Skip to main content

Hi, well I wanted to try out the analog tape experiment, so I went out and bought a Tascam 32 1/4" reel-to-reel mastering analog tape machine and started experimenting. Oh my god it's so perfect! In all my 4 or 5 years of experience recording in the digital domain I've never EVER come across this type of sound, it's like having human ears as a recording system, I couldn't understand it! Now I totally believe now that analog recording is an essential part of the profesional recording process, but It's sure that digital mixing and editing is totally necessary as well. Of course in the experiments I made I was listening directly through the headphone outputs of the Tascam machine, so here's my question... If you track in analog and then import into digital (for example Cubase SX in 32bit), will this process affect my beautiful analog signal in a negative way? if so, how? and how can I work around it?

Comments

RemyRAD Mon, 10/29/2007 - 23:42

Hi, Cosme!

In many ways, listening to analog tape playback is much more like listening to the output of your console than the output of a CD. Like riding in a Rolls Royce! In spite of the tape noise, wow & flutter and analog tape's nonlinearities, there isn't any detrimental filtering. It's that lower sample rate of 44.1kHz and 20kHz brick wall filter that makes digital sound like digital and not like analog. But starting with that analog sound, you get to capture THAT flavor before the horrific brick wall filtering. So analog, transferred to digital, sounds more like analog than a pure digital production. Just listen to all of those hits from the 1970s and eighties, which were later transferred to CDs. You still get that analog glow.

In the land of digital recording, you will approach a more acceptable analog like quality by recording at the higher sample rates of 88.2, 96 and 192kHz. That's because with every incremental advance, the brick wall filter can be moved to a higher frequency. Rendering its adverse effects less audible in the perceptible bandpass. But you get shoved right back to go again when down sampling for CD, MP3, i-pods and pod casts at 44.1kHz sampling rate. Thankfully, you still retain that nonlinear saturated analog sound, along with the horrible brick wall filter.

What I recommend to everybody is if you want that analog sound, don't record analog, rewind and playback. But actually monitor playback while recording and transfer to digital in real time. This will cut your wow & flutter by half. While also preventing "print through".

Now, you didn't mention whether you purchased this machine new or used? Nor, did you indicate what type of tape you set the machine to record with? Analog machines are not like digital machines. And analog machine needs to be properly set up to the type of tape that you are recording upon. This is determined by both the tape manufacturer and the machine manufacturer, depending on bias frequency, head the gap and other factors. Plus, one needs to know whether they like the sound of peak bias or over bias. Both methods present different textures to the sound. One never really wants to under bias. And, if your bass guitar sounds like you recorded it while pouring gravel in the backyard, your bias settings are not optimized for lowest modulation noise. That's the trickiest adjustment of the record bias.

And before one sets the record up, playback needs to be properly calibrated with the use of highly specialized reference calibration tapes and appropriate test gear. I've never recorded on an analog machine out-of-the-box. Aligning an analog machine is an art of science and compromise. Thankfully, you'll not have to align the 24 track analog machine for each session coming in and going out the door for each and every client. Those were the good old days. Or was that the gold ole days? And just remember, if you happen to get one of the playback alignment tapes, DO NOT ALIGN THE LOW-FREQUENCY PLAYBACK EQUALIZER'S TO THE PLAYBACK REFERENCE TEST TAPE. Those are to be aligned only when doing the recording to playback response. To align playback response, you only use the frequencies from 500 to 20,000. The low frequencies are only used in a mono, full track situation due to "fringing effect" on anything with more than a single channel, i.e. 2 or more. Otherwise, you will never be happy with the low frequency response on record to playback. A mistake I've seen way too many times by some actually very fine and educated engineers. A little-known information factoid that frequently had to be handed down from somebody who knows. Even to my engineering manager and chief engineer at NBC radio in Washington DC. Go figure? Thankfully, I'm a fossil.

Stoned again
Ms. Remy Ann David

TVPostSound Tue, 10/30/2007 - 08:21

Plus, one needs to know whether they like the sound of peak bias or over bias

My opinion:

1K peak bias always sounded better on full range recordings like music.
On television programs which much of everything under 80Hz was rolled off,
10K 3 dB over bias worked very well.

Remy,
Dont forget about Azimuth, Wrap, and Zenith!!!!

If you're a fossil, Im a dinosaur!!

