Transferring Cassettes to PC

[Hugh Robjohns]

Yes, it was quite remarkable what they managed to get out of the cassette format, particularly through the 1980s and into the 90s.

[Tomás Mulcahy]

Some cassette decks incorporated DBXII NR which from memory was around 80db S/N requiring more care when setting transfer levels.

Yea, so just do the math. That equates to 13 bits. My Yamaha MT4X running at double speed (which gives a 3dB improvement) and dbx engaged claims 85dB. Still only 14 bits I had so many tapes I decided it was worth my while doing them at 16 bit via a 24 bit converter. YMMV.

On the subject of cassette sound quality, that Yamaha is probably the second best cassette 4 track ever made. Apparently the Marantz was better. The Yammy was definitely a lot better sounding than my Tascam Porta Two. Amazing technology for the time.

I jumped to Cubase VST as soon as I could and never looked back

[Tim Gillett]

Absolute best case for cassette with a top notch deck and NR engaged is around 60dB.

...

... My Yamaha MT4X running at double speed (which gives a 3dB improvement) and dbx engaged claims 85dB...

I said,

Some cassette decks incorporated DBXII NR which from memory was around 80db S/N requiring more care when setting transfer levels.

I did not say, "therefore 16 bits is not sufficient to capture that." Rather, I agreed with you.

[Tomás Mulcahy]

To be clear Tim, one is a standard cassette deck, the other is a double speed four track. It's safe to assume OP is referring to a standard cassette deck, not to a four track.

Regardless, the maths is sound and is all that is required to comprehend the procedure.

[Tim Gillett]

To be clear Tim, one is a standard cassette deck, the other is a double speed four track. It's safe to assume OP is referring to a standard cassette deck, not to a four track.

Tomas, you claimed,

Absolute best case for cassette with a top notch deck and NR engaged is around 60dB.

FYI, a number of standard cassette decks (stereo, 1.875ips) had dbx built in. Teac and Technics ran such models. dbx was not peculiar to double speed and/or four track decks.

Regardless, the maths is sound and (16 bits) is all that is required to comprehend the procedure.

You will notice that I did not say "and therefore 16 bits is inadequate to capture dbx cassette".
Rather I said,

Some cassette decks incorporated DBXII NR which from memory was around 80db S/N ................requiring more care when setting transfer levels.

(my emphasis)

[Hugh Robjohns]

While there is nothing wrong with the sums I'm still struggling to see any practical benefit to recording the cassette transfer at anything less than 24 bit.

Why make the job harder than it needs to be? What realistic benefit is there to transferring at 16 bits? I can't see any.

By all means store the 're-mastered' audio at 16 bits if you want to save some storage space and/or format for CD archive.

But for transferring cassettes with widely varying levels, 24 bits make it so easy to capture the loudest tracks with plenty of headroom, without any compromising the noise floor of quieter tracks. Set once and forget.

And since some re-mastering is inevitable, even if only some realignment and optimisation of levels, the saved interim project will have an even longer wordlength anyway.

[Tomás Mulcahy]

Hugh, I agree. In my case for an archive it represented a 44% saving on storage. YMMV.

Again Tim, the maths trumps everything including pedantry

[BWC]

I don't think it's pedantry, you were off by 20 - 25dB in your, rather assertive, claim. That warrants correction. He never disagreed with the math, or the conclusion that 16b is enough. Both he and Hugh are just saying that while 16b might be sufficient, 24b makes it easy, and there's no good reason to be stingy, and I agree. It's been a very long time since I was worried about conserving data storage space. 44% of? Enough to worry about?

[Tomás Mulcahy]

As I said, I agree with Hugh. Was simply sharing my experience on a particular project. As I said YMMV. It makes no sense to qualify a general statement about average cassette decks to account for uncommon exceptions. OP knows how and why to proceed and can make an informed choice. Let's move on!

[BWC]

It's only because you said, "absolute best case", not "average". But I agree, moving on...

[Tim Gillett]

...The recordings that I’m transferring are very inconsistent in level which makes it difficult to set the gain properly. Obviously this would be easier in 24 bit, but the file sizes would be bigger...

I agree that how to set record levels is not always obvious. If it was, there wouldn't be so many recordings made, even using excellent gear at 24 bits, which are noisy or distorted, but usually distorted! The great thing about transfer of say cassette recordings is there is a limit to how high the level on any cassette recording can be. If we set our DAW record level a few db's above that maximum possible level on any cassette there's normally no need to alter it. James touched on aligning the cassette deck to the signal on each tape. That is the adjustment to be made for best results. It can't be fixed in the DAW, as James mentioned.

[James Perrett]

If file size is an issue then you could use Flac files instead of .wav format files which will give you a 50% reduction in file size. Most of the software that I use can handle Flac files as easily as .wav files (my ancient copy of Adobe Audition is the main exception).

And I'll go along with Tim and repeat my suggestion to check the azimuth - using the correct azimuth for the tape is far more important than whether to use 16 or 24 bits.

[Folderol]

In the past, I've gone a far as setting the azimuth for best results for each track when I've suspected that they weren't all recorded at the same time, or even on the same machine!

Not something I ever want to do again!

[Tim Gillett]

It gets worse. Different azimuths, different tape speeds, different track formats, different NRs, all on the one tape! It underlines the limitations of mass, automated, or even just unattended tape digitization practices. But try and convince some managers of that...

[RichardT]

Making sure the deck is working properly is far more important than the bit depth! That said, I agree that 24 bits is the best option for recording the transfer.

[Tim Gillett]

Making sure the deck is working properly is far more important than the bit depth! ...

Indeed it is, not least because it is a large part of ensuring no damage either mechanical or magnetic is inflicted on the tape. For well preserved 50 year old tapes, their most dangerous time is likely when simply being played. But how do we achieve that in practice? Quality tape machines are likely to be decades old with few spare parts available (eg: the crucially important tape head) and even fewer skilled personnel to service them. How can the average person even test the machine's condition on all the relevant criteria? Specialised tools are also needed. This is another area where 'standards' are difficult in practice to achieve. So much easier to simply insist on the 'standard' of a 24 bit record depth. It looks impressive on paper.

[Hugh Robjohns]

Hmmm.... I think what you're trying to say is that the chosen 16/24 bit word-length doesn't determine the quality of the cassette transfer source, and that the mechanical condition and electrical alignment of the cassette replay machine are critical in extracting the maximum audio information (ie. sound quality) from the tape.

This is all true... although wandering into the realms of professional audio restoration rather than meeting the needs of someone with a casual amateur interest in transferring some vintage personal cassette recordings.

[Tim Gillett]

Yes and amateur or pro, the result can only be as good as the playback allows.

[Tomás Mulcahy]

I wrote a blog post on cassette transfer specifically for the casual amateur. In the hope that kerb has not been scared off/ bored by this debate