|
|
|
11-29-2009, 04:38 PM
|
#121
|
Human being with feelings
Join Date: Jul 2009
Location: Ljubljana, Slovenia
Posts: 3,801
|
Quote:
Originally Posted by jbone1313
The OP in that thread nails it:
I will add that this 'early' problem (glad we have a name for it), is also what screws up the MIDI-Loop-Overdub recording--even with input quantize enabled.
|
I think that having the independent settings for all these recording options
will save many people a lot of trouble.
Chris: glad it at least works.. don't forget to save it as a track preset for
it to be a lot less painful the next time around
Last edited by EricM; 11-29-2009 at 04:45 PM.
|
|
|
11-29-2009, 04:40 PM
|
#122
|
Human being with feelings
Join Date: Oct 2009
Posts: 323
|
Quote:
Originally Posted by jbone1313
I will add that this 'early' problem (glad we have a name for it), is also what screws up the MIDI-Loop-Overdub recording--even with input quantize enabled.
|
I think we have managed to get quite far in this thread, regarding what needs to be done to remedy this problem. I just hope someone important will notice this, and somehow implement ways to help you and others like you jbone1313.
Fortunately I'm in no need for loop recording of midi these days, so in most cases (where I need to overdub) I'll just use midi in quantize. If I need non-quantize I'll just use the method described here.
Quote:
Originally Posted by EricM
Chris: glad it at least works.. don't forget to save it as a track preset for
it to be a lot less painful the next time around
|
Haha. You bet. Its the first thing I did.
Thing is, I've learned a bunch while doing this. Everything is good for something.
|
|
|
11-29-2009, 04:41 PM
|
#123
|
Human being with feelings
Join Date: Dec 2006
Location: CA
Posts: 391
|
"The OP in that thread nails it:"
Finish the thread there is a workable solution there
|
|
|
11-29-2009, 04:52 PM
|
#124
|
Human being with feelings
Join Date: Aug 2008
Posts: 1,144
|
Quote:
Originally Posted by jbone1313
I've tried your other suggestion. It doesn't really work, and if it does, its quite fiddly. (No offense, I really appreciate you guys help and patience.)
Are you able to do overdubbed MIDI recording in REAPER at low latencies (~256) whereby the notes (including the first notes) are captured/recorded accurately and consistently?
If so, how?
|
this is totaly new problem for me,
all the discussion in this OP was, a timing issue,
and the requested solution was a timing one.
no lost notes, nor accurate.
what you said, I'll have to check,
I don't know yet.
|
|
|
11-29-2009, 04:56 PM
|
#125
|
Human being with feelings
Join Date: Aug 2008
Posts: 207
|
Quote:
Originally Posted by Fritz
"The OP in that thread nails it:"
Finish the thread there is a workable solution there
|
Will do. Had to run out and grab some food.
|
|
|
11-29-2009, 04:57 PM
|
#126
|
Human being with feelings
Join Date: Oct 2009
Posts: 323
|
Quote:
Originally Posted by Fritz
"The OP in that thread nails it:"
Finish the thread there is a workable solution there
|
Christ. Now I finally see that input and offset option that I have been japping about all the time. I've looked through the options several times, but I've probably just thought that this isnt affecting midi. Why didnt you pop in earlier Fritz?
I'll save the link from that thread, and progress to hopefully nail this by tomorrow. Thanks to you too as well.
|
|
|
11-29-2009, 04:58 PM
|
#127
|
Human being with feelings
Join Date: Aug 2008
Posts: 1,144
|
Quote:
Originally Posted by jbone1313
I've tried your other suggestion. It doesn't really work, and if it does, its quite fiddly. (No offense, I really appreciate you guys help and patience.)
Are you able to do overdubbed MIDI recording in REAPER at low latencies (~256) whereby the notes (including the first notes) are captured/recorded accurately and consistently?
If so, how?
|
what do you mean by ~256, mili seconds or samples ?
|
|
|
11-29-2009, 05:03 PM
|
#128
|
Human being with feelings
Join Date: Dec 2006
Location: CA
Posts: 391
|
Quote:
Originally Posted by Chris Ihao
Christ. Now I finally see that input and offset option that I have been japping about all the time. I've looked through the options several times, but I've probably just thought that this isnt affecting midi. Why didnt you pop in earlier Fritz
|
Don't hang around here too much as I'm dialed in at this point. Just checking in on the status of things as I'm thinking about upgrading to V3
|
|
|
11-29-2009, 05:10 PM
|
#129
|
Human being with feelings
Join Date: Jul 2006
Posts: 20,859
|
Quote:
Originally Posted by EricM
I think that having the independent settings for all these recording options
will save many people a lot of trouble.
|
i think that there should be no need for the end user to concern himself with this stuff. call me crazy.
|
|
|
11-29-2009, 05:13 PM
|
#130
|
Human being with feelings
Join Date: Aug 2008
Posts: 207
|
Looks like I was watching that thread, and I had actually commented on it. Nice.