RemyRAD Tue, 10/30/2007 - 12:18

Cosme, I think you will not want to monitor with the sync option engaged. This function is that of a multitrack recorder that allows playback through the record head, on an adjacent channel to another that you are recording upon. And, so, when engaged, playback through the record head makes it possible to add additional material to the adjacent track with full-time synchronization. You want to be in "repro" playback mode while recording. This is the setting that we use to monitor the recording after it has passed by the recording head and is then being picked up approximately 80 to 100 ms later (assuming one is recording at 15 IPS). The repro mode is also what we must use when aligning the record section, otherwise, record set up could not be done in real-time. This was a huge problem on the world's first low-cost 16 track 2 inch Scully model 100 recorder. A substantial cost savings was had by only supplying that machine with 2 heads. So on that machine, you were always playing back through the record head in "sel-sync" (selective synchronization) mode. It was fine if you were not always tweaking the machine's record electronics on a daily basis. Otherwise it was a dog to align as it could not be done in real-time.

And of course, another highly educated person in the know has indicated that I had not said anything about physical head adjustments.

Heads required numerous adjustments such as azimuth. This adjustment dictates the angle of the head gap. The head gap should ideally be exactly 90° perpendicular to the horizontal length of the tape. With a multitrack machine, such as a stereo on up to 24. When the azimuth is not correct, terrible phasing issues occur. This is because smaller high-frequency signals will be picked up a few microseconds apart. This sounds dog awful. No offense Davedog as we all know you are a quality engineer. And so Davedog is synonymous with quality as opposed to improper azimuth, which is the most commonly made of all head adjustments as height, Zenith, wrap are usually only acted upon with head replacements.

Zenith is the adjustment of how flat and parallel the head is in relation to the tape. That is to say, if the head is angled back or forward, the top or the bottom of the head will receive greater uneven wear, since it is pushing into the tape unevenly. This can result in the head being worn out on the left track or track 1 and not worn badly on track 2 or, track 24, depending upon the machine type.

Of course there are also height adjustments that when made, will also affect that Zenith adjustment.

Then there is rap. Which is a bunch of guys with awful microphone technique. They should not be confused with wrap that, when also done incorrectly can disappoint your children at Christmastime. This too should not be confused with wrap which is the rotational alignment of the head gap to the tape traveling over the face of the head.

All of these adjustments require the knowledge and dexterity akin to a watchmaker. Otherwise, no cigar.

Thankfully, head technologies have changed over the years from the soft ferrous metals that were once used. To the almost indestructible ferrite glass faced heads in common use today. But those too can be damaged if improper head cleaning solutions are used. The best head cleaner is now no longer available and that was the refrigerant Freon. A completely inert, quickly evaporating, nonconductive, Dow developed chemical. In the absence of Freon, only 91% alcohol should be used. NOT 70% RUBBING ALCOHOL. That stuff contains other ingredients that's not really conducive to head cleaning.

And then there's this important demagnetizing stuff. The erase and record heads are virtually self demagnetizing albeit not completely. The real problem occurs with the playback head. This is because the playback head is only designed to pick up extremely small magnetic fields and after a period of time and/or north to south traveling, the playback head gets magnetized. When it gets magnetized, it begins to wipe off high frequencies on the tape every time the tape passes by the head. So an external de-magnetizer needs to be utilized to de-magnetize the playback head on a regular basis. Especially, just prior to alignment with those expenses calibration tapes by companies like M. R. L. a.k.a. Magnetic Reference Laboratory and STL a.k.a. Standard Tape Laboratories and others that are lesser-known.

I think I need to resurrect my old tape recorder alignment video that I made back in the early 1980s? This is where I described and demonstrate proper tape recorder maintenance and alignment. Not even sure if I even have that video anymore, which may necessitate my production of a new video?

So, how many people here, would like to know proper analog tape recorder and maintenance procedures?

I'm smoking my Montechristo Cuban cigar while waiting for your replies. I do that because it's sexy.
Ms. Remy Ann David

RemyRAD Tue, 10/30/2007 - 21:02

Sure, TVPostSound, the voice on the M. R. L. test tapes is that of Jay McKnight. The founder and owner. The same person who reamed me a new ass hole for calling him up and telling him that his latest series of fast sweep calibration tapes that he sent to me at Scully were all defective because he changed his positive going scope trigger pulse to a negative going trigger pulse. Later to call the next day, apologizing profusely to me and replacing all of our calibration tapes with properly mastered versions.

