Your Vinyl Transfer Workflow (sharing best needledrop practices)*

I'm glad to hear you're in good shape.

Thank you very much. You know, the YT sound quality for free users is only 128 kbps and .opus audio files. I uploaded Genesis LP wav 96/32. The original file listened through ASIO driver and ES9016 SABRE32 Ultra DAC (DNR: 124dB) is amazing.

 

They do make simple plug converters from RCA to LXR, which is the balanced one. TRS is line-level.

Something told me to buy that Scarlett. I was in Guitar Center just a day before the stay-at-home order. They had a tall stack of them at the branch I went to. Now they are on backorder too.

 

That's what exactly a DI box does. Convert an RCA/TS (tip and sleeve) unbalanced input to a balanced XLR/TRS (tip, ring and sleeve) output.Why this? To clean the signal from noise, hum, buzz and EMI, and to reduce the impedance 10x about. Now with the cleaned signal, there is gain in SNR and a decrease of THD+N. And more headroom for the USB audio interface preamp knobs. The balanced signal is weaker but more cleaner. No more noise, hum, buzz, EMI, etc. With balanced signal input in the USB audio interface you gain 10 dB of SNR. Usually, people link their preamp to the audio interface with the unbalanced RCA cables and a TS adaptor and the needledrop quality is low.

The MOTU M2 is better than FOCUSRITE Scarlett 2i2 3rd Gen. Better ADC/DAC, DSP and clock. More outputs including MIDI and the loopback thing.

 

I checked out your needle drops over on YouTube. Nice work!

Could you point (or post a link) to one of the vids where you applied this reverb to the high frequencies? Just want to hear what it sounds like.

Thank you, much appreciated.

 

Thanks. Checks the first minute of this track.

 

A/B'ed against MetalGuruMessiah's '82 remaster and "Roxy Music"'s and I can hear that you've opened up the head space listening on headphones and added extreme clarity to the opening electronic taps that have a dulled or softened timbre on the others. You do have presence! Huge improvement. You really know how to roll off 'S' sounds in vocals without making it too murky.

I have an old Audio Unit plugin (AUmatrixReverb) that comes with a Garageband install on my Mac that kinda' does this but it's just a pain to use on non-synth/electronic style music. I applied it to Chicago's stuff from their Group Portrait 4 disc CD just to get separation and clarity between vocals and brass section. Drums are so upfront (cymbal taps louder than hell) it's just hard to balance all of it.

Thanks for posting the vid sample. First time I've heard non-hit album versions of Roxy Music. Guess I'll have to get some of Ferry's albums.

 

I don't need MIDI. Check this out:



Turns out that the differences are negligible.

 

As an ex-audio engineer of 18 years I can say your workflow looks pretty good. I would upgrade the turntable. A Oracle, Rega, a Lynn or a Project. And I would stick to MM cartridges with a line contact stylus.


Sending you the money right now.......


Seriously though. Too expensive. It is a decent setup but I would upgrade. What are you using to clean your vinyl?

The biggest help I can give is in recording. You might know this fact already but it never hurts to repeat myself.

0VU does not equal 0dbfs. Recording so the signal hits 0 dbfs will not yield the best results. All ADC (especially the Pro ones) are designed to mimic analog devices. For 24 bit 0VU is - 18 dbfs RMS (not peak). And don't peak over - 6 dbfs PEAK. When recording treat - 6 dbfs PEAK like it's 0 dbfs. All analog ADC have input analog preamps and they have a sweet spot and the sweet spot is - 18 dbfs RMS (In Europe it is - 20 dbfs RMS). At this level you will get the: lowest distortion, best stereo soundstage, most detail, etc. Although some of my colleagues in the industry would disagree the normalization should come after the 16/44.1 conversion. But if you get satisfactory results why fix what ain't broken?

Little tip. At the studio we stick to sample rates that are a nice 2:1 ratio down to 44.1 khz. Such as 88.2, 176.4 and the dreaded 352.8 khz sample rate. 24/352.8 is now the studios archival PCM format. Those files Eat data like there is no tomorrow. It is not the issue it was back in 1995 but in the future I would stick to 88.2 khz. When you reduce your files down to 44.1 they will sound better.

Interesting fact: CLA takes all of the Pro Tools fiels he gets (24/96, 24/192) and knocks them down to 48 tracks of 16/44.1 on his ancient 1988 Sony 2 inch PCM 3348 DASH multitrack. Strange but true!

 

This is true if you work at 24-bit. If you capture at 16-bit, go as high as you can and do not do anything to the file after capture except maybe some gain adjustment. Just make sure your DAW does proper dithering.

