A Message Board, Guestbook, or Poll hosted for your website.
Both Sides Now Stereo Chat Board

Register Login New Posts
 
Both Sides Now Publications > Forums > Technical stuff [Moderator: Steve Massie] > Adobe Audition Audio Restoration Tutorial
 
Username:  
Password:  
 
   
 


Thread Tools
Reply
 
Author Comment
 
diamone
Registered: 08/10/08
Posts: 270

    10/08/08 at 02:30 PMReply with quote#1

Several new/beginning kids have PM'ed me about a basic class for Adobe Audition, so I will just replicate portions of my junior-college digital editing class on here, along with a few other relevant topics.

If this feels like it's a little beyond a truly beginning class, I apologize.

If I slowed it down or amplified it much further, the rest of the guys would accuse me of being the Little Nipper Storyteller. ``When you hear the noise (RAZBERRY) turn the page'' (WINK).

But speaking of turning the page, the best way to get full use out of this or any tutorial is to print it out and clip it to a stand next to what you are working on.

Then you can refer back and forth to it while you are working the practice sessions I take you through.

When installing Adobe Audition, when the screen comes up that says Types of Files to Associate with Adobe Audition, Deselect ALL Except .ses and .pk.

When recording into the computer from tape or phono, ALWAYS record in Stereo from a stereo deck or turntable, no matter if the record or tape is Mono or Stereo.

Use the highest resolution recording parameters available, such as 32-bit 192 KHz. This makes all the processes that follow considerably more accurate.

For a tape deck, a simple RCA-to-Stereo-Mini (Headset) plug will do the job nicely.

For phono, rather than using a USB turntable, the soundcards therefrom being VASTLY inferior, send the signal through any one of a number of good tabletop preamps you can get off eBay or Craigslist for a modest sum.

Plug the preamp into the power strip your computer is plugged into, plug the phono into one end and the aforementioned RCA stereo plugs to stereo mini-plug from the output of the preamp to the input of the computer soundcard (unless it has RCA inputs in which case just use a 2nd set of RCA stereo cables).

In addition, if you can find a good linear tracking turntable such as a Studer or Technics use that instead of an ordinary pivot-arm turntable for better sound and close to no wear on the record. The other choice is a Garrard Zero available all over Craigslist and eBay for about $50.

In either case, to improve record tracking and reduce wear even further, try and find a Stanton 780 cartridge with a Shibata-type stylus (MicroRidge, Nude-Square-Shank-Hyper-Elliptic, Tetrahedral, Curly, Larry an' Mo', all different names for the same thing). The DQ (Shibata) stylus also helps recover unworn parts of the record groove, allowing you to retrieve close to mint sound out of  that raggedyass copy of Meco Monardo's Star Wars Disco and Other Galactic Funk or Saturday Night Fever album you ripped to shreds as a kid on a little plastic turntable.

Just remember to use an Absorb-O-Thane mat underneath the record on top of the platter and Absorb-O-Thane feet underneath the body of the turntable to kill all the motor and other noises before recording.

When you set your turntable up, remove the center spindle and set a small round level there instead.

Adjust the feet of your turntable so that the bubble in the level is in the exact center. Replace the center spindle. 

Take a small nylon stylus brush and gently brush off the stylus tip of any dust or dirt.

Do this before and after playing any record.

Now take the blank ungrooved side of a Disco Single and plop it on the platter.

Set the tonearm down on different parts of the blank record.

If it starts to slide around, set your Anti-Skate up so that no matter where you set the tonearm down on the blank disc, it will stay where you put it and not slide back to the beginning or forward to the end.

For tape, try to find a player where you can set the Azimuth and Head Height without having to disassemble everything.

Clean the heads and tape path with a cotton swab moistened with denatured alcohol.

Gently rub across the tape head, guides, rollers and other material which comes in contact with the tape, changing swabs as needed until the swab comes out clean with no brown residue.

To set azimuth and head height (two different screws near the playback head, check your player service manual for specifications) slowly rotate first the azimuth screw until you get the strongest signal, as shown on the player meters, and then do the same for the Head Height til you get the best treble sound.

Now, close all other programs and disconnect from the Internet.

Next, set your recording levels in the computer. Do that by:

1. Plugging the cord from the player into the Line In (green jack) of your computer soundcard, then:

2. On your desktop, double-click on the Start button, then go to Control Panel, Sounds and Audio Devices, click on the Volume tab and click Place Volume Icon in the Task Bar. Once it appears, click the speaker icon next to the clock on the lower right side of the screen.

3. Click on Options in the upper left and then click Properties.

4. Click Recording and select Stereo Mix then take that to two or three lines from the top.

5. Click Line In and set that to two or three lines from the top.

6. Click the Mute buttons one at a time on all but Line In and then re-click (activate) on Stereo Mix.

You need to do this because otherwise every other noise on your computer will end up in your recording.

This will allow Adobe Audition to ``hear'' whatever is playing through the soundcard.

Now find the loudest point on the record or tape you plan to record.

On records, it's easy, just look for the spot where the grooves are the farthest apart and the deepest, without being the beginning or the end.

On tape, rewind the tape to the beginning, reset the counter to zero, play the tape all the way through and watch the little red needles bounce to tell you the level.

Note the counter number at the spot or spots where the loudest and most bass-intense music is.

1. Go into Adobe Audition and click the red Record button in the lower left. You will see a Properties slide appear. Choose Stereo, 32 bit float and 192 (or higher if your soundcard has it)

2. Set the tonearm down or run the tape to just before the loudest part and play.

You should start to see a colored wavy line appear.

The top of the colored wave in the loudest point you noted before should come to just under the little white rails about a quarter inch from the top and bottom of where the recording is taking place. 

