Topic: Nakamichi to Nagra unbalancing leads

[andromedanwarmachine]

Hmm...

this is definitely more 'setup' than 'microphones' question- but not a straight 'interconnects' query really I think...

I need to drag a load of material into my DAW from cassette. I had previously done this using my (doorstop) M-box as the A>D converter, but let's not mince words- that is pretty rubbish. It has occurred to me that a much better set of converters lie in my location recorder and as such I'm going to play to material into my LB and then just transfer the files across into my DAW.

My question is this; is it worth building a transformer stage to go between my Nakamichi DR3 O/Ps and the balanced line I/Ps on the LB?

Also- (and I know this is a massive topic- did it ever get reduced to a standard answer?!) what bit rate and sample frequency would people here classically use for cassette transfer into digital- I need to suck every last detail out of this material...

Thanks

JimP

[John Willett]

I would go from the Nakamichi to the Nagra LB directly, without adding the extra distortion that a transformer would introduce.

Make sure you clean and demagnetise the Nakamichi's heads before you start - and maybe do it again later (and again) if you have lots to transfer.

I would definitely record at 24-bits, but 44.1kHz (CD sample frequency) will be fine as increasing the sampling frequency only extends the top end response and as it's not on the cassette in the first place it's not worth doing at 96kHz.

I hope this helps.

[andromedanwarmachine]

Hello John,

thanks for this- I knew I could rely on you!

It's more of a techie question really, wondering whether a transformer stage will be kinder to unbalancing the LB inputs than just a straight pin 3 to pin 1 short, you know?

I've got proper 1+1:1+1 600/600 Ohm audio-grade Vigortronix trannies- I didn't think I'd get an issue with distortion actually; I wouldn't have thought it could get hot enough for distortion from -10dB output...?- can I ask you to expand on that concern?

And that's very helpful on the sample frequency information; thankyou...

JimP

[DSatz]

I'd disagree slightly with John W. here because the total dynamic range of the very best cassettes (e.g. pure metal tape with optimal recording levels and Dolby S or HiCom noise reduction) is in the range of ~65 dB at the utmost. That dynamic range would be more than covered by a 12-bit digital recording system, let alone 16-bit; by the time your recorder is 10 dB quieter than your source material, adding further dynamic range to the recording device doesn't buy you anything audible.

The most important recordings that I've ever made or will ever make in my life were recorded on cassettes in the 1970s--and I transferred them all at 16 bits.

[noahbickart]

I'd disagree slightly with John W. here because the total dynamic range of the very best cassettes (e.g. pure metal tape with optimal recording levels and Dolby S or HiCom noise reduction) is in the range of ~65 dB at the utmost. That dynamic range would be more than covered by a 12-bit digital recording system, let alone 16-bit; by the time your recorder is 10 dB quieter than your source material, adding further dynamic range to the recording device doesn't buy you anything audible.

The most important recordings that I've ever made or will ever make in my life were recorded on cassettes in the 1970s--and I transferred them all at 16 bits.

But surely you'd agree that there isn't any harm in recording at 24bit right? The extra headroom might make it a little easier to set levels at the A>D stage (e.g. lower), and normalize in post. This all depends on the dynamic range on the source tapes.

[runonce]

Is there any benefit to using 24 bits if you know you'll be doing substantial editing?

[John Willett]

Is there any benefit to using 24 bits if you know you'll be doing substantial editing?

Yes - I would use 24-bits - but I would not use a high sample rate and would stick with 44.1 for this project.

[Gutbucket]

Is there any benefit to using 24 bits if you know you'll be doing substantial editing?

Yes, during the processing, but your source files need not be the same 'bit-depth' as what the editor uses for its internal calculations.  Your editor probably does it's processing in 32 bit floating point numbers, irrespective of the bit depth of the source files.  You just need to make sure the bit depth used when digitizing the transfer is sufficent to capture the dynamic range of the source material.  In this case 16 bits is sufficient, 24 won't hurt other than using more storage space.  Of course, you'll want to adjust your levels correctly in either case.

[andromedanwarmachine]

thankyou all-

yes, Runonce- I think that's what I'm going with following Mr Willet's comments-

certainly the fidelity of the material does not warrant 24 bit transfer, but it IS going to be one piece in an extensive post-production adventure of which it is a key component.

Just about to start my Master's Degree Final Project in Sound Design- deemed to be 600 hours work (write-up & research included) to finish in September. This material is the backbone of it...