So I have these old audio cassettes of my old band playing a gig and I want to transfer them into my computer. The problem is–I need to get hold of a good quality cassette player and I don’t want to buy one just for this project.
I could borrow or scrounge up some sort of player but I want to use a high-quality one. Anyone know if it’s possible to rent one of these nowadays? Anyone have any other idea how I could get a hold of one?
Now for the “fun” part:
One of these cassettes was a used tape at the time of the recording (note to self: :smack:!) and the tape recorder we used apparently had a dirty/faulty erase head. When we played it back it had the recording of our gig along with remnants of the previous recording bleeding through below. Really annoying.
It occurred to me that once I have it in the digital realm I might be able to do some sort of phase cancellation of the unwanted, bleed-through recording. I could play around and maybe figure it out, but I’ve never done this particular audio trick before. Anyone have any suggestions?
The unwanted audio is commercial recordings of common music so it won’t be a problem to get the source material in order to prepare the phase cancellation (or any other nifty trick you guys dream up .)
Oh yeah–any suggestions on pr preparing the cassettes for transfer (cleaning, etc.) would also be appreciated.
There are two big issues I see. One is that even with the source material, you need to get the speed absolutely correct in order to be able to coherently subtract the bleed-through. It won’t just “naturally” be at exactly the same speed, especially if it’s on a different player than it was recorded on. If what you’re hearing has a lot of high frequency content, that will be very hard. If it’s the bass, mostly, it would be easier.
That segues into the other big problem: What’s on the tape will almost certainly have different amplitudes at different frequencies, so you’re not going to be just looking for the correct amplitude to cancel it out, you’ll need a frequency-varying amplitude. One approach might be to band-limit the source material to several frequency ranges, and try to reduce those one at a time.
Don’t let this stop you from trying, though. Audacity is free, and can easily do the steps you need. Just keep those issues in mind, if it seems like no matter what you do it doesn’t work.
I figured I could use a digital editor (probably Pro Tools) to precisely match the speed of the clip. If it’s significantly off, I might try playing it back on the cassette player using the speed change control. I might try running it through a cassette player anyway because that might help match it to the frequency biases inherent in a compact audio cassette
I didn’t know that amplitude is such an important consideration when trying to “phase cancel” an unwanted sound. As I said… I’ve never done that sort of thing before!
I have Audacity and you know what? It is free, but I have some problems with it. I can’t seem to navigate through the timeline with the precession I’d like. Maybe there are better ways to navigate that I just haven’t learned about yet. Could be “operator error”. I was hoping to use Pro Tools (but I don’t have it—maybe a friend will let me use his).
I’ve been using Audacity for a long time now, and generally like it. It isn’t totally intuitive. I had to re-install it on a new computer, and the settings were all wonky. It took me a while of fidding with input volume to get it to work.
I use the Ion turntable, and the companion Ion cassette player. The latter is under $40. I’ve also succeeded in using an ordinary cassette-player (“boom box”) and my computer’s audio-in jack. But the Ion cassette player is a USB device, and works very well.
Hope it works out for you; saving old data to newer formats is a lot of fun.
Audio restoration has been a passion of mine for years and I have gotten quite good at it (material I restored has been issued by Rhino, and fans of live New Order should know of my latest endeavor).
Basically, restoration here via phase cancellation, while excellent in theory, is a complete non-starter at this level with this material.
The samples have to be identical, mirror images, down to the sample level. Otherwise you will end up amplifying what you wish to destroy.
Cassette speed at the best of times is never precisely 1 7/8 ips. Unless your deck is really having a bad day, it should be consistently near 1 7/8 ips within the margin of detectability to the average listener, but there is inherent wow and flutter in the transport. You can attempt to adjust the pitch (speed) digitally, but the tape’s bias signal and other noise in the system - much less the patience to have to manually pitch milliseconds of audio thousands of times per song - is a prohibitive factor. Remember, for phase cancellation to happen the two sources have to be in lock sync with each other, no margin for error.
I would look at other options such as gating (how low level is the bleed-through audio?) or learning to enjoy your mashups.
Yeah… I played around a little with inverting phase in Audacity. When the inverted signal was an exact mirror image it worked perfectly. Total silence. But if i change the speed a tiny bit it started off OK, but gradually got louder. Other experiments with EQing and amplitude changes cancelled some of the audio, but not all. So yeah. I see now just how precise it has to be.
How do you see gating working on this? If the “bad” audio had a low enough amplitude you think I could “gate it out”? That could possibly work. I don’t have a cassette player so I can’t check the levels right now but if I remember correctly, the levels were relatively high (though not as high as the primary audio levels).
I’ll have to play around with that. I think I would most likely lose too much of the primary signal though. Thanks a lot for the idea though–even if it doesn’t work, I hadn’t even thought of gating to get the bad sound out.
It’s possible the bleed-thru is due to head misalignment. Recording #1 was done with a head that was at position A, and when recording #2 was done, a different machine was used with a slightly different head position, B. Since A and B weren’t the same place on the tape, #2 didn’t erase & write over #1 perfectly, but left a little bit to the side.
Now the playback machine is picking up a little of the old, misaligned track. If there is any way you can adjust the head position very slightly, you might find the unwanted sound disappearing.
I would try that method first, as any audio processing would be second best, working with a combined sound.
Another thought: When you get the recording made, try listening to the Left and Right channels individually, and also make the sum and difference of the Left and Right channels and listen to them, to see if the bleed-through sound is stronger in one of them. It might make it easier to isolate the bleed through. To do this, you won’t have to worry about the timing, so it’s relatively easy to try.
(In Audacity, there’s a command to split Stereo into separate tracks. You’d need to get the Sum and Difference using separate mono tracks.)
.)