3. The green bars at the bottom should bounce exactly in unison for a mono program, or close to unison for a stereo program except where one channel is featured prominently over the other. If it sounds lopsided in the headphones, put the headphones on left-for-right and see if the same lop-sidedness occurs in reverse. If it does, your hearing is normal and the original mix is indeed lopsided. To correct for this, doubleclick on your Speaker icon next to the clock, go back into Options, Properties, Recording, click the Line In button and slide the little balance arrow up at the top across from left to right until the track is balanced.

The recording also needs to be tall enough to reach the white rails without going over. If the recording is not tall enough, rewind or reset the music, go back to Recording Properties above and slide the Stereo Mix and Line In sliders up a little and repeat until the levels are correct.

You can also tell if it's too tall by the bars at the bottom of the screen.

If they slam into the right side and blink red, even for a second, go back and take down the levels til you get the strongest signal you can without slamming into the wall.

Don't worry if clicks and pops from the record slam into the wall.

You can tell those from music during your test recording, because the clicks and pops will look like little toothpicks laying on top of the music.

Highlight your test recording you just made, go up into Edit, Cut, rewind or reset your music and do another test recording. Your ``tape'' will now be blank again.

After the test recording, follow the directions below for Removing Center Channel  using the one for Quick & Dirty as discussed below, except do NOT invert one track. If the center channel disappears entirely, fine. If not, repeat the Balance step above until you get the closest approximation you can.

Highlight and delete the test recording as described above.

Now you are set to record.

Highlight the test recording you just made by clicking and dragging the mouse across it, go up to Edit and click Cut. Your ``tape'' will now be blank again, ready for recording.

Click at the extreme left edge of the recording space and then click Record in Adobe Audition, start your tape player or turntable and sit back to watch until the record or tape side is through.

If the recording is the correct level with no skips, go up to File and click Save As, write the title of the record or tape you just recorded, and the side number and number of sides (Side 1 of 3 for example), and then write Raw after it in the title bar and click Save.

After recording the file into the computer from the turntable or tape deck and saving it as the Raw version:

Load the file and play it in selected spots.
When the original recording is supposed to be full stone mono, in the case of tape origin, both the meter bars near the bottom of the screen should bounce in exact unison, both in intensity as well as modulation. If not, repeat the Balance step above.

In the case of mono records, or mono records taken from cassettes, the pops and clicks will be on different channels.

For stereo, the average volume between the two channels should be close, except during passages where the left or right channel is featured prominently over the other.

If one track of a mono program bounces with more intensity apart from clicks and pops and you couldn't get it exactly right from the Balance and Center Channel Extractor steps above:
a). Click and drag the mouse over the top or bottom track whichever is louder to highlight it,
b). Go into Effects, Amplitude and type in or slide over to the minus-dB or fractions thereof you think will match the other track. This may take several tries to get exact.
(Subtraction in digital is always preferable to addition, due to artifacting)

For records, or tapes with records recorded onto them, the first step is

Click on Effects, Noise Reduction, Auto Click and Pop Eliminator.

Choose Heavy Medium or Light sensitivity, or try a few and create your own.

Never erase your original file and always save each file after you do something to it, with your abbreviations as to what you did: blablabla Raw, blablabla EQ, blablabla NR, etc. Also, always click No on Save Changes to File because they mean Do you want to write over your original file? and the answer to that is always No because presumably you just saved it a second ago under its' appended name.

It is recommended that you do your Equalizations and Denoising prior to erasing difference channel information in order to bring out voices, enhance the treble or bass response or bring clarity to one or more areas of the sonic spectrum.

Doing noise reduction before this step is not advised, as some of the treble response will be sacrificed in the noise reduction process.

Equalization is performed by:

Going up to Effects and sliding down to Filters, then slide across to Graphic Equalizer.

Start off by setting the Master Gain to (Minus) -12.

This is a guess and may need to be tweaked later.

This step is necessary because the act of equalization increases the volume of some frequencies in relation to others.

The resulting file may end up going outside the white rails at the top and bottom of the waveform as seen in the view.

Therefore, the final volume must be reduced in order to accommodate the equalizations you are about to perform.

You can also do this by going up into Effects, Amplitude and selecting 6dB Cut or 10 dB Cut whatever will allow the peaks of the music to remain within the white rails.

Bass frequencies are to the left, midrange/vocals are in the center, and treble is to the right.

In order for the noise reducer to do a better job, when you increase the various treble frequencies, say from 5K-12K, when it sounds like a normal modern recording or a close approximation, bump up those frequencies just a little bit to give the Denoiser room to work.

Save the file under an amended filename in case you have to go back and do it again with different parameters.

Next perform the Denoise.

You denoise by:
a.) Finding a 1.6 second blank piece of ``tape'' in your file, usually at the beginning or end of your selection, or between songs.

If there's not enough, highlight what there is, go up to Edit, Copy, move the mouse and click to the beginning of where the ``blank tape'' is, go up to Edit, Paste and repeat the process til you have enough to do the Noise Reduction with.

Instead of cloning the same piece of ``blank tape'' over and over, you can also Edit Copy and Edit Paste other sections of ``blank tape'' from other parts of the file in order to make up your 1.6 seconds.
b.) Highlight a 1.6 second piece of the ``blank tape'' at the beginning or end of the recording with the noise from the record or tape.
c.) Go up to Effects, slide down to Noise Reduction and slide across to Noise Reduction.

Don't get fooled into using Hiss Reduction just because it's tape hiss you are trying to get rid of.
d.) Set Snapshots in Profile in the upper right corner to `9999'.
e.) Set the FFT Size and Overlays to the largest number available.
f..) Select Remove Noise on the buttons directly above.
g.) Select Capture Profile in the upper right corner and wait.

