rbowser
Max Output Level: -10 dBFS
- Total Posts : 6518
- Joined: 2005/07/31 14:32:34
- Status: offline
Reported latency question
Latency seems to be the topic of the hour. I really haven't been concerned with latency for a long time, everything's running smoothly for me. BUT - it's always been a mystery to me why what I'm told in the Sonar Preferences audio settings sounds like I should have a problem when I don't. Here's some info - Maybe an expert here can 'splain to me what the deal is. Here's just one of the current threads where latency is discussed: "Experienced user, still not having it with Cakewalk Sonar products" On that thread, Robert Bone said: robert_e_bone I still think your Total Rountrip Latency of 18 milliseconds is too high for good recording without noticeable lag.
I believe the audible lag detection point is something like 14.5 milliseconds or thereabouts. It was some German guy that figured that out, if memory serves.
I always recommend keeping settings to where for recording purposes, the Total Roundtrip Latency reported by Sonar is around 10 milliseconds, or just under that... Here's what's mysterious to me: --When I've engaged my settings to use while recording, Sonar says that I have-- "Effective latency at 48kHz/stereo 5.3 msecTotal Roundtrip: 53.3 msecs"When I play a soft synth, the response is immediate. There's no detectable lag, so playing piano, playing anything goes smoothly as it should. Then when I'm mixing, I open my Alesis iO|2 interface and move the slider up so that my "effective latency" is around 10 or 11 msec, and that works great for mixing. WHY, with my recording setting, does it say my roundtrip is a whopping 55.3, when Robert says, as everyone says, that roundtrip needs to be under 10, or the delay will be heard and make live playing impossible? Is it because my interface reports differently than most interfaces? All I know is that I've always had a total roundtrip report of around 50 msecs, and I've always just shrugged it off since everything's working fine. What's the story, what's the deal?? Randy
Sonar X3e Studio Roland A-800 MIDI keyboard controller Alesis i|O2 interface Gigabyte Technology-AMD Phenom II @ 3 GHz 8 Gb RAM 6 Core Windows 7 Home Premium x64 with dual monitors
|
CJaysMusic
Max Output Level: 0 dBFS
- Total Posts : 30423
- Joined: 2006/10/28 01:51:41
- Location: Miami - Fort Lauderdale - Davie
- Status: offline
Re: Reported latency question
2014/06/04 11:27:57
(permalink)
WHY, with my recording setting, does it say my roundtrip is a whopping 55.3, when Robert says, as everyone says, that roundtrip needs to be under 10, or the delay will be heard and make live playing impossible?
If one way is about 5msecs and roundtrip is 55.3, then something is really really off. Those numbers do not make sense Also, its probably because you are not Actively recording that track. Latency for recording will not be noticed when track is just armed for recording. You need to actually be recording. CJ
|
rbowser
Max Output Level: -10 dBFS
- Total Posts : 6518
- Joined: 2005/07/31 14:32:34
- Status: offline
Re: Reported latency question
2014/06/04 11:37:08
(permalink)
CJaysMusic
WHY, with my recording setting, does it say my roundtrip is a whopping 55.3, when Robert says, as everyone says, that roundtrip needs to be under 10, or the delay will be heard and make live playing impossible?
