Your Vinyl Transfer Workflow (sharing best needledrop practices)*

The overcompression used in CD's to crank up the volume and reduce the DR to 06-08, brought me into the world of vinyl. This is not music, these are tribal screams.

When I use the Klanghelm "vari-mu" compressor plugin to make the sound better "glued" and more natural, I put the compressor threshold at the same value as the "Max. RMS level" of the material (eg. if the max RMS level is -12.62 dB, the compression threshold is the same, 12.6). Slow attack, medium release, but they are material dependent and oscillates between 0 and these limits. And I use the side channel high pass filter to 160 Hz to not attenuate the LF bass. Finally, the right make-up to push the volume and total RMS to -18 dB (0 VU). There is no Dynamic Range reduction using this settings and the sound is louder but richer, with the previous "silent sounds" boosted.

 

Yet, a lot of product has been released on CD that way. Careless engineers don't align the heads and check the phase. Using a misaligned stereo head stack on a mono tape sounds bad, even if the end-listener never listens in mono.

 

What's wrong with 88.2?

 

Truth be told, I went through a period of also allowing overs for my own projects in the late 90s. I had just learned that compression/limiting was bad, and I also discovered that I couldn't hear the clipping distortion. I soon stopped that practice. I'm sure I still have a couple of things on my hard drive where I did that. The trick to doing overs is to do it in 32-bit. Go over a couple of db, then dither it to 16-bit.

What I do sometimes today is find outstanding peaks, and instead of using a limiting/compression, i'll simply zoom in and lower the volume of the peaks, ensuring that it won't impact on the apparent impact of the sound. From there I can simply raise the overall gain of the file. No compression/limiting or clipping, and I still get a louder file. It is time-consuming, though.

I have a feeling that people like that clipped look because when people load up stuff in a computer for whatever reason, they like to see things going into the red. They like to see that brickwalled look on a waveform. For the average person, that looks "professional". When they play some music and it's so loud they have to turn down the volume, they think that's professional. Wow.

Thank you! I've been saying this for years and years and everyone blows me off. But, it started earlier than that! I bought CDs in 1989 that were brickwalled. Not compressed, just brickwalled. I think a lot of people say it started in the 2000s because it didn't affect music they listened to until then.

 

Sounds good. I haven’t got to the EQ/filter stage of needledropping-something to aspire to...

thanks so much for taking the time.

 

Thanks. That is so much information-I have read the post a dozen times, and still don’t grasp all of it-but I will keep trying...

Again, thanks for taking the time...

 

Cheers![/QUOTE]

I like this version. It has a fuller sound, more body to it.

I have always liked BrillantBob's work but this one fell short for me. It sounded a little thin and the sax sounded distorted.

 

I have always liked BrillantBob's work but this one fell short for me. It sounded a little thin and the sax sounded distorted.[/QUOTE]

I am playing around with my new vari-mu compressor plugin. MJUC. I still learn to find the best settings.

 

What? And no one listened? The bast**s!


I haven't come across any remastered CDs before 1998 that were made loud with compression. They didn't even have the tools for that back then. As for some idiot engineer mixing a band badly with overs and whatnot I have no idea. Could you give me a title of one CD that had this problem back before 1990. I will find the engineer responsible and have him/her/it severely punished.

I can't speak for other studios but this is how we transfered analog tape:

Inspection of the analog tape. We check for bad spices and fix them. Unfortunately some clients have a "no touch" tape policy. Which means you have to get the clients permission to do any cuts or re-splices to the tape.

Let's say for example your studio has the tapes for 24 hours. You have plan to transfer the tapes to: 24/352.8, DSD, and a nice 1 inch analog copy at 30 ips. The tapes are for one studio album. So that's two quarter inch reels. One for Side-A and one for Side-B. Upon inpsection Tape "A" has one bad splice that will probably come apart during playback. And Tape B has 3 splices that have come apart in the same song. Back in 1972 the engineer in question made three distinct mixes of "Baby, I Got Wood For You."
Mix 1 for the verses and chorus. Mix 2 for the Bridge. And Mix 3 for the end part that has the 30 bar solo in 5 time. And then he spliced the three together. No automation in dem days! So you need to fix these 4 splices but the client has a "no touch tape" policy. In this case you have to get permission from the bass player/singer of the band. Unfortunately he went into drug rehab 2 days ago and the clinic has a "no visitor or phone calls" policy for the first week of the program. So what are you to do? You have no choice - you have to fix the splices. This nonsense has happened to use more than once.

We next see if the tape needs to be baked. Brand, tape number, and year of manufactor will tell us that. Otherwise there is a 20 minutes test that will tell us if the tape has SSS.

