Left Channel EQ CHANGES VOLUME?

GoldWave general discussions and community help
Post Reply
petenoak
Posts: 10
Joined: Sun Oct 15, 2017 7:40 pm

Left Channel EQ CHANGES VOLUME?

Post by petenoak » Sun Oct 29, 2017 2:40 pm

Here's another single channel editing question.

I used the Equalizer in Goldwave to reduce the treble in a single channel. I was surprised to see the wave form's size increase. I would have thought it would be reduced in volume.

For example, I have a track length of 7:00 minutes and a left channel, mild treble EQ of
-1 at 2400hz, -2 at 6000hz, -3 at 15khz
resulted in an overall volume increase of 0.43db. I had to reduce that channel's volume by this amount or else the stereo balance would be affected.

So this is my workaround but it does seem odd that treble reduction would INCREASE the size of the wave form. Is Goldwave working correctly? Am I missing something?

pete

DougDbug
Posts: 2044
Joined: Wed Feb 16, 2005 3:33 pm
Location: Silicon Valley

Re: Left Channel EQ CHANGES VOLUME?

Post by DougDbug » Mon Oct 30, 2017 1:54 am

I'm not exactly sure what's causing that, but it's normal and there are a couple of possible explanations -

Filters (both digital and analog) are not perfect and depending on the design compromises the filter may have ripple (bumps) which can boost frequencies in the pass-band.

Some filters also introduce phase shifts. That is, different frequencies are phase-shifted by different amounts. This causes different frequency components to shift-slightly in time, which moves-around and changes the overall waveform peaks, making some peaks higher and some peaks lower. The highest peaks may become higher, especially if the recording was "artificially" compressed/limited (like most modern recordings).

...Phase shifting is the more-likely explanation. There is something called an all-pass filter that introduces phase shifts at different frequencies without normal amplitude-filtering and without affecting the sound. This will often make some peaks higher than before.

You get s similar effect when you make an MP3 or a vinyl recording... The peaks get higher without making the average higher, and that makes a higher apparent dynamic range without actually affecting the sound of the dynamics.

petenoak
Posts: 10
Joined: Sun Oct 15, 2017 7:40 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by petenoak » Tue Oct 31, 2017 11:55 am

Thank you, DougDbug for this explanation. Phase shifts seems to be the key though I wonder how my ear perceives such things? From your explanation, is it possible that applying any kind of EQ could result in slight change in the dynamic range in the track?

I went back and re-tested the EQ process with a stereo track, not isolating a channel and there's the same, slight increase in volume of the tracks. But a closer look shows the peaks shifting! In Goldwave's display, the wave form is above and below the 0.0db center line in each channel. In many of the recordings that I'm working with, the peaks are not symmetrical above and below this center line. Applying EQ appears to shift some peaks from above the center line to below it! They get flipped. But only some do this flipping. Others remain on the same side of center line. Is this the phase shifting you're referring too?

The resulting volume change is slight according to Goldwave's "Match Volume" button and it's hard to hear any volume change. But it's a bit counter-intuitive that the wave form appears to get slightly bigger with both treble reduction OR treble boost. I've always thought that treble reduction would reduce the perception of the volume. Boosting the highs usually makes the music sound a little louder/stronger, and treble reduction would make music sound a little duller/quieter? Perhaps it depends on the source because in my editing with EQ I can recall seeing a very large increase in the size of the wave form after applying treble reduction.

These are recordings from r-t-r tapings either off the air FM or perhaps tape copies of radio stations' own sources. These are Classical Music broadcasts from the 1950's, 60's and '70's. Frequently these old home tapings are digitized on a r-t-r machine with a different speed and/or head gap alignment. So I've got pitch and channel imbalance issues to correct. That's why I have to EQ the channels separately.

So perhaps there's been a lot of phase shifting with the music that has gone through broadcasting, taping and copying?

pete

DougDbug wrote:I'm not exactly sure what's causing that, but it's normal and there are a couple of possible explanations -

Filters (both digital and analog) are not perfect and depending on the design compromises the filter may have ripple (bumps) which can boost frequencies in the pass-band.

Some filters also introduce phase shifts. That is, different frequencies are phase-shifted by different amounts. This causes different frequency components to shift-slightly in time, which moves-around and changes the overall waveform peaks, making some peaks higher and some peaks lower. The highest peaks may become higher, especially if the recording was "artificially" compressed/limited (like most modern recordings).

