Latency & Time-Sync During Overdubbing

[-mjk-]

to avoid through-the-air-sound latency.

What is that?

 

[shredd]

What is that?

Hey @-mjk-
You know I’m no rocket-surgeon, and certainly no studio-wizard…so, yeah, “over-the-air latency” is surely not a real thing, nor does it really specify quite what I mean.

What I’m getting at is: how if you’re recording overdubs by listening to an existing recording on room monitors (that are typically at a distance from the performer) rather than in ‘phones, the m/seconds that it takes for the monitor’s soundwaves to reach your ears, be brain-processed, and signal your playing-fingers or voice to perform is - even though verrry small, time-wise - is enough to absolutely FLOCK your time-synchronization.

I have COUNTLESS recordings of really good originals and covers I recorded in ‘the ol’ days’ before I figured this out, and recorded in an “open room” w/o ‘phones…and while my recording skills were getting better, EVERY recording was ruined by the various parts being out of synch w each other.
You’ll note on my SoundClick page that several of my older originals have recent-dated “remasters” posted - they were the few recordings I still had separate trax for (saving projects - another thing I was too stewpyd to do when I was younger… I thought when a project was “done“, it was really “done“! ) where I fixed a lot of studio flaws (FX, mixing, eq’ing etc etc - AND these time/synch flaws!)

So - that’s what I meant…

 

[-mjk-]

how if you’re recording overdubs by listening to an existing recording on room monitors (that are typically at a distance from the performer) rather than in ‘phones, the m/seconds that it takes for the monitor’s soundwaves to reach your ears, be brain-processed, and signal your playing-fingers or voice to perform is - even though verrry small, time-wise - is enough to absolutely FLOCK your time-synchronization.

Just how far away are you talking? You can still play with a band spread across a stage though.

I listen on recording monitors that are right in front of me.

 

[shredd]

Just how far away are you talking? You can still play with a band spread across a stage though.
I listen on recording monitors that are right in front of me.

We're talkin' "normal" room size - no more than a large bedroom/small living room - and I recall the mon's being somewhere in the distance-range of 10-12 feet.
Doesn't SOUND like much...and yet EVERY recording I did in that setup suffered from enormous, un-ignorable time-synch problems that absolutely ruined them.

Though I'm not completely able to rule out the impact of being young, stewpyd, inexperienced in the studio, and a VERRRRY amateurish player at the time...

 

[mixerizer]

Sounds like the latency you describe may have more to do with how you were monitoring, while recording. The various delays imposed by Isaac Newton n pals can often stack and cause cumulative timing issues that frankly cause a track to lose its rhythmic focus and tightness.

For an easy example, in an analog tape model (which is a useful analogue to the modern digital models, next), you might just record using a mic, overdubbing to tracks.

You would play thru the air into a mic routed thru the system straight to the recorder (no digital or effects yet). When it reaches the recorder it gets on the tape via the record head. However while you were generating the music being recorded, if you are listening to a signal generated by the playback head, the physical distance between the two means the new music recorded on one head will be offset in time very slightly compared to the sound you were hearing from the other head (physical distance between heads + time it takes for all that light speed magic thru the mixer n stuff). Add tracks and that delay stacks. If you aren’t playing tighter than a crab’s posterior, you lose the groove. W tape u do stuff like using the less-hifi REPRO function to monitor off of the record head, but if ur setup is calibrated well and you don’t put monitors too far away from performers, you can play live to the speakers in the room if the band is feeling it and nothing in the room w the speakers is needed to be recorded via the mic on tape. There is some delay of course from the speed of sound in the air of the control room monitors to the musicians’ ears (1.1 millisecond per foot, slightly less in Nepal, Denver and Mexico City), your own body’s ability to respond, etc., but it’s definitely done and doable. Personally, I prefer headphone monitoring while recording my own performances.

W a lot of digital stuff the latency comes from a lot of sources but u can think of some of them similarly to distances between recording, playback, etc.

For example, when the live music hits the mic and the electrons go into the computer, that digitization takes a certain amount of time (a/d conversion latency). Then there’s the time the computer needs to think about every command it executes. Tape doesn’t really have this ‘dsp’ latency the way computers do. The value of computers are often defined in part by their ability to process as many of such commands as they can at once, as quickly as possible. Then when you are playing it back, turning that post-dsp stuff back into sound for you to hear thru headphones or monitors takes another little bunch of time (d/a latency).

Then, when you record, you stack your own accuracy with those two factors (‘time to make the sound for you to listen to + your timing accuracy to what u hear + time for the computer to digitize the overdub you are recording’) and the problem escalates quickly (for example, I found using external time-based effects via FireWire on my DM destroyed my delay timing, especially when I tried to route analog digital delays into my mix using the computer and the FireWire card on the DM. Using TDIF, ADAT, AES-EBU and analog I/o on the DM was much better for the timing of those delays than digitizing through FireWire, even though each also has its own latency and conversion delays to consider.

There are all kinds of techniques, technology, etc to combat these issues (for example, on the computer you have low-latency mode, latency compensation, spending $ for gear designed to have minimal latency, not having a 50’ recording studio room, and getting really good non-verbal communication w your fellow musicians).

