Bias question

[illuminations]

Hi all,

I have a question about record bias. I have tried reading about it and I guess I’m having a bit of a hard time grasping how it’s adjusted. Please bear in mind I’m new to all of this.

From what I’ve read and understand, record bias on a tape deck is not an exact thing? It seems like there is some wiggle room to get it into spec. But I may be wrong here. I guess that’s my main confusion.

I have a Tascam 22-2 recently set up by a knowledgeable tech. I have no doubts he did a good job. Still I’m noticing a slight bump in the high frequencies and the low frequencies are just slightly lacking.

I initially thought maybe the eq adjustments were a little off. But then reading about the bias I’m beginning to wonder if it’s just slightly under biased.

Sorry if I’m not explaining this the best. But hopefully someone will see what I’m saying and can offer some insight.

Thanks

 

[sweetbeats]

When doing record calibration on a machine with adjustable HF EQ, there’s an interrelation between record level, bias, and EQ. It’s almost always best to follow the procedure as laid out in the manual. It’ll usually say if you have to adjust the record EQ to go back and re-check the record level, bias, and then check overall response again. Basically rinse and repeat until you have things dialed in.

Changes to the EQ can effect level.

Record bias is a high frequency signal introduced to the tape by the erase head ahead of the record head. You can read up on the history of it, but in layman’s terms the bias signal prepares the oxide on the tape to better capture the audio bandwidth signal it’s about to encounter at the record head…pre-excites the magnetic material on the tape. Without it the recorded program material has much higher distortion and there is a HF boost in the response curve; thin and distorted-sounding is how I describe it. As you increase the bias level, usually via a specific trimmer for each track (refer to your service manual), the distortion diminishes and the HF response increases, but only to a point…the HF response starts to fall as the distortion reaches minimum. There comes a point where the distortion is at minimum and the HF response should be relatively flat. Adding additional bias signal diminishes the HF response to the point there is a HF cut. This is an over-bias condition. The typical way to adjust bias is to record a high frequency tone (see your service manual for the proper frequency), and increase the bias level while monitoring the recorded level of the tone. As you increase the bias level the recorded level of the HF tone will increase, and then it will hit a knee point and start to drop. Usually the correct bias level is where the recorded level of the HF tone is about -3dB below whatever the peak was (give or take…more or less…YMMV…depends on the tape and the machine and your *taste*…you can actually tune the bias for specific material…like try tracking a kick drum with a little under bias…a little HF boost and a little distortion might just give the kick drum attack a little more bite and presence in the mix…this assumes we are talking about a multitrack machine where the kick drum is on an individual track…but the possibilities are many). The “over bias” method where you dial the bias level up until the HF response reaches a point X dBs below the peak is a little arbitrary because it doesn’t tell you what the distortion level is, which, IMO, is the correct thing to monitor…adjust bias for minimum distortion. But not everybody has a distortion meter. So there are usually guidelines for different tape types with a given machine for how far past the HF knee you set the bias. There are other ways to do it too…there’s the LF modulation method which actually does monitor distortion. You can search for information by searching “bias rocks.”

Anyway, I hope that helps. Bias is there to achieve minimum distortion, the HF EQ adjustment is there to compensate if the lowest distortion level results in a HF response curve that is below the manufacturers spec. And these things effect the record level. So, rinse and repeat.

 

[rob aylestone]

This was exactly how I failed my BBC radio interview for a sound department trainee role in 1976! The guy asked me about the purpose of bias and I tried to blag it, to the deputy chief engineer who had stepped in for the interviews to help. He did though, put me right. Sweetbeat's explanation is on the money, but quite complicated. The BBC engineer's dumbed down explanation to me was simpler. He drew a little graph on his sheet - audio voltage on one axis and the success of the magnetism of it on the tape on the other one. It looked a little like an 'S'. as the audio level went up from zero, the magnetism on the tape barely moved, then it picked up and went upwards in a straight line at 45 degrees, then as it approached it's maximum magnetism level, the curve flattened out again when more volts did not magnetise the tape any more - it was saturated. He told me that line up made sure that the bit of the S shape you used for your recording was ONLY the 45 degree bit - don't go lower, and don't go higher. That way, what is on tape reflects most closely what the audio was doing. If you pushed the levels, you started to move into the compression territory, where more oomph in does not result in much extra level on the tape. At the bottom, the same thing, but it was a fight with noise. Every tape needed different bias current and equalisation to get you on the flat bit.

 

[illuminations]

Thanks so much for the detailed response. The process is starting to come into focus now.