If one way is about 5msecs and roundtrip is 55.3, then something is really really off. Those numbers do not make sense Also, its probably because you are not Actively recording that track. Latency for recording will not be noticed when track is just armed for recording. You need to actually be recording. CJ
Thanks for the reply, CJ - Before posting, I opened Sonar to make sure I had the numbers accurate. That's what it says, 55.3 roundtrip - But what I'm saying is that this is what Sonar has Always said in that audio settings dialogue, and I've had this same audio interface for about 7 years. I'm talking about recording tracks, not just arming them. With those settings I've always been able to record either MIDI or live audio with Zero detectable lag - with "effective latency" reported at 5.5 My theory is that this Alesis interface just isn't being interpreted correctly by Sonar, and that 5.5 is actually my roundtrip figure. When I have my Mix setting engaged, and my one trip is at 10 msecs - of course I can hear the lag then, and it would be impossible to play a keyboard with that. That kind of latency is what people are talking about when they're reporting being unable to play accurately due to lags - but I never attempt to record with the settings moved up like that for mixing. Randy
Sonar X3e Studio Roland A-800 MIDI keyboard controller Alesis i|O2 interface Gigabyte Technology-AMD Phenom II @ 3 GHz 8 Gb RAM 6 Core Windows 7 Home Premium x64 with dual monitors
|
scook
Forum Host
- Total Posts : 24146
- Joined: 2005/07/27 13:43:57
- Location: TX
- Status: offline
Re: Reported latency question
2014/06/04 11:48:11
(permalink)
You are running at 48kHz with 256 buffers according the the "Effective latency at 48kHz/stereo 5.3 msec". These figures are the results of the a simple calculation: Effective Latency = NumberOfBuffers / SampleRate in this case .0053 = 256 / 48000 The round trip latency is the sum of the ASIO reported input an output latencies. This relies on information from the interface in addition to the calculated latency. I would imagine most would find 256 buffers at the edge for monitoring through the DAW.
|
rbowser
Max Output Level: -10 dBFS
- Total Posts : 6518
- Joined: 2005/07/31 14:32:34
- Status: offline
Re: Reported latency question
2014/06/04 11:51:30
(permalink)
scook...The round trip latency is the sum of the ASIO reported input an output latencies. This relies on information from the interface in addition to the calculated latency. I would imagine most would find 256 buffers at the edge for monitoring through the DAW... Thanks for the post, scook - Yes, 256 is exactly where I have things set for when I'm recording. You've put "reported" in bold, because you're suggesting that this Alesis unit isn't reporting figures in a way Sonar expects? Randy B.
Sonar X3e Studio Roland A-800 MIDI keyboard controller Alesis i|O2 interface Gigabyte Technology-AMD Phenom II @ 3 GHz 8 Gb RAM 6 Core Windows 7 Home Premium x64 with dual monitors
|
scook
Forum Host
- Total Posts : 24146
- Joined: 2005/07/27 13:43:57
- Location: TX
- Status: offline
Re: Reported latency question
2014/06/04 12:01:35
(permalink)
I found nothing about the interface so can't say how/what the Alesis reports. The round trip is the sum of the input and output values , each value is reported separately just above the total. The round trip only happens when an audio signal is coming from the audio inputs interface and returning to the interface. When playing a soft synth the audio is originating in the DAW, it is not coming from the audio inputs on the interface.
|
robert_e_bone
Moderator
- Total Posts : 8968
- Joined: 2007/12/26 22:09:28
- Location: Palatine, IL
- Status: offline
Re: Reported latency question
2014/06/04 13:29:28
(permalink)
I haven't seen my name in bold for a long while. Yikes! :) I looked up the specs on the interface: http://alesis.com/io2 Hopefully, I have the correct one. IF so, it looks like this interface does NOT have its own ASIO drivers, and instead relies on ASIO4ALL. If you go to the above link, then click on Docs and Downloads, you will see what I am talking about. SOOOO, I am quite confused. Please tell me what Driver Mode you are running in Sonar, and if you are using ASIO4ALL for your audio drivers for this interface, and also what all of your settings are. Thanks, Bob Bone
Wisdom is a giant accumulation of "DOH!" Sonar: Platinum (x64), X3 (x64) Audio Interfaces: AudioBox 1818VSL, Steinberg UR-22 Computers: 1) i7-2600 k, 32 GB RAM, Windows 8.1 Pro x64 & 2) AMD A-10 7850 32 GB RAM Windows 10 Pro x64 Soft Synths: NI Komplete 8 Ultimate, Arturia V Collection, many others MIDI Controllers: M-Audio Axiom Pro 61, Keystation 88es Settings: 24-Bit, Sample Rate 48k, ASIO Buffer Size 128, Total Round Trip Latency 9.7 ms
|
scook
Forum Host
- Total Posts : 24146
- Joined: 2005/07/27 13:43:57
- Location: TX
- Status: offline
Re: Reported latency question
2014/06/04 13:38:38
(permalink)
I believe the ASIO Reported Latencies are only shown when SONAR is in ASIO driver mode. Running ASIO4ALL may explain why there is such a disconnect between the calculated latency and the reported latency. Don't know for sure though don't use it.
