5 Tricks to Reduce or Eliminate LATENCY when recording
Vložit
- čas přidán 12. 06. 2024
- Recording delay will cause a poor performance, and timing issues. If you are having problems with latency, or recording lag, this video is a must watch. We go in detail about the various causes of latency, and several solutions that can work in any recording setup. This is important to understand for recording studio's, and home recording setups that want professional quality audio recordings. These strategies for reducing latency work with Focusrite, Audient, SSL, Motu, Behringer or any audio recording interface and any daw.
It can be frustrating trying to record a video and find that the audio and video are lagging behind each other. In this video, I'm going to show you a simple way to solve the recording lag problem, no matter what your recording setup is. By following these steps, you'll be able to record videos with no delay or lag!
In this video, we'll show you how to solve recording lag (aka recording delay) with any recording setup. By following this simple guide, you'll be able to fix any recording issue, no matter how big!
By understanding recording lag (aka recording delay), you'll be able to fix any issue before it becomes a problem. Watch this video to learn how to solve recording lag with any recording setup, and be prepared for any recordings you make in the future! - Jak na to + styl
GREAT VIDEO, but the background music is really annoying.
definitely one of the clearest and most concise videos on the subject I've seen yet. nicely done!
Great video! Really appreciate how clearly it's laid out, as well as taking the time to spell out the acronyms. The learning curve is made infinitely more frustrating when everything is explained too quickly and only using acronyms.
Glad it was helpful!
Thanks for sharing. I only wish the background music was much lower. On my laptop it's almost drowning out your voice
I am a beginner (used a portastudio back in the day, lol). I literally quit trying to record anything because of latency. Thanks for your awesome explanation! I think I'll get an inexpensive analog mixer and try that route. I subscribed . . . again, thanks!
Thank you. You have explained it very well and clearly.
This has always been the way, one reason why I love Grace pres. good vid. Interface like Hilo or others with a matrix are valuable.
Very informative video! Thank you!
Excellent....thank you
Thanks for the video. Very helpful 👍
Glad it was helpful!
I've had dreadful tales of woe with laptops - until I found what was wrong. Many of them not only give varying latency with changing temperature they also just throw away information when the music gets to a tacit. The musicians get the blame until, as I say, you figure out what the computer has done. And here's a handy tip once you are working with a halfway decent computer: run a sharp click off one track and pick it up with a microphone, now make a three or four second WAV with the two clicks, you can use that as a little ruler to visually slide all your overdubs back that few milliseconds. Works great, I really recommend it. Thanks for covering the latency issue on youtube, you've cleared up a hell of a lot of points.
I’m glad it helped, and thank you for the kind comment :)
great information.
This is the best explanation I have seen on CZcams in terms of correctness and understandability. And thanks for not being silly or over the top as some CZcamsrs are, often to camouflage their lack of knowledge. The music is a bit distracting , especially for people from the U.S., Liberia, or Myanmar.
Lol… yeah… the music is a bit too loud. I mixed it in using laptop speakers…. Won’t do that again.
Thanks for your comment:)
@@Get.Beautiful.RecordingsI really like your videos and take a lot of information out of them, but I can't see the point of any music background when you're talking. As a musician I always listen intensively to music, it's not elevator sound or music in a shopping mall to put me in a mood to buy stuff.... particularly in a long 19min. long video i have to stop the video after 4-5 minutes. Just because of the music background. Sorry. I would prefer a video with only music when you show a test recording about the topic..
you have geiven me a deeper understanding of latency thank you
You’re welcome. Thanks for commenting:)
reluctantly subscribing cause you blew my ears out with your obnoxiously loud white noise
thanks a lot man
Thanks
great video! i think there lets say in the analog monitoring recording ... agree there is no combing effect at all. But.. there is still the Analog to Digital conversion delay... is there a way to automatically sync it that you know of?
Your videos are more helpful than a lot of garbages around there. Thank you.
Glad you like them!
thank you! :D
Hi Chris, what solutions in 2024 is there for using my lap top with Audio interface for live performance out of the box for 6 mics ? I can’t get rid of the feedback being reprocessed by MacBook Pro. Thank you for your valuable videos and responds.