I was hoping that a bias adjustment would be a stand alone process. But from what I gather now it is something that needs to be dialed in along with hf eq and level. I was really hoping that I could just tweak the bias a little.

So would it be right to assume that if the machine were to better represent the high and lows it is recording, there would be a potential trade off in distortion or level? Is it possible that in attempting to lower the distortion as much as possible the tech slightly under biased? I guess I never realized there was room for variations when calibrating a machine. Is there a sweet spot where all frequencies are accurately represented? Or do you have to sort of make a decision in slight trade offs?

I’m using the machine as a mix down from an 8 track. The tonal differences are subtle. But since I’m spending a lot of time mixing I notice the difference, especially in the lower frequencies. There is a slight lack that I just wish would be there. When I monitor from my multi track the lows are fuller and more detailed. This is also the case when I record a digital master directly from my multi track vs when I record a digital master from the 2 track machine. I feel like it just needs a minor adjustment but I don’t want to go in there and muddle things up.

 

[sweetbeats]

I was hoping that a bias adjustment would be a stand alone process. But from what I gather now it is something that needs to be dialed in along with hf eq and level. I was really hoping that I could just tweak the bias a little.

Well you might well could…in fact of course you can. But the point I want to make is to ensure you understand that bias level is not akin to an EQ knob. If you use the bias level control to increase HF response (i.e. by reducing the bias level), you also increase distortion. So it’s better to set the bias for the tape and machine you have and get the HF response curve dialed in the way you like or within manufacturer spec by using the HF record EQ adjustment, if your machine has one. And you certainly don’t *have* to rinse and repeat. You could just tweak the HF record EQ control and be done with it, but then if you’re a guy like me you can’t sleep at night because you’re now wondering if the record level is now off-kilter a smidge, and so on. That might not be you. But now you’re informed and you can make those decisions on your own. I think getting these things right on a mastering deck is a bit more critical than a multitrack machine, especially one that is just being used by you and your tapes aren’t being sent to other places and played back on other people’s machines. That’s one of the main reasons for electronics calibration is for when studios would, for instance, send a 2” tape from the recording studio to a mastering house. The levels and response curves had to match as close as possible.

So would it be right to assume that if the machine were to better represent the high and lows it is recording, there would be a potential trade off in distortion or level?

You must understand bias has no real impact on LF response. None of what I told you is going to impact a change in your machine’s LF response. And as to your question, and specific to HF response, there’s only a trade-off with HF response if you’re using the bias level control like an EQ, which I already said is not the right way to go about it. Set bias according to the procedure in the Service Manual for the tape you are using, which should attempt to reach minimum distortion. Then use the record EQ control to tweak the HF response. So there shouldn’t be a trade-off. There should just be a process of setting it up.

What tape are you using?

Is it the same tape your tech used to calibrate the machine?

Does the 20 series have a LF response trimmer? Don’t answer that…hang on…[mumble-mumble…] no it does not. I didn’t think so but I thought I’d check.

Did your tech adjust azimuth and do other transport checks and adjustments to ensure proper operation and positive tape-to-head contact?

Have you run your own tone ladders? It’s super simple to do if you have a mobile device and can pay $3 for a tone generator app. And the results are completely quantitative.

Is it possible that in attempting to lower the distortion as much as possible the tech slightly under biased?

Read my description of how bias works again. I am thinking the greater potential is a tech OVER-biasing…as bias level increases distortion decreases, HF response increases and then begins to drop just before minimum distortion. Overbias conditions won’t impact the distortion minimum, but will continue to diminish HF response. But is your concern HF or LF response or both?

Also consider this…with a lot of program material, good HF response helps to bring out LF elements…the frequency spectrum doesn’t live in silos. So your ears might perceive diminished LF response if you have a problem with your HF response.

Tone ladders.

I guess I never realized there was room for variations when calibrating a machine. Is there a sweet spot where all frequencies are accurately represented? Or do you have to sort of make a decision in slight trade offs?

Ummm…not really. I mean within the performance capability of the machine. Like I said above, you dial in the bias for minimum distortion, and then adjust the record EQ if necessary for flattest response or whatever you’ve decided to do with your machine. But typically folks shoot for flat. But it’s never going to be flat. It’s a tape machine. So that’s why the manufacturer gives a range of +/-3dB from some Hz to some kHz…or +/-2dB. Your goal is to get it in the range specified by the manufacturer, so there shouldn’t be any trading off. Minimum distortion, response within manufacturer spec.