A yellow and green line will appear when it's done, profiling your noise selection
h.) Try setting the little Noise Reduction square button to about 75 or so.
i.)  Click Select Entire File and click OK and wait for completion.

If you are doing an entire tape side, this could take awhile.

It also may look jammed, and start to blank out, but unless you are using a computer so old you have to crank it to get it to work, or you are so old yourself you think the CD tray is for coffee cup placement you should be alright.

Eventually it will recover and continue on its' process.
j.) Play the file to audition it once it's done.
The idea is to get rid of as much noise and hiss as possible, without damaging the music.

You can tell if your noise reduction has been too severe and you've interfered with the music if you hear a ``dingley'' or ``metallic mumbling'' noise in the background.

In that case, go to Edit, Undo and start over with a lower setting, omitting all the steps prior to  h.).

Conversely, if the hiss is still too much, and no ``dingley'' noise is heard, you can try a higher noise reduction setting. Save file when complete with NR after the title.

It is recommended that you cut apart the individual selections on your side of tape or LP in order to perform the remainder of the processes.

Do that by dragging the mouse over the first song til you see the first section of blank.

Go up to Edit, Cut and then when that's done go back up to Edit and click Paste to New.

Go to File and click Save As, enter the track number and name and click Save. Repeat for all the rest of the songs on the side.

After all songs are separated and saved, select Close All Files.

Click on File, Open and select the first track on the tape or LP side.

Quick and Dirty Version for L-Minus-R Extraction:
1. Highlight ONE of the two tracks, either top or bottom, go up to Effects and click Invert.

2. When that's done, de-click the highlighted track so that both tracks will be operated on.

3. Go up to Effects, Amplitude and then slide down to Channel Mixer to Vocal Cut and click.

This will erase the Difference track and amplify the Main track at the same time.

If you end up with nothing, you forgot to invert ONE track and not both.

If you end up with no change in amplitude, you clicked on the wrong Extractor.

You can also erase the difference signal more effectively by:

Going up into Effects, Amplitude and then slide down to Channel Mixer.

Slide down to LR to Mid-Side and click.

The center channel (your monaural program) will extract and move to the left (or right), and the `everything else' channel will take the opposite position.

Move the mouse to either the upper or lower channel and play it in isolation.

Repeat for the other channel.

Choose the cleaner of the two, highlight it, go up into Edit, Copy to New and save the now-single-tracked program in still another amended filename.

Go into Edit and slide all the way to the bottom to Convert Sample Type.

Select the same sample rate and the same bitrate as your original file and click Stereo in the top center and click OK.

Save the file. I use ``Blablabla Center Channel or C-Chan for short.

Now maybe you want to Stereoize your selection so it doesn’t sound so flat and lifeless.

Go up to Effects, Delay Effects and slide down to Full Reverb.

Try out all the different reverbs in the list to the right and see which one fits the best.

For stereo sources you wish to expand out into 5.1, there’s only 2 presets that preserve the left and right information.

I forget which ones they are, but you can use a stereo source, find out, and then make your own adjustments.

Then preview your reverb.

You can tweak the reverb by using the Dry, Early Reflections and Wet settings. Since you are going to be mixing the Reverb track in with the Center Channel you just extracted up above, you want as little of the Center (Original Signal) as you can get away with in the Reverb.

Apply your Reverb.

Save the file as Blablabla Reverb.

Close all files.

Say No to Save Changes to Files  because otherwise your original file will be overwritten, erasing all ability to go back and do anything over again.

Go to File, Open and click on Blablabla- C-Chan.

When it’s open, go up to Edit, Insert in Multitrack.

Repeat for the Reverb track you just made.

Click the Multitrack View button to the left of the button of the CD up in the upper left corner

Go up to View and click the following options:

Show View Tabs

Show Pan Envelopes

Show Volume Envelopes

Enable Envelope Editing

Enable Clip Edge Dragging

Go to the little grey square near the upper left corner next to where it says Track 1.

Wave the mouse over the square until the hand disappears and a magnifying glass appears.

Click on the lower edge of the square and push it up so that only your Center Channel and Reverb Channel is visible and all the blank tracks are hidden.

For the Center Channel Track:

Go up to the upper left corner of the track and you’ll see a little white square attached to a green line. Click and hold on it to pull it down to about 75%. Do the same for the white box at the other end of the green line.

For the Reverb Track:

Same thing, and drag each side down to about 25% or so.

Play the file. if it sounds as if it’s swimming in a well, take the reverb track down even more.

If it still sounds dead and lifeless, take the reverb track up a little bit.

When you find the balance you like, give it another 5% or so volume on the reverb track

Next, go back to Edit View and drag the mouse over the top track in the Reverb Track so that it alone is highlighted.

Go up to Effects, Filters, Graphic Phase Shifter and select  +90 Degrees and click OK

Repeat for the bottom track, replacing -90 Degrees for +90 Degrees above.

Then go up to Edit. select Mix Down to File, All Tracks Stereo and Mixdown.

Play the Mixdown. If you like it, save it as Blablabla Mixdown 5.1 take 1 and go onto the next step.

Make sure the Mixdown is inside the white rails at it's loudest point.

If it's not, discard the Mixdown, go back into Multitrack and reduce the volume of both tracks by the same amount and then click Mixdown again.

Save it and try again with different paramters in case you like your next attempts better.

Remember to place the Reverb tracks +/- 90-degrees out-of-phase before mixing down.

Close all files.

Say No to Save Changes to Files or else your original will be overwritten.