...Phase shifting is the more-likely explanation. There is something called an all-pass filter that introduces phase shifts at different frequencies without normal amplitude-filtering and without affecting the sound. This will often make some peaks higher than before.

You get s similar effect when you make an MP3 or a vinyl recording... The peaks get higher without making the average higher, and that makes a higher apparent dynamic range without actually affecting the sound of the dynamics.

DougDbug
Posts: 2044
Joined: Wed Feb 16, 2005 3:33 pm
Location: Silicon Valley

Re: Left Channel EQ CHANGES VOLUME?

Post by DougDbug » Tue Oct 31, 2017 2:22 pm

Phase shifts seems to be the key though I wonder how my ear perceives such things?
You generally can't hear phase shift (of an all-pass filter)* or the resulting peak changes. Although you might hear it if it's only in one channel... For example, if you flip the phase (or polarity) by 180 degrees with the Invert effect you won't hear any difference unless you only flip one side... Then the soundwaves cancel in the air (depending on your position and how the sounds reflect in the room) and you get a spacey/phasey effect. (Of course if you flip the polarity, or time-shift the file, you're not affecting the peak levels.)

There is probably plenty of phase-shifting going-on already with the NAB tape equalization. The same thing happens with RIAA EQ on records... When the record and CD are made from the same master, the often record "appears" to have greater dynamic range. The frequency response of the record & playback filters is specified & complementary, but there is no spec for the phase response.

...And, whatever is causing the problem you're trying to correct may have also introduced some phase-shifting in one channel.
...The resulting volume change is slight according to Goldwave's "Match Volume" button and it's hard to hear any volume change. But it's a bit counter-intuitive that the wave form appears to get slightly bigger with both treble reduction OR treble boost.

...From your explanation, is it possible that applying any kind of EQ could result in slight change in the dynamic range in the track?
No, you usually won't hear a volume or dynamic range change because perceived loudness doesn't correlate well with the peaks... Loudness is more related to the average (or RMS) level, and the frequency content. i.e. Most commercial releases are normalized/maximized for 0dB peaks, but some songs are still louder than others.

If you boost a few peaks hear-and-there, you will increase the Crest Factor, which can be a crude (mostly useless) measure of dynamic range.

There is one possible negative side effect of all this... If your peaks are pushed-up over 0dB, you'll need to re-normalized (run the Maximize Volume effect) and bring the volume down to prevent clipping.

Since MP3 compression often has a similar effect, some people normalize to -1dB or so before making an MP3.



* Of course, "audiophiles" often claim to hear things that normal people can't...

JackA
Posts: 154
Joined: Thu May 24, 2012 5:52 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by JackA » Tue Oct 31, 2017 2:59 pm

petenoak wrote:Here's another single channel editing question.

I used the Equalizer in Goldwave to reduce the treble in a single channel. I was surprised to see the wave form's size increase. I would have thought it would be reduced in volume.

For example, I have a track length of 7:00 minutes and a left channel, mild treble EQ of
-1 at 2400hz, -2 at 6000hz, -3 at 15khz
resulted in an overall volume increase of 0.43db. I had to reduce that channel's volume by this amount or else the stereo balance would be affected.

So this is my workaround but it does seem odd that treble reduction would INCREASE the size of the wave form. Is Goldwave working correctly? Am I missing something?

pete
Excellent question, Pete. I have come across "brick walled" waveforms, increased key frequencies, then Maximized the waveform, the away went the brick walled look! Maybe the dominating high frequency(ies) was holding everything else back from being Maximized, why the (majority of the) waveform appeared to enlarge. This ASSUMES you did a "Maximize" on the waveform.

Jack

JackA
Posts: 154
Joined: Thu May 24, 2012 5:52 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by JackA » Tue Oct 31, 2017 3:09 pm

DougDbug wrote:
Phase shifts seems to be the key though I wonder how my ear perceives such things?
You generally can't hear phase shift (of an all-pass filter)* or the resulting peak changes. Although you might hear it if it's only in one channel... For example, if you flip the phase (or polarity) by 180 degrees with the Invert effect you won't hear any difference unless you only flip one side... Then the soundwaves cancel in the air (depending on your position and how the sounds reflect in the room) and you get a spacey/phasey effect. (Of course if you flip the polarity, or time-shift the file, you're not affecting the peak levels.)

There is probably plenty of phase-shifting going-on already with the NAB tape equalization. The same thing happens with RIAA EQ on records... When the record and CD are made from the same master, the often record "appears" to have greater dynamic range. The frequency response of the record & playback filters is specified & complementary, but there is no spec for the phase response.

