13 posts
Page 1 of 1
I'm looking at the dropout protection tab wondering where is the "off" option?

The minimum dropout protection setting leaves my standard roundtrip at 10.3ms

If I use LLM on a audio channel then I can get 5.94ms... (which is about 1ms longer than the audio roundtrip of my interface)


My question is... does the standard round trip really need to be so large!? has it always been this slow? back when i was using studio one v2 it felt like the same plugin chain reacted faster than in v4 when using normal monitoring

I always thought that the total roundtrip was mostly interface roundtrip plus whatever plugin latency might happen in the daw… Looking at the numbers, its taking studio one over 5ms to take the signal and spit it back out at 44.1 and a 48 sample buffer on my fireface uc (no plugins active)

I ran studio one v2 (32bit) for ages using about 4.2 ms of plugin latency at 44.1, at a 128sample interface buffer... no one noticed any latency in vocal booth headphones.. Since moving off version 2 now up on version 4, i feel like there is extra latency even at the lowest buffer my interface supports (48 samples which my new pc build is running very smoothly with)

I have been trying to get used to the LLM feature in version 4... But just like in my brief effort in studio one v3, I'm finding its clunky and interrupts my supposed to be smooth session!

Things that bug me in LLM mode:

slow duplicating of a record enabled track.
audio cuts out on record enabled track during duplication
plugin meters have all gone dead, which is discouraging me from tweaking settings while recording
enabling/disabling a track to record feels slow

basically while using LLM, a lot of the smooth sneaky ways i was able to work with studio one in a session now come with a pause attached to them, usually indicating to anyone within earshot that something just popped a bit.. Part of what has been so great about studio one was being able to do so many things without interrupting playback or even do some risky moves while audio is recording (I tend to do a bit of editing and tweaking while audio is being captured)

in standard mode (LLM off), i can pretty confidently sculpt the sounds im hearing as i am recording them... all the meters are working so i can dial in settings quickly during the first moments of the artist vocalizing on the track... this is the way i want to run all the time, and in these recent versions of studio one it hasn't felt quite right latency wize...

was the audio engine latency in studio one v2 actually faster than studio one v4 set at minimum dropout protection? I might just be imagining it, but i think v4 is slow
User avatar
by ralfwottrich on Fri Oct 19, 2018 8:14 am
Hello,

I asked for the same when the new audio engine was introduced and learnt there is no "off" option in the concept. Meanwhile I got used to the new concept and like it. You just have to understand that you can simply switch of "Z" in all the outputs and run on minimal protection to get a similar behaviour, and it helps if you don't pay to much attention to the ms displayed but just listen if you feel any latency or not.

While tracking audio (or playing "live") I set the dropout protection to minimal and make sure no "Z" is active on any output channel. That feels quite similar to the v2 and worked for me in V3.5.6 well enough.
(In V4.1 I noticed the internal buffer on minimal dropout protection is double, which is increasing displayed ms numbers, but I don't really feel any difference when playing.)
There is maybe a difference in the displayed milliseconds and I was not happy firstly, but have to admitt it still works well without any audible latency (nobody in the band complained, yet).

While working on the mix I can incease dropout protection to run the mix on a higher buffer with lots of plugins, and still have the option to activate "Z" to record an additional audio track or VST instrument without noticeable latency. In V2 you would need to increase the ASIO driver buffer to run the mix, and adding another recording would be a headache to switch back. With the new audio engine it is quite easy.

I suggest give it some more time, and don't activate any Z in the output channels if you don't need to.
User avatar
by BobF on Fri Oct 19, 2018 10:57 am
Maybe this will help some:

viewtopic.php?f=153&t=25650

-

-------------------------------------------------------------
Studio One Pro - latest / Win10Pro x64 (latest) / i7-6700 @3.8Ghz / 32G / UMC1820 / Event PS8
RC w/Galaxy S7+ / KKS61MK2 / Atom SQ
User avatar
by lawrencefarr on Fri Oct 19, 2018 11:35 am
there is no "off" option in the concept


Dropout Protection set to Minimum is off.

