Metering and gain structure in analog tape decks?

[SouthSIDE Glen]

A little help?

I'm currently working on a web applet describing the details of metering and gain structire in the recording chain (analog and digital). I think I have most of the details straight when it comes to metering and calibration on your average analog tape deck but I'm not positive about some aspects of the metering points inside the deck.

A few questions if I may. These questions mostly deal with the relationship/reconciliation between metering to measure signal strength (voltage) on input to and output from the deck - i.e. standard gain stage metering - and metering measuring the magnetic flux on the record/playback heads - i.e. metering calibrated to an MRL or other tape reference level.

Is it an either/or choice as to whether one calibrates their metering to line level voltages or to reference tape levels?

If one does not calibrate to MRL references, are the default calibrations set to a factory reference tape or to the line signal voltages?

What is your preferred method for setting the proper gain structure in and out of your analog tape deck to get the optimum not only from your tape, but also to ensure proper I/O voltages as part of the overall gain structure?

Do the answers to these questions vary depending upon deck make/model?

Thanks in advance for your help on this. I'll give credit where it's due in the release announcement for the applet (as well as in the chicklet acounts).

G.

 

[evm1024]

both

On the input side:

We apply a reference voltage (-10 dBv or +4 dBu) which will cause the meters to read 0 VU.

The 0 VU level will record a signal (1KHz typical) onto tape at the selected reference fluxivity. This flux is selected based mostly on your selection of tape. You can change the fluxivity to get more headroom or lower noise floor or (most commonly) an optimized level.

The old reference is 185 nW/m. A + 6 tape is +6 dB more than the 185 nW/m (320/370 nW/m). Tapes are +3, +6, +9 generally speaking.

On the playback side:

A tape recorded at your reference fluxivity will read 0 VU and will output your reference voltage (-10 dBv or +4 dBu).

Of course getting all this calibrated to have in, out and 0VU all match is the art of calibration.

Matters not what machine you use. In's = out's at any given fluxitivity on tape. That is the goal.

regards, Ethan

 

[SouthSIDE Glen]

D'oh!

Ethan, thanks for your reply. OK, I think I now have the one stupid and obvious fact that I was mishandling in my head, and need to slap my forehead for:

The calibration entails calibrating the reference line voltage to the desired reference magnetic flux - i.e. ensuring that +4dBu delivers (for example) +6/185 to the record head. That way 0VU represents *both* line level voltage and optimum flux. And then vice versa on the way out.

Do I have that right?

G.

 

[evm1024]

Yup

Ethan, thanks for your reply. OK, I think I now have the one stupid and obvious fact that I was mishandling in my head, and need to slap my forehead for:

The calibration entails calibrating the reference line voltage to the desired reference magnetic flux - i.e. ensuring that +4dBu delivers (for example) +3/185 to the record head. That way 0VU represents *both* line level voltage and optimum flux. And then vice versa on the way out.

Do I have that right?

G.

Yes that is correct. Because the reference flux is on a MRL tape we calibrate the playback side first (head to lineout and head to meters) then we adjust the record level with a line in at the reference voltage to print on tape at the reference flux. easy with 3 head machines, harder with 2 head machines.

--Ethan

 

[FALKEN]

also note that reference flux level is frequency dependent and also each machine is different and is subject to a changing frequency response so there are eq adjustments that go into calibrating a reference level on these machines for both playback and record.

 

[SouthSIDE Glen]

Yes that is correct. Because the reference flux is on a MRL tape we calibrate the playback side first (head to lineout and head to meters) then we adjust the record level with a line in at the reference voltage to print on tape at the reference flux. easy with 3 head machines, harder with 2 head machines.

--Ethan

OK, thinking out loud here, I might be over-thinking and pushing myself down anther stupid blind alley, but it seems like half the trick is in selecting a reference level - and reference tape formulation - ideally suited for the premp circuitry in the tape deck. For example, I could imagine that a +9/185 level might not be suitable for a deck that had a noiser pre becaue you'd have to push that pre too hard to get it up to that level. Am I even close with that one?

Thanks a bunch for the help Ethan, and for giving me a chance to win a bar bet that I know of a forum that has two regulars named Ethan on it .

also note that reference flux level is frequency dependent and also each machine is different and is subject to a changing frequency response so there are eq adjustments that go into calibrating a reference level on these machines for both playback and record.