I have noticed that we are getting more and more 88.2 kHz offerings on hi-rez stores.

Always good to read your posts, John!

 

I record my needledrops in native DSD128 directly from TT and then I convert them to wav 352.8/24. This is the best preserving sound quality in my experience. Then I found this format is the famous DXD. ;-)

Here I remove clicks and pops and I apply some Butterworth filters to remove the UHF and ULF noise debris from DSD conversion. The next step is converting the audio file to some 44.1 multiple (88.2 or 176.4), 24-bit. For further full scale processing, unavaible in my software for the 352.8/24 format. As you said, the 96 and 192 are bad formats for the audio file conversion. These are simple gimmick derived from the video production companies.

Then I make the in-depth de-noise of the file. Noise is everywhere. I use an unusual method, I de-noise and EQ only by phase cancellation. Better sound quality, no coloration and no FFT remains (another annoying noise!). I remove everything beyond my recording ADC capabilities (SNR 118 dB, THD+N <-105 dB). So the real thing is only a 20-bit result even if the file is in the standard format of 24-bit.

 

Is this gross overkill for a needledrop with only a 16 kHz frequency response at best, and a S/N of about 68db?

 

Indeed beware of phase issues.

Using the wrong EQ curves can also induce phase issues and other anomalies.

Many of us use a High Pass Filter on our needle drops and I believe its a necessity. However, don't brickwall it, it will cause phase issues. Use a nice gradual curve.

 

The JICO 1 SAS/b stylus grabs everything from the grooves. I still hear the music in speakers on my unprocessed needledrop at -74dB to -80dB total RMS. After I use the inflator and the LA-2A VSTs, I add some air and presence and the SNR and the Dynamic Range are boosted. Usually my processed needledrops has DR 14-16 at total RMS -18dB. Sounds very good in my speakers + sub.

 

I was thinking the same, not disparaging BrilliantBob's hard work as a waste of time though. It looks exhausting what he's described.

I'm curious whether he did a simpler method and skipped certain steps like capture at lower rez and maybe simplify noise reduction not just hiss but all the other stuff mentioned in acronyms and A/B'ed the results to see if it was worth it.

 

Is it common for people to add reverb to needledrops? I get filtering out turntable noise and doing small repairs on things, but isn't adding extra reverb (even just to boost S/N ratio or DR) defeat the purpose of trying to preserve each particular LP's sonic signature?

 

Smart. 98 / 192 khz has no business being used.
It isn't a major issue like it was back in 1992 but it still makes sense to stick with 88.2 / 176.4 or 352.8 khz for a nice even ratio down to 44.1 khz. Glad to see some members have some common sense. The higher the better mentality has unfortunately taken over the audiophile world. And many think that because 192 khz is higher than it must be better.

32/352.8 are massive huge files. All the analog multitracks I transferred in the last 3 years were done at 24/352.8. 32 bits just adds 8 more zeros and is unneccessary and a waste of hard drive space. All professional editing software (even Wave Pad Editior) handles audio files at 32 bit. So nothing will be lost if you record your vinyl at 24 bit.

There was one project we had a few years back. Some well know corporate Rock band wanted all their analog multitracks (1975 - 1988) transferred to Pro Tools. 11 albums including one double live LP. It was a massive transfer. All the songs from an album are not on two 24 track tapes. They will often be many takes of a song. Sometimes many different versions. Alternative versions etc. They wanted ALL of it. Many of their albums were recorded to 2 synchronized 24-track machines. That is 46 tracks per song. (Minus 2 tracks for sync pulse. Beep! Beep!)
And then all the quarter inch mixes. One song can have 50 mixes easy. They wanted every mix of every song. Even the mono demo mixes. That is a lot of tape! And each track is recorded at 24/352.8. Imagine 46 tracks of 24/352.8 × 200 songs! Feel nauseous yet!

Most of the albums we get to transfer comprise of 6 to 10 reels of 2 inch tape. This project was 400 reels of 2 inch tape. We usually burn the files to 2 -5 dual layer Blue-ray disks. Then it is Fedex off to the client. And of course we make a back up. Usually 100 - 250 GB at most. This project was to big for Blu-ray. We transfered all the files (almost 6 TB) to two 3TB portable hard drives and sent that off to the client. I hope they are f*!$#n happy! And we had to make a 6 TB back up. There is a reason I seem to be annyoned or irritated on here sometimes. Now you know why. That project took 6 months and it made us $350 000. Went to pay our debts. We still owe.
Oh, and we gave them our special rate. $*%!@*!!@#

So be cautious about saving everything as a 32/352.8 file. It may just come back to haunt you one day. (Sound of spooky horror music plays)


Don't hold me on this. It was either: Survior or Night Ranger or Nazareth or somebody. Some well know band like that.

