Ampex 631 / 641 tape - deciphering the specs

[cjacek]

OK, here's what I know.... It's a so called "voice grade" 185nanowebber red oxide tape, no backing, 1mil / 1.5mil tape, introduced in the 1950's (as far as I know) and I'm currently recording with it and really, really dig the tone. It's fairly limited in high / low frequencies and adds a certain "old" character but I'm really curious about its specs... like what is its true frequency response and optimum bias among others... I've been only able to find this:
http://recordist.com/ampex/docs/apxbrochures/tape/600tape.pdf

Page 3 lists the "Response" but I don't understand it. Any help would be appreciated. Thanks guys!

 

[Tim Gillett]

Daniel, I've not seen it expressed like that before. I assume they are quoting the response at a particular wavelength rather than a frequency at a particular tape speed, to make it independent of tape speed, which is probably a truer way to express it.

This is because tape doesnt care what speed it runs at. But the faster you make it travel, the higher upper frequency it can record and play back. The only meaningful way to express its response as a frequency is to also quote the tape speed used.

Even these old tapes would be capable of well over 20khz so long as the tape speed was fast enough. That's how video tapes were able to capture video signals way above audio frequencies. The tapes themselves werent necessarily any different from ordinary audio tapes but they ran them past the head at very high speeds.

I guess to get a frequency response figure for a particular tape speed here you would divide the tape speed by the wavelength quoted.

As you know the beauty of reel to reel tapes and machines was their various tape speeds and track widths. In a way you couldnt really say one reel of tape was "voice grade" or "music grade" because the sound quality greatly depended on the tape speed and the track width. The same reel of tape could store data at low or high quality depending on what was required. It was always a trade off between quality and economy.

Same with digital files today where you choose your sample rate ( sort of like tape speed) and bit rate ( sort of like track width) depending on the quality of program you are trying to record.

BTW, no recorder, whether digital or analog, has an "infinite sample rate",(infinite tape speed) and thankfully doesnt need to either.

Cheers Tim

 

[Dr ZEE]

Hi, Daniel... took me some time to make this diagram, see attached.

Now I am no expert on this stuff and honestly it gives me headache when I'm trying to dig into it.
Also, As , Tim, pointed out already, I also was thinking about, but I don't really know how to put it "right" in words and terms, nor I have any clue what "exactly" those characteristics represent.
As I understand, when you record AC signal on tape there's no really such thing as "frequency", but rather there's "wavelength" - that's the space on tape that is being taken by full wave cycle of an AC signal (see from yellow-to-red-to-yellow-to-green-to-yellow). So on tape there's only "wavelength" of a recorded signal, and it will be "translated" into frequency only as the tape runs at certain speed.

 

[Tim Gillett]

Well put, Dr Zee, and good, clear diagram.

Cheers Tim

 

[cjacek]

Tim, Mike, thanks so much for your words and, Mike, that's one hell of a good diagram. I really appreciate the time that you put into it. Yeah, I still have to figure it out 'cause it is indeed a bit confusing, their way of expressing those specs. I guess I wanted to put on paper, in numbers, what I was hearing from that tape but....... heck, if it sounds good, then it sounds good.

 

[Dr ZEE]

Daniel, as I "see it" (I may be wrong, just guessing), the "idea" here is to lay out a characteristic to compare just tape to tape, instead of comparing tape at speed A to tape at speed A.
Also wavelength is not measured, but is calculated from measured frequency and measured/calculated speed to begin with. heh heh

It would be cool to actually find out what the hell exactly did they "expess" there in that table. All I see there is that all the 600 series are the same compare one to another in respect to "relative response" , so? arghhhhhhhhhh, it must be GOOD, since it's been published in a colorful promo-paper

 

[analog aaron]

I bought

quite a bit of this tape from a guy on ebay. He said he hadn't any complaints about it and that he's sold tons of it. His feedback looks good. I was told the tape came from Ampex under government contract and was in a temperature and humidity controlled warehouse until it was shipped to me. I received the tape and it looks fine. I haven't had the chance to use it yet. I wanted to set the bias on my TASCAM 388 for this tape (for a little while anyway, since I've got soo much of this stuff) Would anyone be able to give me any advice or instruction on this? Thanks Guys,
Aaron

 

[cjacek]

arghhhhhhhhhh, it must be GOOD, since it's been published in a colorful promo-paper

You got that right!