When that nonsense is finished I play the master/copy in our Ampex quarter inch half track machine. I will align the height azimuth of the playback head for maximum treble and bass for each song. Most studios don't do this. They will put the tape in and then align the playback head with the first song. This is good but it is not good enough. Here we decide if the master needs some help.. But our EQ never changes the tonal balance of the album. For example: 40 hz + 1.5 db, 3 khz + 1 db and 14 khz + 2.5 db. Nothing you would notice. And it adds a little zip to the master without changing the signature sound of the album. Mastering Eq Should never be more than 3db. They are exceptions. Older 60's albums that have a real honk in the lower midrange (500 - 800 hz). We do this EQ during the AD conversion with our Manlely Massive Passive all tube passive mastering stereo EQ.

Unfortunately too many mastering engineers do too much. One of the popular foolishness is reducing the cloud. (90 - 250 hz). This takes up most of headroom so modern engineers hate it. If your 1976 mix has a good amount of cloud they will pull it down. This is where most of the detail in bass is by the way! How does - 5 db @ 150 hz @ a Q of 1 sound to you? We don't touch the cloud. Not on a remaster. Some 18 year old's over bassy love ballad mix is another matter? Oouch!!

Then comes the level. We don't hunt for peaks. That would take forever. We set our ADC level by the 1 khz test tone that is at the beginning of most tapes. Real easy. The 1 Khz tone is recorded at 0VU and for 24 bit 0VU = - 18 dbfs RMS.
Note: For test tones RMS is the same as PEAK.
This is how they did back in 1986. But back then it was 16 bit. (0VU for 16 bit = - 12 dbfs RMS) there was no normalization then. This is why old CDs are so low in volume.

Then in Pro Tools we: edit, sequence, do fades, sample rate conversion, NSD and normalization.
After all that is done the file is reduced to high definition 24/176.4 and CD 16/44.1 version. All the files are saved to a USB Flash Drive. And that is sent off to the CD/Blu-ray plant for pressing.


We don't try and make any album or box set loud to compete; ruin it with OVERS, add distortion (saturation) or pull down the cloud.
That would be like chopping off my right hand because one armed men are the rage these days!

We might add some real tube sound to a digital master that sounds cold and lonely. We have an ancient Ampex 1959 tube preamp at the studio. If needed we can run stuff through there. But we would do this during the AD conversion.

There have been occasions where we wanted the CD to sound just like the record. In one case we: summed the bass to mono below 85 hz @ 24 db per octave. Added the same compression that was used in cutting the record. We even dropped the sample rate down to 36 khz to simulate the upper cutting peak of vinyl. And we went one step further. We took the average frequency characteristics of the most popular MM carts used between 1975 - 1984 and applied that as a EQ curve over the master file. Looking back now that might have been to much. The idea was the full vinyl Monty but without the vinyl. We only went that far once.

Engineers doing "overs" has got to stop! Even if you don't hear it as PCM distortion you will hear a certain harness the music.

And that is my Canadian nickel.....

 

88.2 khz will give you gout.

 

No, it wasn't a remaster, just a small series of oldies comps by the young Priority label when they were still trying to find their niche.

 

Believe it or not, I am quite familiar with all the things engineers do to produce a CD form master tapes. It sounds ridiculous to a lot of the purist forum members, but this is what is done. This is why the industry doesn't like to tell us what is actually done. Some people would freak out.

 

Wait! Has anyone jumped off a roof yet? I know a few so called "popular" mastering engineers that will add fake tube distortion to an old master tape because they think it sounds better. Sorry, "saturation." And do aggressive EQ. High passing all the bass at 40 hz is very common. There is even an award winning (We know ya stole that trophy laddy!) mastering engineer who high passes everthing at 50 hz. And don't get me started on the high pass madness in mixing! I didn't bring my bucket with me.

 

Ahhh, you scared me there for a second!

 

After years of doing this and using this thread as a great resource. This is my workflow for needle drops:

Cleaning Record Stage

This is an extremely important step and makes life easier down the track. Unfortunately, I haven’t got an expensive ultrasonic or suction cleaner. However, I’ve found a relatively cheap and very effective process.

· Firstly, I’ve discovered Record Revirginiser. I got it for about $45 Australian, its great stuff. Massage it into the record, let it dry, peel it off and it effectively cleans the record very well. You could clean about 30 albums with a bottle of it.

· Sometimes I’ll use my Knosti Antistat (similar to a Spinclean) in conjunction.

Equipment

· My humble Audio Technica LP120 TT, modified/improved. USB Pre amp removed; heavier counter weight and old rubber mat added.

· Cart AT 440MLa with AT150SA stylus.