I agree; we shouldn't really have to think about all this stuff.
|
|
|
11-29-2009, 05:14 PM
|
#131
|
Human being with feelings
Join Date: Oct 2009
Posts: 323
|
Quote:
Originally Posted by Fritz
Don't hang around here too much as I'm dialed in at this point. Just checking in on the status of things as I'm thinking about upgrading to V3
|
I sure hope you understood that I were "kidding" . I have high hopes for this one. Just one point, I have about 24 ms ahead on the mid (just by measuring with reapers time ruler, and taking an average). Isnt this quite much, considering you guys are talking about 6 ms and so forth? I have 5.3 ms asio latency, for the record. I just find it strange is all.
You write "Now reaper has for my latency for recording is the amount the card is reporting+6ms-6ms+98 samples. The 98 samples is for my external converters the card has no idea about the latency of."
Where do you factor in these 98 samples? Maybe there is some overly agressive compensation in play, like we talked about early in this thread?
I sure appreciate your input.
|
|
|
11-29-2009, 05:17 PM
|
#132
|
Human being with feelings
Join Date: Aug 2008
Posts: 207
|
Quote:
Originally Posted by whatsup
what do you mean by ~256, mili seconds or samples ?
|
Samples.
|
|
|
11-29-2009, 05:20 PM
|
#133
|
Human being with feelings
Join Date: Aug 2008
Posts: 1,144
|
jbone, see my post above, the question.
anyway, if your problem is lating notes,
you will need something to get them more quickly,
and that couldn't be done with a plug, even with pre-delay,
because that mode, won't get your notes earlier, it just get other track notes
to be late.
anyway this thread was surely effective, for us, and maybe for future users.
we learned something
|
|
|
11-29-2009, 05:27 PM
|
#134
|
Human being with feelings
Join Date: Aug 2008
Posts: 207
|
Ok, after reading the other threads and some testing, it seems that enabling "Preserve PDC Delayed Monitoring in Recorded Items" combined with input quantize seems to resolve my issue.
Occasionally, REAPER misses the first note, but its not as bad. Of course, that could be my timing. On the other hand, Ableton always gets my first note. I might be able to get it tighter by adjusting the latency settings.
I'm doing to add this to the other thread I have open about loop recording.
Thanks to all for all of your help and patience.
Last edited by jbone1313; 11-29-2009 at 05:34 PM.
|
|
|
11-29-2009, 05:31 PM
|
#135
|
Human being with feelings
Join Date: Dec 2006
Location: CA
Posts: 391
|
Quote:
Originally Posted by Chris Ihao
You write "Now reaper has for my latency for recording is the amount the card is reporting+6ms-6ms+98 samples. The 98 samples is for my external converters the card has no idea about the latency of."
Where do you factor in these 98 samples? Maybe there is some overly agressive compensation in play, like we talked about early in this thread?
I sure appreciate your input.
|
What is happening here is that my firewire interface reports the latency back to Reaper or whatever program for the analog I/O. Since I mainly use a better external converter I have to compensate for the extra latency of the external converter hooked up to the adat ports on my interface.
The 98 samples goes in the input samples box to compensate the audio tracks
The +6ms and the -6ms go in the input and output ms boxes to compensate the midi. These numbers cancel out in the latency calculation so the midi moves without effecting the audio tracks
If you do not have a problem with audio offset, then you could just enter numbers for midi offset but there would have to be a value of + or - something so that audio offset would not be messed up
|
|
|
11-29-2009, 05:46 PM
|
#136
|
Human being with feelings
Join Date: Aug 2008
Posts: 207
|
I have to wonder if all of these problems could be avoided if REAPER used the time stampes provided by the MIDI interface differently. (I think REAPER does use the time stamps, but I'm not sure that REAPER uses it to precisely set the location of the recorded/captured note.)
Note that Schwa talks about time stamps in this post: http://forum.cockos.com/showthread.php?t=34533
It seems like what should be happening (and is not happening), is that REAPER should take the provided time stamp, and use that to place the note on the time line.
I've noticed that MIDI timing is tighter in Ableton Live (notwithstanding the delays Ableton adds to the captured notes to record what is "heard" rather than what is "played").