Yeah, I am aware of the individual frequency and overbias calibrations and such. I was just having a disconnect (read: brain fart) in the relation between line voltage measurement and magentic flux measurement in terms of gain structure, totally missing the obvious point of the exercise; namely to make the measurements coincident.

G.

 

[evm1024]

Kudos to Falken, I skipped over frequency and bias effects.

In answer to your question RE optimal flux vs preamp etc. the answer is that the maker of the machine did that for you. They made some trade offs in preamp cost, noise etc and selected a target tape and fluxivity. This is more true in consumer machines and less so in prosumer machines and much less so in pro machines.

Many machines do not have enough range of adjustment to go much away from their designed fluxivity. Older machines did not have the option of higher MOL (max operating level) tapes.

On an A3440 for example you might be able to get the record amps to print at 370 or more nW/m but the reproduce amps did not adjust to a low enough gain to read 0 VU at that flux (might read + 2 or 3).

Others like my Ampex ATR700 (a quite old) could handle +6 and + 9 tape. And mt Tascam BR-20 will handle just about anything.

More expensive/more pro oriented machines have better/quieter preamps and very good heads.

Regards, Ethan

PS I too am on 2 BBS with Ethans

 

[Beck]

Glen,

Ethan covered it well.

I’ll just add that a good way to visualize it is by remembering the main calibration steps (as Falken mentioned there is more to a full calibration but these basic steps are most pertinent to the question). The relationship of the systems in an analog machine is most apparent while hunched over a machine with tweaker in hand. Keep the Advil close by.

- Apply a 1kHz tone at -10 dBv or +4 dBu (whichever is specified) to the input jacks so that you have the same level at the output, while at the same time the meter is reading 0 vu. This is the line operating level. On a machine using the so-called semi-pro level of –10 dBv your 1kHz input and output would measure roughly .316 volts.

- Play a standard calibration tape and adjust the meter to 0 vu. This is a separate adjustment that won’t change what you accomplished in the first step. This is the flux operating level.

- Record a 1kHz tone onto a blank tape @ 0 vu reference the machine’s meter.

- Play back the recording and adjust repro level so that your meter reads 0 vu and your output agrees with your line level (-10 dBv or +4 dBu).

So you have two things that people call operating level, but one is line level and the other is flux (magnetic energy) level to tape. They are independent of each other, but you use one to reference the other so the deck as a system is in agreement with itself.

It’s helpful to remember in what way the two levels are independent when you see that a 1kHz tone recorded @ 0 vu on a machine with a +4 dBu line level will play back @ 0 vu on a machine with a -10 dBv line level…. Given that both machines are set for the same fluxivity (for example 250 nWb/m) and are adjusted for the same tape (for example Quantegy 456).

You are on the right track in your thinking about hot tapes and preamp circuitry, particularly opamps, but the limitation is more in the output capabilities of the circuitry and distortion levels.

There is an art to finding the best tape and flux combo for a given machine. Even if your machine can ramp up to +9 tapes, crosstalk at the heads can be a limiting factor. So someone looking for a clean sound on something like the TASCAM 38 without noise reduction should limit flux level to 320 nWb/m even though the circuitry will do more.

I posted a couple things a while back that are somewhat related to this topic and somewhat unrelated, but I’m going to be lazy here and post the links rather than risk being too redundant or over-explaining.

https://homerecording.com/bbs/showpost.php?p=2136216&postcount=23
https://homerecording.com/bbs/showpost.php?p=2083032&postcount=12

 

[SouthSIDE Glen]

So you have two things that people call operating level, but one is line level and the other is flux (magnetic energy) level to tape. They are independent of each other, but you use one to reference the other so the deck as a system is in agreement with itself.

Ths was the key on which I brain farted. I'm all over all the rest of it, but I just didn't make the connection that the calibration process aligned the meter calibration to BOTH the line voltage and the magnetic flux. I had it in my head that it was an either/or thing for some reason. Too many years sniffing solder fumes....

Thanks all for the kick in the head and for the extensive help . Back to the development machine...

G.

 

[ofajen]

Just two little things to add to the excellent discussion so far:

1) There may actually be two operating voltage levels of interest: one for recorder input and one for recorder output. Typically, the tape deck's VU meters are calibrated to a standard operating level. For example, my 3M's meters are calibrated so that a 0 VU output level will meter at +4 dBu, without regard to whether it comes off the heads and through the playback electronics or from the input section. However, the operator (me) can choose to use any operating level on the input side. This is set with the recorder gain. Back when I used my humble Mackie CR-1604 preamps to drive my 3M inputs, I would routinely crank the recorder gain so that a lower input level (-2 dBu or sometimes lower) would produce 0 VU on the tape (of course this would still give +4 dBu on the output side). The reason was that the Mackie direct outputs are unbalanced and this operating level on the recorder input side gave me more headroom and cleaner sound. The Mackie direct outs clip at no more than +16 dBu and I tried to stay well below that point, so I would often set the input operating level down around -6 dBu.

