Vinyl Record Cutting Dynamic Range Compression

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello !

I hope someone out there can give me some advice.....

I am transferring my old record collection onto cd to save wear on my turntable.

In researching the way music was transferred from tape to vinyl, I have found that a number of things were done:

1) Low bass was cut back in the inner grooves.

2) Treble was also cut back.

3) Compression was used to accomadate the lower dynamic range of the vinyl.

Does anyone know any specifics about the three points mentioned ? Regarding what frequencies, what compression ratios etc.

I purchase a dbx 3bx expander and a dbx 1/3 octave eq. for the transfer and would appreciate any information about how to set them up.

It seems that the dynamics and frequency response are better in the outer grooves and that one set of values for the eq and expander is not correct for the entire side of the disc.

Perhaps there are some retired cutting lathe engineers out there who would care to help !!!!!

Thank you in advance for any help

oldheathkitphi
 
Thank You Hugo,

I would like to try and recover some of the original dynamic range that was lost when the transfer from the master tape to the vinyl cutting lathe was made. The master tape in general should have had a dynamic range of about 65 db whereas the vinyl at best was about 50 db.

Also Treble and high frequency rolloff and bass compression was used to compensate for the lessening ability of the stylus to track properly in the inner grooves.

I am hoping as I transfer a record to cd to adjust the expander and equalizer to compensate for this, but I need to know a few
specifics about how much expansion to use and what frequencies
to raise.

Thank you very much for responding.

oldheathkitphil
 
Hi oldheathkitphil,

I wonder if you gain something by that. Compression is also done by the sound engineer as an “artistic” tool as part of the complete sound balance. Also some processing is done to keep phase difference between left and right within limited margins (otherwise the needle will jump out of the groove)..

Since you have no knowledge about what and how much processing is done it is almost impossible to reverse process it. To only thing you can do is to fiddle around experimentally to suit your taste.

;)
 
Ex-Moderator
Joined 2003
oldheathkitphil said:
I would like to try and recover some of the original dynamic range that was lost when the transfer from the master tape to the vinyl cutting lathe was made.

You try to recreate something that is lost. Unless you can get your hands on the master tapes you will not succeed. Real good equipment will get the most out of the grooves.
There is no such thing as a difference in equalization between the inner and outer grooves, assuming your record collection is post RIAA standards. Before these standards, almost anything was possible and I would agree that experimenting with equalization would be necessary.
On modern records, over the entire surface, the bass is cut back and the highs are boosted within a well defined curve. The phono preamp compensates for that in the opposite way.
(RIAA curve – Inverse RIAA curve)
You do have a proper phono preamp I assume?

/Hugo
 
Thank you Piotr and Pinkmouse fro responding.

Over the weekend I found a book in the public library entitled
Sound Recording by John Engle 2nd edition. In the chapter discussing vinyl disc cutting, he describes the various forces on the cutting stylus while it cuts a master disc. There is a graph that shows the natural treble rolloff as the cutting head
goes toward the center of the disc. He also states that this rolloff
can not be compensated for as the cutting head for various reasons would not be able work properly, things like overheating th cutting head coils and waveforms being too small to cut due to the diminished circumference as the head tracks inward.

That was the the general information I think I was looking for and is not the same thing as the RIAA characteristic.

Basically it shows that there is rolloff at about 10kz of 5db by the time the pickup stylus has passed the 5 inch from the center radius point and this rolloff continues to about 8 db at the innermost groove distance of about 3 inches.

Can anyone out there verify this.

Basically it is saying that as one transfers a lp to cd one has to slowly increase the treble to compensate for the rolloff which increase as the stylus tracks into the inner groove area.

Also, along with dimished treble response, compression has to increase so that mistracking of the pickup stylus will not happen when the record is played.

That is why I think if i slowly increase the treble with the equaliser
and slowly do some expansion as the pickup travels inward that I should get a better quality transfer.

Can anyone verify ????

Thank you again in advance for any responses

Oldheathkitphil
 
May i offer you, instead of a technical solution, a more philosophical one? It will take the sorrows away, let us all chill quite a bit, and best of all, it includes redwine and women.

So you have vinyl and want to hear them digital? Fine.
Vinyl won't last forever, this is true. And since you can't take'em with you, once you die, you could sell them aswell.

Since i just recently are a converted analogist, saved from the deamons of the digital domain, i may sound a little enthusiastic here :D.

