Tape head alignment is relatively! easy ...
The below article (from the late Central Florida recording engineer, Bill Vermillion) used to be on-line, but I can't find it anymore. I kept a copy as it highlights one difference in modern recording, with how it used to be. My guess is that each engineer had their own protocol, and their own 'best way' of doing things.
For my part, the only things I never miss in consumer hi-fi are open reel tape/machines, and FM tuners. May they both burn in hell.
________________________________________________________________
This is an article on tape alignment that comes from doing this more times than I would care to admit, on 1, 2, 4, 8, 16, 24, and 32 track machines!
Of course be sure to clean the tape path before starting.
De-magnetization (every time) is optional in my opinion. There are some that insist it be done every time, but in our studio we felt that it was easier to introduce problems with it. We also checked the machines with a magnetometer, but typically we thoroughly demag'ed the machines only about once every 1 to 2 months.
Some machines were re-aligned several times in one day. This was because we would be working with source tapes recorded on different media, in different studios at different operating levels.
I had one machine that I did a complete re-alignment on three times in the course of one day. Two different 24 track tapes, one Dolby, and a 16 track setup. After doing this that often we got to be able to set up a 24-track machine in 30-40 minutes, which is about the length of time it took me to do my first alignments on two track machines.
Be sure to use a good monaural alignment tape. (They should all be that way.) There are some multi-track tapes that are recorded monaurally and then have guard bands erased after recording. This will make a difference on low speed playback EQ - but I use a method that eliminates any problem with low-frequency fringing effects.)
The steps used to align are as follows and
are to be done is this order:
1. Physical head alignment (azimuth alignment)
2. Playback level
3. High frequency playback response
4. Bias adjustment
5. Record level
6. Record high frequency response
7. Low frequency playback
Standard procedure for record head alignment was to align play head, and then adjust record head while during recording watching the playback head. Most multi-track machines and a good many two-track machines now have the ability to put the record-head into the playback circuit (often called Sel-Sync*). This is a much more accurate and preferred way to do physical head alignment.
The first tones on the alignment tape are HF alignment tones. If you have a scope you can adjust by watching the phase. If not available you watch for maximum output at this frequency. I had the luxury of having phase meters at the studio, this makes it all so much easier.
Playback level is then adjusted. Alignment tapes are manufactured to different recording levels. Assuming you have a tape for the level you wish to record, adjust the playback levels to a 0 reading on your meter. To be precise however you should look at the tables for fringe effect that should be included in the manual with the alignment tape.
Fringe effect is a higher output caused by playing tapes that have a wider record track than the width of the playback head. At our studio we compensated for this as we had standard two-track machines, and we had "stereo" machines. The latter use the European spacing for 1/4" two channel tapes of .75mm. The standard two-track spacing is 2mm. If compensation was not made the tapes will vary in playback when played on a different two-channel format machine. Most studios ignore this, however at the higher speeds it is important. At 30 ips the fringe effect will give a 1db too high a reading at 1000Hz! This will throw your alignment off by 1db at the high end of the spectrum.
Once playback level is adjusted it is important to MAKE SURE that this is not touched during the remainder of the alignment procedure.
At this point you can then make your HF playback adjustments to give you flat(est) response from your alignment tape. These playback calibrations are critical as all the rest of the alignment depends on them.
Now that playback is done the next step is bias, which will affect level and record response.
For bias I ALWAYS adjust for minimum modulation noise as that to my ear is the worst part of mag tape.
To do this pick a nice low frequency tone (I always use about 7Hz) and listen to the noise components when playing back. Cut the bottom end of your playback amps if you must because you only want to hear the noise components.
As you start below the bias point you will heard hf noise pulsed at the 7Hz modulation frequency. As you increase the bias the noise will decrease to a point and then start increase with a change in the "tone" of the noise. Go back to the minimum point. If you do this, you will find that you can set the bias more accurately by ear than you can doing the typical 1-3 db overbias at 1kHz or 10kHz (depending on your choices).
The only tape I have found that is cannot be done on is the Agfa 468. Minimum modulation noise will overbias the tape 6 to 9 db on this tape.
I had a machine shut down as I overloaded the bias amps when trying this.
(*NOTE - Since this article was written, new tapes such as the Ampex 499 have become available. I do not know if this procedure will work properly with them. wjv)
Many recommend 1db overbias at 1kHz. I find that you should use 10kHz for anything less than 15 ips, and for 30 ips use 20kHz. Typically you overbias 3db at these frequencies, but if you have the chart for your particular brand of tape you can find the exact point. However, except for 468 I always use minimum modulation noise. Agfa recommends a 14Khz audio signal for 468, and I don't remember the exact amount of overbias. They will furnish to you if needed.
After the bias is set, the next step is record level. All you do at this point is set the output to be 0 vu by matching the output level to the level you had previously set on playback of the alignment tape.
At this point you can calibrate the record meters if your machine has record calibration controls. Since the PB level has been matched you go to the record metering side (up to this point you have been watching playback metering).
You then adjust the record CALIBRATION (not level) so that the input levels read 0.
