Go Back   Cockos Incorporated Forums > REAPER Forums > MIDI Hardware, Control Surfaces, and OSC

Reply
 
Thread Tools Display Modes
Old 07-28-2012, 02:38 AM   #41
viscofisy
Human being with feelings
 
Join Date: Oct 2011
Location: Dalriada
Posts: 13,294
Default

Ok,just done what brainwreck suggested,using much the same 6 tracks as mentioned above.

I changed the track time indicator to samples.
Zoomed to where the difference between transients is readable.

I got 560 samples.

560/96000 = 0.0058333


The Reaper indicator says 2.4/3.4.
I take it that's in/out?
viscofisy is offline   Reply With Quote
Old 07-28-2012, 02:58 AM   #42
gofer
-blšnk-
 
gofer's Avatar
 
Join Date: Jun 2008
Posts: 11,164
Default

Quote:
Originally Posted by viscofisy View Post
Both of them gave lower results for 96kHz than for 44.1kHz ?
Is this because the buffers clear at 2X the speed,or something?
Yep, that's pretty much it, I think. An amount of samples at 96kHz is representing a shorter time than the same amount of samples at a lower rate. So your buffers must be processed faster. Of course you buy the speed with higher CPU and hard disk usage.
At least, that's another layman's way to look at it .

Personally I am pretty insensitive with latency, I can comfortably play at around 10-15 ms, even when I begin to sense the indirectness. Maybe it's because I'm a guitarist used to play far away from my speakers? I'm not doing live gigs anymore, but back then I sometimes went into the crowd and there timing was rather done by feeling the fingers/plec hit the strings. Listening to the guitar's sound from the speakers was more for validating whether I did it right after the fact and readjust if I was slipping off sync with the rest of the music .
Of course that was not inside comfort zone, but I think the practice has helped me to be able to play at higher latency now.
gofer is online now   Reply With Quote
Old 07-28-2012, 09:04 AM   #43
tls11823
Human being with feelings
 
tls11823's Avatar
 
Join Date: Aug 2010
Location: Harrisburg, PA USA
Posts: 1,405
Default

Quote:
Originally Posted by gofer View Post
Maybe it's because I'm a guitarist used to play far away from my speakers?
I was in the high school marching band. Our practice area was between the school and a bus garage. The garage was right beside our practice area, and the school was across the parking lot, a few hundred feet away. The "goalpost" ends of the area had nothing for quite a distance. So the latency you heard from your instrument varied from very short to several seconds, depending on where you were on the field and which way you were facing. Of course, being part of a marching band meant you were constantly moving and changing direction.

That was the practice field. When it came to game day, everything was different because of the configuration of the stands and everything else near the field.

So I just got used to feeling the music and keeping an eye on the director rather than listening. Now a few milliseconds of latency means nothing to me, although I know it drives my friends crazy when they're trying to record.
__________________
We act as though comfort and luxury were the chief requirements of life, when all that we need to make us happy is something to be enthusiastic about.
--Charles Kingsley... or maybe Albert Einstein... definitely somebody wiser than myself--
tls11823 is online now   Reply With Quote
Old 07-28-2012, 11:52 AM   #44
brainwreck
Human being with feelings
 
Join Date: Jul 2006
Posts: 20,840
Default

Quote:
Originally Posted by ivansc View Post
Trouble is, when all is said and done it STILL comes down to "what does it feel like to ply with" doesn't it?
Yep, that's what it boils down to. I also look at it in the respect that there is no downside to being able to run an interface at the lowest possible latencies, given that it's stable. The lower the latency, the less chance that it is having any effect on performance. As an experiment, you might try a little exercise. Bring the latency up, say in the 40 ms + range, and play a drum sampler to a click. You might notice how difficult it is to stay locked in time. Now gradually bring the latency down, and observe what happens to your performance and the effort that goes into trying to stay on the click. I have done this, and all sorts of other little experiments with latency, just to satisfy my own curiosities. My conclusion is that latency can't be low enough. Call me crazy, but I long for the day when digital audio latencies are in the microsecond range.
brainwreck is offline   Reply With Quote
Old 07-29-2012, 01:53 AM   #45
ivansc
Human being with feelings
 
Join Date: Aug 2007
Location: Near Cambridge UK and Near Questembert, France
Posts: 18,941
Default