For the most part, this video is very helpful. However, I do have a bit of an issue with the Thunderbolt vs. USB discussion towards the end. Thunderbolt is in fact faster than USB, and in a way that impacts latency, not just the number of channels you can run simultaneously. Thunderbolt is basically an external PCIe connection, and it has direct access to your computer's memory. It's not just a little more efficient in how it routes data to memory, it gives your audio interface the ability to directly read from and write to the memory inside your computer, without having to go through the whole USB subsystem, etc. USB 2 is also a half-duplex connection, whereas Thunderbolt is full-duplex. What that means is that a USB connection has to switch back and forth in 1ms intervals, taking turns sending and receiving data, while a Thunderbolt connection can send and receive at the same time continuously. The result of this is that for companies that are still pursuing Thunderbolt in their interfaces, latency is continuing to drop. The NTP DAD AX Center, for example, has a round-trip latency of 0.67ms over Thunderbolt 3 at 96 kHz with a 32 sample buffer. The very best USB interfaces and drivers have hit a wall at about 2.1 ms RTL at 96 kHz, 32 sample buffer. Due to the half-duplex nature of USB-2, it's basically impossible to achieve sub-2ms latency over that protocol, whereas latency on Thunderbolt interfaces keeps improving well below that threshold.
One area where the total bandwidth could be a real consideration is this: 32-bit float is starting to be available on more and more interfaces. I have an interface from 2017 that works in 32-bit float natively, and there are some good reasons to want to use it. It gives you enough dynamic range that you can recover from bad gain settings after the fact. However, with 32-bit float, each sample takes up 33% more bits than 24-bit. When you stack that up in a multi-channel setup, you start to run into bandwidth limitations. A 32 channel system (not all that uncommon in studios, and way too small for many touring setups) would completely saturate a USB 2.0 connection in 32-bit float (even though the link speed would suggest otherwise, the whole half-duplex thing comes into play here, too).
RME has done a really good job of selling people on USB, due to the fact that they've written some solid drivers, and they make hardware that people love. They also have a great track record of providing long-terms support for their products. Unfortunately, they are also perpetuating technological views that are not based entirely in fact, and I think that's doing a disservice to the industry. There are really only two substantive reasons why I've seen some interface manufacturers moving away from Thunderbolt: ongoing chip shortages, and not wanting to have to pay software engineers to maintain both Thunderbolt and USB drivers when cheaper USB-only interfaces sell much better. However, Thunderbolt is persisting in the high-end of the interface market, because it has concrete advantages.
what if you are using a USB 2 interface ( antelope orion 32) and have the option to use a MADI PCIe card with it? will the MADI card reduce the latency of a single signal ? or is it only good for providing more throughput ?
Thank you very much for sharing your hard earned lab results, and I don't mean just this video. This is what guys and girls need to cut through the bs. I commend you
Thank you so much !
Thanks for the info. So if I understand this correctly, for guitarists who want to hear the effects they're playing with, there's no way to eliminate audible latency without connecting a mixer or a reamp box for analogue monitoring? I'm still not sure how a reamp box would help with latency, as I thought the purpose of a reamp box was to convert the line level signal from the interface into a Hi-Z/instrument-level signal that an instrument-level input, such as a guitar amplifier, expects. How does that reduce latency? PS do you have a video showing how to freeze tracks in a DAW?
what a legend!
I have an honest question. I am aware that Min may say it subjective but around 12 minutes and 17 seconds you mentioned using the higher-quality microphone for tracking in the lesser quality microphone for monitoring in anyone opinion which would be the higher quality between an audio technical AT4040 on GTZ103 clone?
thank you again
Your welcome :)
I was really pleased to see my MacBook Pro only had 2.2 ms latency from pro tools with 128 sample rate. I had to stop using my PC because my graphics card was causing noise in the signal chain for some reason.
Great video, really helpful however I would like to point out that to my knowledge the BIOS stands for Basic Input Output System!
This video was one of the best videos i have seen regarding this issue. Thumbs up!
I am using a behringer interface (umc202HD) myself with direct (analog) monitoring exclusively when i`m recording my gitar track in a project that has a finished midi drum track that i hear from the daw when i record.
It works fine, but my question is this: I hear myself playing the guitar instantaneously through my headphones, but how can i measure how long does it take to actually record the guitar signal in the daw, and how to correct the offset?
I assume there has to be a slight recording delay between the drum track that i play back from the daw and the track that i record with the interface wether i use direct monitoring or not. I don`t use any plugins on the target recording track. I just record the dry guitar singal through the interface.
