Skip to main content

Hi,
There is something I've always wondered about that I've never been able to find any info about.

On many recordings (particularly I guess 60s jazz recordings) I hear this thing happening where, if there's a very quiet or silent section followed by a loud entry (such as a trumpet coming in), I'll hear the trumpet line come in early and very quiet, before it comes in normally about 1 second later.

Does anyone know why this is? I used to think it was something weird about something I was doing, until I realized it was happening everywhere, and on different recordings of the same source material, making me think this must be in the source. Does it have something to do with stereo tape recording or something like that? I've always wanted to know!

If anyone knows what I'm talking about, please enlighten me.

Here are some examples:
Eric Dolphy - Out to Lunch - "Straight up and Down" at 7:30
Miles Davis - Miles in Tokyo - "My Funny Valentine" (all throughout the intro -- great example since it's a lot of nearly unaccompanied trumpet)

Thanks!
Dave

Comments

anonymous Sat, 01/29/2011 - 13:21

I would wildly conjecture that if you are listening to vinyl, or if the recordings were taken from vinyl, perhaps you are hearing a mechanical bleed through from the record groove cutting process. It does not seem impossible that the imprint of strong transient material could affect an adjoining silent or relatively quiet groove during cutting, such that bleed through occurs. The fact that you say the delay is one second is a clue. Perhaps you can more precisely measure the interval. If it turns out to be about 1.8 seconds, this would seem to indicate a 33rpm album as the source. With a stopwatch and several trials averaged out, you should be able to get a more precise value. Or better yet, transfer the most obvious example into your DAW, and just read the interval.

Maybe this kind of bleed through or ghost image can occur on certain studio tape machines running at slow speed, where heads other than the main playback head anomalously feed or pick up signal. Again, I would guess a precise measure of the interval would provide key information to those with experience in using tape.

anonymous Sat, 01/29/2011 - 13:41

Hi,
Thanks for your response, I realized after posting that I should have indicated the sources.

In all the cases I've noticed this, it's been on digital media, either CD's or FLACs or MP3's made from CD's. But that's not surprising since that's how I listen to >90% of the music I listen to.

In the case of Out to Lunch and Yusef's Lateef's Eastern Sounds (another recording I've noticed this on) I've heard it on multiple CD pressings including the highly regarded Rudy van Gelder remasters of each. Both albums were originally recorded by Rudy in the 60s in his Englewood studio. I'm guessing the remasters were made from original tapes? Or copies of original tapes? Certainly not from vinyl though, right?

I'm sure Rudy wasn't recording on junk, but would the technology back then have been susceptible to this sort of thing? It's certainly interesting, because these are two albums that are considered among the best ever, and recorded by what many would agree is the greatest engineer in jazz.

Thanks,
Dave

anonymous Sat, 01/29/2011 - 13:53

I have tremendous respect for the sheer technical facility of many jazz players, but I must confess I am not intimately familiar with the genre. Using albums as a mastering source is not unheard of, and it is my understanding that vast swaths of older music must be re-rendered from vinyl, since this is the only (best) source that exists for many recordings. I suppose I made the rough connection of the jazz you were listening to, to blues, wherein again, some CD's of today would have been rendered from vinyl.

I do not think it is a question of quality. I trust someone familiar with tape will jump in and clear things up quickly.

I too find these kind of anomalies and puzzles fascinating.

anonymous Sat, 01/29/2011 - 13:58

Cool ... haha btw, to explain why I said what I said, I somehow misread your phrase "studio tape machines" as "stupid tape machines" ... :)

Anyway, yeah I'll be interested to hear what a tape expert says. I was kind of surprised that I couldn't find lots of info online on this topic, since as I said, these are well known records.

Thanks again!
Dave

anonymous Sat, 01/29/2011 - 14:01

Actually, I guess this is the answer: [[url=http://[/URL]="http://en.wikipedia…"]Print-through - Wikipedia, the free encyclopedia[/]="http://en.wikipedia…"]Print-through - Wikipedia, the free encyclopedia[/]

I suppose I didn't quite know what to search for ... but this seems to correspond to what I'm hearing, and I would guess it's there on the original tapes.

Big K Sat, 01/29/2011 - 14:07

That bleed often comes from the 1/4 inch master tape the mixes have been recorded to. Especially when it was forgotten to store them " tail out ".
It is a slight crosstalk of the magnetized tape with music signal onto an empty or quiet tape strip, mostly before the song starts.
It could be removed when re-mastering (digitally & analog), but sometimes it is not for various reasons.
That effect is not existing with digital tapes. It would be out of synch and is being ignored, filtered by several stages...

anonymous Sat, 01/29/2011 - 14:14

Tail out? Can you elaborate? Are you saying the tendency to imprint depends on which way the tape is wound on the reel? If so, wouldn't that mean the ghost imprint would instead be stored after the transient instead of before, and that it might still be exist in silent or quiet passages if they follow quickly after loud transients? But then again, in this configuration, the loud transient might render our hearing momentarily threshhold shifted such that it would be less able to detect the trailing ghost sound than if it had occured in front of the transient.

Big K Sat, 01/29/2011 - 15:01

It has more then one good reason: Tail-out storing is minimizing the tape tension during storage and ensure an even tape pack.
After listening to the tape, don't rewind it, because fast winding usually puts more tension on the tape.
The tighter the tape pack, the higher is the chance of a cross magnetization or print-through.
Also Tail-Out gives you a post echo rather then a more annoying ghost echo before the start.
As well, the magnetizing effect is lower since the magnetic energy has to travel through the carrier material first which weakens the crosstalk quite a bit.
And then, the magnetization ist strongest in the first few minutes when stopped and gets lost to the amount of 6 dB when you rewing the tape to start.

All this makes storing tapes tail-out a good idea when dealing with valuable tapes and cassettes.
Mind you, it is no guaranty that there wont be any crosstalk, anyway...
Some tapes we had to bake before transfering them into the digital domain showed evel crosstalk eventhough they were stored tail-out.

anonymous Sat, 01/29/2011 - 15:32

Thanks for the interesting and helpful answer. I highly doubt tape multitrack is in my future (due to the convenience of digital recording) but I have thought about getting a simple reel to reel for mono or stereo tracking, just to learn more about tape and its characteristics like saturation, etc., in which case, that information will be most beneficial.

Big K Sat, 01/29/2011 - 16:13

LOL..
Saturation you don't want to know about other then to avoid it as good as ever possible.
What you are after is tape compression...
If you are already in the digital domain, try the new UAD Studer A 800 tape simulator...
That gives you excellent idea of what tape sound is all about without getting caught in kilometers of tape.
Oh, how I hated this editing of hour-long audio reports recorded onto quarter inch tape.

RemyRAD Sat, 02/05/2011 - 04:02

The echo phenomenon is referred to as "Print Through". That's because it basically prints through from one layer to the next. This causes both pre-& post echoes but it's generally the pre-echoes you hear most of the time since it occurs before the opening note. And the more heat that the tape is stored in, the greater the print through can be.

Now there are also artifacts within highly compressed audio files, such as MP 3, flac, others. These artifacts can appear like underwater echoplex echo and other lovely effects such as flanging where there was none intentionally added before. it really sounds great on string quartets.