Quote:
Originally Posted by brainwreck View Post
Yep, that's what it boils down to. I also look at it in the respect that there is no downside to being able to run an interface at the lowest possible latencies, given that it's stable. The lower the latency, the less chance that it is having any effect on performance. As an experiment, you might try a little exercise. Bring the latency up, say in the 40 ms + range, and play a drum sampler to a click. You might notice how difficult it is to stay locked in time. Now gradually bring the latency down, and observe what happens to your performance and the effort that goes into trying to stay on the click. I have done this, and all sorts of other little experiments with latency, just to satisfy my own curiosities. My conclusion is that latency can't be low enough. Call me crazy, but I long for the day when digital audio latencies are in the microsecond range.
+1.
But we already HAVE true zero latency monitoring, but of course at a price. No FX on your input signal as monitored.
and we are almost back to the old "split your signal to another track and apply fx for when you are recording" paradigm from tape days! But at least we didnt gt latency that way unless using a really really long lead in a BIG studio! (grin)
ivansc is offline   Reply With Quote
Old 07-29-2012, 02:59 AM   #46
gofer
-blšnk-
 
gofer's Avatar
 
Join Date: Jun 2008
Posts: 11,164
Default

We also didn't get to play virtual instruments. A price I wouldn't accept to pay for some millisecs, personally.
gofer is online now   Reply With Quote
Old 07-29-2012, 05:29 AM   #47
ivansc
Human being with feelings
 
Join Date: Aug 2007
Location: Near Cambridge UK and Near Questembert, France
Posts: 18,941
Default

Quote:
Originally Posted by gofer View Post
We also didn't get to play virtual instruments. A price I wouldn't accept to pay for some millisecs, personally.

+1000

DAW recording is a slippery slope.
I dont think I could ever go back to a pile of boxes with wires hanging out of them in a huge, heat generating 19" rack.

Incidentally, anyone want to buy a classic heay - duty Zimmer frame style rack unit on castors?
Very cheap to UK Reaperists before I bay or gumtree it.
ivansc is offline   Reply With Quote
Old 07-29-2012, 10:55 AM   #48
viscofisy
Human being with feelings
 
Join Date: Oct 2011
Location: Dalriada
Posts: 13,294
Default

Ivan,what kind of buffer or mix control comes with the Saffire 6 usb?
There's a guy asking for advice on lowering latency in this thread :
http://forum.cockos.com/showthread.php?t=107621
viscofisy is offline   Reply With Quote
Old 07-29-2012, 11:06 AM   #49
ivansc
Human being with feelings
 
Join Date: Aug 2007
Location: Near Cambridge UK and Near Questembert, France
Posts: 18,941
Default

I already replied in the thread but have gone back and amplified little. It is almost certainly something he has set up wrong in his audio settings. For instance I am pretty sure the saffire6 wont go above 48/24...may have to dig the manual out if it comes to that
ivansc is offline   Reply With Quote
Old 07-29-2012, 03:11 PM   #50
Banned
Human being with feelings
 
Banned's Avatar
 
Join Date: Mar 2008
Location: Unwired (probably in the proximity of Amsterdam)
Posts: 4,868
Default

At typical conditions, sound waves travel through the air at roughly 300 meters per second, so that makes 1 ms about 30 centimetres, and 10 ms about 3 meters.

How far does a bass player stand away from the drum kit? That must be about where the limit is for really tight grooves then, no?
__________________
˙lɐd 'ʎɐʍ ƃuoɹʍ ǝɥʇ ǝɔıʌǝp ʇɐɥʇ ƃuıploɥ ǝɹ,noʎ
Banned is offline   Reply With Quote
Old 07-29-2012, 07:46 PM   #51
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default

Quote:
The lower the latency, the less chance that it is having any effect on performance.
exactly!!! so things like 'to play guitar a latency of xxx ms is enough..' is meaningless.... THE LOWEST THE BETTER.

Now.....from MY tests and from reading MANY forums and people that seems to know what they are talking about, i' d say that OVER 8-9ms of total round trip latency you start to feel it...

most interfaces give higher values at the lowest setting. this is what i complain about. manufacturers NEED TO CHANGE this. 5ms round trip latency should be the norm. not to go OT or beat the dead horse, but that is one the the very few advantages on mac, where CORE AUDIO is very efficient and you usually get very decent RTL

Danny Bullo
bullo is offline   Reply With Quote
Old 07-29-2012, 09:47 PM   #52
Andy Hamm
Human being with feelings
 
Andy Hamm's Avatar
 
Join Date: May 2011
Location: Canada
Posts: 1,371
Default

The easiest way to see if latency is detectable is to use headphones or IEMs. If a performer doesn't notice a delay, then it is neglagible and can be considered non existent.
Andy Hamm is offline   Reply With Quote
Old 07-31-2012, 07:43 PM   #53
chucky5p
Human being with feelings
 
