been forever buried in the very back of the storage room, at the very bottom of the box pile (yes, getting them out WAS a lot of work!). I'm curious if there are superior methods off noise reduction when doing cassettes. For example, turning off Dolby NR on the tape deck and instead doing it manually in a program like Adobe Audition or similar program which allows you to capture a noise profile. Or would this be too damaging to the sound file?

That's an interesting question! I've had some pretty yucky experiences with software-based NR, so I tend to stay away from it. Of course, I'm also old-fashioned and would rather just play the cassette back the same way I recorded it. So, it's conceivable that one could turn Dolby off and use software to correct the frequency response and remove noise, but I wonder if that would be worth all the extra work?

Certainly, proper tape bias will have to be set at the deck, or you have another set of problems: metal tape played with FeO2 or CrO2 bias settings sounds screechy and thin, at least to my ears.

Bias is only active when you record a tape, not when you read it.
Dolby must be on or off according to the state of the tape. If it was recorded recently, you might read with the same Dolby used during the recording. If the tapes are old, you'll get too much pumping effect in treble. Then leave it off.
See the FAQ for a thread about recording tapes onto digital.

For MP3, the compression artefacts are completely independant of the tape sound. You can hear them compressing any tape into MP3. So the recommended setting should be -V2.

Ok, but here's a question for you then: how come if I record a metal tape and play it back with normal FeO2 bias setting on the player, there is an increase in treble? The same seems to be true for recording a normal bias signal on a metal tape, like in a cheap portable recorder. The tape sounds thinner and has a treble boost compared to a metal tape recorded and played back with the correct settings.

In fact, my car tape player has this problem: tapes recorded at normal bias sometime comes up as metal, and sound dull (and this is without Dolby NR on).

The bias adjusts the premagnetisation current during the recording. It should be different for each tape. Cheapest devices have three positions : Type I, Type II, and Type IV. Normal devices have a fine tuning button that allows you to adjust the bias exactly for each tape model (it's not exactly the same for TDK SA-XS and Sony UX-S, for example, while they are both Type II). On high end recorders, a signal generator can adjust it automatically.

On playback, you have a completely different thing labeled "EQ", that can have only two positions : Type I, or Type II/IV.

Both settings, bias and EQ affect treble. Since you can't record one tape while playing back another, the same button is used for the two settings (when there is a button).

Thanks, Pio; that explanation makes perfect sense.