From the Ableton Live Manual - Chapter 32 - MIDI Fact Sheet
Quote:
Live needs to know exactly when those events were received from the MIDI keyboard. But Live cannot receive them directly they must first be processed by the MIDI interface’s drivers and the operating system.
To solve this problem, the interface drivers give each MIDI event a timestamp as they receive it, and those are passed to Live along with the event so that Live knows exactly when the events should be added to the clip.
|
RE Recording what is heard vs what is played:
Quote:
When monitoring is enabled during recording, Live adds an additional delay to the timestamp of the event based on the buffer size of your audio hardware. This added latency makes it possible to record events to the clip at the time you hear them not the time you play them.
|
I'll bet a lot more folks are using Live for loop recording, and if Live didn't MIDI things in this manner, people would notice very quickly. They probably had to address that problem early-on, since they are very "loop oriented."
|
|
|
11-29-2009, 05:50 PM
|
#137
|
Human being with feelings
Join Date: Aug 2008
Posts: 1,144
|
you know that, don't you.
|
|
|
11-29-2009, 05:55 PM
|
#138
|
Human being with feelings
Join Date: Aug 2008
Posts: 207
|
Quote:
Originally Posted by whatsup
you know that, don't you.
...
|
Oh, Lordy. I think you're right. I'm going to try QPC combined with my aforementioned solution. Will report back.
Edit: Its even more accurate with QPC turned on. It still seems not quite right though.
Last edited by jbone1313; 11-29-2009 at 06:24 PM.
|
|
|
11-29-2009, 06:02 PM
|
#139
|
Human being with feelings
Join Date: Oct 2009
Posts: 323
|
Quote:
Originally Posted by Fritz
What is happening here is that my firewire interface reports the latency back to Reaper or whatever program for the analog I/O. Since I mainly use a better external converter I have to compensate for the extra latency of the external converter hooked up to the adat ports on my interface.
The 98 samples goes in the input samples box to compensate the audio tracks
The +6ms and the -6ms go in the input and output ms boxes to compensate the midi. These numbers cancel out in the latency calculation so the midi moves without effecting the audio tracks
If you do not have a problem with audio offset, then you could just enter numbers for midi offset but there would have to be a value of + or - something so that audio offset would not be messed up
|
Thanks for that. However, I just did some testing, and even with a bunch of ms's added to the output offset, the midi doesnt actually move considerably compared to where it was placed before this. I just tried -200.00 ms output and 200.00 ms input and no dice. What the heck am I doing wrong now then. Bah. Never an early night.
|
|
|
11-29-2009, 06:22 PM
|
#140
|
Human being with feelings
Join Date: Dec 2006
Location: CA
Posts: 391
|
"Thanks for that. However, I just did some testing, and even with a bunch of ms's added to the output offset, the midi doesnt actually move considerably compared to where it was placed before this. I just tried -200.00 ms output and 200.00 ms input and no dice. What the heck am I doing wrong now then. Bah. Never an early night."
I'm stumped works for me here. Well like I said earlier I'm still on Version 2.58.
|
|
|
11-29-2009, 06:34 PM
|
#141
|
Human being with feelings
Join Date: Oct 2009
Posts: 323
|
Quote:
Originally Posted by Fritz
I'm stumped works for me here. Well like I said earlier I'm still on Version 2.58.
|
Hmm. I take it back. It certainly IS doing something. I just tried out merely leaving them at 0.00, while turning off the "use audio driver reported latency" and voila, now the notes are bouncing between getting a few ms's ahead and behind, like they "should". I saw a suggestion from someone who found that this this was preferable in that thread you mentioned.
Now its usually between 5-10 ms ahead or back, while before it was basically ALWAYS 20-30 ms ahead, no (maybe one or two) notes behind. THAT is an improvement. Now I havent tried audio yet, but that'll come soon enough.
Maybe the reason why I didnt get results is because I had the "use audio driver reported latency" on most of the tests, and that my card somehow is reporting this completely wrong. Dunno. What I KNOW is that if this is it, I'll just call it a day and begin recording instead of wasting more time on theory. Hehe.
|
|
|
11-29-2009, 06:43 PM
|
#142
|
Human being with feelings
Join Date: Dec 2006
Location: CA
Posts: 391
|
Quote:
Originally Posted by Chris Ihao
Hmm. I take it back. It certainly IS doing something. I just tried out merely leaving them at 0.00, while turning off the "use audio driver reported latency" and voila, now the notes are bouncing between getting a few ms's ahead and behind, like they "should". I saw a suggestion from someone who found that this this was preferable in that thread you mentioned.