Yes about that tiny Mr. Weber stuff.... What's your favorite flavor 185? 250? 320? Or should I say Scotch 111 is to 185. Like Scotch 206 is to 250. Like Scotch 250 is to 320. Or substitute your favorite Ampex or Agfa equivalents. Frankly, I couldn't stand the smell of Ampex tape. I liked the consistency of the slit of the Agfa 468. But that 250 at 6 DB over 250 nano-Webers, when biased for lowest modulation noise utilizing a 10 hertz sinewave while monitoring repro/playback, through headphones at 30 IPS was fabulous. No bias rocks and superclean bass tracks. OK, OK so it pretty much worked out to 3 DB over 10kHz at 15 IPS. But watch out for that print through that's where the Agfa 468 was really super over both Scotch and Ampex.

It was pretty cool to see, at the AES in NYC this year, the couple of new 2 inch tape manufacturers. Both have their special new secret recipes with 11 secret nerds and Spice Girls. Rolling barefoot without nose reduction. Finger picking good! Mmm mmmmmmm.

Oh yeah Cosme, you really can't make this work in an overdub situation. This would be fine if you're tracking initially. Say you were recording a violin demo with piano accompaniment in your studio to 2 track. This would work in real-time since the performers are not wearing headphones listening to play back while playing.

In that respect, one could roll 2 inch 24 track in playback/repro while recording the basic rhythm tracking sessions. You route to your 24 track analog machine, through that to your 24 track digital machines input. That's where you really get the advantage of the saturation mostly on drums and instruments with significant transients.

Otherwise, you could glean some advantages by simply taking your purely digital production and instead of using saturation emulators like the new Neve thingy, you have the real deal instead. Basically, you would in effect be "re-amping" your digital signal, decoding your stereo digital mix to its analog output, then on to analog tape input and re-encoded back to your digital 2 track input, for obvious purposes and assorted release formats.

I actually met Colonel Sanders when I was a kid. Love that chicken, from Popeye's also since I also like spinach. Even though they don't serve any. Just remember kids, don't play with recording tape if you're eating fried chicken. The tape already has plenty of silicon lubrication and doesn't need chicken fat. But anything with chicken that is good so, disregard the part about playing with the tape with greasy fried chicken fingers.
Ms. Remy Ann David

TVPostSound Tue, 10/30/2007 - 22:29

Agfa? Really?? That stuff fell apart on us after 2 years!!
But the Scientologists loved it, and always brought their own!!
I hated the smell of Agfa.

I used 456 on an MM1200 for so long I could not get happy with anything else!! 250nW +4 EOL, 15 ips, 1K peak bias, Dolby A.
The last analog was 499 on an MTR90II, then we stepped up to 320nW +6EOL 15 ips, 10k +3 overbias, again with Dolby A, then SR

Wow, I remember all this stuff!!

And the only Scotch I liked was 1/2" or
the stuff that was aged in an oak barrel!!

So Remy, I was taught to edit tape with a dull blade by a guy who owned/operated a truck in the D.C. area 80s/90s.
But did post on his productions here in LA. He showed me how it could soften the edits. Did you ever run into him??

If you dont remember, Ill tell you later!!

RemyRAD Tue, 10/30/2007 - 22:58

I think the remote truck from the mid-late 1970s you're talking about was either Sonority Recordings truck, where one of the partners, Roger Byrd, was also from NBC who trained me for Meet The Press but I don't think that's who your talking about. Or, if it was the late 1970s, early 1980s then Eddie Eastridge of BigMo Recordings. Eddie had a custom built audio console which utilized the Allison/Valley People trans-amps. The Sonority truck had a Neil Muncie S. S. I. (Suburban Sound Inc.) Melcor/API console with rotary faders and was in a step van. Where Eddie had a Ford front-end and a customized 24 foot box.

The best remote truck in the DC area with a vintage Neve
Ms. Remy Ann David

anonymous Wed, 10/31/2007 - 05:58

demagging how often

This is a good topic. More please, Remy.
It's getting harder to find engineers to calibrate machines, especially here in the UK.
How often should demagging be performed? I'm using a Tascam msr 24s, and recently I was doing some rough mixes. I got a great snare sound on one track. I had to set up for another track and lost the mix (no recall or flying faders). When I returned to the original track I just could not get that snare sound.
After some thought about it I decided it was time for a demag. As soon as I did that the snare came back to life.
It wasn't just top end that had died, there was depth missing from the whole track, and I was getting good peak readings on the VU meters but it didn't sound as big and loud.
I know mine is a budget machine but I'm getting really nice results. I've just got to remember to regularly demag, because it happened quite quickly.
So how often should it be done?

bent Wed, 10/31/2007 - 08:09

That seems a little complicated I didn't quite get it, I should use a delay after the machine's output signal? Wouldn't that make the latency even longer?