Repeat for anything you want to convert from mono or stereo to 5.1 surround

Enjoy.

diamone
Registered: 08/10/08
Posts: 270

    10/20/08 at 05:53 PMReply with quote#2

(Continued from Above)

One way of using Adobe Audition is for creating De-Constructed (or Digitally Created) Stereo (DCS) or Digitally Extracted Stereo (DES).

Includes a brief History of Professionally Recorded Sound to aid the newcomer in comprehension of that which he is trying to accomplish.
The question was:

Is Adobe Audition the right program to make a DES/DCS with?

I'm trying to start doing it and nobody will tell me how and there's no tutorials.

Adobe Audition is indeed one of two widely-preferred programmes with which to perform audio restoration, the other being Diamond-Cut Pro, generally used in conjunction with one another.

Digitally Created Stereo (also known as De-Constructed Stereo) or it's somewhat-easier city-slicker cousin Digitally Extracted Stereo is both a labor of love. You will never get rich from it, your girlfriends will run away because of it and in addition to providing you with vast amounts of knowledge as you progress, it will also make you fat.  Which fits in just fine with the majority of the rest of us in the audio restoration community, so you might as well get used to it now.

A lot of threads and boards cross in this topic, the DCS/DES board, the QuadraphonicQuad forums and the antique Victrola forums at http://sonoraman.proboards107.com.

Most guys have their fingers in all the pies, so you might want to get a little familiar with all three ``dialects'' as they have a lot of crossover.

There is a lot of speculation on the various boards as to what constitutes Extracted and what constitutes Constructed, even though both methods use a considerable amount of the other's engineering techniques in order to get together something pleasing.

And now for a short tutorial for the beginning audio restorationist:
Locking Up to Bias

Every magnetic tape when it's first recorded has a supersonic steady tone recorded above the limit of human hearing. This tone, called the Bias can be retrieved from all tape and used to lock up all the original production elements, (discussed in greater detail below) making DCS and DES a snap compared to the way we do it now.

Unfortunately, when tapes are transferred to CD, this information is irretrievably lost unless you go back to the original session tapes and do a new extremely high-resolution transfer. By ``high resolution transfer'' understand that most music is mastered to the CD rate of 16-bit 44.1 KHz sample rate or the DVD-Audio rate of 24/96.

To capture the bias tone along with the music, a special head is needed for playback which is very expensive and delicate, and requires frequent cleaning in order to maintain performance.

In addition, when recording it is necessary to record the audio program at 32-bit or greater and 13.75 MHz sampling frequency. That's over three thousand times faster. To give you an idea how much data that is, at that speed, just the Basic Tracks reel of the three reels used for original production of one song (discussed below) would fit onto a Blu-Ray, so for 3 tracks of 3 reels, for each song you'd need 3 Blu-Ray discs to store it when you were transferring. And then you couldn't play off the Blu-Ray anyway, you'd have to dump them all to a video-capable hard drive and work off that

A video-capable hard drive is something akin to a Seagate Cheetah, running at 15,500 RPM (twice that of a normal 7200 RPM drive) with a data transfer rate in excess of 15 MHz. But that's only if you have original session tapes and you're trying to lock it up to itself using the bias tone. You can still use the Cheetah Drive to store on though, because once you get used to the speed, your normal 7200 RPM Sam's Club hard drive is going to seem so slow you'll begin to think you'd be able to do it faster manually.

But back to reassembling tracks.
For one of the more difficult titles people challenge themselves with, look up all the forums on The Beatles' She Loves You or the Beach Boys' Good Vibrations on this and other boards.

These and other similar titles are so difficult because, due to the recording and transfer techniques described below, the original production elements that remain do not sync up in and of themselves. Original twin-track session masters for She Loves You have been either lost or recorded over, so what people do is get hold of the twin-track German version of Sie Liebt Dich which features a quasi-isolated music track without vocals. For those of you familiar with postproduction practices, you then correct as best you can for timing errors and perform a layback of the instrumental track of Sie Liebt Dich against the monaural composite track of She Loves You with vocals and all, and then extract and remix tracks as you are able.

Which is not as easy as it sounds.

Unlike digital or film recordings, analog tape has extremely small continuous speed variations throughout the recording that cannot be corrected for after mastering to CD.

Everytime a session is mixed down, the miniscule continuous speed variations are multiplied from those on the master session tape, to those of the pre-mix (described above) and on to the final composite mix.

So the first thing you need to do is go and read all the Good Vibrations and She Loves You posts, before you go off and plunge yourself into DES or DCS hell.

You also might want to know how records were ``built'' in those days, so you can get a better idea of how to ``reconstruct'' them for stereo. The following primer may be of use.

History of Record Production, 1955-1965

In those days, everybody recorded together all at the same time, rather than record everything onto its' own track all by itself one at a time where the musicians are listening to a click track to keep in sync as we do today. Musicians, singers, soloists and sidemen were all there, recording on 2-track quarter-inch or 3-track half-inch tape all at once. If you were going along and somebody messed up halfway through, you had to stop and start again.

As each recording section and each recording selection has its' own unique characteristics, maybe the piece you had there was the exact kind of beginning, middle or end you wanted, and it happened totally by accident, so there'd be no re-creating it later on purpose. So you save it.

Also, there was no such thing in those days as ``drop-in'' or as they say down in the States ``punch in'' where one would perform a portion of the piece, an engineer would capture a few bars in the middle of the partial re-performance and that would become part of the final record.

It was all or nothing.

So in your three-hour session in which you might do as many as four ballads, four jump tunes,  four string numbers or four brass songs, you might have two or three takes that you think you like. These are marked PB for Playback. One of those might be really close to what you are looking for, and that would be marked HT for Hold Take as in we might use it later.