|
robert_e_bone
Moderator
- Total Posts : 8968
- Joined: 2007/12/26 22:09:28
- Location: Palatine, IL
- Status: offline
Re: Reported latency question
2014/06/04 14:43:59
(permalink)
Scook - yes, you are correct. ASIO reported latency values are only posted in Preferences>Audio>Driver Settings when the Driver Mode is ASIO. I too suspect that the culprit here is again ASIO4ALL - possibly not reporting things correctly, or who knows what. Bob Bone
Wisdom is a giant accumulation of "DOH!" Sonar: Platinum (x64), X3 (x64) Audio Interfaces: AudioBox 1818VSL, Steinberg UR-22 Computers: 1) i7-2600 k, 32 GB RAM, Windows 8.1 Pro x64 & 2) AMD A-10 7850 32 GB RAM Windows 10 Pro x64 Soft Synths: NI Komplete 8 Ultimate, Arturia V Collection, many others MIDI Controllers: M-Audio Axiom Pro 61, Keystation 88es Settings: 24-Bit, Sample Rate 48k, ASIO Buffer Size 128, Total Round Trip Latency 9.7 ms
|
Jim Roseberry
Max Output Level: 0 dBFS
- Total Posts : 9871
- Joined: 2004/03/23 11:34:51
- Location: Ohio
- Status: offline
Re: Reported latency question
2014/06/04 14:46:50
(permalink)
Round-trip latency is the sum of the following: - ASIO input buffer
- ASIO output buffer
- A/D D/A converters
- The driver's hidden safety-buffer
So if you have your audio interface's ASIO buffer size set to 64-samples at 44.1k... The ASIO input buffer = 1.5ms The ASIO output buffer = 1.5ms The A/D and D/A converters = ~1ms The driver's hidden safety-buffer is the X FACTOR Better units use a small safety-buffer (RME, MOTU, Lynx) Some units use a large safety-buffer If your audio interface uses a large safety-buffer, there's not much you can do about it. Most safety-buffers are hidden from the end-user... so there's no means of adjusting it. As a point of reference, the best audio interfaces yield ~5ms round-trip latency at a 64-sample ASIO buffer size/44.1k. If you're playing a soft-synth, you're not dealing with round-trip latency... but rather one-way (playback) latency. To effectively overdub (monitoring thru software while recording), you need to keep round-trip latency as low as possible. I find anything above 6ms uncomfortable. Anything above 10ms feels like you're playing thru molasses. Don't believe it? Here's a simple test (assuming your audio interface has low round-trip latency): Insert a delay processor set 100% wet. Adjust the delay time to 50ms (to make the effect very obvious). Try to play/monitor in realtime thru that 50ms delay. Drop the delay time to 20ms and repeat. Drop the delay time to 10ms and repeat. Drop the delay time to 5ms and repeat. Finally, disable the delay plugin and repeat. If you're playing/monitoring high-transient instruments (bass, drums, guitar), you can really feel the lag of the extra latency.
|
rbowser
Max Output Level: -10 dBFS
- Total Posts : 6518
- Joined: 2005/07/31 14:32:34
- Status: offline
Re: Reported latency question
2014/06/04 15:07:04
(permalink)
Excellent posts - Thanks Jim, Robert (oh no, your name's in bold again!), and Scook for replying.
I'll tell ya, I haven't worried about latency in so many years since all is working fine for me. But the current posts about latency got me curious, since my round trip report has never made sense to me.
But you can see from my posts - it's been so long since I've thought about latency, it slipped my mind when I started this thread that of course round trip isn't what's involved when recording MIDI.
HOWEVER - I never have any trouble recording vocals. When I record audio, a mike is the most common source. I put on my headphones, listen to track playback, and sing away without a problem. I thought there was delay compensation involved or something - all I know, there's no delay.