Now its usually between 5-10 ms ahead or back, while before it was basically ALWAYS 20-30 ms ahead, no (maybe one or two) notes behind. THAT is an improvement. Now I havent tried audio yet, but that'll come soon enough.
Maybe the reason why I didnt get results is because I had the "use audio driver reported latency" on most of the tests, and that my card somehow is reporting this completely wrong. Dunno. What I KNOW is that if this is it, I'll just call it a day and begin recording instead of wasting more time on theory. Hehe.
|
With mine it works great with the use driver reported latency. Are you sure you are not putting the ms in the samples box? Anyway I'm done for the evening too
|
|
|
11-29-2009, 06:50 PM
|
#143
|
Human being with feelings
Join Date: Oct 2009
Posts: 323
|
Yeah, I'm sure. It does indeed work now that I've turned off above mentioned setting though. Only thing remaining is to dial in the right numbers. Night Fritz, and thanks again.
Edit: Now, having run a full circle around this subject, I'm now practically unsure if I'm even having a problem with this, except my own human timing that is. More info here: http://forum.cockos.com/showpost.php...8&postcount=50
It IS useful to have the ability to fine tune the offset of midi recording though, due to the nature of recording in this manner (and small differences between audio and midi latency), so the offset settings should come in handy any way.
Last edited by Chris Ihao; 11-29-2009 at 08:10 PM.
|
|
|
11-30-2009, 02:13 AM
|
#144
|
Human being with feelings
Join Date: May 2008
Posts: 678
|
Quote:
Originally Posted by jbone1313
Oh, Lordy. I think you're right. I'm going to try QPC combined with my aforementioned solution. Will report back.
Edit: Its even more accurate with QPC turned on. It still seems not quite right though.
|
Yeah, with my Echo Mia this thing made a difference too. I set it to QPC and used that "preserve PDC delay while monitoring". I just have to test it more but so far so good.
|
|
|
12-02-2009, 05:10 AM
|
#145
|
Human being with feelings
Join Date: Oct 2009
Posts: 323
|
Conclusion
Just to round this up for me. Yes, I need to dial in some offset to get it right. I let the "use audio driver reported latency" stay checked, then I put in -8.00 ms output offset and 8.00 ms input offset.
This way I in effect let Reaper manage the audio offset while I fine tune the midi. I may increase the output/input offset a few more ms, to move the notes a tad "back", but its fine now at 8.00 ms.
Thanks to everybody who contributed to this thread, and especially to Fritz who burst the bubble about how to adjust the midi offset through Reapers own settings. I think there should be something about this in a sticky, as this really isnt that obvious. Fritz post at http://forum.cockos.com/showpost.php...1&postcount=37, and the thread its posted in, should be a great reference to others.
Anyhow, lets make music!
|
|
|
01-04-2010, 10:27 AM
|
#146
|
Human being with feelings
Join Date: Feb 2009
Location: Reaper HAS send control via midi !!!
Posts: 4,031
|
Quote:
Originally Posted by EricM
Reaper records only "what you play", so in case when using
hardware synths everything stays in time, as it predelays
the MIDI that needs to get to hardware on playback.
|
Predelaying, yes, but for which amount? Using which MIDI, MME or Direct-X? What if some midi tracks are set to MME and some others to Direct-X?
Do we have all together different optimum solutions, based on what we are using?
1. Using only hardware synths + hardware efx
2. Using only hardware synths + software efx
3. Using hardware and software synths + software efx
4. Using only software synths + software efx
If yes, how do such optimum solutions look like? Or does Reaper care already about everything automatically? How can we measure the "unpreciseness amount" of Reaper regarding midi timing?
|
|
|
01-04-2010, 10:37 AM
|
#147
|
Human being with feelings
Join Date: Nov 2006
Location: Columbus, Ohio
Posts: 2,028
|
Quote:
Originally Posted by brainwreck
i think that there should be no need for the end user to concern himself with this stuff. call me crazy.
|
Um, agreed. I am glad I stumbled upon this thread as I have been going bonkers with early MIDI myself. I KNOW my timing isn't that off. I had sort of written off as my USB MIDI interface... but it seems everyone is wrestling with this.
Why is this not a bug report or is it? MIDI sync should not be an issue or even anything to have to think about ever.
|
|
|
01-04-2010, 01:30 PM
|
#148
|
Human being with feelings
Join Date: Sep 2008
Location: Sweden
Posts: 7,416
|
MIDI time travel...
I searched the forum the other day but did not find this thread...
I did MIDI loop back tests, and I get early MIDI events with ASIO4ALL.