In addition to all that, you'll have all kinds of  False Starts (FS) (intros and a few words) Long False Starts (LFS) (everything but the end, or beginning and end but somebody messed the middle up), Pickups (PU) meaning you liked an LFS but you want to see if you can get away with just doing the missing piece, all of which might have something in it that you liked.

Then you had to pick one take that was going to be used for overdubbing. This was labeled M for Master, whether it was going to be spliced into or used in its' entirety.

So an album back then had roughly twelve songs or about half-an-hour of material, 15 min per side. You spend three hours and do a half-dozen takes of your four songs and do that for four or five days. That equals 16 or 20, because you always had to record more songs than you were going to use on the final LP in case you didn't like how some came out.

This is where all these ``posthumous'' albums with their ``unreleased'' songs come from that comes out after an artist passes away. They don't sound like an album in their own right because the songs are three from this year two from that year, one we had a couple tries at and we just put it together the best we could, etc.

At the end of the recording week, you mix all your Basic Tracks tapes down to mono, whether your final album is going to be stereo or not. This is where it starts to get hairy.

Sync-Ups 

Since there's only three tracks and no way to sync them all up since we are not using film, there's only one other way to do it so that you have room to add in your soloists and sidemen over the weekend.

That is, mix all your basic tracks (underdubs) to mono, and record that mono mixdown onto a second 3-track leaving two open to add in your sidemen or featured players.

So now you have two reels of 3 track:

1. 1st reel with bass and drums, on track 1, piano on track 2 and rhythm guitar on track 3.

2. 2nd reel with all that mixed to mono on 1 and your sidemen/featured players on 2 and 3.

In a vocal piece, the background singers take the place of the sidemen/featured players.

Now you need a lead vocal or solo instrumentalist. But you still only have three tracks.

So you have to do it all over again.

1. Take your original mono-mixdown of the Basic Tracks off track 1 of the 2nd 3-track reel

2. Mix the 2 tracks of sidemen/featured players/background singers into mono.

3. Record both of those mono mixes onto track 1 and 2 of the third 3-track

4. Record your lead vocalist, featured soloist or other prominent player on track 3 and mixdown.

And mix all THAT down to mono and/or stereo on Sunday morning all groggy from being up all night recording other people. So that is also why some ``stereo'' songs of mono hits sound weird. They were only able to find the 3rd three-track as described above. One track of mono band, one track of mono background and one track of mono lead. Mix the lead in the center, have the main track in the left and the background track in the right.

Poof.

Stereo.

Ugh.

Add to that the fact that in those days, after mixdown, producers would splice the beginning of one take, to the middle of another take, to the end of another take in order to end up with the final 45-RPM single you hear on the radio. Now you see why it's so difficult to sync sources, because of the miniscule errors in length in the physical splices, which were never perfect.

After mixing down on Sunday morning, you'd  then grab coffee and a donut in the Donut Factory, run down Santa Monica Blvd one way to get your record cut at the mastering plant, wait an hour and a half for both sides of your LP to be done,  go to lunch, run up to Sunset to get your masters plated in the afternoon so stampers can be made for the record press, wait three hours for that, go to dinner, run that down Santa Monica Blvd the other way to the record pressing plant so you can catch the 8 PM Sunday night press run, wait til the wee hours of the morning for that and then have your test/audition records ready for the suits at the crack of dawn Monday morning. There was none of this business like you have now of `record for a year when you feel ``in the mood'' mixdown for six months and then master for a month'.

----------------------------------------------------------------------------------------------------------

So AA and DCPro are the tools to have. AA has the only Center Channel Extractor, but watch out, there's like three different kinds. Since each one has their own uses, it can get confusing.

The first one, the quick and dirty version just finds the common-channel information and extracts it as it is with no parameters.

About the only thing that's good for is extracting the lead out of a matrix-encoded or discrete-modulated quadraphonic LP. That is mostly because in quadraphonic, mixing the lead in the center is done with much more care than in stereo, because you don't want any lead vocal or soloist in the surrounds. Therefore, no matter whether your LP is discrete or matrix, the lead instrument has very little bleed through to the side channels and is therefore a lot easier to extract.

The three Vocal Removes are as follows:

1. in Edit View under Favorites Options Vocal Remove.

This removes vocals placed exactly in the center with no parameters, leaving the stereo background. Good for Quad LP's discussed elsewhere

However, this one does not work in reverse to enhance the vocal so you can extract it to use later

2. under Effects, Amplitude, Channel Mixer.

This removes vocals placed exactly in the center with no parameters, leaving the mono background. Works in reverse to enhance the vocal, and is good for restoring mono records to extract the performance out of a bed of hiss, pops and crackles.

In addition, see my post under Adobe Audition Restoration

To restore:

1. Hold down the mouse and select the bottom waveform only and drag the mouse across.

2. Go up under Effects, and click Invert.

3. Deselect the bottom waveform by clicking within the song.

4. Go up to Effects, Amplitude and choose 6dB Cut.

5. Go back up to Effects, Amplitude and this time click Channel Mixer

The monaural program will enhance, leaving many of the clicks, ticks and pops behind.

To allow parameters:

1.Go up to Effects, Filters, Center Channel Extractor and choose the maximum FFT and Overlay.
2.Set both the Center Channel Level (plus or minus) and the Crossover.

3.Crossover settings allow extraction of center channel audio which is not exactly in the center.

You can also do this by increasing or decreasing the left or right channel independently in order to center the vocal first, and then perform Center Channel Extraction.

