Okay, let's take a different approach:

- What plugins are you using and what is the reported latency in Studio One? Not the latency you get when changing the buffer, the latency all the way at the bottom of the project where you can see CPU load.

- You said you use 16 samples in SO and 32 in Ableton because that is the lowest it can go. Why are you surprised 16 samples craps out faster than 32 samples? Why not use 32 samples? A 64 buffer in my Audient only has about 8ms of latency. The latency you have seem to be self-inflicted by the plugins you're using.

- Maybe ditch the Scarlett and get an Yamaha AG series or a Steinberg if zero latency with DSP is a must?

 

Yes, exactly @ shinjiya, I like your approach best: It's best to use high-quality plugins with very low latency and another high-quality audio interface or a PCI sound card from RME.

 

Personally, I use gig performer upstream of my Daw for plug-in chains and I have no latency problems because the effect chain is used before impacting the latency of the daw.
On S1, you can also insert plug-ins on the recording inputs. But you'd better use a standalone host software to route your plugins to your daw's input.
Otherwise, I also have two preamps and two hardware compressors that I use in direct monotoring without latency with universal control software and a reverb in the daw. I use a Presonus Quantum HD 2 sound card whose latency and preamps are good.

 

Last edited: Jun 11, 2025 at 11:48 PM

That still won't fix his issue. There's no magic solution that will make latency disappear if you move it elsewhere, you're just shifting the latency from one place to the other. It is simply a workflow issue: most plugins will add samples of latency because they have to get their CPU cycles somehow. How do you cut latency? You either make the payload smaller (reducing buffer), or make a compromise in signal quality (simplifying the algorithm to make it require less CPU cycles - most often removing aliasing filtering). This is just how it works, if you want to slap two compressors, a preamp emulation, some kind of boutique EQ, etc, you're getting latency.

 

Just try it and you'll tell me, please.

 

Yep, I'll surely plan for a new interface after a few months.

I'm using all the zero latency plugins only when recording vocals. So, Reported latency for this chain in S1 is 0.00 ms.

Mono Channel: Pro-Q4 > UA LA2A Silver > Pro-MB/C6 > CLA76 > Pro-Q4
Valhalla Vintage Verb/Valhalla Plate (BUS)

 

Well, since it's reporting zero latency, I'm pretty sure the CPU is crapping out, that's a lot of plugins to be running in 16 samples. I think you should work on using less plugins instead? Or maybe just track through a different DAW instead if Studio One is really the problem.

 

Other DAWS handle this great like really better than S1 I can have multiple instances of this exact vocal chain & everything works great throughout the record session, no spikes or clicks.

S1 definitely has something going on with their Audio Engine, something doesn't feel quite right as it gives pops & clicks way too easily compared to other major DAWS.

 

this is why many vocalists use audio interfaces with external dsp support. You get near-zero latency, and direct monitoring with fx already applied. If external analog gear isn't an option, it's the next best one. It does not absolutely need to be UAD either, but you are still going to be in the close to 1k range. Running at such a low buffer is no longer an issue because you really do not need to. That's why most people never record anything with a 16 sample buffer setting.

 

The Studio One audio engine works excellently. Check your settings in Studio One. In your first post, you expressed your distrust, which is completely unfounded.I think your problem lies in your head; you want to find errors. And when you think like that, Studio One seems flawed.

Tuning and Optimizing Computers for Digital Audio
https://support.presonus.com/hc/en-...ng-and-Optimizing-Computers-for-Digital-Audio

How to Improve Computer Performance for Digital Audio
https://support.presonus.com/hc/en-...mprove-Computer-Performance-for-Digital-Audio

 

Last edited: Jun 12, 2025 at 8:18 AM

Nope, That's not the case! I already said other DAWS perform better in low latency settings like you can have multiple vsts running + recording in low latency & ableton or logic will not give a single click or a pop in audio but S1 can't match their performance it's not even close, S1 starts to give pops & spikes like way to early.

If you don't trust me then go & do the tests yourself & find out the truth yourself or better share the results.

I haven't said studio one is totally trash in this conversation. I am only stating the problem i've faced for many many years, It was the same in windows & now same in mac. And I have done all the tweaks that you mentioned above but no luck.

You may have great equipment that's why you may have not faced any of the issues i'm stating.

My audio interface is somewhat a potato. It's Scarlett Solo 2nd gen, The 1st gen was bad & 2nd gen is somewhat okay'ish not great i know but the fact that other DAWS are able to squeeze performance & low latency that's why i am puzzled about S1.

I know there are many solutions to this like changing audio interface to something better or simply going analog. But I can't afford those solutions right now.

I am here on the forum from years but just a month ago I made an account & started talking with y'all as I found you guys quite informative & insightful.

Cheers!

 

I'm not really trying to pull your leg here, just trying to reason a bit.

What about:

1- Why 16 samples? Unless I got it wrong, you seem perfect fine with 32 samples? I'm a pianist and I feel no delay when playing at 64 samples. I know you said it has less latency in voice, but I'm really hard pressed that someone can differentiate 16 and 32 samples.

2- have you tried to change bit precision to 32bit instead? Don't get it confused with project bitdepth. Might help you mitigate the clicks and pops (that's usually just the DAW reacting to the CPU spikes). This option is in the audio tab in the settings, close to the sample rate and buffer. A lot of DAWs won't do 64bit processing like Studio One does, maybe you're making the comparison unfair because of that.

