Mastering for a compilation in vinyl - All coming from squashed CD's

[Gomez]

I did a search and nothing relevant came up.

I've been asked to master 14 songs for a double vinyl compilation. I'm familiar with the process of mastering for vinyl. However, in this case all the songs come from very loud (rock) CD masters. It's all a bit last minute so getting uncompressed 24bit wavs is not possible.

I've got a pretty clear picture of what to do EQ wise, but not 100% sure on how to "restore" dynamics. I was wondering also how much dB's will you guys drop of a heavily squashed wav and what peak level will you send these through when they're peaking at about -9 RMS

What would you gurus do in this case? Any tricks or suggestions would be appreciated.

[Cellotron]

Lower the peak level of the tracks that are most smashed to match the level of the more dynamic tracks. The peak level of these is essentially irrelevant as this will be set as needed when the program is transferred to the cutter head anyway.
EQ if/as needed (without "dumbing down" any of the tracks) so that spectrally tracks are slightly more uniform. Possibly tame any sibilance or overly bright percussion as needed.

And then hand it off to the person who will actually be doing the mastering - the engineer running the lathe - who is in fact in a better position to determine what the tracks may or may not need in order to allow them to translate to the vinyl master.

Best regards,
Steve Berson

[Surbitone]

I've found Voxengo's Gliss and HarmonEQ to be really effective for bringing back a little movement into a squashed track and also for subtly de-essing/harshing whole mixes. I used this combo recently for a clipped / limited reggae track that was destined for 7", and it did the job nicely.

[lostintime]

For restoring dynamics CuteStudio's SeeDeClip seems like an excellent choice, especially for the price.

[Odeon-Mastering]

Originally Posted by Gomez

What would you gurus do in this case?

I am definetaly no guru but I would probably do exactly as Steve (Cellotron) suggested

[Gomez]

Hey Steve, thanks a lot for the advice. Much appreciated.

[Paul Gold]

Adjust relative levels and make the disc.

If there is something way off like terrible sibilance or wacky EQ, fix it. Other than that I would try not to change the masters that have been approved.

[Gomez]

hey steve and others

I just thought about your post, the lowest peak i have in ALL the tracks is 0.0dB. Crazy. Do you guys think that dropping the level to say, -3.0 dB of the least squashed masters and match all the others to the rms of that? That of course, will mean that the max peaks will be different throughout. What's more important in this case? To match to average volume or get the all to peak at the same? I'm inclined towards the first...

Originally Posted by Cellotron

Lower the peak level of the tracks that are most smashed to match the level of the more dynamic tracks. The peak level of these is essentially irrelevant as this will be set as needed when the program is transferred to the cutter head anyway.
EQ if/as needed (without "dumbing down" any of the tracks) so that spectrally tracks are slightly more uniform. Possibly tame any sibilance or overly bright percussion as needed.

And then hand it off to the person who will actually be doing the mastering - the engineer running the lathe - who is in fact in a better position to determine what the tracks may or may not need in order to allow them to translate to the vinyl master.

Best regards,
Steve Berson

[Paul Gold]

Always level match by ear. Anything else is just plain wrong. The overall level doesn't matter to the cutting engineer. It will be adjusted for the best cut.

[Waltz Mastering]

Originally Posted by Gomez

hey steve and others

I just thought about your post, the lowest peak i have in ALL the tracks is 0.0dB. Crazy. Do you guys think that dropping the level to say, -3.0 dB of the least squashed masters and match all the others to the rms of that? That of course, will mean that the max peaks will be different throughout. What's more important in this case? To match to average volume or get the all to peak at the same? I'm inclined towards the first...

Having the peak or rms levels consistent song to song imo is not as important as having the overall perceived loudness consistent song to song.

Might be a good move to lower the songs from full scale a couple dB, and see if expansion and eq can tame any of the harshness as you even everything else out by ear.

[Stereophonic]

Originally Posted by Paul Gold

Always level match by ear. Anything else is just plain wrong.

Agreed. But in addition to accurate VU metering.

Originally Posted by Paul Gold

The overall level doesn't matter to the cutting engineer. It will be adjusted for the best cut.

I have a master that is headed to you. Q:IF I recreate all the fades in the record, I can send you a 2488 master, otherwise I have the 2444 ready to roll, client approved. Is it worth my time to assemble a double sample rate master? I know it's splitting hairs considering many engineers cut only from 1644 cd audio.

[Paul Gold]

Hi Timothy,

I am format agnostic. You can send whatever you think is best. Send them as either single files that when lined up recreate the sequence or a single file per side with a running log. If there are crossfades I suggest a single file per side.

[Gomez]

Ok guys, I think session went pretty well today.

I lowered the most dynamic track by -3.0dB and used it as reference for loudness. Then I did processing to the tracks as needed (some had some crazy HF and some had some crazy LF amongst other things).

