Your Vinyl Transfer Workflow (sharing best needledrop practices)*

I started capturing records on my 2011 Mac Mini straight out of the Arcam rphono through the 3'5 jack input in Audacity at 24/96. So far so good. When I open the files in RX7, they sound good, but I can't see any music data above 21500khz.
I tried to capture music in an old laptop running Ubuntu 14 with Audacity (3'5 jack input again) and I can see data above 22050khz (not much, but still).
Do you think the Mac Mini jack input is limited? What would be a good and cheap turnaround (Mac/Windows compliant)?

 

I'm not familiar with Mac's but there may be an audio setting in your operating system that needs to be set to allow for higher resolution. See if you have some soundcard settings that would allow you to do that.

 

I prefer an analog EQ applied before the signal is captured to digital. I saw a link here to a really good one that was not very expensive. I'll find it.

 

The audio input on my Mac Mini was stuck on 24/44.1. Now that I've found where to look, I know how to go to 24/96! But I must rip a lot of records again!
Probably gonna buy a USB ADC, like Behringer 202HD....

 

When I finally went 24/96 with my drops I redid just a few of them, the rarer nicer ones. So many of the 16/44 drops from earlier sound rather fine anyway.

 

To be frank, I doubt there's much musical information above 22050khz, at least on record. I looked earlier at spectral views of 24/96 Beatles remixes in RX and there's hardly anything above that point. I'm beginning to understand why Universal released the 2009 remasters at 24/44.1. Anyway, next thing on the agenda is a USB ADC.

 

The BTR magnetic tape recorders used at Abbey Road were capable of recording from 30 Hz to 16 kHz with a S/N of 65 dB. Even with modern noise reduction, 24/44.1 is more than adequate for this.

 

One more question. I know the answer is hiding in this thread but I can't find it. What sort of cable should I buy to connect my phono pre out to the ADC's XLR/jack plugs? RCA <-> 6.35 mono jack or RCA <-> XLR ?

 

These are the cables I use with my TASCAM which is similar to the Behringer:
https://www.amazon.fr/Hosa-Technology-CPR-202-Câble-Noir/dp/B000068O17/ref=pd_sbs_1

You may want to send an email to Behringer to confirm RCA-to-TS will work. Their literature says TRS, but that's for a balanced (e.g. XLR) connection; your phono pre-amp is RCA out, right? Not balanced.

 

To save space? 96 kHz files are very large...

 

I use a balanced XLR. At the moment, I am conditioning an XLR made by Audio Envy. I will A/B that cable against one I have more experience with, a Grover Huffman Empress XLR. So far I really like the sound on this Audio Envy cable, more neutral with terrific transparency as heard in soundspace.

 

DJM 900nxs has its own sound card and can USB straight into the computer correct? Then just use any audio software to record it.. Audacity should be fine. doesn't matter if the gain is high as long as you arent clipping anywhere in the chain. If it is still not loud enough then consider putting a light compressor on it just to utilize the gain boost and avoid clipping







192.168.100.1 192.168.1.1 jpg to pdf

 

I loved you workflow. One comment. Ahhh, you are normalizing to - 0.03 dbfs PEAK.
Everyone knows you should normalize to - 0.029 dbfs PEAK....Joking.... seriously though it is good that you are avoiding normalization at 0 dbfs. In fact I know a few well known pro engineers that normalize not only to 0 dbfs but frequently to 0.5 dbfs and beyond. I won't name any names. Hint: they all work in LA. and Vancouver.

The reason for not normalizing to 0 dbfs is because played back on consumer equipment the song will clip. There is a long detailed explanation for this. I will not go into it here. Survice to say it will not clip played back on Pro equipment. I would stick to - 0.1 dbfs or lower. Unfortunately - 0.01 dbfs is just to close and might as well be 0 dbfs.

We stuck to - 0.15 dbfs at the studio. For 5.1 mixes it was - 0.5 dbfs. But that is another long story that requires more math than I can comprehend today. Something about the Dolby Digital processor when converting PCM files to Dolby Digital. It is actually in the manual. I haven't done a 5.1 Dolby Digital mix for televison show/movie in 4 years.

Stick to - 0.1 dbfs.

 

John

Thanks!

I can't believe this thread is still rockin!

Probably time for me to update some of these steps to meet my current 2020 specs.

Cheers

VP

 

Wow, someone that actually knows what the real specs for 60's tape machine were. At least someone beside me knows this stuff. However, detail and frequency response are two separate things. I can show you songs from 1966 that have way more detail in them then some over compressed rubbish from 2018. Going up to 20 000 hz will not automatically mean more detail but ahhh it probably will. Modern songs will have more detail because the mixing boards are capable of more detail. Still, this is a complex issue. So I think your point stands for the most part.

Interesting fact: Up until 1970 all multitrack tape machines were pretty well stuck at 30 - 15 000 hz +1/-2 db. (Or 20 - 16 000 hz +1 / -3db). The Studer J-37 had a signal to noise ratio of 66 db 'A' weighted. How good is this? It is the same spec as a 2 inc 16 from 1975. Damn good. The later used 3M M23 1 inch 8 track is another matter.
Some may say, "Just a conservative spec." Nope those machines back then had a big problem with the last half octave regardless of tape width. Proof: The Ampex test tapes up until 1970 were 30 - 15 000 hz.