 

Hello @blinkitspenguin, I've been making music on my PC since 1998. I've tried all the DAWs and then decided on Studio One. The workflow suits me very well. I previously worked with PCI sound cards and then switched to the Audient iD14 USB audio interface.

I have an "ASUS - ROG STRIX B550-F GAMING" motherboard and an "AMD - Ryzen 7 5800X3D" CPU. I understand your financial situation, but I can't understand how you're criticizing Studio One here. First, you're worried that Studio One will go bankrupt because of Fender—a fear issue, in other words.

Then you test the same mixer chain/project with other DAWs and draw the wrong conclusions.
Why you insist so persistently that Studio One is faulty is beyond me.

I don't understand why the Scarlett Solo 2 is a flop either. It's one of the best-selling mid-range audio interfaces, and you can certainly achieve usable results with it. Perhaps you should discuss your problem directly with the manufacturers of Studio One and Focusrite.

Please run the following program and post the results here.
Screenshots of your Studio One settings and the Scarlett settings would also be very helpful.

LatencyMon 7.31
The LatencyMon software analyzes and identifies which device, which driver or which service is causing audio dropouts or malfunctions. Real-time audio suitability checker for Windows 10/8.1/8/7/2016/2012/2008 (x86 and x64)
Scroll Down to: System Monitoring Tools https://www.resplendence.com/downloads

 

Buffer of 64 samples should give you 1.333..ms of latency @48k or 1.45ms @44.1. I run everything at 48k/256 samples which gives me 5.3ms of latency and it's hardly perceptible even when finger drumming. . How to calculate latency: buffer size/sampling frequency. 256/48=5.333... Of course I could run it much lower, but CPU usage starts to logarithmically rise from 256 samples down. 512 is only marginally better than 256, so I found the sweet spot between speed and CPU usage to be 256.

No matter which CPU you have, it's always the same. The only difference between high-end and low-end is that with a high-end you can run more plugins at very low latency before it starts to crackle, but you could still get better usage of that CPU at higher latency setting. Go as high as you can bear it (it's like weed ).

p.s. my ADDA converters' feedback loop has 82 samples of latency @48k = 1.708333...ms. A-D 0.61ms and D-A 0.4ms + 0.38ms of digital processing delay. My output latency is 5.333+0.4+0.38=6.11333ms... and input 0.61ms higher - 6.32333ms. I have adjusted Reaper audio driver I/O to compensate for those 82 samples so I can resample through hardware without thinking too much. It's the only DAW that allows it AFAIK.

 

Last edited: Jun 12, 2025 at 5:33 PM

Being part of the Beta for 7.2 I can say they worked very hard to get the Apple silicon support up to snuff, there were close to a dozen beta cycles over 3-4 months.

If you have persistant bugs, I recommend you re-make your templates. I found I was using a template from V4 that had been carried through all the subsequent versions that was becoming unstable.

 

lol 16 samples for vocals ...
I can shred 80's guitar at 64 samples or more.

"At a sample rate of 44.1 kHz, a latency of 5 milliseconds corresponds to 220 samples."
5 ms (total) being for the most sensitive AND fastest guitar players on earth.
Everything else is in their head.

Even Dragonforce use digital MFX, digital in ear and digital console in their live sound chain.
Around 10ms ... minimum. And they shred like crazy ... live.

For vocals, less than 20ms is MORE than efficient.

If you hear slapback delay, it is more than 20ms-30ms.
And so NOT 16 samples : your plugin chain is ruining your latency, WHATEVER setting, computer or soundcard you will use.

So there is only solution : DRY monitoring without effects ... or low latency effects.
One good solution is monitoring through specialised hardware vocal effect.

You can run at 16 samples on the best computer ever, if your plugin chain add 30ms or more, you are screwed.

 

Let this sink in: it takes approximately 30 milliseconds for you to hear the speaker 10 meters away (like on stage). 1m away = ~3ms.

The denser the air (fog, rain, humidity), the slower the speed of sound.

Speed of sound, and speed of light even more... two absolutely fascinating things. Can make your mind go eeeek!

 

Last edited: Jun 12, 2025 at 5:47 PM

Reduce latency in the DAW settings

The length of the latency, i.e., the processing time incurred in this process, depends primarily on the processing buffer (called buffer size or device block size), which can be adjusted in the settings. The rule applies here: "High buffer = smooth playback even under heavy load = high latency," and vice versa: "Low buffer = risk of audio dropouts during playback and recording = low latency."

High buffer size equals high latency


Low buffer size equals low latency


Low Latency Mode in Studio One

If you set a buffer size in Studio One one step lower than the one specified for Dropout Protection on the "Processing" settings page, you'll also be offered the "Low Latency Monitoring" function, which you can activate by clicking the large "Z" in the bottom right corner of the main bus. This significantly reduces the input and output latency for your recordings and allows you to record with integrated plug-ins, provided they also have their own latency settings.

 

Maybe I'm doing something wrong or it's my audio interface that's giving me unnecessary audible latency. At this point, I'm not sure why it's like this, as many of you stated that my situation is somewhat "odd".
I can only confirm stuff when I bring a new audio interface after a few months.

 

Do you guys have any recommendations for a good audio interface for lowest latency around 300$ range? I consider Presonus ES2 as it has lowest latency sometimes on par with RME's low-latency. In some articles, people said it has Electronic Interference issues...Is it true?