2) While not really related to tape decks per se, I still mix my tape tracks on little mixers that don't really like high levels, so I typically run each tape output through a fixed resistor pad that drops the level by 6 dB, just to avoid overloading the mixer inputs.

I hope that doesn't confuse things too much.

Cheers,

Otto

 

[MCI2424]

OK, thinking out loud here, I might be over-thinking and pushing myself down anther stupid blind alley, but it seems like half the trick is in selecting a reference level - and reference tape formulation - ideally suited for the premp circuitry in the tape deck. For example, I could imagine that a +9/185 level might not be suitable for a deck that had a noiser pre becaue you'd have to push that pre too hard to get it up to that level. Am I even close with that one?

Thanks a bunch for the help Ethan, and for giving me a chance to win a bar bet that I know of a forum that has two regulars named Ethan on it .Yeah, I am aware of the individual frequency and overbias calibrations and such. I was just having a disconnect (read: brain fart) in the relation between line voltage measurement and magentic flux measurement in terms of gain structure, totally missing the obvious point of the exercise; namely to make the measurements coincident.

G.

GP9 tape (high bias) is only useable on decks that have enough gain to attain that high of a bias. Older decks and most consumer decks cannot use anything other than standard 456 type tape (the "old' high bias tape). These decks crap out anywhere above +4. The high bias tapes push the noise floor way down to the point of not having to use noise reduction. If a deck can use GP9, it is great tape and in a well setup and maintained machine will make excellent recordings.

 

[evm1024]

A little confusion here

GP9 tape (high bias) is only useable on decks that have enough gain to attain that high of a bias. Older decks and most consumer decks cannot use anything other than standard 456 type tape (the "old' high bias tape). These decks crap out anywhere above +4. The high bias tapes push the noise floor way down to the point of not having to use noise reduction. If a deck can use GP9, it is great tape and in a well setup and maintained machine will make excellent recordings.

I'm sure that MCI2424 knows what he is talking about but just has mixed his words up. So just for a little clarification.

GP9 has a high MOL (max operating level) which refers to the magnetic level in nW/m on tape (nano webers per meter). It takes the right preamps andhead driver amps to put that level onto tape (and get it off, the inverse problem). Many driver amps cannot adjust that high and many heads cannot take that much drive. Also, many head preamps cannot adjust low enough to allow calibration and optimal noise (hot tape means hot signal to the head preamp).

GP9 also is a high bias tape which means that the high frequency "carrier" signal that is also applied to tape needs to be at a high nW/m. and many bias driver amps cannot adjust to that nW/m.

 

[SouthSIDE Glen]

For example, my 3M's meters are calibrated so that a 0 VU output level will meter at +4 dBu, without regard to whether it comes off the heads and through the playback electronics or from the input section.

I highlighted that last part as the part I'm not sure I follow you on. if you're monitoring the output as passed through from the input section - basically sounding like the deck is just sitting idle passing the signal thru with the transport in stop mode and record circuit disengaged - how can you set the proper input levels for recording? Are the meters switchable between input and output, or do they automatically switch when the record circuit is engaged?

Of course, if the machine is calibrated all the way through, then it really doesn't matter, as 0VU is the Goldilocks level at every gain stage anyway.

2) While not really related to tape decks per se, I still mix my tape tracks on little mixers that don't really like high levels, so I typically run each tape output through a fixed resistor pad that drops the level by 6 dB, just to avoid overloading the mixer inputs.

Yeah, there's always going to be different personalities and characteristics in analog hardware that cause one to want to deviate from "standard" gain stage levels in order to get the best overall gain structure. Which is an important principle to get across in my applet.

The difficulty in my task at hand is that I can't really go into an extreme amount of detail on all points. 95% of the details of what we all have discussed throughout this thread is beyond the detail scope of what I'm working on. For example, I have the space to discuss the input, tape, playback, and output gain stages and how tape calibration is important and relates to them, but not to actually get into a description of the actual calibration process itself (there's plenty of recources on the net that go into that in good detail anyway)

In that light, I have a duty to (repeatedly) mention that there are always going to be reasons for those of us experienced with our hardware and in gain structure principles to push some levels at this stage or reign them back at that stage in order to get the most out of our gear. And while I do mention some general examples, I simply can't go into listing every exception to every rule, as the document would wind up filling the Internet .