I’m using the machine as a mix down from an 8 track. The tonal differences are subtle. But since I’m spending a lot of time mixing I notice the difference, especially in the lower frequencies. There is a slight lack that I just wish would be there. When I monitor from my multi track the lows are fuller and more detailed. This is also the case when I record a digital master directly from my multi track vs when I record a digital master from the 2 track machine. I feel like it just needs a minor adjustment but I don’t want to go in there and muddle things up

Well you’re talking LF response again so it doesn’t have anything to do with bias or your record EQ adjustment.

1. I’m a broken record here but tone ladders…see if your machine is recording and reproducing within what it was designed to do. If you haven’t done it, do it. If you have, and it’s working to spec, don’t expect more out of it. But…

2. You understand tape machines have all sorts of sonic anomalies to them right? And you’re talking about a “subtle” difference when mastering to your halftrack. You’re going from analog to analog. That’s another analog generation vs mastering to digital. I think you may just be expecting it to do something it’s not going to do, and it’s sounding the way it’s going to sound, and you just aren’t quite pleased with the sound. What mixing console are you running your mix through, and have you considered when mastering to the halftrack, if the LF response is just subtly lacking you dial in a little more LF EQ at the console? But…

3. Before you do that ^^^^^, you should see for yourself quantitatively what the machine is doing response-wise.

 

[illuminations]

Thanks for you patience and for walking me through this. The tech did a full set up. He actually put on new record and playback heads and calibrated and adjusted everything. I'm using the same scotch 207 tape that he used. So the machine is in top shape.

2. You understand tape machines have all sorts of sonic anomalies to them right? And you’re talking about a “subtle” difference when mastering to your halftrack. You’re going from analog to analog. That’s another analog generation vs mastering to digital. I think you may just be expecting it to do something it’s not going to do, and it’s sounding the way it’s going to sound, and you just aren’t quite pleased with the sound. What mixing console are you running your mix through, and have you considered when mastering to the halftrack, if the LF response is just subtly lacking you dial in a little more LF EQ at the console?

This might be the real issue. I know the machine is set up to spec as the tech I worked with has extensive experience with these machines and is pretty meticulous. But maybe what is called for is just different than what I want from the machine. Maybe a slight emphasis on the high frequencies per tascam? I'm wondering if there is a slight bump in the high frequencies, then maybe the lower frequencies would sound less pronounced. The reason I thought it might have been under biased is because I had read another thread where there was a low frequency bump on the same machine and it was put out there that it was over biased, so I was just working in the other direction. I'm mixing down from a tascam 388. I've thought about just bumping the low frequencies of my mix but I kind of have everything keyed in. I'm not sure I'm confident enough to make little tweaks to compensate for what I get when I monitor from the 388 to where they would translate to an even mix on the 22.

But I'll take your advice and try the tone ladders. I do admit I'm not sure exactly what the process is. I'd venture to guess it's recording a series of frequency tones to the tape to check the response? Would you be able to walk me through it or possibly point me somewhere where the process is explained? I would definitely be curious to see how it's done, and as you mentioned it would give me important information about what the machine is actually doing.

But typically folks shoot for flat. But it’s never going to be flat. It’s a tape machine. So that’s why the manufacturer gives a range of +/-3dB from some Hz to some kHz…or +/-2dB.

One last question. You mention here the +/- range. This is in reference to the frequency adjustment? But if you're only option is a single eq, would you just test different a few different frequencies and try to get them both within range? The manual mentions 400 hz and anything over 15 kHz.

 

[rob aylestone]

One thing. Tweaking bias by just listening is potentially plain luck. Your tech has tone generators, scopes and probably nowadays, analysers to see what your adjustments are doing. If you get it wrong you will have distortion, and possibly weird spikes and dips in the frequency response. By ear you only hear these when they get bad. If you paid the technical guy to align your machine to standard, moving away from that seems a bit of a waste of money, if you want an effect, not quality? If you have bought the test gear, then the world is open to you to tweak and fiddle, because you can always return to perfect and start again. Have you ever attempted to tune a piano? I know I can detune one for effect, I did it. Could I get it back again? No.

 

[ecc83]

Not in nearly the same company as Mr Beats! But, I re-biased a lot of cassette decks back in the day and these rarely had any kind of adjustable EQ. The bias process therefore was to record 333Hz* and 10kHz at -20 VU (or Dlby B level) and check that the two came out at the same level withing a dB or two. If 10k was a bit lower, less bias and if a bit higher, a tad more. Since most machines were only two heads this took a while!