Can you help me with this please? That would be very much appreciated!
If there’s any offset, it’s less than 1 millisecond. It’s a good question, and I don’t know the answer. The daw will compensate for any latency caused by the drivers and buffer, but I’m not sure if it compensates for the latency caused by the a/d converters, which will be close to nothing, and it’s not something I’ve ever even remotely considered to be an issue.
@@Get.Beautiful.Recordings Thanks for your reply! Very much appreciated. I did a little experiment yesterday with my interface. I generated a click track with audacity. I imported it into my daw software. Then I begun to play it back and routed it into one of my inputs and recorded it. I then analyzed the waveforms of the audio files in audacity again. It showed that the recorded audio file is in 1.09ms of delay compared to the original. It means that as you said the daw actually corrects the latencys of the drivers, buffers etc. but has zero knowledge how long does it take to convert the input-output signal (AD/DA conversion). Now I know thanks to your video! Thanks! Subbed. :D
Wouldn't it be enough to align phase for that vocalist issue?
3 KM of cables!! LOLLLL - I have a quick question, what does a computer needs to handle reduced samples and avoid glitches and dropouts? More RAM or more processing power? or both?? Thanks and great vid BTW ;)
Chris, which mic are you using and whats your setup, you sound so good !
I’m using a Telefunken tf-11 fet.
Thanks for the kind comment :)
very good information but you really should have turned down the music in the back ground - we want to listen to you not some back ground music :)
Yea… I agree, I made it a bit too loud. Thanks though :)
Great Vid! SUPER helpful! THANK YOU!
BTW, BIOS = Basic Input Output System and "sits" below the OS (Windows in your example) not a contributor to this...
Oh darn, this is a revelation for me. I new something was not right but I could not figure out what it was, my DAW latency is very low. Me and the lead signer when recording our band live (all analog no monitoring), had almost perfect pitch and that even whit a crazy loud drummer (can't complane he is one of the best). But when recording in our little home studio it was a complete different ball game, we were always a bit off tune no mater what. The worst is that, I have all the equipement for monitoring direct.
I will change my set up and try it in the days ahead.
Thanks a millions Chris.
PS Sorry for my bizard accent ;-)
I didn't even hear the accent :) I'm glad the video was helpful.
I work with many singers who like using auto-tune during tracking, even for subtle pitch adjustments. While I'd pick the Apollo x8 again, the Aurora N converters do sound better. I wish the Lynx system had DSP like auto-tune. Personally, I'd prefer direct monitoring for lower latency, but I stick with DAW monitoring for efx.
Yeah… auto tune is definitely a challenge, I think the Apollo interfaces are one of the only solutions for this
Absolutely! I found your video quite enjoyable. You raised some excellent points!@@Get.Beautiful.Recordings
Two mics for a single vocal track. One goes into my old analog Mackie mixer, which has a send and return from an old Quadraverb. My DAW has an optional output that also feeds this mixer. The mic I use for actual recording into my DAW isn’t monitored during tracking. It’s a great, no latency, with fx, solution!! So I do this with a small condenser mic attached to my SM7. For monitoring, you also use a 57 or 58, instead of the condenser.
I just use and inline mic mixer shouldn't be too hard to implement this for anyone.
12:50 for electric guitar & bass I always go through my DI boxes so I can record amp and Direct signal concurrently. is that the same as a mic splitter and will cause signal deterioration?
Yes. The deterioration is slight. If the di box is passive, then you can’t get something for nothing, so it splits it into 2 signals that are weaker. There’s also a slight loss of fidelity just from passing through the transformer.
If the di box is active, then it routes the signal through an additional gain stage, and gain stages always reduce fidelity just a little bit.
@@Get.Beautiful.Recordings thanks for the quick answer! iThey are 2 active BSS Audio AR133, phantom powered through my interface - so it's good to know if I want really high quality I should use the Hi-Z on my interface (Steinberg UR 824 + Behringer ADA8200, soort of "prosumer" grade equipment).
Although at the Moment I am quite happy with the quality, there is always a tradeoff of convenience vs. ease of use... Come to think of that I don't think my interface has a Hi-Z... so I might have to use a different solution. I feel like going through my cheapo Mackie ProFX12 Analog mixer may even be worse as regards losing quality, just in the Analog domain... I will try some A/B comparisons and see what's better. One thing that bugs me about recording in Windows (Cubase Pro) that is hard to mix and match interfaces, it would be so much nicer if we could just add a dedicated guitar interface on top of the 16 channel "workhorse" that I use daily. Or maybe I can get an ADAT extension with Hi-Z ins?