· Phono Pre Amp, Rega Fono.

· Audio Interface A/D Convertor, E-MU0404USB

· Software for capture and processing, Izotope RX5 Advanced.

Recording and Preparation

· Capture at 32 (float)/192, set levels for highest peak at -6dB

· Trim file and set regions as tracks.

· Invert Phase 180 degrees, I need to do this because the E-MU0404 inverts the phase when sending the signal to my PC via USB.

· If it’s a mono recording, sum to mono.

Analysing Files

· I then start listening to the files, at first I’m listening to hear where record noise is audible (quiet passages) or if it’s just audible in the (side) S Channel. I mark those areas as regions.

· I then split the file into the M (middle) and S (side) tracks using Channel Ops. For those regions I declick, denoise etc., the M & S Channels separately. I then resum the file back to stereo and continue declicking, denoising etc. the rest of the file.

Digital Cleaning

· I do not arbitrarily run the declicker at a prescribed setting. I proceed through the file bit by bit, declicking at the needed rate. This is the difficult part, you have to make sure you’re not taking out transients when you declick. It may sound nicely cleaned up, but if you A/B properly, you may notice that the ‘clank’ of the snare drum has lost its ‘…k’.

· If I don’t hear clicks in an area, I don’t declick there. I’m usually declicking where I do at light settings. I deal with big clicks individually and zoom in, where you can use stronger settings.

· Denoising is only used where I hear record noise, mainly in-between songs and quiet bits. I also use decrackle sometimes.

· Once I’ve finished all that, I save the 2 sides of the record as 32(float)/192 Wav files for archiving.

Audio Processing

· Using the Corrective EQ from RX5, I apply a High Pass Filter. Generally at 30Hz with a 12dB/Oct slope. For a handful of records (mainly from the later 70s or early 80s) I used 25Hz with a 12dB slope and for one record I had to use 20Hz.

· I use the digital setting. It’s important to use the right slopes and be careful about your Frequency Precision settings. The middle ground, 12Hz usually works best but not always.

· Usually that’s the only EQ I apply, however sometimes further EQ is necessary. On occasions I’ve had to cut some highs because of shrillness, usually around 6000 to 8000 Hz, only about a 1dB reduction. And on occasion a cut of about 1dB has been necessary at around 120 Hz. I’ve never been inclined to increase frequencies.

· I then adjust the gain, I don’t normalise or worry about peaking at -0.3 or -0.7 db. I have found that increasing the gain too much actually makes somethings you don’t want louder, louder. You can even lose clarity in the top end. Having captured at -6dB some increase is necessary. I usually increase gain so the RMS is between -15 and -12 or even -9. That sometimes results in a maximum peak of -0.3dB or sometimes -2.0 or -3.0dB.

· If I do get overshoots after that process (happens occasionally), I mark those peaks, zoom in and reduce their gain before increasing the gain of the file.

After Processing

· I export the files as tracks (Wav files).

· Using the Saracen software, I dither the files to 24 bits using a standard TPDF dither.

· I then tag the files using DBPoweramp and then finally convert them to 24/192 FLAC using dBpoweramp.

 

Excuse me, 51 and not getting any younger but what is "16 core processors?"

 

They are both insane Grant.

But they have a secret to getting 24/352.8 quality out of their vintage DASH machines. Every morning Chris and Tom take a belt and beat their DASH machines. I heard the story somewhere...... LOL

 

Yes. Intel and AMD make 16 core processors. Where have you been hiding? And, Intel has their Xeon W-3175X with 28 cores! $3000 baby!

 

Modern CPU are basically several CPUs in one. Even cheapo laptops have 4 cores (4 CPUs in one), I think the bleeding edge in the PC world these days is 64 cores. It makes them faster, they can do several things at the same time as if they were 64 computers and not just one.

 

Oh! If only things were so simple.

 

Didn't grow up with computers. Never had my own until 1999. Not everyone is up on computer terms.

 

Don't large studios have tech people whose job it is to maintain these machines? Or are they just standing around? Even when I was 15 I knew the right way to copy a mono mix on cassette to another stereo cassette deck. I couldn't align any heads but I would take the loudest signal (left or right) from the tape and split it with an adapter and plug that into the copying deck.

 

That's the claim anyhow. Bus speeds are in fact just as important. If I tried to burn a CD and rip some music to my cell phone at the same time it would slow to a dead crawl or not work at all. So much for doing a bunch of things at once. LOL

 

Hello John,

I used the vari-mu MJUC MK3 compressor plugin with threshold -26 and no saturation and the Dynamic Range boosted from 11 to 13 !!! How it's possible? Stronger transients?

 

Don't know. But the plugin looks very useful.