Studio One Technical Support | QA
------------------------------------------------
Issue Reporting Add-On - Version 4.3.0
------------------------------------------------
Kontakt 6 | MOTU Mach 5 | Spitfire BBCSO Core & Albion One | Cinematic Studio Solo Strings | Vienna Synchron Prime | Cinesamples Musio | NI Session Horns Pro | Toontrack EZ Drummer, EZ Bass,& EZ Keys | Relab LX480 Complete | Soundtoys Little Plate & Echoboy
User avatar
by collinsylex on Sat Oct 20, 2018 10:54 am
thanks for the replys everyone!

minimum = off... i get that its now the lowest possible... and from one of the above posts i also understand that the minimum dropout protection latency in v4 is now double what it was in v3.5...

so from my own breakdown of the numbers i understand that for me to now be running on my interfaces minimum buffer (48 samples) and dropout protection at minimum, in standard mode (llm off) i am getting similar latency to when i was running a 128 sample buffer in the old studio one v2 days...

being that i just upgraded my pc to a bit of a monster, this is a little disappointing! I now have the power to run at much tighter latency figures and its now being diluted by the software unless i decide to run using the LLM which i dont feel is ideal for my workflow...

its a little disappointing that at a 48ms buffer, i tried playing my guitar over the speakers the other day, and i could definately feel the latency delay between the impact of my pick to what was coming out of the speakers... with LLM enabled it did feel very good! When i am tracking myself playing some instruments i will be sure to use to use LLM... but when doing my usual work all day every day with clients i dont want to run that way... between takes the client is trying to communicate with me from the live room thru the mic... in those same moments between takes i am trying to copy what they just did to an identical settings track and get ready to record again... in LLM with the pausing of audio caused by what i need to do in the moments between takes, some of what the artist is trying to tell me from the booth is cut off and its distracting/annoying!

i read somewhere that studio one once had close to the lowest latencys out there... now its pro tools and reaper currently working the fastest at less than 2ms on a focusrite chart i came across... im seeing studio one with my fastest possible interface settings (at 44.1) its more like 5ms... as much as i love studio one, i'm watching reaper close for their next update which is supposed to support ara2...

i kindof wish studio one v2 would still be a good option for me, but the latest version of melodyne doesn't work in it anymore so i'm kindof stuck moving forward
User avatar
by Jemusic on Sat Oct 20, 2018 1:03 pm
Studio One's latency depends on the interface and the port you are using. One of my interfaces is a Focusrite Clarett 2 Pre and over thunderbolt I can set a 16 sample buffer and get 2.2 mS round trip! Yes round trip, so that is roughly 1 mS either way. It is so fast I can software monitor everything while tracking. There is no perceivable delay.

I don't use drop protection at all. I find I get the fastest and best response by just setting it to minimum and leaving it at that while tracking.

Even on my main Win 7 machine using an RME PCI interface I am getting 3.5 mS round trip as well. Still fast. PreSonus Quantum also provides blistering low latency as well.

Specs i5-2500K 3.5 Ghz-8 Gb RAM-Win 7 64 bit - ATI Radeon HD6900 Series - RME HDSP9632 - Midex 8 Midi interface - Faderport 2/8 - Atom Pad/Atom SQ - HP Laptop Win 10 - Studio 24c interface -iMac 2.5Ghz Core i5 - High Sierra 10.13.6 - Focusrite Clarett 2 Pre & Scarlett 18i20. Studio One V5.5 (Mac and V6.5 Win 10 laptop), Notion 6.8, Ableton Live 11 Suite, LaunchPad Pro
User avatar
by collinsylex on Sat Oct 20, 2018 1:36 pm
Jemusic wroteStudio One's latency depends on the interface and the port you are using. One of my interfaces is a Focusrite Clarett 2 Pre and over thunderbolt I can set a 16 sample buffer and get 2.2 mS round trip! Yes round trip, so that is roughly 1 mS either way. It is so fast I can software monitor everything while tracking. There is no perceivable delay.

I don't use drop protection at all. I find I get the fastest and best response by just setting it to minimum and leaving it at that while tracking.