...And, whatever is causing the problem you're trying to correct may have also introduced some phase-shifting in one channel.
...The resulting volume change is slight according to Goldwave's "Match Volume" button and it's hard to hear any volume change. But it's a bit counter-intuitive that the wave form appears to get slightly bigger with both treble reduction OR treble boost.

...From your explanation, is it possible that applying any kind of EQ could result in slight change in the dynamic range in the track?
No, you usually won't hear a volume or dynamic range change because perceived loudness doesn't correlate well with the peaks... Loudness is more related to the average (or RMS) level, and the frequency content. i.e. Most commercial releases are normalized/maximized for 0dB peaks, but some songs are still louder than others.

If you boost a few peaks hear-and-there, you will increase the Crest Factor, which can be a crude (mostly useless) measure of dynamic range.

There is one possible negative side effect of all this... If your peaks are pushed-up over 0dB, you'll need to re-normalized (run the Maximize Volume effect) and bring the volume down to prevent clipping.

Since MP3 compression often has a similar effect, some people normalize to -1dB or so before making an MP3.



* Of course, "audiophiles" often claim to hear things that normal people can't...
Always thought of Phase being related to time. Polarity reversal is when you incorrectly hookup stereo speakers incorrectly. Why I always like to check for waveform inversion (one stereo channel) with speakers, because it's sometime difficult to hear/detect with headphones.

Jack

JackA
Posts: 154
Joined: Thu May 24, 2012 5:52 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by JackA » Tue Oct 31, 2017 3:20 pm

petenoak wrote:Thank you, DougDbug for this explanation. Phase shifts seems to be the key though I wonder how my ear perceives such things? From your explanation, is it possible that applying any kind of EQ could result in slight change in the dynamic range in the track?

I went back and re-tested the EQ process with a stereo track, not isolating a channel and there's the same, slight increase in volume of the tracks. But a closer look shows the peaks shifting! In Goldwave's display, the wave form is above and below the 0.0db center line in each channel. In many of the recordings that I'm working with, the peaks are not symmetrical above and below this center line. Applying EQ appears to shift some peaks from above the center line to below it! They get flipped. But only some do this flipping. Others remain on the same side of center line. Is this the phase shifting you're referring too?

The resulting volume change is slight according to Goldwave's "Match Volume" button and it's hard to hear any volume change. But it's a bit counter-intuitive that the wave form appears to get slightly bigger with both treble reduction OR treble boost. I've always thought that treble reduction would reduce the perception of the volume. Boosting the highs usually makes the music sound a little louder/stronger, and treble reduction would make music sound a little duller/quieter? Perhaps it depends on the source because in my editing with EQ I can recall seeing a very large increase in the size of the wave form after applying treble reduction.

These are recordings from r-t-r tapings either off the air FM or perhaps tape copies of radio stations' own sources. These are Classical Music broadcasts from the 1950's, 60's and '70's. Frequently these old home tapings are digitized on a r-t-r machine with a different speed and/or head gap alignment. So I've got pitch and channel imbalance issues to correct. That's why I have to EQ the channels separately.

So perhaps there's been a lot of phase shifting with the music that has gone through broadcasting, taping and copying?

pete

DougDbug wrote:I'm not exactly sure what's causing that, but it's normal and there are a couple of possible explanations -

Filters (both digital and analog) are not perfect and depending on the design compromises the filter may have ripple (bumps) which can boost frequencies in the pass-band.

Some filters also introduce phase shifts. That is, different frequencies are phase-shifted by different amounts. This causes different frequency components to shift-slightly in time, which moves-around and changes the overall waveform peaks, making some peaks higher and some peaks lower. The highest peaks may become higher, especially if the recording was "artificially" compressed/limited (like most modern recordings).

...Phase shifting is the more-likely explanation. There is something called an all-pass filter that introduces phase shifts at different frequencies without normal amplitude-filtering and without affecting the sound. This will often make some peaks higher than before.

You get s similar effect when you make an MP3 or a vinyl recording... The peaks get higher without making the average higher, and that makes a higher apparent dynamic range without actually affecting the sound of the dynamics.
Only way I know you can Phase Shift with Stereo, is by advancing or retarding (time) one of the two stereo channels. You mention Stereo (actually, you did not), yet you say radio broadcasts from the 50's. If Mono, why not delete one of the two waveforms?

petenoak
Posts: 10
Joined: Sun Oct 15, 2017 7:40 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by petenoak » Tue Oct 31, 2017 5:46 pm

JackA wrote:
Excellent question, Pete. I have come across "brick walled" waveforms, increased key frequencies, then Maximized the waveform, the away went the brick walled look! Maybe the dominating high frequency(ies) was holding everything else back from being Maximized, why the (majority of the) waveform appeared to enlarge. This ASSUMES you did a "Maximize" on the waveform.

