One man's opinions ...

Regardless of what frequency range an individual is capable of, the main POINT of the speaking voice 
is audibility of the words, i.e. what the person is SAYING. It really has nothing to do with 
fidelity (see a hundred plus years of Bell Labs research). It has to do with knowing what 
frequencies are key for audibility/intelligibility and make sure those frequencies are emphasized 
and/or frequencies that interfere with intelligibility are eliminated. You can do a nice 
flat-response high fidelity transfer if you want it for the archives, but actual users are going to 
want maximum intelligibility, which often involves high- and low-pass and/or band equalization.

Also, as for the 3.75IPS speed for reels, most decks from back in the day operating at that speed 
had terrible wow and flutter specs, and the kind of users who would be recording music at that speed 
(i.e. non-professionals) were likely using tape that wasn't slit all that accurately. So, not likely 
the recording results are very good. There are exceptions of course ...

Given the sound quality of most slow-speed spoken-word source material I've gotten from clients, I 
am not at all shy about suggesting sometimes aggressive measures to improve 
audibility/intelligibility and thus user enjoyment and overall usefulness of the transfer. Being a 
purist isn't a good idea when the source is junky-sounding.

-- Tom Fine

----- Original Message ----- 
From: "Richard L. Hess" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Monday, April 16, 2012 1:09 PM
Subject: Re: [ARSCLIST] Slow Reel-to-reels

> Hi, Mike,
> I think if you can make 16 kHz flat within 1 dB off a test tape (with few major bumps below that) 
> at 3.75 in/s and still have respectable 20 kHz response (say within 2 or 3 dB) we've got a 
> good-enough head. Few recorders back in the day were able to do that. that demonstrates the 
> wavelength response. Maybe some 1.88 in/s tapes can benefit from going the extra mile with ultra 
> narrow-gap heads, but it's a budget thing, too. I've used some 50 µin heads and don't see a whole 
> lot more coming off the tapes than with 100 µin heads.
> Studer produced heads with gap lengths of 79 µin (ReVox A77/B77) to about 146 µin (A80). The 
> A810/A807 repro heads were about 118 µin and the special 3.75 in/s head for the A80 RC was about 
> 98 µin.
> Resonating the head was common practice to increase its high-frequency response and Jay McKnight 
> has a calculator on the MRL website to help with that.
> The target audience of your message should be the interviewers and the researchers who use(d) the 
> $19.95 drugstore cassette (or reel!) no-name recorders and the equally no-name Bargin-o-rama tapes 
> to record important interviews.
> With something like the Zoom H1/H2n now there is just no excuse (and thankfully (in the long run) 
> no backlog of work being created for the likes of me) to have anything but good-quality interviews 
> ... well unless they put the unit across the room and under a pillow. With the H2(n) I urge people 
> to use the four-channel mode for most oral histories and I've had people use two of them to record 
> 10-16 person panel discussions where everything had to be intelligible.
> Oh, and if you're interviewing in a house on a busy truck route, try to keep the windows shut and 
> put the dog out as well (even in the country).
> Cheers,
> Richard
> On 2012-04-16 11:11 AM, Michael Biel wrote:
>> From: Nigel Champion<[log in to unmask]>
>>> The drawbacks with this method are:
>>> 1. It's only suitable for low-bandwidth material (ie speech and not music)
>> Is there SOME way to rid the audio and video world of this concept that
>> speech does not need wide-band recording and reproduction???  The human
>> speaking voice has the widest frequency range of just about any musical
>> instrument and do not assume that the only reason to record someone
>> speaking is intelligibility of the words.  I grew up reading all the
>> tape recorder ads that said use 3 3/4 for voice and 7 1/2 for music.
>> Enough already.  That concept should have been ignored in the 1950s and
>> should be ignored now.
>> IF the speech recording you are working with does happen to be low
>> bandwidth, that is one thing.  But don't make it low bandwidth by bad
>> reproduction techinques because "what the hell, it's only speech".  Some
>> 3 3/4 speech recordings DO have wide bandwidth.
>> Mike Biel  [log in to unmask]
> -- 
> Richard L. Hess                   email: [log in to unmask]
> Aurora, Ontario, Canada           (905) 713 6733     1-877-TAPE-FIX
> Quality tape transfers -- even from hard-to-play tapes.