Now that you have the record level set you can adjust the HF record response. The standard that many use is to set 10kHz to be at the same playback level of the 1000Hz tone. I personally disagree with this. I try to make the 20kHz level 0 IF the 10Khz region does NOT increase too drastically at this point. In other words, if I can get 20hKz to zero and keep 10Khz no higher than 1.5 to 2db I'll go that way. I would much rather have SMOOTH wide response, than FLAT response that falls off at the high end.
Depending on the ability of your machine, you MAY have to make the high frequency record alignment adjustments at a lower level. Because of the high frequency pre-emphasis some machine's record amps may not take too kindly at 0db at 20Khz before being overdriven. On the other hand, some pro machines, Studer in my experience, have absolutely NO problems at recording at 6db over standard level at 20kHz and above when running at 30ips. The slower the tape speed the more critical this becomes.
Since we have now calibrated the record level against the play level, if your machine has problems in this area, we can now change the levels. Turn the record level down about 10db. Turn the playback up so that your output at 1kHz is zero, and then proceed to do the HF record EQ as this level. If your machine requires this you should also make sure that you are careful not to "go into the red" when recording on this machine.
One point I disagree with in most the procedures I have seen, is that many say to touch up the alignment (mechanically) after this you have completed these steps.
I have seen that you can change the physical alignment of the playback head or record head at this point and further peak the output (or bring the phase close if monitoring phase).
However, since you have aligned with a stock tape originally I disagree with this. (And this is my own method that we used at the studio - and it seems to work. However I have not had the time or opportunity to test this theory of mine so I may be way off base).
After all this check the phase response at 20kHz, and carefully adjust the bias on one of the record channels to bring the phase to as close to 0 deviation as possible. You may have to touch up rec eq just a bit because of the bias change, and you may have to do this once or twice.
I attribute this phenomenon (In my head at least) to the "bias bubble". The signal is recorded on the trailing edge of the record gap, and the amount of bias will affect just exactly where the trailing edge "seems" to be. This varies with the frequency being recorded. Changing the bias while watching the HF phase will show that this does affect the phase response.
EQ doesn't enter into it. Changing bias will affect the HF record response however, but this should only be level sensitive. The only thing I can think of that would account for this is the slight displacement in the track caused by bias. (As I say - this is my own "theory" and I have never had the time/resources to check it thoroughly - so I may be way off base, but record/play sounds really great done that way. I guess all of us a permitted to have at least one eccentricity
).
After you have done the final hf rec eq, you do the lf pb eq. Never set the LF playback eq from the alignment tape. Except for a very few machines, you have NO control over the low frequency record characteristics of your machine and you want to align your LF playback to your LF record.
Doing the above on an ATR102 - I could get the machine flat within +- 1db from about 30Hz to 20kHz. Tweaking the bias, as I mentioned above, I could typically pull in the 20Khz with less than 5 degrees phase shift. I really loved having a phase meter, as opposed to trying to interpolate on a scope. On a Studer A-800 at 30ips, I could get within about +- 1db from 30Hz to well past 20kHz. The Studer 3db down point was at 33kHz. (One hell of a machine!)
If you align your machine very carefully you will hear great differences between tape brands/types.
I have not had a chance to hear the new Scotch high output tapes, but of the rest of the tapes these would be my choices.
For such things as strings, but without any high level peaks, Scotch 250. Wonderfully quiet. Has more print than I'd like. That's why I stay away from this tape on music with large peaks or big endings. You will get echo on the first playback. There is very little modulation noise on this tape.
For voices, horns, acoustically generated music, Scotch 226. Doesn't print like 250, very low modulation noise, minimal asperity noise. Good tape.
For pop/rock. (Things that don't have big gaping holes (rest) or low level vocal tracks) Ampex 456. Seem a bit "brighter" or "harder" than the Scotch tapes. Great rock 'n roll tape. But tends to have more asperity noise (but only noticeable in quiet passages).
All the tapes measure the same (in frequency response on a given machine) but all have slightly different sound characteristics.
(For those who are not familiar with the term "asperity noise", this is a low frequency noise component. We called it "rocks" in the studio because it is what you would imagine large -really large- boulders to sound like hitting one another. This is cause by slight uneven-ness in the oxide coating. You hear it when you have such things as a soft vocal group - or soft horns. Virtually inaudible unless the music has pauses. More noticeable the higher the recording speed).
And to explode one "myth" here. Worn tape head don't ALWAYS show up by having a degraded HF response. In pro machines the heads have a very deep gap. As the head wears the depth of the metal is less while the gap is the same width. The tip-off here is that you INCREASED HF output.
I found this out when fighting a problem on an Ampex MR-70 and Ampex tech support pointed this out to me. (The MR-70 has to be my favorite all time tape recorder - followed by the AT102 for 1/4" and the Studer A-800 and Stevens 821-B for 2". The latter is a true 'hackers' machine. You have to KNOW the machine to love it, otherwise you'll hate it.)
So while the above procedures may not always follow the book, they are the ones we adopted in the studio and the machines always sounded good.
A lot of this information came from trial and error, but to give credit where it is due, some of the best information I have received working in the business came from conversations with John Stevens, who built an amazing tape recorder, John French, of JRF in New Jersey, who remanufactures and builds magnetic tape heads, and Gordon McKnight, of Magnetic Reference Laboratories, who make MRL alignment tapes.
* Sel-Sync is a Trademark of Ampex
(Copyright 1991 & 1995 by W.J. Vermillion.)