Now, 10kHz might seem pretty low-fi these days but there was still a decent output at 15kHz as a rule. The problem was you see, the decks were often biased for a mid priced type 1 tape as delivered and the more discerning customers wanted better so I tweaked them for TDK AD or SA. Of course with Dolby B you really need to keep a flat response to at least 10kHz because any deviations are exaggerated.

Sooo long ago!

*In case anyone wonders? Not 1kHz as you would for a HS OR machine because the cassette time constant is at a much lower turnover point.

Dave.

 

[sweetbeats]

Thanks for you patience and for walking me through this.

Sure thing. Hopefully it’s helpful to you and others down the road.

The tech did a full set up. He actually put on new record and playback heads and calibrated and adjusted everything. I'm using the same scotch 207 tape that he used. So the machine is in top shape.

Okay. Fair enough.

I know the machine is set up to spec as the tech I worked with has extensive experience with these machines and is pretty meticulous. But maybe what is called for is just different than what I want from the machine. Maybe a slight emphasis on the high frequencies per tascam?

There’s no reason you can’t do that. A machine doesn’t HAVE to be setup to factory spec. As you get more advanced in knowing and using your machine you’ll find that’s one of the hidden beauties of using tape…between tape types/formulations, fluxivity standards, standard operating levels, bias settings and response curve tweaks there is a lot you can do with a tape machine. But it starts with understanding what all of those different facets and settings do, and once you have a grip on those things and get used to your flow and result you might go “I want more ____” and tweak your setup to lean toward that result. So there’s no reason you can’t nudge the HF record EQ, but, again, what you DON’T want to do is do it blindly. You need to establish a quantitative basis for where the machine is at and make incremental adjustments. Fortunately for you, you have a three-head machine, so you can monitor off the reproduce head in real-time as you make adjustments and listen.

I'm wondering if there is a slight bump in the high frequencies, then maybe the lower frequencies would sound less pronounced.

That’s the opposite of what I was suggesting. I was suggesting diminished HF response might bring the perception of a coincident diminished LF response because often there are HF elements to LF content, and without the HF elements prominent in the mix the LF content falls back in the mix.

The reason I thought it might have been under biased is because I had read another thread where there was a low frequency bump on the same machine and it was put out there that it was over biased, so I was just working in the other direction.

Well I don’t know who said what, but every tape machine has a LF “head-bump”…you can do a web-search on it to get the background of why it exists, but it’s part of why many people like tape, because there’s a natural LF boost, and some tape machines have it in a good spot for particular music. It’s not something you tune. It’s a fixed element related to the head design, tape speed, etc. And, again, bias level impacts distortion and HF response…LF response is not a factor relative to bias level. Period.

I'm mixing down from a tascam 388. I've thought about just bumping the low frequencies of my mix but I kind of have everything keyed in. I'm not sure I'm confident enough to make little tweaks to compensate for what I get when I monitor from the 388 to where they would translate to an even mix on the 22.

But…that…is the whole thing about mixing…that’s mixing…make yourself some cue sheets so you can document your baseline mixer settings (to be able to revert to them), run a mix, monitor off the reproduce head, tweak away! Hear your changes in real time off the reproduce head. Play. Experiment. Forget confidence. If you document your base mixer settings you can always go back to them like an undo button. And by monitoring off the reproduce head you don’t have to run a mix and play it back and try to fix what isn’t right…my goodness that’s the best part about mixing is experimenting. Try it out. Have fun.

But I'll take your advice and try the tone ladders. I do admit I'm not sure exactly what the process is. I'd venture to guess it's recording a series of frequency tones to the tape to check the response? Would you be able to walk me through it or possibly point me somewhere where the process is explained? I would definitely be curious to see how it's done, and as you mentioned it would give me important information about what the machine is actually doing.

Running tone ladders is just the same thing you do during the calibration step of checking the record frequency response. This assumes your tech already checked playback frequency response which you do by playing a tone ladder from a calibration tape, and verifying the playback electronics are reproducing within the range specified by the manufacturer. So actually, I’d ask your tech what their notes say about how the machine did with the playback frequency response check, verify it was within spec, and ask what tones were reproduced and what the levels were. They should be able to tell you the tones because they have the cal tape with the tones on it. And if they can’t remember specific levels per tone I would hope they’d have notes or recall generally how it did: “Yeah it was within +/- 2dB across the specified frequency range, head bump was around x Hz and it started to drop off at around x kHz”…something like that. But with that you know what you might expect by recording your own tone set to the tape you’re using and checking the response curve. You can just use the same tones your tech used. Take a blank reel of tape, inject each tone into the inputs, and while monitoring the tone generator signal at the input before recording, set the record level to 0VU. Then record maybe 10 seconds of each tone. At the end of the process you’ll have a “ladder” of tones recorded to tape. Now rewind to the beginning and reproduce the tone set and document the level of each tone represented on the meters. Do this for both the sync head and reproduce head. Feel free to share the results here. Does that make sense? And the quality of the tone is not critical, so you don’t need a low-distortion oscillator or anything. For this kind of thing I use an app called Tone Generator on my mobile device. Cheap, does varieties of noise as well as tone, different waveforms, nice interface for output level and frequency.