Cosme,

You're right, I've been trying to wrap my head around this and cannot come to a logical conclusion (they have yet to invent a machine that knows what you are going to play before you play it). :roll:
The latency would increase, I gave you some serious misinformation there.
Sorry for confusing you!

bent Wed, 10/31/2007 - 19:34

If you track in analog and then import into digital (for example Cubase SX in 32bit), will this process affect my beautiful analog signal in a negative way? if so, how? and how can I work around it?

Cosme,

I hope this helps, being that I totally misguided you with my last couple of posts:

I think that tracking onto the tape and dumping it into Cubase will be perfectly fine. I've done the same myself and have experienced no truly negative results, as far as sound coloration and what-not.

The only thing I'd caution against is - if you are going to track onto the tape, while playing along with something you've already recorded in the digital world, you should keep in mind that more than likely the tape is not going to playback at the exact same speed that the digital tracks do, which means that you may need to use a plugin to adjust the timing of the analog tracks once you dump them into Cubase.

-side note:

Being that it's Halloween, it seems that the board is a bit dead at the moment...

IIRs Thu, 11/01/2007 - 06:58

RemyRAD wrote:
Oh yeah Cosme, you really can't make this work in an overdub situation.

Yes you can. All the DAWs I know allow recorded audio to be automatically offset by a predetermined amount to compensate for the latency of the audio interface: you simply need to set this to include the additional latency from the tape recorder.

JoeH Thu, 11/01/2007 - 09:21

If you're tracking in/out or to/from the tape machine to/from the DAW, via the input to repro heads, you don't have speed issues per se.

You're simply using the tape-printing as an effects loop, but you're essentially only dealing with whatever timing issues may be involed in such a tiny distance between the heads, running at 7.5 ips. You're not storing anything on the tape per se, so there wouldnt be speed issues. you're simple going in and out of the analog tape world for as long as it takes the tape to move from one head to the other. As long as you're printing it back into the computer digitally, there's only the actual latency of the proces itself (plus the tape path travel time) to deal with. You can (as others have already pointed out) simply do an offset with the new "warm analog" track after the fact.

Ah, the smell of warm 486 in the morning........ :twisted:

anonymous Thu, 11/01/2007 - 10:01

Yeah the problem really comes out when you're overdubbing, the 100ms difference is noticible when you're playing along, that's the actual problem, imagine that you're recording a guitar track trough the tape machine and you're using a pre recorded drum track as a monitor, you play the guitar and 100ms later, it's recorded through the DAW because of the tape path, so you hear your guitar signal 100ms later and it's recorded 100ms later. That's the actual issue, and the only solution I can find is reamping, the offset option is the one I don't really understand, but it seems insteresting

IIRs Thu, 11/01/2007 - 11:32

Cosme wrote: the offset option is the one I don't really understand, but it seems insteresting

Your DAW already has to cope with latency when overdubbing. eg: lets say it takes 2 ms for any digital audio sent to the audio interface to emerge as an analog audio signal, and lets say it takes another 2 ms for an analog signal hitting the inputs to arrive at the software as a digital audio stream... if this is not compensated for then the musician's will play 2 ms late, and what they play will be reorded another 2ms late, so the resulting audio clip will be 4 ms later than it was intended to on the timeline.

The only difference when using an analog tape machine is that the input latency is now more like 100 ms than 2 ms.

Tracktion has an auto-detect feature to help you set your time adjust parameter correctly: you physically connect a cable between an audio out and an audio in, and Tracktion will play a series of test pulses and measure the delay. But you can do this manually if you need to: loop an audio output back to an input, then play some audio on one track while recording it back to another. The time difference between the recorded track and the original is the latency that you need to adjust for.

RemyRAD Mon, 11/12/2007 - 11:22

Rosemary, thank you for the comments. I haven't seen you here lately.

I'm trying to work out the time to sit down and come up with a new tape recorder maintenance and calibration video that will utilize some common audio software such as Adobe Addition with its built-in analyzers and tone generation capabilities to utilize as a simple test jig set in lieu of tone generators, volt meters and oscilloscope's. And because I still have some NOS full track mono head stacks for my custom-made Scully 280B, I'll probably also offer a simple calibration test tape that will be much more affordable than the STL or M. R. L. test tapes. Still, a fairly time-consuming project but obviously some call for. Hopefully I'll try to have it ready by Christmas? If not early in the new year?

I'll let you all know when it's ready.
Ms. Remy Ann David

x

User login