WHY You Should Calculate and Set a Record Offset in PreSonus Studio One
HTML-код
- Опубликовано: 12 сен 2024
- In this video, I demonstrate how to calculate (and set up) a Record Offset for your audio interface in PreSonus Studio One. In addition, I also talk about WHY it's important.
Blog Article - marcus-huysken...
Facebook - / marcushuyskensmusic
Twitter - / marcushuyskens
IG - / mhuyskens
#PreSonus #StudioOne #Tutorials
@presonus
Bro! YOU SAVED MY LIFE! I always heard about the latency issues, but never had this problem of "anticipated record" until now with my new audio interface... I was on the verge of going crazy, I even questioned my ability to play ... I never conceived that it could be a technical problem. Until I gave myself the task of searching thoroughly, most of them talk about latency, which is perfectly normal, thanks that I found your video, I have an anticipation of 47 samples! Already corrected and everything is great, although it seems too much to me, but thanks, the important thing is that I can play well! THANKS!!!
So impressive and helpful, as always! PreSonus, please hire Marcus to make more of this type of in-depth, valuable content!
He already did a bunch of training videos on Studio One for Groove3. Check them out if you like his style. www.groove3.com/browse/by-author/marcus-huyskens
A lot more useful than some of the people PreSonus have hired for that.
ruclips.net/video/lOI1Thjc1Ts/видео.html tho
Generally latency like that is because output audio is buffered and processes a little more than input audio in device drivers. It also depends on the bus the audio interface is connected to and the chipset running that bus (thunderbolt vs. 1394/Firewire vs USB). Additionally some input drivers will attempt to correct for signal offsets as they process the buffers being sent.
So this is a awesome way to demonstrate that issue and how to address it! Thank you!
Thanks for your videos, Mark! You're my nº 1 Studio One mentor on that side of the Atlantic, alongside Gregor in Europe!
Hi Marcus,
1 point, you recorded the metronome before you removed your 2 smple offset that was already in place . Therefore when you reset it to 0 samples afterwards and then recorded the second time, surely that will have influenced the outcome, by 2 samples would it not?
And if 2 samples takes it forward, then removing them would take it back, so when it came up as - 2 samples, surely it would have been 0 samples if the offset hadn't been in place when you recorded the metronome the first time if the offset had been 0 to begin with!
Excellent. And for a non english speaker like me, If you appear in the video is very helpful. Greetings from Argentina
Fantastic. I am so glad I found your videos.
Dude that’s brilliant! Thanks! Been awhile since I recorded, but jumping back in I notice there is so much garbage printing happening. I’m dead accurate with timing, but have been pulling my hair out cause I can’t hit the beat, or rather the system won’t print accurate. Used PT prior, but decided I’d give Logic a go… 🙄 They all have their + and - but recording analog SUCKS! Just downloaded the S1 demo, and it was 1 billion times better. Sub’d and it’s downloading like 40g. I’m doing this FIRST THING tomorrow!
I’m left with more questions than when I started.
Hey, man - I enjoy your channel, what questions do you have, maybe I can help.
Cheers, M 🍻
Me too my friend.
Your tutorials are awesome!
I will do this with my Scarlett 6i6 1st gen. Great video Marcus!!!
very nice video 👍 Marcus
Another great video! I did all this with my Apollo Twin and the result was... everything aligned perfectly. Is that odd for it to require no adjustment? Or was that expensive Apollo worth it? ;)
Thanks Marcus... Great tip
Thank you. This is genius.
Hey Marcus i have a question....
I just bought the new PreSonus Quantum HD8.
Is it still necessary to do this loopback test procedure on this interface since it already has low latency drivers??
If people want to compare. My very old FW-1884 is 0 to 0.2 samples accurate (eg 0 offset) in 48 KHz/24 bits/256 samples (on the interface), 48 KHz/64f(loat)/256 samples, 48 KHz/64f/32 samples.
At 96 KHz, 64 samples is the minimum for it, 96 KHz/64f/64: 8 samples offset. 96 KHz/24 bits/64: 8 samples.
First off, I love your videos and am a subscriber. But I'm confused by something here. A few things actually. When I measure my offset with a loopback and then adjust for it, it doesn't change the Pipeline ping. So it seems that pipeline is unaffected by the record offset which could screw things up, right?
Or does it not matter because pipeline as an insert is by definition only dealt with recorded material.
But the bigger question is that I was hoping you would address the "why" here but you kinda only addressed the "how." Why does a 2 sample offset matter? Mine was either 28 samples or 43 samples depending on the device, but both less than 1ms.
So I could definitely set my offset to -28, but what impact will this have? I'm always worried about changing things and causing a downstream problem and then not remembering that I changed some esoteric setting like this.
Thanks again!
In my setup, rendering the click in Studio One places the audio 7 samples late on the timeline. Any ideas? Thanks.
Hey Marcus, will the offset also apply to tracking vocals?, or is that a different measurement having to consider the input latency?
I'm not 100% sure about this, but I believe this applies to any audio signal. The cool thing is you can test it: record yourself singing something and then re-record it, testing it the same way he describes in the video. If you've already adjusted it accurately, then the vocals should also be spot on. Hope this helps!
when i follow the vedio and do the same settings it works! But somehow when I change the setting like Z bottom it still give 2 sample of latency. What happens? And I want to know why I can’t record same shape of click it was?
Does this sample change happen differently on different buffer size? Do you get same 2 sample change with every buffer size option and different sample rate?
I would guess you can leave your buffer at a high setting like 512 if you get the timing right it doesn't matter.