Anyway, if you work to ensure your timing is good as a musician, and you center the rhythmic integrity and tightness of your recordings by simplifying and mastering your gear enough to learn and track down/eliminate the sources of your latency, you can definitely record multitrack stuff that sounds together and powerful.

 

[shredd]

OSSUM analysis/explanation, @mixerizer !

In MY case: the analog aspect isn't applicable - I used digi-corders exclusively, starting with Fostex MR-8's and moving into TASCAM 2488's. Also - other than a very brief experimentation with DAW/'puter recording (which, despite being done on a purpose-built, well-tuned, and fully-dedicated PC, failed miserably) - I NEVER used a 'puter for recording.
I also had very little in the way of outboard FX in those days to introduce latency - IIRC, the ONLY one in use was DigiTech GNX4 for git'rs. Any other FX was internal to the MTR's.

Based on your scenarios, I'd have to speculate that the horrific timing/sinc problems with my overdubbing recording sessions was a combination of/compounding of: those small A/D-D/A delays; the physical distance sound had to travel (source-to-mic, and also monitors-to-ears); the infinitesimal brain-processing of hearing a playback and the brain instructing the fingers or voice to perform; and juuuuuust maybe being a raw amateur bedroom plank-spanker with neither awareness of these problems, or the skills to avoid them or compensate for them. Hail, I barely understood how to effectively utilize a click-track back then...not to mention that by comparison, my vox woulda made Bob Dylan sound like Freddy Mercury.

EDIT - just wondering: this thread was originally about my ree-donk-yoo-luss gear scores...this segment has strayed into the recording/techniques vein...maybe our mod's can move these posts (#'s 20-26, as of this post) to the Recording category? Title it "Latency/synch'ing" or something?

 

[-mjk-]

@mixerizer great summary. Compound errors.

 

[Phil Tipping]

Interesting thread.
Fwiw this video measured a 1.3ms latency when feeding an output back through an input, but this would be swamped by mixerizer's 1ms/foot figure... notwithstanding Denver and other places

 

[shredd]

I can't say with 100% certainty this sound-traveling time issue was necessarily a factor, as the period of time I'm discussing was going on 20 years ago...a lot was different then - my gear, my musical skill, and my studio abilities (my understanding/knowledge of ANYthing to do with recording was limited to knowing what "pan" meant...).
But it's the best explanation I have for the disastrous recordings from that phase.

I recall once recording really a batch of top-quality "backing" tracks of popular-song covers I was doing live w/a terrific girl singer I knew at the time (open mic's and such). I took them to her house and they were played back over her PA (in an open room, no 'phones) and she sang to them, recorded on my ol' Fostex MR-8.
The time-lag was SO BAD that the recordings were literally unusable. Even trying to time-sync them in 'post' didn't work, because her vox were timed to what she heard in the room, and everything was off (entrances, call-and-responds, etc)...never got them to line up.
A real shame - she moved away that summer, and I never performed/worked w/her again. This girl sang like an angel with a record contract!!!

But it doesn't help either that my hearing is so damaged - even back then - that it has always affected me, both playing live and in studio...nowadays, I probably wouldn't hear it if I was recording vox and a nuclear explosion went off in my yard. I have tinnitus so bad that Pete Townsend sends me "get well soon" cards...I can't even hear myself speak.

PS: MODS - this discussion, from post #20 on, should surely be in the Recording area...

 

[mixerizer]

Well then I’ll sum it up by saying great headphone score I’m cool w this little nugget of audio quasiscience tangent here in the great gear snag thread!

I’ll add for context that one of my biggest audio jobs was time-correcting 32-track live stadium gigs, to combat the alignment issues when the rear-of-house mics are 300 feet distant / 330 milliseconds behind the mics at the front of the stage. Nudging clips 1ms back and forth and training to hear the effects of timing, phase and latency in those early days (DA-88’s, Pro Tools on Mac OS9, multiband dynamic compression and mastering on the TC M5000) became my jam, and has always been a joy for me (no idea why).

Sounds like you have passion for what you do, and that will carry you thru all kinds of adversity. Looking forward to hearing your next tracks!

 

[shredd]

Thnx @mixerizer - as the original gist of this thread was about: I'm as big a gear-hawr as any of ya - and I've probably spent enough on git's and gear in my lifetime to hire Elon to scrub my toilet.
But I just LOVE scoring super-deals on great stuff. The original post here (the Roland GP-16) was a ree-DONK-yoo-luss steal, and has been serving with distinction in my studio ever since.

Sadly, I've been sort of on hiatus/"sabbatical" for a while, due to a variety of factors, not the least of which is health. So I haven't even picked up my acoustic for more than 5-10 minutes at a time in at least a year.
Yes, it sux; yes, it bums me out, yes, I miss it horribly...I literally have nothing else I enjoy in life other than enjoying music (the playing/making/recording of which is primary!). SO - I hope that ends soon. And if it does - the results will surely show up on my SoundClick page...I useta do vids, but I now despise y/tube at the same level as instag'm and tik-schlok. If it weren't for Rick Beato and The Lexington Lab Band, I'd be petitioning Congress to ban it.