 

I would agree but that would make me a hypocrite as the studio PCM archrival standard is 24/352.8. I felt silly transfering some 1968 1 inch 8 track tape at 24/352.8. And more silly transferring some 1959 half inch 3 track tape at 24/352.8. 62 db A weighted signal to noise ratio if you are lucky. Noisy as hell. Hiss away!! Ssssssssssssssssssssssss..... better to be safe than sorry....Right?

 

BrilliantBob added reverb to only the high frequencies which are highly directional (they get to your ears first and quicker). I've added reverb remastering thin sounding CD music but in order to provide separation of vocal choruses and other sounds that blend into each other. But I have to go back in with an EQ to restore the lower mid-range like drum tim-tom hits.

Sounds great on headphones but play back on my big box speaker home system I can't get too close to the speakers (sweet spot equidistant triangulation around 4 feet) because it pushes up front lead vocals and other instruments slightly back stage. The only solution is to increase volume and listen about 12 feet away. Distance and size of speakers is a major influence on reverb perception. And some songs have the reverb cooked into the song but can't be heard listening at comfortable volumes like songs from Chicago's Kath years.

 

Yes. But believe or not some fool engineer at Motown added reverb to the 2001 remasters of J5 albums. (First 4). You probably won't hear it's there but it was a pointless excercise none the less.

No. Never add reverb. That is done at the mixing stage. I would be weary about adding Eq as well. Any EQ changes should be done through SPEAKERS NOT WITH HEADPHONES. If you do you will over Eq everthing. NEVER master with headphones. Just ask Steve.

 

Interesting info on noise with needle drops and high rez. Just curious but have you found in editing hi-rez vinyl transferred audio that you're able to push noise into higher frequency ranges where it's so high pitched the noise is inaudible? And/or have you found dithering helps as well? I don't capture audio. I just work on thin sounding CD's and notice the reason why they're thin is that the hiss tape noise is so high pitched it can be heard when cranked up or it's just not as noticeable.

 

I don't have a choice. I live in an apartment so headphones I have to rely on. I do avoid most of the time adding even a bit of reverb (learned my lesson listening to the results on my home system) even though it increases the sense of head space and stereo imaging on my cans. I A/B first to the original and I've trained my ears to know what it's going to sound like on my home big box speaker system.

I did just discover stepping farther back from big speakers it improves the front stage but who listens to big speakers 12 feet away?

 

Thank you. Although I had a meltdown recently.
It was a good melt down. Mmmm...Yumm! Yes, I threaten to blow up certain studios if they didn't start mastering at 88.2 and 176.4. (Joke)

If it was 1989 then yes. But 16 ADCs are not the horrors they use to be. If recording at 16 bit I would still record at - 12 dbfs RMS and don't peak over - 3 dbfs PEAK. Even 16 bit ADCs are designed to mimick analog components. The worries of the signal falling into the lower resolution of the lower bit ranges are a thing of the past. I hope! Back in the 1980s when all the other mastering engineers were recording at - 12 dbfs RMS ( based on the 1khz tone on the tapes) into their 1610. (Oouch!!) Steve Hoffman was ignoring the rule and slamming his ADC to the wall. According to some members he was going way over. Not audible overs but over none the less. A Sony PCM 1610 / 1630 was about $5000 USD. And those DCC disks did sound a lot better than most of the CDs that were coming out at the time. So who knows.....

If you pop in any CD Before normalization you can see it hovering around - 12 dbfs. But seriously, who is using 16 bit these days?... Oh yea, CLA and his silly ancient 1988 2 inch Sony PCM 3348. He takes all his clients nice 24/96 and 24/192 files and wacks them down to 16/44.1 (Jesus Christ!) For his super-duper- awesome sounding- DASH machine. Chris doesn't jam up the signal to 0 dbfs. But he has a professional $250 000 USD DASH 3348. Mmmmm......I give up.

 

my goal when needledropping is not to alter the sound, but to make a faithful recording of the LP. So not really any post processing (though I do use a Meridian 518 to adjust the digital level and you can apply some very subtle nosie shaping if you wanna). Fairly high end system, properly set up, clean records, good ADC. that's about all is required IMO...

 

No argument there....

I am sending each member a Visa credit card loaded with $50 000 USD on it. These funds should allow you to buy a really good record playing system. Well, maybe not a real good one but a decent one. Now, no excuses! I except everyone to have a MM with a line contact stylus.

Please insert Poe's Law.

 

thanks, waiting for the $$

MC only for me though