Talk:Print-through

I'm not convinced that the cited passage of Led Zep's song is a real example of print-through. Given that it's only audible in that portion and affects only the vocals (despite that there's a very loud burst of guitar in there), I'd be more inclined to believe it was an intentional effect. In fact I'm not even sure the pre-echo matches the main vocal, the delay between phrases seems to change. Chris 07:12, 27 January 2007 (UTC)

---

Hey, how can I prevent "pre-echo" (due to print-through) on cassette tapes? I understand the pros store their reels "tails out", but what's the cassette equivalent?

On a multitrack cassette tape, should I rewind all the way or fast forward all the way before storing?

I think the answers to these questions would make fine additions to the article.

---
 * I don't think there's anything you can do about it with cassettes. It doesn't matter whether you rewind it or not, the problem is caused wherever the tape is in contact with itself on either take-up reel. Thicker tape is less prone, so use C60s (assuming they do in fact contain thicker tape - I wouldn't mind betting that these days it's the same tape as a C90, cut from the same base material to save manufacturing cost). Make sure there are no stray magnetic fields that can act as a copying field (tricky, since the earth's magnetic field will have some small effect over time, and screening it out altogether is expensive!). Basically, if you want to preserve a recording, use another medium. Graham 23:02, 18 December 2005 (UTC)

--- That's interesting!!! I thought (I guess incorrectly) that one reel would cause a pre-echo and the other reel would cause a post-echo. But *both* reels would produce a pre-echo?? Weird!


 * Not really weird. The printing effect simply causes the signal to be copied across adjacent layers of tape, in either direction. Which signals end up lying on top of each other is a matter of chance - signals nearer the end of a tape will have a shorter pre- or post-echo when the reel circumference is smaller, longer when it is larger, so which end you wind it to allows you only to select the length of the echo delay, not its magnitude, nor whether it represents a shift forward or backward in time. To visualise this, image a tape blank except for a short burst of signal. Imagine that signal colours the tape red at that point. Now imagine that red block in the middle of a reel of wound tape. The colour will "bleed" to adjacent layers in both directions, weakening rapidly as it moves further out. When that tape is unravelled again, you will encounter weak red blocks in advance of the real block, and then again following it. You'll get both pre- and post-echo in other words, and that will be there no matter which reel it's wound on to. It may be that longer delay times are more tolerable for your particular music, which might explain the "tails out" winding, but it seems to me that's more superstition than science! In addition, cassettes can be turned over, whereas most open reel tapes are used in one direction only, so whichever end you plump form the other side of the tape will have the opposite. One other factor is the coercivity of the tape - I seem to recall that high coercivity tapes such as metal will have much less print-through, since the signals embedded in the adjacent layers simply isn't strong enough to have any effect on the tape next to it, so that might help. Graham 10:14, 19 December 2005 (UTC)

Very very interesting!!! Thank you so much for your insight. You really know what you're talking about and should expand the article to include this info. ALSO, are you a Usenet fan? You should bring this up in rec.audio.pro and discuss this because they seem to believe in the superstition we're talking about (pre vs. post echo).

--

Hey I just thought of something (to add to this post/pre-echo discussion). What if there is some kind of protective layer on one side of the tape? I'm sure reels must do this; I think TDK cassettes used to, but I don't know if they do anymore. Would this make a significant difference? Breaky McWind 20:35, 30 December 2005 (UTC)


 * I've not heard of a layer designed to reduce print-through specifically. Such a layer would have to have some magnetic shielding properties (like mu-metal). I'm not sure how this could work with a flexible and very thin tape. However, even if it existed, it wouldn't change the effect of the print-through in terms of pre- or post-echo, though it might effect the amount of it that occurred. The reason is - if you think of a stack of tape in a reel, the "protective" layer of a particular signal would prevent signal transfer in one direction, but the protective layer of the tape overlaid on the next turn would have exactly the same effect in the other direction. So any protection would work exactly the same both ways. It might reduce the signal print-through, but it would do so symmetrically assuming it worked simply as a barrier. I'm skeptical that any sort of shielding material exists (made from unobtainium perhaps?) but if it does, then it's almost certain it wouldn't have a one-way shielding property.Graham 00:57, 31 December 2005 (UTC)

Paper
Printing on paper, print-through is the visibility of the opposite page, a common issue in newsprint. That's probably the first use of the term, shouldn't it appear ? PolBr (talk) 08:44, 20 December 2018 (UTC)

Description?
There’s a wealth of information here about cause, physical description, and solution, but not one example of what the phenomenon sounds like. Can someone with expertise please provide a description or, better, an audio clip?  PacificBoy  01:34, 24 March 2019 (UTC)

Software Removal / Approaches?
Aren't there any approaches yet for plugins or software filters to eliminate the print-through copy effect? Once you define the time shift manually and also the approximate level, phase inversion could cancel the echo within a limited range. Of course, the time shift varies as the wound tape diameter increases, hence it might take K.I. algorithms to track the ghost copy. I can't believe there is no R&D for this? 95.116.33.226 (talk) 19:05, 23 March 2023 (UTC)