As for noise, Dolby A came out in May of 1966. But in 1969 the single channel 361 unit was $1500. That is $10 000 in today's money. I don't know if Abbey Road was using Dolby 'A' by Abbey Road but I doubt it. They were by DSOTM in 1973. Abbey Road was always running behind in technology. They always had the best stuff but it was always out of date. And this was the case even up to 1986. Just ask the music composer for Aliens. I doubt whether Abbey Road studios would cough up the $12 000 USD (£8 000) for 8, 361 Dolby A units back in 1969. But maybe some super smart member can tell me otherwise.

Most 70's 24 track ** tape machines had a signal to noise ratio of 64 db A weighted. By using Dolby A (which they needed real bad!) could knock it up to 74 db - 79 db. Still many an audio engineer hated any kind of NR and refused to use it. This why some of the songs from 1969 - 1984 are dead quiet and some appear to be really hissy. Ever wonder why some 1 inch 8 track from 1970 was quieter than a 2 inch 16 track recording from 1976? Now you know.

Actually the Beatle tapes were transferred at 24/192 not 24/44.1. After they transferred them to Pro Tools (just a digital recorder guys!) they did some digital EQ, clean up, editing, etc. These were then D/A to add some analog tube limiter and whatnot. They were finally fed back into Pro Tools at 24/44.1 Why? I have NO idea... Makes no sense....Why go to all that trouble of transfering them at 24/192 and then knocking these songs down to 24/44.1?


For the last 23 years 24/96 has been the mastering standard. I don't think anyone here believes that the Flack 24/44.1 USB Flash drive is the sound equivalent of the original 24/192 files.




** exceptions:

The Stephens 2 inch 40 track running at 30 ips. (1973)

The famous Studer A800-24 Mark 1 (1973)

The Ampex ATR124 (1979)

 

If I needledrop a record I don't care that much about, i'll just do it in 32/44.1. and dither it to 16-bit. If I really don't give a crap, i'll do it in 16-bit, clean it up in 16-bit, and move on. But, I do all 45s in 32-bit/44.1.

Very important (to me) records get done at 32-bit/88.2 or 96kHz.

Yes, I know, 32-bit is good for exceeding full scale zero without clipping because of it's theoretical dynamic range, but I like to work without having to worry about that until it comes time to finalize the work. It's just 24-bit with an 8-bit mantissa. And, all of my audio programs process internally at 32- or 64-bit.

 

If all you're doing is Beatles records, yeah. But, with some select recordings, with the right cart and preamp, you could possibly exceed 22,000 cycles. But, in most cases, you're not going to see much of anything past 16,000 cycles on a good day.

 

I sometimes avoid using a limiter or band compressor in favor of other methods to get a loud file. But, it is time-consuming and requires a lot of concentrated listening. I reduce loud peaks by isolating them in Audition and lowering the gain to a certain decibel. Then, I play the passage back to see if I can hear the reduction and if it adversely affects the dynamics, and I do like dynamic transients.

Anyway, once I get all of my peaks down, I then use unity gain to bring up the file so it's louder without using any compression. I think it's well worth the effort.

Or, what i'll do in some cases is use a band compressor. I'll do extensive listening to ensure that the sound isn't adversely affected. I do this if I want a slightly fatter, fuller sound.

And, of course, if I just don't care all that much about the project, i'll use a limiter for two or three db.

 

So with what John and Grant are saying can someone please explain why when I record needle drops from most any time period the highs generally go beyond 48K. Some you can see what looks like the engineer used a brickwall filter at 20k and many others show what looks like a natural decay being clipped at 48k.

I had always thought symbols and other percussion instruments decaying. I may have even heard someone post that this is distortion, which never made sense to me. After read Johns post a light bulb when in my tiny little brain. These high frequencies must be generated by the phono cartridge some sort of reaction high frequency. It just seemed odd to me because on many well recorded albums with a lot of detail you can clearly see patterns in RX well above 20K.

Mind you I know all to well I can't much above 12K and now my tinnitus makes it even worse.

 

Yeah.

 

I capture at 192 and there usually appears to be genuine information just above 50k. Don’t know how precise the spectral is on RX5.

 

This has been my assumption but if John is correct and the recording devices of the time could only record to 15k it not possible to have audio above 20k.
I first questioned this some time ago when this was discussed in a thread Dire Straits Brother's in Arms was recorded and mastered in 44.1 I have a 180g reissue and it appears to have sound above 48k.

 

If anyone on Windows 10 is having trouble recording anything and is getting the “MME Internal Error”. I found the solution.

1. Windows Key
2. Search: Settings
3. Go to privacy
4. Go to Microphone
   
5. Allow your program to use your mic.

I don’t know why this stopped working, but if this fix helps you like it helped me, im glad to share it.

 

But the point of using high sampling has nothing to do with more high frequencies.