Open source digital vinyl system timecode experiement

[oligau]

I'm registering with lathe trolls because I'm experiencing with my first dubplate pressing. I'm working with http://www.xwax.org team to create a fully open source timecode vinyl for xwax digital vinyl system that runs on linux. For non DJs, this is a system that permits to play digital content in sync with a turntable. With low-latency sound interface it really feels like the sound is pressed on the disc. Commercial solutions all use vinyl timecodes now, Rane Serato is 1khz, Native Instrument Traktor Scratch is 2khz, MixVibes is 1.3khz. I tested our custom 1khz timecode using CDDJ, Pioneer CDJ-400 and it works as good as Serato's.

I have a master ready to send to the shop and I wanted trolls opinion on
my mastering decisions before committing to my first dubplate.

I attached a picture of waveform of my 10" 10 minute 33 rpm master that I'm about to get cut on acetate.
Timecode is -6dbfs when created by xwax software. I made 5 2 minutes tracks with different gains to test which groove will work best with the DVS software. B side will be a -6dbfs 45 rpm cut to compare between 33/45.

What should I say to the dubplate shop? Don't touch my audio, no eq, no compression, simply cut it as is?

[ejemmons]

IMNSHO, that would be precisely what I'd tell 'em! Pretty easy cut, seems to me. Like a bi-phase encoder on vinyl... Best of luck!!!

[opcode66]

Posting a link to the audio would be ideal. You can't tell energy from that volume map. Volume and Frequency together give us a picture of the energy involved in the cut. This is a primary consideration for anyone with pro grade cutterhead. However, you did in fact tell us the frequencies, which seem to be around a standard 1KHz tone. That is a piece of cake to cut (even at high levels) as ejemmons already accurately stated. You run into issues with volume on sustained tones that are 4KHz and above.

The other thing I can't really tell from the zoomed out volume map that you have in the screenshot is stereo separation. I don't know if the tone (ie data) is the same on both channels or not. That is another consideration when cutting. But, again, if the tone is 1K then out of phase data isn't particularly an issue. 300Hz and below yes. But, not 1KHz or higher. Tones in the upper frequency range when out of phase don't generally cause the ramps that bass tones do. Those ramps (pinching of the groove) cause the tonearm to jump out of the groove and therefore cause a skip in the audio playback.

It seems like these manufacturers did their homework. They picked a frequency that was truly easy to cut as well as interpret digitally. Interesting.

[ejemmons]

I am willing to bet that the data is identical on both channels, save for a 90 degree phase offset (quadrature). This enables the system to tell which direction of rotation the disc is currently experiencing. I would guess that the signals are zero-cross detected, and thus speed and direction are extracted in precisely the same way they are on an optical shaft encoder. I found this out on a gig with Serrato and a dead left channel: it knew how fast, but not which way... If you went from 1kHz to say 2kHz, the system would think the disc was revolving at 66.666 RPM instead of 33.333, and the timecode would advance at double the rate.

[ejemmons]

Hmmm.... I suppose one could cut slew-rate limited SMPTE code (in mono), and get an absolute time ref, and the word geometry would tell the system which way we're going.... [square waves don't cut well - any more than they did when striping TC to tape] You could use an old Fostex synchronizer to both generate the time code AND read it AND lock up to video and/or audio tape with the record as TC master! Vitaphone ho! You could sync up the projector and the audio right from the 16" disc.... Oh Boy!

[drdub]

if its like the serato code , cutting is no problem. did this several times..
cheers mex