When working with more than one ``reel'' (the Good Vibrations stereo tracks from the box set laying back to the mono single off the Greatest Hits CD for example):

Lay them both in the Multitrack in Adobe Audition, magnify the view and scan the song visually making note when peaks and valleys begin to go out of step with one another. Correct by making sure the beginnings of the songs, i.e. the downbeats are both in sync . Once you start seeing sync problems, you have to see if it's a major speed error or a minor speed error.

For major speed errors, line the two tracks up as before, Enable Envelope Editing, Enable Clip Edge Dragging, Enable Pan Envelopes and Enable Volume Envelopes.

Find the longer of the two incarnations of the song and note its' exact running time down to thousandths of a second. Go to Effects, Time/Pitch Stretch and push Resample and through trial and error, see which setting under Compress Wave down to three decimal points most closely matches the longer wave with the shorter wave.

After that, you have to tweak sections, noticing again when the two sets of waves taking out a few samples in the longer track allowing it to match up more closely with the shorter track.

By ``tweaking sections'' I refer to the fact that the 1st reel of Basic Tracks are generally going all the time in all channels with no rests, breaks or dead spots. Therefore one must consider the Basic Tracks the Master and sync everything else to that.

Utilizing the multiple reel format discussed above, it is therefore possible to sync up the 1st overdubs (2 tracks of background singers or sidemen from the 2nd reel) to the Basic Tracks, simply because the background singers or sidemens’ tracks will have considerably more space between their featured sections than the Basic Tracks.

All you have to do therefore, in addition to re-timing each phrase of the background singers or sidemen to the Basic Tracks, you will probably also have to edit out miniscule sections of silence between each sidemens’ or background singers’ phrases in order to keep them in time to the Basic Tracks.

Now, you still have a lead vocal or featured instrumentalist to sync up. Since no analog instrument can be played 100% continuously, the same holds true for the lead. Simply repeat the above steps you just performed on the background singers or sidemen for the lead.

Do that until you turn blue in the face, you get a headache, your eyes cross and you pass out from too much concentration.

 Then keep going tomorrow and the next day and the next until you get a result you are happy with.

Which be advised you will never totally be, because as of this moment, you have officially been bitten by the Restoration Bug.

Hup. Too late. Can't go back now.

Have fu-unn.

theoxrox
Registered: 10/31/06
Posts: 738

    10/20/08 at 06:23 PMReply with quote#3

Well, diamone, just wanted to let you know that at least one BSN'er is reading your tutorial here and trying to understand it, as it's largely wayyyy over my head.

Please keep adding to it as you get the inspiration!
Peterelliott
Registered: 09/20/06
Posts: 1,028

    10/21/08 at 07:21 AMReply with quote#4

Thanks for this Diamone. Have saved a copy of your tutorials since they will come in very useful next time I do some audio work. I use Adobe Audition but only certain features. I also use Sound Forge and use both programs to create my own DES and DCS mixes.

The biggest problem was working out how to use the multitrack in AA to synch up tracks, so your tips on how to do just that is very handy.


PaulBigelow
Registered: 03/23/04
Posts: 3,651

    10/21/08 at 07:29 PMReply with quote#5

Thousandth of a second!  Gotta do better than than that so one can introduce phase cancellation or instruments drifing from one side or the other.  Just kidding.  Sort of.  I've had to work at the bit level to synchronize and fix things in Audacity.

How does Adobe Audition Compare to Audacity?  Is it subjectively or objectively better?  I'm always looking for better ways as long as it is better rather than just different.

Cheetah drives aren't cheap, either.  Even the Velociraptor would blush.
diamone
Registered: 08/10/08
Posts: 270

    10/23/08 at 11:50 AMReply with quote#6

You'll notice I recommend both methods in succession:
1. use the Compress tool in order to get as close as you can sync-wise
2. go in at the sample level and correct manually where it needs it.

As far as Adobe Audition and DCPro vs Audacity, Audacity has no Center Channel Extractor which you need if you are doing post-sync extractions in order to remix. Adobe Audition has three different ones. All the tools available in Audacity are also available in Diamond Cut Pro in considerably higher-resolution formats.

IMHO Audacity and Sound Forge and all the other ``poorman's'' digital audio workstation programs are mostly for people doing quick and dirty LP-to-CD transfers that they don't plan on doing much with restoration wise.
The resulting recording in Audacity or Sound Forge or Roxio or any one of the other ``entry-level'' programs simply pales against that of DCPro or Adobe Audition.

Most people have both AA and DCPro, especially if they have to lift off vinyl or tape sources for the other half of their sync-up. DCPro has all the inverse vinyl and tape equalization curves, and Adobe Audition as I've said has all the rest of the features in considerably higher accuracy than the ``entry level'' programs.

JRNelsonSr
Avatar / Picture

Registered: 03/02/05
Posts: 1,671

Contact using Yahoo

    10/23/08 at 01:30 PMReply with quote#7

Quote:
Originally Posted by diamone


IMHO Audacity and Sound Forge and all the other ``poorman's'' digital audio workstation programs are mostly for people doing quick and dirty LP-to-CD transfers that they don't plan on doing much with restoration wise.
The resulting recording in Audacity or Sound Forge or Roxio or any one of the other ``entry-level'' programs simply pales against that of DCPro or Adobe Audition.


Considering that SF is a mere fifty bucks less than AA I'd hardly consider it a "poor man's workstation", but it's a lot more flexible (and difficult to use) than cursory needle drops and declicking let on. I generally prefer it for the actual transfers, and the ability to preview edits (combined with an increased ability to hone in on the precice location of the cut) makes it far more useful than AA for cut and paste editing. I find there are some functions one job does better than another, and you pretty much have to save a temporary wav file and switch back and forth until you get exactly what you want.

diamone
Registered: 08/10/08
Posts: 270

    10/24/08 at 08:19 AMReply with quote#8

What about when all you have to work with is a mono song? What can you do with it? I put a slight delay & I think a slight pan & it sounded kind of cool but I totally don't know audition..Let me know more of what you can do with a mono only track.