Join Date: Apr 2011
Location: Canada
Posts: 454
Default

Quote:
Originally Posted by brainwreck View Post
...The lower the latency, the less chance that it is having any effect on performance... My conclusion is that latency can't be low enough. Call me crazy, but I long for the day when digital audio latencies are in the microsecond range.
+1000

Note: RTL=Round Trip Latency

Personally, I long for the day when total MAXIMUM RTL will be < 50 microsecond (that's 100 times smaller than typically available in 2012). Why should it be that small? Because at such small latencies, PHASE cancellation effect (combing/flanging) between the original acoustic source and its processed result by the audio interface/DAW are almost non-existent or at least inaudible (1/.000050 = 20 kHz)

One thing that I almost never see mentioned about latency issues in digital audio processing (recording, monitoring, etc) is this: even if a soundcard’s latency is low enough to prevent any "playability" issues (i.e. with a RTL of say 1 msec, nobody would ever complain that it’s too big-even drummers) but that same 1msec is still big enough to cause severe and very audible phase cancelation issues with original acoustic sound sources. I won’t go into lengthy details here as to why this happens as this would take a few pages to explain fully (see link below), but suffice it to say that the biggest impact these phase cancellations have are on musicians who have no choice but to listen to both the acoustic version and the processed version of their instrument simultaneously (i.e. singers). Delays (latency) in the processed version compared with the acoustic version can produce large flanging effect which may in turn hinder the musician’s performance. The most affected by this are vocalists (whose instrument are their own voice-they feel it directly in their head), followed by other type of sound source (i.e. drummers, saxophonist, etc), and for electric instruments (i.e. electric guitars, keyboards), phase cancellation issues with the original acoustic source (guitar strings, keyboards keys) are for all practical purpose non-existent.

You can read more about phase cancellation (flanging) here:
http://en.wikipedia.org/wiki/Flanging

Ah well…I guess I’ll have to wait quite a “few” years before this happens. NOT!

Chuck
chucky5p is offline   Reply With Quote
Old 07-31-2012, 08:33 PM   #54
ivansc
Human being with feelings
 
Join Date: Aug 2007
Location: Near Cambridge UK and Near Questembert, France
Posts: 18,941
Default

I remember giving up on wireless guitar a few years back because of latency issues experienced when I strayed too far away from the P.A.

Strangely, I recently started using a wireless rig for off-stage excursions again & this time - no issues.

Mind you, this is over a distance of no more than 50 feet from the PA. Not sure what level of latency is involved over those distances but it cant be ll that much, I guess.
ivansc is offline   Reply With Quote
Old 08-01-2012, 04:58 AM   #55
chucky5p
Human being with feelings
 
Join Date: Apr 2011
Location: Canada
Posts: 454
Default

Quote:
Originally Posted by ivansc View Post
I remember giving up on wireless guitar a few years back because of latency issues experienced when I strayed too far away from the P.A.

Strangely, I recently started using a wireless rig for off-stage excursions again & this time - no issues.

Mind you, this is over a distance of no more than 50 feet from the PA. Not sure what level of latency is involved over those distances but it cant be ll that much, I guess.
The precise math is: Distance from the source/Speed of sound in the air at 20C (BTW, the speed of sound in water is 4.3x faster)

So in your case 50ft / 1126 ft/sec = 0.0444 seconds or 44.4 milliseconds

Note: A very fast and rough way to calculate this is for every foot =~ 1ms (in your case 50 feet =~ 50 ms)

Also note that this measurement doesn't include any additional delay caused by speaker design, audio processing(digital or not), etc.

Chuck
chucky5p is offline   Reply With Quote
Old 08-01-2012, 11:41 AM   #56
chucky5p
Human being with feelings
 
Join Date: Apr 2011
Location: Canada
Posts: 454
Default

Quote:
Originally Posted by brainwreck View Post
Really? We had good latency with some of the lower cost interfaces running under XP, but for Vista/7, latency has gone to shit again. And I'm sitting here with an onboard soundcard/asio4all, getting half of the latency of my usb interface, the same interface that got half the latency on XP, but doubled when moving to 7.
Are you referring to the Fast Track Ultra 8R? If so, I know a trick allowing the 8R to get 5.2ms RTL latency under W7/Vista (64 samples @ 48 kHz). You need a motherboard that can disable HPET in its BIOS (High Precision Event Timer). With HPET disable, youíll be able to use the old 8R XP drivers in W7 and select ďHigh Performance ModeĒ which allows for 64 samples, et voila, it works really well! My main setup still uses XP but I used this BIOS trick on my Acer Netbook (pre-loaded with W7). Originally, the Netbook didnít have the HPET option in its BIOS but I was lucky enough to find a hacked BIOS that did.