If you are faced with this 2 scenerio which would you prefer or go for. Option 1 is tracking with an Apollo twin x using universal audio prepamps and compressors for traking in or Option 2 Using an Audient ID44 mkii with real analog compressors as inserts for tracking. Which would you go for if you are tracking vocals, saxophone, bass and guitar and mixing 100% in the box without UAD plugins
In this scenario, I would suggest the Apollo Twin.
If you're on Linux in particular, you should pay attention to number of periods per buffer, the default is very high and changing it is not very obvious in DAWs like Bitwig, especially when using Pipewire/Wireplumber. The default is quite high iirc which causes high latency, the rule of thumb is to set it to 2 or 3, maybe 4, iirc 3 works best for most people, me included. This is a kernel setting for handling block devices, configurable through Alsa (or indrectly through e.g. Wireplumber or when using Jack with qjackctl), it also exists on Windows but there it is set to 2 by default and doesn't cause any issues there.
And yes, your MOBO's DAC plays a part here too, if you want to monitor through DAW, you can always buy a separate DAC that connects to the PCIe port, but for most MOBOs it's not necessary to get acceptable level latency for live performing even with digital effects like EQ, gate, compression and chorus if your CPU can handle it, especially if your interface allows for direct monitoring which is always preferable.
A nice feature of some interfaces (like Behringer UMC404HD) is being able to hook up analog effects after the preamp and still be able to directly monitor the signal, so if you're performing live or don't mind having some effects in your signal before it hits the DAW, you can directly monitor yourself with stuff like compression and reverb with no delay, really useful feature if you don't mind buying actual analog effects for your mic.
I have volt 1 interface i never had lateency delay untill now idk why all i settings are right
Thanks for this information. Let's say that I want to record 2-part harmony. I record the drums, guitar, and first voice, then obviously need to hear this recorded material in order to be in sync while recording the harmony. At first I was worried that using direct monitoring would prevent me from hearing the already-recorded tracks. But I think I understand now -- direct monitoring allows me to hear the harmony that I'm currently singing with no latency, while at the same time hearing what has already been recorded. I may want to mute the track that I'm recording to, so that I don't hear myself with a latency-caused echo.
You are exactly correct.
You will hear everything that was already recorded. You need to mute the track you are currently recording so that you don’t hear it doubled and with a bit of latency.
Thank you, @@Get.Beautiful.Recordings. Very helpful videos and feedback.
@@design1569 There's even a special expression for that: "Mix-minus" or N-1.
What is the best way to send from an outboard pre into your interface? Is ideal to go into an XLR in? I find I’m often going from XLR to 1/4” and that probably means it’s unbalanced
At that point, right?
No, 1/4” can be balanced or unbalanced, depending on the cable. If the preamp output is xlr (most are) then converting from xlr to 1/4” is the correct setup, then plug into the 1/4” line input of the interface.
You can verify if your 1/4” connector is balanced by looking at it. If there are 3 contact points (called tip, ring, sleeve) then it will retain the balanced signal. If there are only 2 contact points, (tip and sleeve) then it is unbalanced.
Latency is affected by protocol packet rate. I don't know the time between each packet compared to Thunderbolt vs USB. Probably something that doesn't really affect anything but is technically there.
Impressingly good instructional video, - one of the best I have watched. Two suggestions; Skip the background music and allow Yourself "time to breathe" in the narration - the pace is too high and gets stressful. You could perhaps use some simple graphics to accompany the "talking head" sequenses so as not to make it too monotonous. Anyways; Top Notch content - I'm a subscriber now! Best regards
Awesome suggestions, I will take them into account with my next video.
Have a great day :)
@@Get.Beautiful.Recordings Great video, but I agree. Either kill the background music, or dial it way down, it's just distracting and makes it hard to concentrate on what you're saying. And yeah, no reason for the no-pause, machine-gun delivery, that also makes it harder to understand. Thanks for the info though! I'm having major latency issues, using an SSL12 & Mixbus32C, with a superfast brand new Dell laptop (i9 processor, 64GB Ram, SSD drive, etc) , so I hope this helps me solve my problem
Had no problems
The most efficient way I've found to capture a guitar D.I. signal is to have the guitarist plug into a D.I. box and have an XLR cable go to the interface, and the connect a TS cable to the throughput and send that to their rig.