CLEARING UP THE ALESIS DRIVER QUESTION:
--This unit (supplanted by the "Express" version of the same interface) does have its own ASIO driver, and that's what I use primarily. However, huge caveat, the driver doesn't work in 64 bit. As you discovered from your search, Robert, Alesis just points users to ASIO4ALL - IF they're working in 64 bits. So, when I have 64 bit Sonar open, ASIO4ALL does kick in. But I'm still doing most of my work in 32 bit, so use the unit's native driver.
Jim, you refer to "the driver's hidden safety-buffer," and how the user often doesn't have access to it. Maybe that's what I'm looking at when I open the Alesis control panel?
Since more info was asked for, here's everything that seems pertinent when opening Sonar's Preferences: --Input Drivers - ALESIS IO|2 USB ASIO Input (when in 64 bit - ASIO4ALL v2io|2 1)
So, Playback Timing Master and Record Timing Master are both the ALESIS IO|2 USB
Sampling Rate 48
Mixing Latency - 5.3 msec - 256 samples Playback and Recording
ASIO Use Multiprocessing Engine (6 core machine)
MIDI Devices
io|2 for inputs and outputs - and my A-800 Pro keyboard/controller
MIDI
Number of Buffers 64
Playback- Prepare Using 500 Millisecond Buffers (as per the oft repeated advice here on the Forum)
ASIO Settings in ALESIS panel:
Buffer Size 256
Number of Buffers 8 - Jim, is that the "hidden safety-buffer?"
SO - all I know is I have no problems. Ever since I've had this interface, and that's Years, Sonar has always hovered around showing 50 msecs as the roundtrip. Shrug. I experience no lag no matter what I'm doing, so I'm happy. I'll just have to accept the mystery of it all!
Randy
Sonar X3e Studio Roland A-800 MIDI keyboard controller Alesis i|O2 interface Gigabyte Technology-AMD Phenom II @ 3 GHz 8 Gb RAM 6 Core Windows 7 Home Premium x64 with dual monitors
|
robert_e_bone
Moderator
- Total Posts : 8968
- Joined: 2007/12/26 22:09:28
- Location: Palatine, IL
- Status: offline
Re: Reported latency question
2014/06/04 15:14:45
(permalink)
I think it's possessed :) Regardless of the numbers being reported, if it is performing as needed for you, have a blast. And thanks for clearing up the drivers and driver mode weirdness. Bob Bone
Wisdom is a giant accumulation of "DOH!" Sonar: Platinum (x64), X3 (x64) Audio Interfaces: AudioBox 1818VSL, Steinberg UR-22 Computers: 1) i7-2600 k, 32 GB RAM, Windows 8.1 Pro x64 & 2) AMD A-10 7850 32 GB RAM Windows 10 Pro x64 Soft Synths: NI Komplete 8 Ultimate, Arturia V Collection, many others MIDI Controllers: M-Audio Axiom Pro 61, Keystation 88es Settings: 24-Bit, Sample Rate 48k, ASIO Buffer Size 128, Total Round Trip Latency 9.7 ms
|
scook
Forum Host
- Total Posts : 24146
- Joined: 2005/07/27 13:43:57
- Location: TX
- Status: offline
Re: Reported latency question
2014/06/04 15:16:47
(permalink)
When recording audio, is it possible the audio being recording is directly monitored ? If so, this would explain the near zero latency monitoring while recording audio.