AFAIK, HPET is THE reason the 8R crashes under W7 when using High Performance mode. (High Performance mode and HPET are incompatible) And since M-Audio/Avid didnít want to redesign their driver to work under W7 with HPET enabled, they simply removed the High Performance Mode option! The reason why the 8R works great in XP with HP mode on is simply because XP doesn't use HPET, so it doesn't matter whether your motherboard BIOS has the HPET option or not when using XP.

Even though I found this hack, it still makes me angry that Avid didnít release an updated 8R driver for W7 with a working HPmode. And donít count on it because if they did, the 8R would then become in direct competition with their new MBox Pro 3 (which allows 64 sample under W7) but at a lower price. Basically, as far as latency is concerned, you could say:

FastTrackUltra 8R (XP) = MBox Pro 3 (W7)

BTW, in most cases Asus motherboards donít have the HPET option but AFAIK all Gigabytes MB do.

Chuck
chucky5p is offline   Reply With Quote
Old 08-01-2012, 02:23 PM   #57
brainwreck
Human being with feelings
 
Join Date: Jul 2006
Posts: 20,840
Default

Quote:
Originally Posted by chucky5p View Post
Are you referring to the Fast Track Ultra 8R? If so, I know a trick allowing the 8R to get 5.2ms RTL latency under W7/Vista (64 samples @ 48 kHz). You need a motherboard that can disable HPET in its BIOS (High Precision Event Timer). With HPET disable, youíll be able to use the old 8R XP drivers in W7 and select ďHigh Performance ModeĒ which allows for 64 samples, et voila, it works really well! My main setup still uses XP but I used this BIOS trick on my Acer Netbook (pre-loaded with W7). Originally, the Netbook didnít have the HPET option in its BIOS but I was lucky enough to find a hacked BIOS that did.

AFAIK, HPET is THE reason the 8R crashes under W7 when using High Performance mode. (High Performance mode and HPET are incompatible) And since M-Audio/Avid didnít want to redesign their driver to work under W7 with HPET enabled, they simply removed the High Performance Mode option! The reason why the 8R works great in XP with HP mode on is simply because XP doesn't use HPET, so it doesn't matter whether your motherboard BIOS has the HPET option or not when using XP.

Even though I found this hack, it still makes me angry that Avid didnít release an updated 8R driver for W7 with a working HPmode. And donít count on it because if they did, the 8R would then become in direct competition with their new MBox Pro 3 (which allows 64 sample under W7) but at a lower price. Basically, as far as latency is concerned, you could say:

FastTrackUltra 8R (XP) = MBox Pro 3 (W7)

BTW, in most cases Asus motherboards donít have the HPET option but AFAIK all Gigabytes MB do.

Chuck
I have the Ultra, not the Ultra 8r, but I would think that it wouldn't matter. Maybe you know something about that? I'm looking now, and I do see that there is an unlocked bios for my machine. I'm not sure if HPET can be switched in the unlocked bios, though. Thanks for the info.
brainwreck is offline   Reply With Quote
Old 08-01-2012, 08:57 PM   #58
chucky5p
Human being with feelings
 
Join Date: Apr 2011
Location: Canada
Posts: 454
Default

Quote:
Originally Posted by brainwreck View Post
I have the Ultra, not the Ultra 8r, but I would think that it wouldn't matter. Maybe you know something about that? I'm looking now, and I do see that there is an unlocked bios for my machine. I'm not sure if HPET can be switched in the unlocked bios, though. Thanks for the info.
Yes the Ultra and Ultra 8R have the same basic hardware and use the same driver, so you can use the same HPET trick with the Ultra.

Be careful though about updating your BIOS firmware; you need to be absolutely sure it will work with your particular motherboard and that the HPET mod works, otherwise you could "hose" your MB. Also, an "unlocked" BIOS doesn't necessarily mean that the “modders” have unlocked all the options available in the BIOS. Sometimes a particular option is simply not available. You could always ask for help on the DigitalLife forum, it's dedicated in modding BIOSes. (Disclaimer: I will not be responsible for any mod you do on your PC!)

http://forums.mydigitallife.info/for...-Requests-Only

BTW, another nice side effect of turning HPET off is that your DPC latencies will (in most cases) go down to their pre-XP level. If you have tested for this, you probably have noticed that your PDC latencies went way up when you switch from XP to W7 (mine went from ~ 10usec to ~ 300usec) Again the culprit is HPET. Note however that in many cases it is preferable to leave HPET on because W7 was specifically designed to use it instead of RTC in XP. But if you want to keep the High-Performance Mode/5.2msec of the Ultra in W7, there are no other options AFAICT.