Jack
Hello Jack,
Thank you for your reply. However, I do not use "Maximize Volume." I don't fully understand its function. I guess "Maximize" is a way to the signal level up to a standard level? I hope it does the same to both channels equally, I've channel balance issues in these old tape recordings. I do use "Change Volume" sometimes if I feel the signal level is too low.

Let me just repeat that the original reason for this post is my challenge to get the left and right channels from an old tape recording to be reasonably equal. I frequently encounter a bright left channel and dull right channel. Sometimes it is just a spectral difference, but there can be volume differences, too. Not easy to distinguish between these two problems. Sometimes I start by turning up the treble and try to use my ear to judge the difference in the background hiss from the original tape and/or the FM broadcast. I also use my ear to judge the timbre solos instruments of the orchestra that are usually more or less in the center.

Anyway, left channel treble reduction causing a volume increase seems to be working against my attempt to get a proper stereo balance.

pete in oakland, ca

JackA
Posts: 154
Joined: Thu May 24, 2012 5:52 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by JackA » Tue Oct 31, 2017 6:50 pm

petenoak wrote:
JackA wrote:
Excellent question, Pete. I have come across "brick walled" waveforms, increased key frequencies, then Maximized the waveform, the away went the brick walled look! Maybe the dominating high frequency(ies) was holding everything else back from being Maximized, why the (majority of the) waveform appeared to enlarge. This ASSUMES you did a "Maximize" on the waveform.

Jack
Hello Jack,
Thank you for your reply. However, I do not use "Maximize Volume." I don't fully understand its function. I guess "Maximize" is a way to the signal level up to a standard level? I hope it does the same to both channels equally, I've channel balance issues in these old tape recordings. I do use "Change Volume" sometimes if I feel the signal level is too low.

Let me just repeat that the original reason for this post is my challenge to get the left and right channels from an old tape recording to be reasonably equal. I frequently encounter a bright left channel and dull right channel. Sometimes it is just a spectral difference, but there can be volume differences, too. Not easy to distinguish between these two problems. Sometimes I start by turning up the treble and try to use my ear to judge the difference in the background hiss from the original tape and/or the FM broadcast. I also use my ear to judge the timbre solos instruments of the orchestra that are usually more or less in the center.

Anyway, left channel treble reduction causing a volume increase seems to be working against my attempt to get a proper stereo balance.

pete in oakland, ca
Maximize Volume looks for the highest Peaks and sets the upper Amplitude limits.
Forgive me, I'm still lost with what you are working with. You mention 50's, 60's and 70's homemade recorded radio broadcasts(?). But I question the "FM" radio part - that didn't happen until late '60's, why I question why 2 tracks (if not stereo). If there is no need for two tracks, have you tried just rendering both tracks to Mono? I would like to see or hear what you're talking about, even if a snippet, so I can evaluate.

You could always copy one track, and under "Edit" (dropdown), I think, there should me a "Mix" option where you can mix (or blend) one track with another, so you could hear what both tracks sound like when blended together.

Sounds like a neat project!

Jack
p.s. Side note to Chris, would like option to Maximize BOTH stereo tracks individually. No biggie.
p.p.s. Tape Hiss noise. My layman type explanation. Let's say you have 10,000 recording tape magnetic particles to make a 20 Hz. sine wave. You lose a few, no big deal. BUT, lets take that same density of magnetic particles and apply it to 20,000 Hz. We have 1000x (times) less magnetic particles to work with. We lose a few and you've lost about 1/3 of your 20,000 Hz sine wave. What do you then hear, you hear silence where there should be sound. In other words, there is no real "hiss", it's not added, it's actually subtractive, you're just hearing silence mixed with the sound, simultaneously.

petenoak
Posts: 10
Joined: Sun Oct 15, 2017 7:40 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by petenoak » Wed Nov 01, 2017 11:59 am

OK, Jack, I have to clear up my overly general statement about how old my stereo sources are. Sorry about that.

I know that FM stereo broadcasts didn't arrive till some time in the 1960's. What I have is some live material that was probably broadcast in mono but exists also (in private hands) in stereo. One rumor is that the Boston Symphony gifted orchestral members with these private, live recordings.