Records and Redwine and Women share a common denominator : Saving them up for future times is a bad, bad thing. :angel:

It gives me headaches imagening myself sitting infront of a very fine cabernet/women - determined to preserve it/her for futuretimes.
I can instantly comeup with a few dozen people, declaring me a blasphemist :smash:
 
Ex-Moderator
Joined 2003
The cutting head theory is something I was not aware of. Unfortunately I can't find anything from Mr. Engle on the net.
It has never occured to me that treble was weaker towards the inner grooves and 5 to 8 dB is quite a lot so I'm interested to learn more about the subject.

/Hugo
 
The relationship between seconds of music per meter of groove changes dramatically from the outer tracks to the inner tracks, like a 3:1 ratio. Vinyl should have been made with "constant lenght speed" rather than "constant angular speed", like CDs and DVDs, that spin slower for the outer tracks and faster for the inner ones due to the constant data density. I suppose they just didn't had the technology for dynamic speed change at the time.

I have very little experience with vinyl, but the facts suggest that a substantially reduced sound quality has to be expected in the inner tracks. This is comparable to recording a tape 3 times slower than the rated speed: You pack more data in less length at the expense of a lower trebble cut-off.

In any case, you should rather get good "verbatim" 96/24 wave files from the vinyls first, and then do the processing by trial and error in the digital doman. For example you can compare (by FFT or by ear) the beginning and the end of the tracks and try to match them. There are tools for noise reduction too...

BTW: It's a big shame that, now that we have plenty of digital storage media featuring huge amounts of bit-exact data forever, we are doing the worst ever mastering jobs just to make it loud on boom boxes.
 
Thank you all for responding.

The book I referred to is Sound Recording, second edition, by John Engle, ISBN # 0442225571, Van Nostrand 1980.

On page 314 he has a diagram showing the treble rolloff with respect to distance from the outer edge to the inner edge. It generally show that the physics of the cutting stylus has to go down about 5-8 db at 10 khz from outer to inner groove, and that this is not compansatable.

Also Bass compression is implied as the stylus would mistrack
on the inner grooves compared to the outer grooves if it was not.

Its sort of like recording onto a tape recorder that is constanly but slowly, slowing down to half its speed as the recording time increases. the frequency response of the system and its dynamic range goes down and one set of values can not be used for the entire transfer.

Therefore it would seem that in order to get a better transfer from disc to cd one would have to slowly increase treble and expansion as the record was played.

What I was looking for was some specific values of frequency and compression to start with.

Also, the manual with the dbx 3bx expander explains quite well how a 3 band expander wont allow cross talk breathiing like single band expander can.

Again thank you for responding and look forward to any ideas that you or others may have.

Hopefully, there are some retired cutting lathe engineers out there who can share their experience

oldheathkitphil
 
this puts some light on it:
http://weeklywire.com/ww/10-06-97/austin_music_feature1.html


pinkmouse said:
I suspect any compression done during the cutting process was "on the fly" as the engineer balanced track dynamics with available groove space on the vinyl, and as such will differ from album to album, and from engineer to engineer.

Yep, and I am afraid..., every album is unique.
Maybe the quest for the ultimate "Original" was allways a academic quest. Liveconcert/Opera: Once the final note has echoed out - that was it! Digital or analog , just a interpretation of technology. A interpretation thats good enough ot be excited.
But oldheath, this does not mean you cant work on the treble and dynamics to get a wavefile that is pleasing your expectations.


Some interresting from the link above:

...High frequencies, like the "S" sounds in vocals or a drum kit's cymbals, generate so much energy that they can damage the lathe's cutter by forcing it too deep into a lacquer's soft acetone material. Instead, the machine's internal safeguards automatically reduce high frequencies to levels similar to the rest of the recording. The way around this is called "peak limiting," the process of controlling groove-by-groove the odd dynamic peaks created by occasional percussion or loud guitar string slaps...

...Unfortunately, the nature of records is one of limited space. A 12-inch record has 86 inches of recording space, and can only fit 225 grooves per inch. Furthermore, there are only so many grooves that can be cut into a side before those indentions begin colliding into each other. That resulting mess, called an overcut, causes mistracking and can actually kick a needle out of the groove. The formula used by many large mastering houses, calculated mostly by computers, simply plugs the time of a side into the size of a record thereby giving the mastering engineer a hard number of grooves per inch. "It can get the music to vinyl in a technically proficient way," says McNair, "but it doesn't take into account what the producer or artist are looking for in regards to volume, bass, treble, and compression."