Also long hi-Z cables don't cause latency, but they do cause signal degradation. This shows up as not as a delayed signal, but a lower volume and less clarity in the higher/lower frequencies of said signal.
Yes… this is true, however, there will be slight signal degradation from splitting the guitar signal with a DI box.
@@Get.Beautiful.Recordings If the cost of getting a great performance is a "slight degradation" of the signal, I'll make that trade 10 times of 10. The end goal is to get the best performance at the highest quality possible. Unfortunately those scales can seldom be balanced perfectly.
Very good video! One correction though: BIOS does have nothing at all to do with PC performance since at least Windows XP. Things that have an influence: processor power and RAM (more plugins, more CPU power needed), energy profile and (if you're not on Linux but on Windows) using ASIO low-latency drivers.
Great info. Thanks :)
Hi
Can you suggest me good audio interface for EDM and Trance equipped home studio..? Btw 300 to 600 dollars the best one.
I have a recent video “best and worst audio interfaces of 2023”, where I explain the pro’s and con’s of several interfaces. Generally, my top recommendation is the SSL 2, however, it depends on your unique setup and preferences.
I just narrate audio books. Does latency matter for me? I have latency of 23 ms. I haven't done anything to try to lower it.
No, the latency doesn’t matter in your situation, because you don’t need to hear yourself in headphones, or line up your recording in time with a backing track. You can just speak into the microphone like you’re speaking to another person in the room. I hope it goes well for you :)
I directly monitored my guitar and noticed that it gets recorded with a delay. Does that mean all audio recordings have to always be edited afterward to get them in time with the rest of the music?
It’s always a good idea to listen to the relative timing of each track. Sliding the track a little is pretty easy.
There is one more thing... The distance of the microphone to the sound source. And...Singers or horn players hear themselves through the head. If they use in ear monitoring or headphones, they'll hear two signals. One without any latency through their head and another through the phones. When these two overlap, even the smallest latency will change the sound they will hear because of comb filtering. That is something only the artist can hear. And that is why he or she should always have the option to EQ themselves in the phones for example with an iPad App or so.
Good points :)
Does recording at higher sample rates increase file size considerably?
Yes. Twice the sample rate will have twice the file size.
Which Audio interface are you using in your studio
I use a Metric Halo ULN-8 and Lio-8. I’ve chosen these because they are capable of high channel count, and have excellent routing versatility.
@@Get.Beautiful.Recordings is this not an old interface. Are the converters better than the Apollo twin x
It seems counterintuitive that higher sample rate would lead to reduced latency.
Yes, a higher number is lower latency. Think of it like driving a car, a higher number of the speed you are going will reduce the time it takes to get to your destination.
Read about niquist frequency to understand this
Man I watching this entire video. I'm the only one that can still not get my latancy fixed
I send the headphone output to my boss gt1 and the reverb there! Problem solved😮😮😮😮
The information presented is really good and seems very accurate, so i hate to be this guy... but i died a little inside when you said to snip a wire in a balanced mic cable bc you didn't want stereo/unbalanced signal from the headphone output. Maybe it might be helpful if you gave more explanation, but it really looks like you proceed to plug a balanced mic into a balanced input with the cable you just snipped one of the balanced lines on. If this is the case, you will get half of the signal strength from the mic, and you just killed the nose rejection ability of the balanced system. 🤔 What gives?
…analog??
headphone warning at 4:40
Stay away from the SSL2 ... mine crapped out within a year.
That’s not good. I just picked up the 2+. My only issue is it creates what sounds like a groundloop or interference of some sort through the speakers. It could be because I’m using Mackie HR 824.
@@TheBluuHouse That's what mine was doing too. They're pieces of junk. I bought into the hype about them. I went ahead and bought a Focusrite Clarett ... have yet to hook it up. I thought it might've been a problem with the Mac mini M1 not being grounded. But the more I see online it sounds like a common problem with the SSL2s. The buttons are fuzzy and made a terrible humming noise ... also only one channel worked. People should stay away from this item ... Plastic junk with gimmick branding is all these things are.
Too simple
At start of the video... A lot of bold claims from this guy
OMG🤯Could this have been any more longwinded and boring ❓😒🙄