Three performances like these slipped out onto the Classical Music label, Music&Arts which released a CD of the great conductor, Pierre Monteux conducting the Boston Symphony in Richard Strauss tone poems in genuine stereo. These were live performances dating from 1960 and 1962! This CD had to be withdrawn for contractual reasons but you can still find it on the Used market. If these concerts were broadcast it would have been in mono. The only explanation I can think of to explain the existence of stereo captures of these live concerts is that perhaps the Boston Symphony and/or their radio station were experimenting with stereo recording as far back as 1959.

I've come across more of this kind of material from private sources and it's amazing to hear stereo live performances from so far back. They're in rather good sound. It's really the home recorded FM stereo broadcasts that need some editing help.

pete

[quote="JackA"][quote="petenoak"][quote="JackA"]

Maximize Volume looks for the highest Peaks and sets the upper Amplitude limits.
Forgive me, I'm still lost with what you are working with. You mention 50's, 60's and 70's homemade recorded radio broadcasts(?). But I question the "FM" radio part - that didn't happen until late '60's, why I question why 2 tracks (if not stereo). If there is no need for two tracks, have you tried just rendering both tracks to Mono? I would like to see or hear what you're talking about, even if a snippet, so I can evaluate.

JackA
Posts: 154
Joined: Thu May 24, 2012 5:52 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by JackA » Wed Nov 01, 2017 12:59 pm

petenoak wrote:OK, Jack, I have to clear up my overly general statement about how old my stereo sources are. Sorry about that.

I know that FM stereo broadcasts didn't arrive till some time in the 1960's. What I have is some live material that was probably broadcast in mono but exists also (in private hands) in stereo. One rumor is that the Boston Symphony gifted orchestral members with these private, live recordings.

Three performances like these slipped out onto the Classical Music label, Music&Arts which released a CD of the great conductor, Pierre Monteux conducting the Boston Symphony in Richard Strauss tone poems in genuine stereo. These were live performances dating from 1960 and 1962! This CD had to be withdrawn for contractual reasons but you can still find it on the Used market. If these concerts were broadcast it would have been in mono. The only explanation I can think of to explain the existence of stereo captures of these live concerts is that perhaps the Boston Symphony and/or their radio station were experimenting with stereo recording as far back as 1959.

I've come across more of this kind of material from private sources and it's amazing to hear stereo live performances from so far back. They're in rather good sound. It's really the home recorded FM stereo broadcasts that need some editing help.

pete
[/quote]

Pete, I'm finally catching on! Thank you for clarifying! I seem to recall hearing that FM Radio initially started out broadcasting Classical music. And even though stereo wasn't very popular back in the late 50's, you can find stereo jukeboxes being advertised in Billboard Magazine! I'm not heavily into Classical music, but early stereo is interesting!!

If you need a second set of ears to judge your work, just search Uncommon Top 40, and you'll find me!

Thanks,

Best,
Jack

petenoak
Posts: 10
Joined: Sun Oct 15, 2017 7:40 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by petenoak » Wed Nov 01, 2017 9:04 pm

"Pete, I'm finally catching on! Thank you for clarifying! I seem to recall hearing that FM Radio initially started out broadcasting Classical music. And even though stereo wasn't very popular back in the late 50's, you can find stereo jukeboxes being advertised in Billboard Magazine! I'm not heavily into Classical music, but early stereo is interesting!!

If you need a second set of ears to judge your work, just search Uncommon Top 40, and you'll find me!

Thanks,

Best,
Jack"
---------------------------

Hi Jack,
Well thanks SO much for the offer! You mean search this forum for "Uncommon Top 40"?

pete

JackA
Posts: 154
Joined: Thu May 24, 2012 5:52 pm

Re: Left Channel EQ CHANGES VOLUME?

Post by JackA » Thu Nov 02, 2017 10:27 am

petenoak wrote:"Pete, I'm finally catching on! Thank you for clarifying! I seem to recall hearing that FM Radio initially started out broadcasting Classical music. And even though stereo wasn't very popular back in the late 50's, you can find stereo jukeboxes being advertised in Billboard Magazine! I'm not heavily into Classical music, but early stereo is interesting!!

If you need a second set of ears to judge your work, just search Uncommon Top 40, and you'll find me!

Thanks,

Best,
Jack"
---------------------------

Hi Jack,
Well thanks SO much for the offer! You mean search this forum for "Uncommon Top 40"?

http://uncommontop40.org

:-)

pete

Post Reply