Ampex Tape, Track Attenuation

john_dennis wrote on 6/17/2015, 9:06 PM
For those who think this bug isn't real...

Before

After four passes through Vegas Pro 13

Upside: I get to hear the audio.

Downside: The noise floor went from - 58 db to ~ -42 db.

Luckily, this was a recording of a Las Vegas "bag of tricks" lounge act. There wasn't 3ms of silence in the whole 45 minute set.

Comments

musicvid10 wrote on 6/17/2015, 10:07 PM
I see you've got Track FX on the timeline. Could that be part of the problem?
If they're the three default fx I'm thinking of, try removing them from the chain completely.

john_dennis wrote on 6/17/2015, 10:48 PM
In later versions of Vegas there are three fX automatically added to each new track. The effects "should" be neutral.

Here's the view with all the fXs disabled.
rraud wrote on 6/18/2015, 10:31 AM
As I recall, 'all' versions of Vegas added the audio plug ins. Annoying for sure. Fortunately, the default can be changed.
Chienworks wrote on 6/18/2015, 12:20 PM
I don't think i see what's being discussed here. What bug are you referring to? Why did it take four passes "through" Vegas and what did those passes do? What's the relationship to Ampex Tape? I'm a bit lost. ;)
john_dennis wrote on 6/18/2015, 3:50 PM
"[I]I don't think i see what's being discussed here.[/I]"

The issue is the withering away of the signal over the 45 minute set. Generally, live performances stay the same or get a little louder near the end.

It's been a long time since I touched the tapes, but as I remember, some tapes would self-erase. Certain formulations of Ampex tapes were particularly bad. This was an Ampex tape.

"[I]Why did it take four passes "through" Vegas and what did those passes do?[/I]"

I measured the RMS level for the last ten minutes of the set and compared it to the first ten minutes of the set. There was a ~17 db difference. Assuming the attenuation was linear, (not a good assumption at all), I applied successive 6 db volume envelopes (the Vegas maximum) until the RMS values were as close as 1 db.

All the individual songs are being very slightly re-leveled anyway since these are live performances. I'll leave the performance warts, but I do want people to hear all parts.
Geoff_Wood wrote on 6/18/2015, 4:25 PM
Still not clear if you are suggesting that there is a problem with Vegas, or not.. Or just showing what you achieved (levelling) that took 4 passes at +6dB to achieve in the method you used !

Unfortunately the standard volume envelopes (even in Sound Forge !) max out at +6, and the alternative approach (attenuating the louder start parts down to match with the ending) stuffs the s/n for the whole piece, rather than just the end.

Alternative would be to use an automate-able Volume plugin that does more than +6dB gain. Plenty out there, even for free !

geoff
john_dennis wrote on 6/18/2015, 4:46 PM
I was just showing a problem that I solved in Vegas. If Vegas had a wider range, it would have been quicker, but considering what a sink-hole these projects are, it wouldn't have changed much.

After Kelly questioned what he is seeing, I questioned it too. I believe this self-erasure may be the inside wraps of a tightly-wrapped tape. Of course this is one of the few tapes that I've ever seen the problem manifested. The giveaway is that both tracks are affected [I]on a quarter-track stereo tape[/I]. When I get a chance I plan to go back to the previous set to see if it was quieter near the beginning of the tape.

Though I didn't notice it at the time, it could be.

I checked. It's not obvious to the casual looky-lou.
Chienworks wrote on 6/18/2015, 8:36 PM
Ahhhh, i see now. There just wasn't really any indication in the original post that this had been captured from tape.

One suggestion to reduce the number of passes is to start at -12dB, finish at +6dB, and set the track volume to +12dB. That would be the equivalent of 3 passes going from 0 to +6.

If you really want to go extreme, add graphic EQ set to flat, and use it's output gain to add another 20dB. You could then run the envelope from -32 to +6 for a whopping 38dB change from beginning to end in one single pass.
Chienworks wrote on 6/18/2015, 8:39 PM
I wasn't questioning what i was seeing, i just didn't know *what* i was seeing.

"the inside wraps of a tightly-wrapped tape"

My gut feeling i that it shouldn't make any difference whether the part of the tape in question is from near the hub or near the outside. No matter where it is on the reel (except the first & last few inches) there are still layers above and layers below. I've encountered print-through in some cases with very thin tapes, but never the gradual diminishing shown here.