That's one way. Or you can do the bass-and-treble split some people do, or mains panned hard left and reverbs panned hard right. or follow the ``Stereoize your Selection'' from the other post, talking about Mono Sources, where you create your reverb and then place the left reverb track 90-degrees out of phase left (+) and the right reverb track 90-degrees out-of-phase right (-) and then mix back in. This will also give you a nice pseudo-surround as well, especially if you set the `Original Signal' parameters to Low when it comes to Dry (Original) Signal. Then you'll follow the cardinal rule of No Leads in the Surrounds as best you can.
MusicTrax
Registered: 04/28/04
Posts: 4,854

    10/24/08 at 11:09 AMReply with quote#9

Quote:
Originally Posted by diamone
Or you can do the bass-and-treble split some people do, or mains panned hard left and reverbs panned hard right. or follow the ``Stereoize your Selection'' from the other post, talking about Mono Sources, where you create your reverb and then place the left reverb track 90-degrees out of phase left (+) and the right reverb track 90-degrees out-of-phase right (-) and then mix back in.

The main problem with this is that fake stereo sounds bad.

No matter how you do it, no matter how sophisticated the phase or delay techniques... it's not a replacement for discrete, directional stereo. I also don't believe it sounds better than clean, straight mono.

Just because you can manipulate sound to make it drastically different than the original musicians and producers intended doesn't mean you should do it. I think what you're proposing is a bad idea.

I'm all for experimenting and coming up with new stereo mixes, provided the people doing it have access to additional channels or vocal tracks, and have the ability to mix them in a way that sounds like real stereo. But altered mono tracks like what you propose do not sound good. People have been trying to do this for almost fifty years. It was bad in the 1950s, in the 1960s, and the 1970s, and it's still bad today.

--Marc W.
diamone
Registered: 08/10/08
Posts: 270

    10/24/08 at 05:07 PMReply with quote#10

I disagree.

Remember in those days, a lot of recording studios were fairly anechoic and dead.
I don't drown an original performance in processing, I try to preserve its' integrity as best I can and then update the sound a bit by giving it a little space.

I feel TOO MUCH reverb or any processing is as bad as too little, but done sparingly and properly, the process can give a LITTLE air, presence and ambience where before there was none.

I too am absolutely not a fan of ``traditional'' synthesized stereo that has been around since I was little. The (e)Stereo version of the Brigadoon RCA  cast album comes to mind, or the (e)Stereo King & I Decca cast album.

I hate the (e)Stereo versions of both because as everyone knows, in the case of the Decca the right (reverb) channel is swimming in space and in the case of the RCA, the sends of one have the returns in the other and vice-versa.
As you noted, neither works very well.

I also have the original Columbia (e)Stereo reel to reel of the Best of Jo Stafford from 1959 with all her mono early 50's Columbia sessions.
For that, they split treble to the left, bass to the right, and both channels have a ``vocal range'' on them giving the impression of a center channel seperate and discrete from the rest, when in reality there is none. In addition, this album and others like it retain the ``returns'' of the bass channel ever so slightly in the treble channel and vice versa, supposedly giving a little air.
Again, that doesn't work either.

However:
I suggest the appropriate reverb whichever fits the track, as I said using little to none of the original or `source/dry' signal when rendering the reverb track.
I also suggest mixing in said reverb trackss S-P-A-R-I-N-G-L-Y into the original mono mix after the 90/90 phaseout, since the phaseout kills a little bit of the reverb.
Most of the time I use a 15/75 mix of reverb tracks with no or little source material, just enough to give it a little bit of space, making it sound as if it was recorded in a big live room like 30th Street in New York or Radio Recorders in Hollywood.

For example:
I took the 2-track reel to reel of the Decca King & I cast album reverb on the right track and mono original on the left track and no crossover. I split the reverb track into two, dehissed it a little bit, and tossed it 90/90 out of phase and mixed it back in to the mono original. The reverb on that was so strong and so muddy I only needed 10% of their reverb to get something pleasing with a 85% setting on the mono original.

Then I redid my own reverb in Adobe Audition, and there since my reverb isn't so heavy-handed as Decca's in 1959 I set mine to 22.5% to an 85% mono original to mix down.

I gave both to my Mom & Dad & their various friends, and most of them said the ``doctored original'' sounded like singing in a well, not much different from the original LP's in the late `50's, but that my own recreated stereo synthesis let the original performance come through, just gave it a little air and space.

You can PM me if you want examples, say of the Jo Stafford, their split frequency version live from the 2-track (dehissed etc), the original mono from a safety dub of the master tape some friends of mine had, and my split-surround version.


MusicTrax
Registered: 04/28/04
Posts: 4,854

    10/25/08 at 12:33 AMReply with quote#11

Quote:
Originally Posted by diamone
Remember in those days, a lot of recording studios were fairly anechoic and dead.

That's all well and good, but why are you second-guessing the creative decisions the original producer, engineers, and artists made? You really think you know how to present their music better than they did?

Quote:
I don't drown an original performance in processing, I try to preserve its' integrity as best I can and then update the sound a bit by giving it a little space.

I don't think you're preserving the integrity at all. I think you're changing the recording for the sake of change.

To me, this is as lame as buying the Mona Lisa and thinking this gives you the right to paint a mustache on it.