This post: http://forum.cockos.com/showpost.php...2&postcount=24
And others in this thread: http://forum.cockos.com/showthread.php?p=430514
Note that I do not get early MIDI events with my other ASIO driver (Lexicon Lambda ASIO), so I use this one for MIDI.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
|
|
|
07-09-2010, 05:39 AM
|
#149
|
Human being with feelings
Join Date: Aug 2008
Posts: 207
|
Now that MIDI's getting alot of attention, maybe this issue can be fixed?
|
|
|
08-23-2010, 07:27 AM
|
#150
|
Human being with feelings
Join Date: Jun 2009
Location: South, UK
Posts: 14,214
|
2010 bump..
anyone else still finding the built in metronome to be out?
Is there a FR request for this too?
(I know it's a necro but it seems quite vital to have a solid metronome)
|
|
|
12-05-2010, 12:11 PM
|
#151
|
Human being with feelings
Join Date: Dec 2010
Posts: 1
|
Metronome is off
I have run into this issue also (3.72), very simple to check, just record the metronome click (at 200-300 BPM) with a mic on an audio track and you will hear the issue, the clicks should fall in the same place. The recorded one is out of sync with the playback click. I don't use midi and am all acoustic. Had the same issue with Logic (and Cakewalk years earlier), but Logic has a sample delay where one can adjust the output to recording (both a plus and minus adjustment).
This is a show stopper for me.
Yes the metronome is off where if you record to the metronome, it will feel rushed, any help?
Thanks
TC Konnect 6 Firewire interface
This is a latency issue. Fixed this issue by un-clicking "Use audio driver reported latency" in Preferences -> Recording Settings, then on two different audio cards, recorded the click and set a time in "Output manual offset". It is precise now. Had to do this on both my M_Audio Firewire 610 and the TC Konnect.
Last edited by timberlinemusic; 12-10-2010 at 04:44 PM.
Reason: Fixed issue
|
|
|
12-05-2010, 05:42 PM
|
#152
|
Human being with feelings
Join Date: Jun 2006
Posts: 1,082
|
I hadn't noticed this before because I have been doing a lot of freetime recording without metronome.
I just started working on a piece that needs strict time and hit this wall.
I input a pattern with the mouse and then played a 2nd part.
So I now have 2 parts.
I turned on the metronome for some reason to do something else and things sounded out of whack.
I eventually found that I have to move the midi 3-5 128ths ahead to get the midi in sync with the metronome.
Since these were orchestral strings I figured it was just the nature of the samples I was using.
Now it looks like it is a problem with the metronome.
|
|
|
03-11-2014, 10:46 AM
|
#154
|
Human being with feelings
Join Date: Mar 2011
Posts: 61
|
You are talking about: Human Latency compensation!
Hi Chris,
I've read through the whole Thread. Your Workaround with a self prepared 25ms to 35ms delayed Metronome Track is great!
I've found the nudge function ( type [n] for it) as a workaround for this Problem.
Therein you can shift tracks by a chooseable value in milliseconds.
Now what is the reason for this Behaviour?
It is not Reaper! Not the Soundcard reporting wrong Latency times, neither the same issue with the MIDI Driver. And you, the musician is also not responsible to that.
The reasonis, that you are hearing, that there is a small amount of latency while playing with a VSTi Plugin. You hear a timelag between your played notes and the click. Now what happens is that you automatically play a little bit earlier in time, so that your played sound and the click are both in time. You are anticipating the played notes.
But normally you do not record the audio signal but the midi signal. And MIDI is placed there where you've pressed the Key and not there where your sound comes to your ear.
The funniest Thing is, that you've measured exactly the same time lag as I did. 25 to 35ms. I myself measured values between 28 to 34ms depending on the used virtual Instrument.
The time you've played the notes ahead is the sum of all latencys on your Computersystem.
1. Soundcard Latency (on my system 3ms)
2. sampled sound latency (normally around 1 to 8 ms)
3. MIDI Latency (can be between 1 to aprox. 6ms)
4. DA Converter time (1 ms)
Maybe this information is helpful for you. Best Regards, Musinux
|
|
|
03-11-2014, 10:58 AM
|
#155
|
Human being with feelings
Join Date: Mar 2011
Posts: 61
|
Hint : have a look at this thread
http://forum.cockos.com/showthread.php?t=136362
Here I am discussing the same problem.
The big question is the difference between 15ms (sum of all known latencys) and the measured 28 to 34ms.
Anyone with Ideas here?
Best Regards, Musinux.
|
|
|
Thread Tools |
|
Display Modes |
Linear Mode
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
All times are GMT -7. The time now is 04:31 AM.
|