Please feel welcome to ask any clarifying questions.

One last question. You mention here the +/- range. This is in reference to the frequency adjustment? But if you're only option is a single eq, would you just test different a few different frequencies and try to get them both within range? The manual mentions 400 hz and anything over 15 kHz.

Hopefully it makes more sense after my last paragraph. It’s related to the tone ladders. Basically, on the playback side (from a cal tape, where the tones on the tape are at a precisely known fluxivity level), and on the record side (where you reproduce tones you recorded onto your tape), the manufacturer says the deck should be able to reproduce each tone within a +/- dB range of zero. In other words from x Hz at the low range to x kHz at the high range it should be able to reproduce from a cal tape or from tones recorded on your own tape no higher or lower a level than what went in (or is precisely at a known same level on a cal tape) than what is specified. Let’s see what the 20 series manual says…

Okay it says at 15ips:

40Hz - 22kHz +/-3dB @ 0VU
35Hz - 25kHz +/-3dB @ -10VU

That brings another question…you’re running your mixes at 15ips, right?

So let me know what questions you have, but you should be able to record tones at 0 or -10VU across the ranges specified above and the results on playback should be no more or less than 3dB above or below 0 or -10VU respectively. I don’t want to get too far into the weeds but note the service manual indicates to measure playback levels using an AC voltmeter measuring at the output jack for the respective track. Why? I suspect it’s because it’s more accurate. The ballistics on the Teac meters aren’t great at high frequencies. Like they can be as much as a dB off above 10kHz. But if you want to test things quick and dirty you can just use the meters, it will give you a relative sense of where the peaks and dips are and where things roll off…just be aware at high frequencies if the meter says -2VU that might be more like -1VU. And if you have a decent AC voltmeter (one that is accurate across the audio band), then you can always use that. But use a ‘Y’ cable at the output so the output is loaded (one part of the Y connected to equipment, the other being used for your meter measurement).

 

[illuminations]

Well, I've had a busy month with some family stuff I had to take care of, and unfortunately I haven't had the time to dig into this. But I'm ready to dig into this now that I have some extra time.

I am waiting on an adapter to do the tone stacks. Hopefully that will arrive tomorrow.

But I was able to check a few things earlier today. And now I have another question. Does the high frequency EQ only apply to the recording process? Or does it apply to the input stage (or more likely the output stage) on the 22? The reason I am asking is because I am noticing the high frequency while monitoring off the machine without any tape on the deck. I have a millivoltmeter (leader lmv181) and I just wanted to see how it would work passing signal to the machine and to the voltmeter. I used the tone generator app that you mentioned sweetbeats and was getting some interesting results. I initially tried running the tones to just the meter, then just the tape machine. The reading I was getting on both separate was pretty close across the spectrum (and pretty close to zero, though occasionally a few dbs higher or lower). However, when I ran the tones through the inputs of the 22 and used the outputs to send to the voltmeter the results were pretty different. They were consistent through the lower frequencies, but starting at around 2khz there is a noticeable jump (~3-5db) and when I got closer to the 13-15khz the jump got even bigger ( close to 10 db). I also recorded some music to the computer, then ran it through the tape machine first (just in and out, no tape) and into the computer. Again, I was noticing the emphasis on the high end. Maybe this is what the high frequency eq is supposed to correct for? I just figured it was applied during the actual recording to tape. But if you have any insight it would be greatly appreciated. I still want to do the tone stacks tomorrow. It just seems strange that when I run the tones to the inputs, the tascam's vu meters are pretty stable, but from the output section there is such a difference.

 

[sweetbeats]

The record EQ is just that…in the record signal path.

 

[ecc83]

I have mentioned it elsewhere but it might be moot to this section to mention the free comprehensive test tones made available in the May issue of Sound on Sound magazine? There are also several pages of how to use them.

Dave.