Now, I tend to send vinyl masters in a DVD with folders for each side containing its songs. IE:

Side A --> track 1, track 2, track 3
Side B --> track 5, track 6
Side C --> track 7
Side D --> track 8, track 9.

I also add a text file with information for the cutting engineer in the same disc.

I don't really assemble the tracks the same way I would have done for the CD masters, IE: I just send the individual tracks and the leave the cutting engineer to do the track markers - if there's something really specific, I write it down on the text file and hope for the best.

All I'm getting to with this is: is there a "better" way of doing it? Sometimes I'm concerned about the engineer leaving a too long pause in between songs etc.

Thanks again people! thumbsup

[soundroid]

Originally Posted by lostintime

For restoring dynamics CuteStudio's SeeDeClip seems like an excellent choice, especially for the price.

from my experience declippers do little to nothing in terms of restoring an over compressed or limited track. they do make the waveform look less limited!

[lostintime]

Originally Posted by soundroid

from my experience declippers do little to nothing in terms of restoring an over compressed or limited track. they do make the waveform look less limited!

Judging from things the users of SeeDeClip say it can take audio which is fatiguing to the point of being unlistenable and output audio which is actually pleasant to listen to. SeeDeClip will make the waveform look less limited but it makes much more than just cosmetic improvements.

[soundroid]

Originally Posted by lostintime

Judging from things the users of SeeDeClip say it can take audio which is fatiguing to the point of being unlistenable and output

have you tried it? do you also share these users' view?

[Adam Dempsey]

Originally Posted by Stereophonic

IF I recreate all the fades in the record, I can send you a 2488 master, otherwise I have the 2444 ready to roll, client approved. Is it worth my time to assemble a double sample rate master?

I'm obviously not Paul, but seriously, why submit lower res if you can avoid any totally unnecessary conversion/DSP? It's what you leave out that's as important (and sometimes more important) than what you add. If I were supplying the cover artwork I'd be sure to send a high res file rather than a 72dpi jpg...

Originally Posted by Stereophonic

I know it's splitting hairs considering many engineers cut only from 1644 cd audio.

I, for one, wouldn't call it "splitting hairs", I'd call it "best practice", assuming the better source is available. I'd avoid any cutting places that say they cut only from 16/44k CDs (though I don't know of any, either).

[lostintime]

Originally Posted by soundroid

have you tried it? do you also share these users' view?

No I haven't, have you? But it does appear that SeeDeClip does what it claims to do well which makes it something that the OP could've found useful. They have a demo version for download which will process up to 30 seconds of audio.

[DSPaudio]

Re SeeDeClip,

Jesse Graffam has posted positively about it here in the past and he seems sane.

I tried the SeeDeClip demo twice and think it could be useful in a situation where you have a heavily clipped/limited master that you need to further process.

As for the user quotes on the web page, I wouldn't put much faith in those. Consider how much an untreated room changes the sound when you shift your head 3 degrees to the left. Above all that, SeeDeClip with default settings lowers the volume of the output file. The lower volume would automatically translate to less harsh regardless of what the software did on the declipping front.

Really it deserves a spectral analysis and Null test to determine how useful it is to a mastering engineer. Life is short, perhaps the author of the software would post audio examples, null outputs and spectral before/afters... I tried to write him about it just the other day but my mail bounced. He mentions FFT plots on the web page but they aren't anywhere to be found?


Cheers,

[Paul Gold]

FWIW the DSP denoisers and declippers I've tried can make the audio sound better when listening to the file. For reasons that I don't understand files processed with this stuff often cut horribly. Massive distortion. I don't understand it but I'd be careful about leaning on that stuff if it's being cut.

[Gomez]

Originally Posted by Paul Gold

FWIW the DSP denoisers and declippers I've tried can make the audio sound better when listening to the file. For reasons that I don't understand files processed with this stuff often cut horribly. Massive distortion. I don't understand it but I'd be careful about leaning on that stuff if it's being cut.

I have to say that I'm very suspicious about these "de clippers" and other stuff... How can you restore something that IS NOT there in the first place? Sounds too much like a gimmick to me. I generally find that those "miracle" plugs or software leave a great first impression and sad face in the long term - IME.

And as for what Paul says above, I believe it, I think these thing may just introduce some weirdness... IMHO.

[lostintime]

Originally Posted by Gomez

I have to say that I'm very suspicious about these "de clippers" and other stuff... How can you restore something that IS NOT there in the first place?

By making a good context-based guess.

Sounds too much like a gimmick to me. I generally find that those "miracle" plugs or software leave a great first impression and sad face in the long term - IME.

If it sounds good it is good.

And as for what Paul says above, I believe it, I think these thing may just introduce some weirdness... IMHO.

If you don't try it for yourself you'll never know for sure. They have a demo available.

[Gomez]

Originally Posted by lostintime

If you don't try it for yourself you'll never know for sure. They have a demo available.

You're quite right here.