|
robert_e_bone
Moderator
- Total Posts : 8968
- Joined: 2007/12/26 22:09:28
- Location: Palatine, IL
- Status: offline
Re: Reported latency question
2014/06/04 15:19:14
(permalink)
I was thinking that very same thought. I too am wondering if that is what is happening. Bob Bone
Wisdom is a giant accumulation of "DOH!" Sonar: Platinum (x64), X3 (x64) Audio Interfaces: AudioBox 1818VSL, Steinberg UR-22 Computers: 1) i7-2600 k, 32 GB RAM, Windows 8.1 Pro x64 & 2) AMD A-10 7850 32 GB RAM Windows 10 Pro x64 Soft Synths: NI Komplete 8 Ultimate, Arturia V Collection, many others MIDI Controllers: M-Audio Axiom Pro 61, Keystation 88es Settings: 24-Bit, Sample Rate 48k, ASIO Buffer Size 128, Total Round Trip Latency 9.7 ms
|
Jim Roseberry
Max Output Level: 0 dBFS
- Total Posts : 9871
- Joined: 2004/03/23 11:34:51
- Location: Ohio
- Status: offline
Re: Reported latency question
2014/06/04 15:40:19
(permalink)
Hi Randy, Are you monitoring via hardware or software? If you're monitoring via the onboard features of your audio interface, then that's near zero latency (regardless of the ASIO buffer size). In this case, the audio interface is taking what arrives at it's input... and immediately shooting it to an output. This would explain the high round-trip latency figure... while still being able to comfortably monitor. Round-trip latency is really only a factor if you play/monitor in realtime thru software based EFX/processing. ie: You're playing a DI electric bass/guitar while monitoring in realtime thru an AmpSim plugin. No, the number of buffers has nothing to do with the driver's hidden safety-buffer. That's just it... it's typically hidden (you have no means of adjusting it). Some audio interfaces use a large hidden safety-buffer with the logic that it reduces odds of problems upon playback under less than ideal circumstances (ie: a poorly configured machine). But... that extra buffering comes at the expense of higher (much higher in some cases) round-trip latency.
|
rbowser
Max Output Level: -10 dBFS
- Total Posts : 6518
- Joined: 2005/07/31 14:32:34
- Status: offline
Re: Reported latency question
2014/06/04 16:44:58
(permalink)
I think my head's full! Thanks for all the info, tips and conversation - As you said Bob, things are performing as per my needs so, fine.
This will be helpful for me to reply to: Jim sed - "Round-trip latency is really only a factor if you play/monitor in realtime thru software based EFX/processing."
I never use any plugins while recording. Dry bone, raw audio is all I want, and certainly know from experience that many (not all?) plugins are going to give one instant latency - why live with it? I know many musicians, probably especially guitarists, find it a complete drag to not hear FX while playing - quite a challenge to do a great lick and feel how it's going to be in the mix, if all you're hearing is a live, unprocessed "plunk." But I'm not a guitarist, and for what I'm doing, using lots of virtual instruments, it's no hardship to just focus on the notes, the music itself, and then have fun later adding me reverb and whatever else seems good.
Answering another thing brought up:
This interface, like most, has a Direct/USB Monitor Mix knob - it's called different things on different units. But it's very rare for me to turn it to Direct - Yes, that gives me the straight signal before any trips to and from the computer - but that's all that's heard, the new signal going in - If I wanted to monitor the tracks already in Sonar, I'd need to get into a monitor send, external mixer etc --not into doing that, no need for what I do. So the knob stays at the USB position. I listen to the tracks, I sing, I'm happy.
Thanks again.
Randy
Sonar X3e Studio Roland A-800 MIDI keyboard controller Alesis i|O2 interface Gigabyte Technology-AMD Phenom II @ 3 GHz 8 Gb RAM 6 Core Windows 7 Home Premium x64 with dual monitors
|
scook
Forum Host
- Total Posts : 24146
- Joined: 2005/07/27 13:43:57
- Location: TX
- Status: offline
Re: Reported latency question
2014/06/04 17:00:52
(permalink)
I would imagine most never turn the direct/usb knob all the way to direct. This is a mix knob controlling the ratio of direct to usb signal. The knob is there to eliminate the need for a mixer to hear the DAW while using direct monitoring. But as you say, it is all academic.
|
Cactus Music
Max Output Level: 0 dBFS
- Total Posts : 8424
- Joined: 2004/02/09 21:34:04
- Status: offline
Re: Reported latency question
2014/06/04 17:03:58
(permalink)
Latency is very confusing topic and I don't think half the people out there using an under $300 interface even realize the unit probably is like Randy'sv with a TRTL of 25 ms and up. Like my Focusrite Scarlett, people rave about them, but they actually suck at RTL. Mine is 28ms on a good day. But as Jim said, this matters not if your not listening to your input via the outputs. There is the Centrance test you have to run to find out but appearantly Sonars reported RTL is good for most but might not report the so called hidden buffers.
|