Waveform of Gordon Lightfoot Baby Step Back

Here's a waveform image file of Gordon Lightfoot's Baby Step Back taken from the Gord's Gold Vol. 2 CD from 1988 and the recent Rhino remaster of the Shadows album.

 

You will notice that Inglot applied a lot of clipping on the Rhino remaster that wasn't there on the Gord's Gold 2 remaster of the same version. This is one out of 4 original recordings on Gord's Gold Vol. 2.

 

What differences did you hear between the discs that led you to check the waveforms, Bradley? Something bothered you about the sound?

 

Interesting. Inglot's certainly raised the peak level (which was very low on the original CD) a great deal. I don't see a terrible "buzz cut" here, though. OTOH, it does look as if more of the peaks are at the same level than was the case in the original mastering, so I wouldn't be surprised if the overall sound was less dynamic. Are all those 0db peaks flattened (clipped), or just some of them? It's hard to tell without zooming in on the waveform.

Also, can you find the peak and average RMS levels on these files, Bradley, and post them for us?

And I'll echo Bob's question: how do they sound to you when you compare them?

Thanks for posting these.

 

Yeah, I'm seeing roughly the same thing as Gardo. On the 1988 version, the absolute peak was still well short of 100%, which means a lot of potential resolution was being wasted. The new one looks like it has a couple of the biggest peaks knocked down a bit, but overall the dynamics look similar. It's just using more of the available bits.

Without zooming, you can't tell if things are just maxed or actually clipped. Things can *look* clipped when viewing the whole 3:50 at once that really aren't, since the waveform is squished together to fit on the screen.

I've certainly seen a whole lot worse.

Ryan

 

Gord's Gold Vol. 2
Left Right
Min Sample Value: -19000 -17618
Max Sample Value: 18088 14763
Peak Amplitude: -4.75 dB -5.27 dB
Possibly Clipped: 0 0
DC Offset: .08 -.728
Minimum RMS Power: -61.97 dB -46.81 dB
Maximum RMS Power: -14.31 dB -15.94 dB
Average RMS Power: -22.3 dB -24.71 dB
Total RMS Power: -21.6 dB -23.95 dB
Actual Bit Depth: 16 Bits 16 Bits

Using RMS Window of 50 ms

Shadows remaster
Left Right
Min Sample Value: -32766 -32766
Max Sample Value: 32766 32766
Peak Amplitude: 0 dB 0 dB
Possibly Clipped: 0 0
DC Offset: -.001 -.001
Minimum RMS Power: -82.92 dB -82.97 dB
Maximum RMS Power: -7.11 dB -7.39 dB
Average RMS Power: -14.65 dB -15.63 dB
Total RMS Power: -13.94 dB -14.93 dB
Actual Bit Depth: 16 Bits 16 Bits

Using RMS Window of 50 ms

 

Here's the clipping info on the Shadows remaster. Gord's Gold Vol. 2 has no clipping whatsoever.

 

Despite the slight clipping, it's a close call on what the best CD version of this 1982 recording are. Definitely Inglot used more of the available bits but overall, he did a great job, even with the clipping.

 

BTW, the Rhino remaster of Shadows is the first time the album appeared on CD. The vinyl version of Gord's Gold Vol. 2 squeezes 8 tracks per side for a total of 16 tracks while there are 2 tracks on the CD not on the LP, but the one otherwise unreleased track, If It Should Please You, is not omitted from the vinyl.

 

When I restored the clipping, the RMS info has been changed to:
Left Right
Min Sample Value: -30788 -29253
Max Sample Value: 32767 25831
Peak Amplitude: 0 dB -.99 dB
Possibly Clipped: 6 0
DC Offset: -.001 -.001
Minimum RMS Power: -86.48 dB -86.48 dB
Maximum RMS Power: -11.1 dB -11.39 dB
Average RMS Power: -18.65 dB -19.63 dB
Total RMS Power: -17.94 dB -18.93 dB
Actual Bit Depth: 16 Bits 16 Bits

Using RMS Window of 50 ms

 

Now here's the waveform of the clipping removed.

 

Interesting stuff. Thanks. Did you notice an improvement after "removing" the clipping?

Ryan

 

You're welcome. The best remastering by far is the Rhino version, just "restore" the light clipping.

 

The clipping removed actually improves it so that it doesn't sound as loud as the Rhino remaster itself and the VU meter doesn't go into the red as much as before.

 

Updates

After further clip restoration, the following RMS data has occured
Left Right
Min Sample Value: -28051 -27241
Max Sample Value: 27887 25832
Peak Amplitude: -1.35 dB -1.6 dB
Possibly Clipped: 0 0
DC Offset: -.001 0
Minimum RMS Power: -74.81 dB -75.18 dB
Maximum RMS Power: -11.11 dB -11.39 dB
Average RMS Power: -18.71 dB -19.65 dB
Total RMS Power: -18 dB -18.95 dB
Actual Bit Depth: 16 Bits 16 Bits

Using RMS Window of 50 ms

And now the graphic:

 

The sound is better yet and the VU meter doesn't go in the red at all.

 

Chuck, when you "restore" clipped audio in Cool Edit, you choose how much to lower the level. You can set it to reach 100% or not.

 

I know. BTW, my name is Bradley.

 

Thanks for all that interesting data, Bradley. I don't use Cool Edit Pro so I can't comment on the clipping removal, but given all the other data it looks as if Inglot did compress the dynamic range a bit (from around 17.5 db on the first CD to about 15 db or so on the remaster). Your listening comments, though, indicate that he did a better job of it than most do.

With the clipping removal, it almost looks as if the dynamic ranges are roughly the same, but again, I don't understand this function enough to know exactly what I'm seeing here.

 

The engineers who supervised the GG Vol. 2 remaster of this track are Ken Friesen (the original co-producer and engineer of the Shadows album), Peter Norman and Kathy Birkett.

 

For those Lightfoot fans who do want to listen to this track on CD flat from the master but done at a lower level, Gord's Gold Vol. 2 is still THE source for it, but still, Shadows does have several gems throughout the album and is one of his best albums.

 

Oh goodness! Sorry. Don't suppose you had anything to do with Watergate either....

 

I had nothing to do with Watergate

 

First All,

Inglot/Hersch did not use compression. They used LIMITING. This keeps getting confused with COMPRESSION.

Sometimes compression is used to make the apparent volume lower while lowering the peaks. Then a volume boost is used to raise it all higher than before while using limiting to prevent CLIPPING. Why the meters show clipping when played does not actually mean the wave is clipped. What it most surely shows is that they did not leave any headroom.

Having stated that, limiting has the EFFECT of sounding like COMPRESSION, particularly when overdone.

Under some circumstances, a wave can have it's dynamics restored by the de-clip tool in either Cool Edit Pro or Sound Forge 5 & 6.x.

From the CEP help files:

 

They may have used compression as well. I'm not sure we can tell merely by looking at a waveform.

I know there's a difference between limiting and compression, but it is true that compression (a narrowing of the dynamic range) occurs when a limiter is inserted in the signal chain and the average levels are brought up. The limiter keeps the peaks at a steady level, but now the average level is up, and the result is that the dynamic range is compressed. That's why some consumer outboard boxes (like my little Alesis from long ago) are called "compressor/limiters."

Thanks for the information on the de-clip tool. I can see that this would make the square-wave at the top of a clipped waveform into something more musical and less distorted, but I don't see how this restores the dynamics. It looks to me as if it simply removes the flat-top and thus the distortion.