I have no problem with people who try to clean up old vinyl, particularly for songs never before released on CD, but I do have a problem with those who excessively over-process music simply because they have the ability to do it. That strikes me as a love of technology more than a love of music.

BTW, most of the Jo Stafford Columbia-label hits are available in real mono on this CD:

The Columbia Hits Collection
Corinthian COR-115-CD

--Marc W.
Unclebob
Avatar / Picture

Registered: 11/30/04
Posts: 388

    10/25/08 at 09:14 AMReply with quote#12

MusicTrax seems to think he owns the music, that the rest of us that also own the music can't do whatever we want with it.  I use the SRS Sandbox because it does a wonderful job of manipulating the sound...as it plays.  The original is never changed.

 

Capitol installed an underground echo chamber about 1949 or so at the tower.  Sun records used tape delay to create the echo on their records.  They had no money for chambers.  RCA use a stairwell to get the echo on Heartbreak Hotel.

 

So if I wish to add a little reverb to make a really dull old track cut in a 2 by 4 studio sound better, why not? If I want to spread the stereo with SRS, why would you care?  

 

Finally, the radio stations processed those records that you heard on the air.  The cheap record players rolled off the top and bottom.  Early compressors and limiters introduced distortion as well when the 45 was cut.

 

The one thing you can't remove with all the fancy processing is distortion, especially intermodulation distortion. 

 

So, according to your thinking, if I buy a car I would to leave it "stock" with no aftermarket stereos or better tires, GPS unit added, etc.

 

You really need to rethink your position. How about all those people that crank the bass on car stereos.  Should they get a ticket for ruining the music by changing the balance between high and low?  No.  But they should get a ticket for ruining my peace and quiet.

 

MusicTrax
Registered: 04/28/04
Posts: 4,854

    10/25/08 at 03:20 PMReply with quote#13

Quote:
Originally Posted by Unclebob
MusicTrax seems to think he owns the music, that the rest of us that also own the music can't do whatever we want with it.

No, you only own a COPY of the music. This doesn't give you the right to change the mix, especially to take mono and turn it into fake stereo. I only own a copy of it as well, but I respect it enough not to screw it up and turn it into something less than it already is.

Let me make a few more analogies:

1) you buy a new novel off the newsstand. You write a new chapter, match the typeface of the original, print it up, rip out pages from the book, and then meticulously paste it back into the book. Has it made the book better?

2) You buy a DVD of a recent blockbuster film -- let's say, Iron Man. You go into your backyard with a camcorder, shoot a neighborhood kid in a cardboard Halloween costume as the superhero, then cut those scenes into a copy of the movie. Does this make the movie better?

Again: I'm all for people trying to legitimately remaster poorly-recorded original music. If you can eliminate ticks & pops and carefully remove excessive noise, that's fine. Or if you have a group of songs, and some of them are much louder than others, I have absolutely no problem going in and tweaking their volume so that everything matches -- which is one of many things a mastering engineer does.

When it comes to stereo, if you have access to additional tracks -- say, through bootlegs or unreleased material -- and have the ability to remix a mono release to convincing stereo, I can see doing some experimentation that way. Or if you have the time and talent to extract new information from the mono tracks (which I'm still not convinced works very well), I have no problem with hobbyists who try to create something that at least sounds like real stereo.

But adding reverb, EQ, or phase tricks to mono just results in gimmicked-up fake stereo. It doesn't sound better than the original clean mono, nor does it reflect the wishes of the original musicians, the producer, or the record label. It also insults those of us who enjoy true, discrete, directional stereo.

Fake stereo is an abomination. It was originally created just as a marketing tool so that stereo record labels like RCA & Capitol could falsely label their albums and sell them to an unsuspecting public in the 1960s. It does not make the music sound better. It's a cheap, crappy gimmick.

As far as I'm concerned, all you guys are doing is just pee!ng in the punchbowl "to make it your own," as we say in Hollywood.

--Marc W.
Peterelliott
Registered: 09/20/06
Posts: 1,028

    10/25/08 at 03:58 PMReply with quote#14

Quote:
Originally Posted by MusicTrax
Or if you have the time and talent to extract new information from the mono tracks (which I'm still not convinced works very well), I have no problem with hobbyists who try to create something that at least sounds like real stereo.

But adding reverb, EQ, or phase tricks to mono just results in gimmicked-up fake stereo.


Thanks for clarifying your thoughts on this Marc because I know quite a few of us here do spend a fair bit of time experimenting with DES and sometimes the results can be pretty good. I was under the impression you were completely against this so am pleased you aren't. Yes, the results can be variable but it's interesting trying to extract some kind of stereo.

I definitely agree with you that adding reverb, EQ and phasing tricks to mono material does not get great results and I go out my way to avoid that thanks to Dexterised Beatles albums or Capitol "duophonic" monstrosities!

Nobody's ever heard my work and it will likely remain that way. It's something I do to please myself. I do enjoy spending time trying to extract info from mono to try and create stereo... sometimes it's good, sometimes it isn't and I'm always still learning. It's a private hobby I greatly enjoy and indulging in it actually allows me to learn more about the track in question and study what parts can be separated and get some kind of insight into how they were originally recorded and mixed. No way would I ever use my remixes and "release" them or give them priority when I play my music... the original mix still is THE mix.

ded
Registered: 02/02/04
Posts: 723

    10/25/08 at 04:38 PMReply with quote#15

Let's not let this turn into a name calling session.  Marc stated his points very well, and quite clearly. All one has to do is listen to the old RCA, Capitol, Mercury, & Columbia efforts at creating fake stereo and you can hear his points quite nicely. They didn't sound good then and certainly don't sound any better today.

Dave

Page 1 of 6   1  |  2  |  3  |  4 >  » Last
Reply