Even on my main Win 7 machine using an RME PCI interface I am getting 3.5 mS round trip as well. Still fast. PreSonus Quantum also provides blistering low latency as well.


nice! those numbers are great... tell me, when you have your clarett set to 16 samples, and you look at the dropout protection tab set to minimum... the roundtrip including the daw is really that low?? for me at 48 samples on my rme fireface uc its like 2.x ms either direction plus about 5ms extra for studio one totalling about 10ms

perhaps the time is coming for me to next update my usb interface... i was thinking of getting an rme adat card but i guess its gonna be best to avoid the usb digiface... i might have to look at getting the thunderbolt upgrade for my new motherboard... but im gonna take a guess that it will really only shave down about 3 ms max from the performance i have right now... running at 48 ms like i am right now should be really really tight! unfortunately the minimum dropout protection buffer is forcing an extra 96 samples of latency into the equation! can you believe running a 16sample buffer just to have studio one throw a minimum 96 sample buffer right on top of that!?

so whats that total roundtrip including the studio one software on your rig? (total roundtrip for standard audio on the dropout protection tab)
User avatar
by danam2 on Sat Oct 20, 2018 3:15 pm
@collinsylex:
I don't know what your clients need to monitor through S1, but I suggest to create monitoring tracks where you engage LLM (green Z) all the time while disabling monitoring on the tracks you actually record to. That should also solve your LLM problem with copying tracks.

There are other ways to handle the monitoring problem. Please read the final chapter of the PDF that is provided in the thread that BobF linked to.

BUT: The most annoying problem of LLM is IMO the fact that you can only tweak plugin parameters that can be automated. Many guitar amp sims don't offer that by default. You would have to configure the automation parameters. For such situations I hope your audio interface can handle multiple clients. Because the faster solution would be to run S1 (where you don't monitor at all) and the desktop version of the guitar amp sim where you monitor your playing. But as someone already mentioned: Use drop out protection and LLM only when the progress of the project demands for it.

Anyway, calculating the roundtrip latency can be done from within S1. The latency for LLM or minimal dropout protection is calculated (in samples, not ms) at least on my system as follows:
Audio:
input latency + output latency + device block size
Instruments:
a) device block size is set to 64 or 128 samples: Output latency + device block size (or is it the process block size (?), it is the same actually)
b) device block size < 64 samples: Output latency + device block size + process block size
c) device block size > 128 samples: 2x Output latency + device block size
This differs from the explanation in the PDF, but that is what I concluded from the values on my S1 4.1 and Audient iD14 on Windows 10, CPU: i7 6700k.

My latency values with minimal dropout protection for 44.kHz and 32 samples device block size are:
Audio: Standard 8.34 ms / LLM 5.44 ms
Instrument: Standard 4.67 ms / -
As a former RME UCX user I guess that your device may provide slightly better values.

If you do the math you will also see that for VSTi monitoring (LLM active or not) the device block size is best set to 64 samples. For audio the lowest your system allows.

User avatar
by garybowling on Sun Oct 21, 2018 7:02 pm
EDIT: In rereading some of the posts. I suspect you are using soft synths or guitar fx (amp sims, etc) requiring you to go through the DAW for monitoring. I'll leave my notes below in case other people who don't do that could gain some benefit from the information.

-----------------------------------

Since you have an RME interface I don't even understand why you're using the DAW for your monitoring set ups. I just use TotalMix. TotalMix basically allows you to route anything to anything..

In TotalMix, select the outputs you're using for your monitor outputs. Then mix to that output whatever inputs (Mic/Guitar/whatever) and software (DAW tracks) you want. Turn off the blue "monitor" on your S1P tracks.

The only downside to this is you don't get FX from S1P, like reverb or whatever on your live monitored tracks. I have an RME UFX and it has compression, EQ, reverb, and delay DSP effects that you can add and not record. Not sure if the UC has those or not.