When I release this applet on the IRN website (give me another couple of weeks to finish it along with other additions to the site I'm working on concurrently), I'll make an announcement thread in the Recording Techniques forum on this board. That would probably be a pretty nice place to get into some further discussions of great stuff like what Otto described here where one needs to make such special gain stage decisions and details beyond what are described in the applet.

G.

 

[ofajen]

I highlighted that last part as the part I'm not sure I follow you on. if you're monitoring the output as passed through from the input section - basically sounding like the deck is just sitting idle passing the signal thru with the transport in stop mode and record circuit disengaged - how can you set the proper input levels for recording? Are the meters switchable between input and output, or do they automatically switch when the record circuit is engaged?

Ah, but the input monitoring path is more than just a straight wire affair. In both my M-79 and M-23, the input signal runs through a transformer and then through a record level pot and a gain stage in the record card. Pretty much all machines work that way, i.e. the signal input is brought up to internal line level and then either routed direct to the signal output or over to the record electronics and through the heads and back through the playback electronics.

So, even if you are listening to the input signal at the output, it has already gone through those input stages. The M-23 record level control was designed to allow a line input level range from -20dBm to +8 dBm. My only point was that the operator may send a lower (or higher) level signal in and then adjust the recorder gain accordingly. The line input level is not dictated at a single value, but can be chosen according to need.

Also, the M-79 series has output levels fixed at +4 dBm calibration, but on the old M-23 you have a choice of a fixed, calibrated +4 dBm output setting or an output setting through the repro level control, allowing adjustment of the line output level, as well. I think this is also common, but obviously not universal.

Cheers,

Otto

 

[SouthSIDE Glen]

Ah, but the input monitoring path is more than just a straight wire affair. In both my M-79 and M-23, the input signal runs through a transformer and then through a record level pot and a gain stage in the record card. Pretty much all machines work that way, i.e. the signal input is brought up to internal line level and then either routed direct to the signal output or over to the record electronics and through the heads and back through the playback electronics.

So, even if you are listening to the input signal at the output, it has already gone through those input stages. The M-23 record level control was designed to allow a line input level range from -20dBm to +8 dBm. My only point was that the operator may send a lower (or higher) level signal in and then adjust the recorder gain accordingly. The line input level is not dictated at a single value, but can be chosen according to need.

Also, the M-79 series has output levels fixed at +4 dBm calibration, but on the old M-23 you have a choice of a fixed, calibrated +4 dBm output setting or an output setting through the repro level control, allowing adjustment of the line output level, as well. I think this is also common, but obviously not universal.

OK, I'm pretty sure we're on the same page as I understand and agree with all you said here. For some reason I was reading something different into it the way you said it before, though honestly I can't now say just what that was .

P.S. I keep intending to look up just where 92W by 39N exactly is. I suspect it is somewhere around the Quad Cites/Dubuque area? I wouldn't mind finding a nice piece of land somehwere just northeast of there myself.

G.

 

[ofajen]

P.S. I keep intending to look up just where 92W by 39N exactly is. I suspect it is somewhere around the Quad Cites/Dubuque area? I wouldn't mind finding a nice piece of land somehwere just northeast of there myself. [/SIZE]

G.

I'm a ways south and west from there, in a different state. I've rounded the figures to the nearest degree... I'm actually slightly further south and west, but it gets you within about 30 miles.

Cheers,

Otto

 

[MCI2424]

I'm sure that MCI2424 knows what he is talking about but just has mixed his words up. So just for a little clarification.

GP9 has a high MOL (max operating level) which refers to the magnetic level in nW/m on tape (nano webers per meter). It takes the right preamps andhead driver amps to put that level onto tape (and get it off, the inverse problem). Many driver amps cannot adjust that high and many heads cannot take that much drive. Also, many head preamps cannot adjust low enough to allow calibration and optimal noise (hot tape means hot signal to the head preamp).

GP9 also is a high bias tape which means that the high frequency "carrier" signal that is also applied to tape needs to be at a high nW/m. and many bias driver amps cannot adjust to that nW/m.

Yes. The Master bias oscillator on my big deck is set to 225kHz. More than enough for GP9 (which I use and love).