BTW, thanks again.

I wanted to set the bias on my TASCAM 388 for this tape (for a little while anyway, since I've got soo much of this stuff) Would anyone be able to give me any advice or instruction on this?

Well, it's a lot more time consuming on a deck without a dedicated playback head but here's one way of doing it... Dial in the radio, in between stations, so as to get a steady white noise or you may even get a signal generator (can be found online free) to output a white noise. You do a line out from that and into your 388 and record a portion of tape, maybe 10 or 20 sec, on track 1. Then play it back (from tape), listening carefully to that and also to the source. If the signal from tape is different, adjust the bias pot a bit. Then, record again and see if it's getting closer to the source. If not, adjust again. This tape actually takes a lot less bias so you probably would be turning the pot to the left (bit by bit) and seeing if you can get the tape / source to be virtually identical. Yeah, it's record, playback, adjust, record, playback, adjust etc.......... (and that's only for track one). You have to repeat for all the remaining tracks. Hope this helps.

 

[dodgeaspen]

Good info. Thanks.

 

[jpmorris]

Well, it's a lot more time consuming on a deck without a dedicated playback head but here's one way of doing it... Dial in the radio, in between stations, so as to get a steady white noise or you may even get a signal generator (can be found online free) to output a white noise.

I don't think I've ever heard of white noise being used to calibrate bias before. What's the advantage over a simple test tone?

 

[cjacek]

I don't think I've ever heard of white noise being used to calibrate bias before. What's the advantage over a simple test tone?

Well, I'm not saying one is better or has an advantage over the other but, rather, it's a more basic way of at least getting toward a ballpark for a given tape, as far as bias is concerned. In essence, if the source / tape is the same, you have the bias set correct for a given tape. Obviously, depends also on your ears but still..... When you really listen (when adjusting bias) and can tune it so that the source white noise and off tape white noise match up, then I'd say you got it dialed it. White noise, if I'm not mistaken, is all the human range audible frequencies mixed in equal amounts so if you can set the bias that way, it could potentially be a better / more accurate way than going by the text book "keep turning the bias until it peaks and then drops off by 3db.....". I just think it makes sense, especially when no such info is available for Ampex 631 / 641.

 

[cjacek]

Oh, BTW, Aaron, I'd turn the dbx OFF and record at no more than -10db (on the VU meter) when performing this biasing (on the 388).

 

[Blue Jinn]

I just got handed about 3 dozen reels of 641 in black boxes and most of it looks unused. Anyway, saw this thread but d idn't appear to be a consensus about this tape, I found the same 1964 sheet from recordist.org, but also another thread here that 6x1 is "voice grade" and 6x2 is "music grade." Not really sure what that means. The specs seem to be different, the "music grade" has about 310/1100 coercivity/retentivity and the voice grade 260/900. (Akthough another thread gives 1100 as the coercivity for 651....)

If I'm understanding things correctly, hiss and noise would be a problem and I can't hit it at 0VU (on a 456 calibrated machine) but if I stay at -3 and use dbx I don't see too much of a problem for knocking about. And am I correct that bias shouldn't be too much of an issue at 15ips? That is where I get lost in the specs, and I'm not about to attempt to rebias a machine for this tape. I'd note that my Otari has a selector for l m h levels, l being 185 nW/m.

The brochure seems to suggest a 59dB S/N ratio, at a "broad range" again not sure what they mean. OTher info is referenced to wavelengths in mils which I'm too tired at the moment to try and convert...

But with NR giving me 10-20 more dB I don't see hiss being too much of a problem.

 

[ofajen]

631 is good tape. Recordings will last... probably 100 years. You have to live with lower levels.

Cheers,

Otto

 

[Blue Jinn]

I'm going to give it a shot on my 22-4. I have a hacked DX-4D unit to go with it. Also got a bunch of NOS Maxell UD from the same guy.