If not, the upgrade would be to the RME UFX+ or UFX II. That would give you the FX in TotalMix and you really don't do any monitor mixes in your DAW. Gives you very low, almost zero latency. The monitor mix doesn't even go through your computer, just your hardware interface. I typically set my latency to 1024 samples in my RME settings, which gives me nice smooth operation in the DAW and use TotalMIx for my monitor mixes with basically zero latency.

gabo

ASUS laptop (AMD 5900HX), 32G, 2x2TB SSD, Win11-64, RME UFX & BabyFace, Studio One Pro 6, Addictive Drums2, Izotope 10, Soothe2, Waves, many plugins, Melodyne Studio 5, all versions updated frequently

The Moderns,
https://open.spotify.com/artist/1x6Fd133GftlRyRYl0xgjf
User avatar
by ralfwottrich on Mon Oct 22, 2018 12:24 pm
Hello,

to somehow demonstrate what collinsyslex (and me) are talking about: I have a RME Fireface UC a swell and I have S1 in the versions 2.6.5, 3.5.6 and 4.1.0 installed in parallel. Please see the reported latencies in the audio settings.

The driver settings are the same. Obviously there is a change from 3.6.5 to 4.1.0 in setting internal block size to 96 (why?). I never had any issue with 48 samples in 3.6.5 and found it quite perfect for me.

btw. any VST based looper (Gloop, Guitar rig looper) will only run in Standard mode.

edit: Just to make clear, the difference here is only the software version, I just started the version one after the other on the same computer with same RME UC attached.

Attachments
S1 Latency.jpg
User avatar
by danam2 on Mon Oct 22, 2018 4:12 pm
@ralfwottrich: That is odd. What are your device settings exactly? I just tried some other (virtual) asio drivers I have on my system and it seems the process block size with minimal dropout protection is never less than the actual device block size. But on my system this is the same on 3.5.6 for all ASIO drivers including my "real" device, the Audient iD14. The only difference I can tell from your screen is the blue Z in 3.6.5 and the green Z in 4.1. Since my audio device does not offer ASIO direct monitoring (= blue Z) I have no clue what could be different :-(

User avatar
by collinsylex on Mon Oct 22, 2018 5:57 pm
thanks for posting those screenshots ralf!

the fireface uc doesnt have any dsp effects like the ufx or even babyface/ babyface pro... way back in the day i used to record just monitoring off a hardware mixer or the driver mixer, but for the past several years i monitor pretty much everything thru the daw... that allows me a high level of consistancy between playback and recording settings for the artists, and allows me to be very flexable with what the artist wants to be hearing in their heaphones... i record mostly rappers and about 60 percent of my sessions people want to hear autotune in their headphones... i stuck with autotune 5 (32 bit with zero latency version) and studio one v2 32bit for as long as possible... but eventually i was finding myself hitting the wall alot, running out of ram after getting melodyne in there to fix up the pitching of the vocals... i needed to move to a 64 bit daw... so i moved to studio one v3 64 bit and also upgraded to autotune 8 which has a low latency mode...

with studio one v3 i had a few people notice the latency difference but it didnt seem too bad... the few times i personally went and tested autotune on the mic i noticed it seemed a bit delayed compaired to the old version... but low latency autotune 8 actually adds a couple ms of latency which i thought was the big problem... so those couple ms for the new autotune plus a couple ms extra for studio one 3 vs 2 was probably more the of the issue... so now, moving to studio one 4 its even gotten worse..

now i have built a new pc and got my driver latency way down, but i'm still no better off in the latency department than the old days of v2 or even v3 running at 128 buffer... Yes, there is the low latency option, but i kindof feel like using LLM is forced on all studio one users now since standard mode latency has been nerf'd to this extent... kindof annoying as LLM doesnt fit evey workflow, and doesnt exactly function in all situations as we have seen from other posts in this thread...

i really like studio one v4, but i would love to see an option for "off" or at least an option to get down to 48 or less samples in the dropout protection tab

i am trying to decide on a good way to chop a few ms of plugin latency in my recording template as a work around
User avatar
by ralfwottrich on Tue Oct 23, 2018 3:41 am
danam2 wrote@ralfwottrich: That is odd. What are your device settings exactly?


Yes, the RME driver is set to 48samples buffer and was not touched, just starting on after the other software version and take the settings using the snipping tool.
If I remember correctly, in 3.6.5 the green Z is not available with protection on minimal.

For recording I still use mainly 3.6.5 with minimal dropout and all Z to off because I regularily use Jamorigin's Guitar to MIDI VST as well as GLoop VST.

13 posts
Page 1 of 1

Who is online

Users browsing this forum: No registered users and 78 guests