I set my buffer to 512 and aligned both audio and midi and set it as a template.
Thanks, I'll also try this 👍
@@RahulSimaro I also didn’t use this method I used Gregor method.
Using Pipeline XT makes it easy to measure. Quick Tip.
Pipeline adds latency so it will mess up your numbers.
Hi Marcus,
Having originally used Gregors way of doing this I had a -27 sample offset. Using yours, I have a 0 sample offset!
Would this be because all I did was loop in and out of my audio iinterface an RME HDSPe AES 32 PCIe, which doesn't included my mic-pree, which I can't loop through any way?
Looking at the difference, I can only think that last time, I managed to do this including the interface, which isn't impossible to setup!
Or would the sample amount be any different depending on whether you were recording at a different rate, as I now always record at 96K 32F, rather than 44.1 or 48K unless I have too?
Using hardware monitoring Studio One gives me a round trip of ~0 which as standard would be 8.73ms/838 samples at 96K 32F which is eqivilent ofmy low level latencey.
When I did your example, I ended up routing out through a sub and back in, which I felt was werd as I had it routed as you did but the signal wouldn't return as a direct loop, so I must have messed up somewhere on that.
So shoud I be including my mic-pre as this will affect the sample offset, though all incoming externally will take the same amount of time, but bouncing etc will take far less, like 0 samples which is what the result was! It so long since I first did it, that I can't remeber how I came to -27 samples. Thanks in advance.
After you made the changes watching Gregor's video, did you test it again to see if it was accurate?
@@realto619 Hi, as this is going back some two years, I can't rightly recall. Now though with my new system I have a round latency of less than 1ms. So I'm okay. Thanks though for asking, a good point to remember.
Realizing Studio One 4 only allows offset by whole ms increments....
Yeah, I noticed the same thing and mine was off by 2.5ms so I could only compensate for it so much. But I'm thinking the 0.5 is imperceptible. Except for my dog. He still looks at me funny when I play back audio recordings....😁😜
Will this offset be the same for all Sample Rates and Buffers?
For me yes, I got 256 samples on every samplerate
This only applies if you plan on recording something that you are playing out of your DAW correct? Like for using external processing gear like external EQs or compressors. Right? Is setting this going to screw up recording realtime audio like vocals?
also doesnt this value change depending on your buffer setting?
You printed the Click Track to audio with the 2-sample record offset in place. Then deleted it. Then found a 2-sample pre-delay.
Is that just coincidence, or was that a part of your demonstration, or am I missing something?
@marcushuyskensmusic don't you have to do this for midi timing as well?
Yes. Gregor’s video covers that.
@@WetSoilStudioLymm Gregor's video covered this offset issue too
Yep - using the sine wave sample looks easier too
interesting...but is the sample delay in this exsample still 2 if you have recorded 10 tracks ?
Short answer, number of tracks shouldn't matter. Doesn't stack like plugin-latency. You'd notice an increased offset if unadjusted and re-recorded back through multiple times, but setting it correctly means you shouldn't ever have a discrepancy to begin with. Essentially, you set this once and can forget about it more or less. 10 or 100 separately recorded tracks should be affected all the same as the single click track exemplified here.
Correct me if I'm wrong, but isn't this adding your TOTAL audio roundtrip latency to the recording print. For example, let's say you, in this video, had a total roundtrip latency of 200 samples (100 in, 100 out). Thus, you do this loop to calculate that and then you would set your record offset to -200. Cool. The problem is (if it even is a problem...please correct me) when you grab an actual instrument/mic to play and record, it only takes 100 SAMPLES for your instrument/mic to get processed and recorded in time. But because we set a record offset of -200, it will ultimately place your live recordings 100 samples ahead of the time you intended to record.
...So Is this how it works? Shouldn't the record offset be based on input latency only?
That offset in the Options menu is related to Delay Compensation. When you record anything, Studio One automatically compensates for the input delay, effectively pulling back the audio in the exact amount that the audio driver tells the delay is. But sometimes that value is incorrect (in the driver), and the compensation is not so accurate. We then measure the amount of the offset it would need to be really perfectly in sync, and then set it in the Options menu. Studio One would then pull the audio back and subtract that offset value from the delay reported by the driver.
Yes.
I am so frustrated! I'v done the workflow Marcus describes in my set up, but nothing happens!? I monitor trough my soundcard and hardware, so rec. latency is not an issue. Therefore, I can rec. in 4096/96 to reduce stress on my 2011 full specked iMac. However when I do, recorded sound is printed 14 samples ahead of time using the loop back workflow! Can somebody please explain how sound can travle ahead in time? Also; when I offset my preferences by 14 samples, nothing changes. Please; I don't want to go back (in time) to Logic. My soundcard is a Zoom TAC2 Thunderbolt.
Fixed it! However; why a 4096/96 setup put's audio ahead in tine, when a 32/96 setting do as expected and places audio later in time puzzles me.
Is this the Same Problem in Protools too ?
This is a universal DAW issue and can be corrected the same way in those DAWs.
midi are recorded earlier almost 40 sample but i set offset 100 sample nothing happened . Lol
Meanwhile me offset sample -9700 to get it perfectly synced.
Sorry, but this is HOW to set record offset, not WHY.
because that's all the time he had available for today
@@redbull741852 ok, so see you tomorrow 🤔
I too was waiting for the actual "why" explained.
The why is pretty obvious... Because it's more accurate...
@@ianjohnson3546 right, but is it detectable in a final mix?