The artistic approach is based more on trial and error than it is mathematics. The formula may still be used, but the difference lies in the test cuts -- practice cuts of an album's loudest or most bass-heavy section that are used as a sample representation of the rest of the record. Those cuts, made on test lacquers, can then be listened to on a phonograph or examined under a microscope for overcut. At that point, groove depth and size can be manipulated until the overall record meets the producer's expectations. The difference between this method and that of large mastering houses is that Dalmasso will listen to the same piece of music over and over, tweaking small instrumental passages or whole songs until the recording isn't just reproduced on vinyl, but rather becomes a separate and distinct vinyl work.
 
Administrator
Joined 2004
Paid Member
Don't make the mistake of confusing dynamic range with frequency response.. I understand the reduction mentioned is in terms of actual dynamic range not response flatness after equalization is applied. The response is still flat, the problem is that the dynamic range across the face of the disk drops from a high of perhaps 60dB or so at the lead in grooves, to something around 50dB at the lead out grooves. (The inner grooves cannot handle the same modulation velocities due to the lower effective linear tracing velocity.)

I play a lot of vinyl and a have a system with a fairly high level of resolution and notice no significant change in tonal balance across the face of the disk - I do notice audible increases in distortion towards the inner grooves on some disks. (I also have hi-rez digital both sacd and pcm 24/96 to compare against.)

Incidentally a dynamic range of 40 - 50dB is about the limit of tolerability for most music in a domestic environment, wider dynamic range requires either very low domestic noise levels or crescendos that are excessively loud.

While I think they are now carrying compression to an extreme on a lot of recorded material (the loudness wars) some compression is frequently required on material with "excessive" :devilr: dynamic range for home playback.

For me a comfortable dynamic range in my listening room rarely exceeds 30dB or so, and before you flame me think about it.. :D

IMO The kindest thing you can do to this material is use no processing whatsoever and make the highest possible quality transfers to wav files - perhaps even 24/96 which you can downsample for cds. Apply the processing after the conversion and keep the originals. You won't be able to apply any sort of complementary equalization or expansion because you have no idea what was done at the other end.
 
Thank you Kevinkr , Eva and Elvenear

Yes. sometimes I feal I might be like Don Quixote and chasing windmills, but I would think that somehow it should be possible to get back to the sound on the master tape, if one knows where compression and rolloff were in principle probably used.

Most of my records are older classical performances that will never be released on to cd and if they were probably would be victims of as you said the "Loudness Wars"

I am just trying to get another 10 db or so of range into the performance.

I do initial transfers with no compensation at all and then make a copy of the file and work on that

Again thank you all and would appreciate any further ideas.

oldheathkitphil
 
May it be to much asked if you told us what hardware you use/is involved to create your wavefiles? I sense from your last post that there is much passion involved in your project. And i am sure everyone here understands. Maybe, there is a way of figuring out a procedure for each record to set, for instance , a expandersetting.
 
OK, as far as we have now understud from the practice of mastering during the cut of the lathe, its all about the timelenght recorded on side A ,and B. Up to about 12 minutes there are no compromises, except the very high frequency instruments and the intense bass signals. So, if we look up what record we have, say its a 25minute liverecording, thenwe should expect that there must have been done some compression to the sice of the groove. Hmm. but even then, the Person doing the mastering might have adjusted the compression during the silent and the loud parts. Hmm i feel like going in circles here. Anyone have some new ideas??
 
Whilst what has been said about modulation depth is certainly true, on a good recording collaboration between the producer and cutting engineer is used to decide on the order of tracks (where possible), so that those nearest the centre of the disc are less demanding than the outer ones. Groove spacing can also be adjusted dynamically to allow heavily-modulated passages sufficient 'space'.
 
A good example of what can be done on vinyl in terms of dynamics and recording time is "Powaqqatsi", a film music composed by Philip Glass (Nonesuch Records 79192). Nearly 74 minutes of acoustic music with incredible dynamics, it is worth listening to. I read somewhere that they used a computer-controlled groove density control, that is the grooves with higher bass content are spaced more apart, and grooves with smaller amplitude are cut at higher density. This is controlled from moment to moment within one turn of the cutting turntable, so they get extreme high groove density and still high dynamic range.
 
Like kevinkr said, I think you should record it to wav without any extra processing. I suspect that the recording lathe would have some eq tied to the cutter position to keep the frequency response constant, so you shouldn't have to worry about that.

There must be some dynamic range expander plugins you can try during playback.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.