Note: you may have read on another post of mine that I still use XP on my main setup, that’s because it doesn’t have the HPET option in BIOS and I couldn’t find a hacked version (HP A1720N) so W7 on that machine not an option, ~11msec is way too high for me.

Chuck
chucky5p is offline   Reply With Quote
Old 08-01-2012, 10:09 PM   #59
brainwreck
Human being with feelings
 
Join Date: Jul 2006
Posts: 20,840
Default

Quote:
Originally Posted by chucky5p View Post
Yes the Ultra and Ultra 8R have the same basic hardware and use the same driver, so you can use the same HPET trick with the Ultra.

Be careful though about updating your BIOS firmware; you need to be absolutely sure it will work with your particular motherboard and that the HPET mod works, otherwise you could "hose" your MB. Also, an "unlocked" BIOS doesn't necessarily mean that the “modders” have unlocked all the options available in the BIOS. Sometimes a particular option is simply not available. You could always ask for help on the DigitalLife forum, it's dedicated in modding BIOSes. (Disclaimer: I will not be responsible for any mod you do on your PC!)

http://forums.mydigitallife.info/for...-Requests-Only

BTW, another nice side effect of turning HPET off is that your DPC latencies will (in most cases) go down to their pre-XP level. If you have tested for this, you probably have noticed that your PDC latencies went way up when you switch from XP to W7 (mine went from ~ 10usec to ~ 300usec) Again the culprit is HPET. Note however that in many cases it is preferable to leave HPET on because W7 was specifically designed to use it instead of RTC in XP. But if you want to keep the High-Performance Mode/5.2msec of the Ultra in W7, there are no other options AFAICT.

Note: you may have read on another post of mine that I still use XP on my main setup, that’s because it doesn’t have the HPET option in BIOS and I couldn’t find a hacked version (HP A1720N) so W7 on that machine not an option, ~11msec is way too high for me.

Chuck
Don't worry about the bios stuff. I have been there a time or two. After reading a bit about disabling HPET, although it is said to be good for some realtime stuff, it is also said that MIDI accuracy in 7/Vista is dependent upon it. Not that I have to worry about any of that, because I don't have the option to explore it, thanks to a limited oem bios, even the unlocked version. Yea, I installed it earlier, and it seems to run fine. I still may hit up one of those bios mod forums, though. Btw, in case I do find something for disabling HPET, do you have any info about running the XP driver for the FTU under 7? I had a look around, but I didn't find any good info about it.

If I could get all of the drivers that I need, I wouldn't have any issue going back to XP myself. I should have just built a desktop this time around, and now I'm paying for my decision of buying an oem laptop, again. And you're right. Dpc latency did go up, despite running much faster hardware, across the board, than my last machine. And my interface latency doubled...ugghh...I'm going to have to spring for a new interface, which is a shame, because I like everything about this one except the latency under Windows 7. And....the driver crashes if the computer goes into sleep mode with the interface connected, and M-Audio are showing no signs of ever fixing that issue.

Yea, 11 ms is too high for me, too, especially considering that is the minimum that I can get with the FTU. If only more people were critical of interface latencies, we might not be dealing with this stuff.
brainwreck is offline   Reply With Quote
Old 08-02-2012, 10:48 AM   #60
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default

Quote:
Personally, I long for the day when total MAXIMUM RTL will be < 50 microsecond (that's 100 times smaller than typically available in 2012). Why should it be that small? Because at such small latencies, PHASE cancellation effect (combing/flanging)

only if you mix the original (non delayed signal) with the computer out signal with is delayed by latency. but topically you dont want that. i.e: using your daw to monitor you voice with compressor, eq, and reverb....there is no point to add the original unprocessed signal
bullo is offline   Reply With Quote
Old 08-02-2012, 10:44 PM   #61
serr
Human being with feelings
 
Join Date: Sep 2010
Posts: 8,292
Default

Fun fact:
The drum kit (with all the sharp transients) is usually 10' - 14' behind the PA speakers. Looks like I'm running with perfect time alignment here doesn't it.

If you have a problem with playing 10' away from someone then I guess this will be a problem for you with in-ear monitors. I'm not saying that it isn't perceivable on some level but I honestly don't think this is an issue for most.

And again, if you're hearing an actual slap delay then your latency is really much higher than 10 or 12ms.
serr is online now   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 03:05 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.