Topic: Edirol R-44 - 4 Channel Recorder (Part II)

[leddy]

I received R44 number 2 today.  My main concern was how well they sync up. 

Plugging identical mics in each unit, I recorded some hand claps.  The claps were just over 4 milliseconds apart when I examined them in Wavelab.   

Obviously, you would be careful to keep pairs of mics within the same unit.  I think I'll also record some claps or snare hits at the start of a file, and maybe at the end to check drift.

[Ozpeter]

The sync provision is not so much to make them start dead together but to keep them together, according to the manual.  So you would have to manually line the two up, but then they should stay locked together for an indefinite period.  The only way to test is to do a start handclap then further ones every (say) five mins for an hour, and look at that.  Banging two bits of wood together might give a sharper transient or use a dog clicker.

[BayTaynt3d]

The sync provision is not so much to make them start dead together but to keep them together, according to the manual.  So you would have to manually line the two up, but then they should stay locked together for an indefinite period.  The only way to test is to do a start handclap then further ones every (say) five mins for an hour, and look at that.  Banging two bits of wood together might give a sharper transient or use a dog clicker.

What he said. My understanding is that it is known that they won't start/stop completely synced, the more important question is drift. Initial sync will be easy to fix in post, especially if you know you need to (since you can add a clap if you want or other slate-like device). I'd like to run this test personally: Start both machines recording, add a slate (hand clap, clapper board), let it roll for an hour, then add another slate clap. Then in your DAW, sync up the first slate, and then see how far (if any) the tail slates moved (or didn't). That's the real test I'd like to know the result of because if the only thing that sync cable does is somewhat sync the pressing of stop/record but isn't syncing word clocks at all, then it's pointless since you can do that manually right now with any two recorders. However, if they keep sync, then that's a major plus. I've been dying to know that ever since I saw that you can link two decks together. Something tells me we'll find out soon enough.

[leddy]

The sync provision is not so much to make them start dead together but to keep them together, according to the manual.  So you would have to manually line the two up, but then they should stay locked together for an indefinite period.  The only way to test is to do a start handclap then further ones every (say) five mins for an hour, and look at that.  Banging two bits of wood together might give a sharper transient or use a dog clicker.

What he said. My understanding is that it is known that they won't start/stop completely synced, the more important question is drift. Initial sync will be easy to fix in post, especially if you know you need to (since you can add a clap if you want or other slate-like device). I'd like to run this test personally: Start both machines recording, add a slate (hand clap, clapper board), let it roll for an hour, then add another slate clap. Then in your DAW, sync up the first slate, and then see how far (if any) the tail slates moved (or didn't). That's the real test I'd like to know the result of because if the only thing that sync cable does is somewhat sync the pressing of stop/record but isn't syncing word clocks at all, then it's pointless since you can do that manually right now with any two recorders. However, if they keep sync, then that's a major plus. I've been dying to know that ever since I saw that you can link two decks together. Something tells me we'll find out soon enough.

I should have time tomorrow to do it. 

[penguin]

I received R44 number 2 today.  My main concern was how well they sync up. 

Plugging identical mics in each unit, I recorded some hand claps.  The claps were just over 4 milliseconds apart when I examined them in Wavelab.   

Obviously, you would be careful to keep pairs of mics within the same unit.  I think I'll also record some claps or snare hits at the start of a file, and maybe at the end to check drift.

I've been using my two R-44s in slave mode for a few of rehearsals and one gig at a bowling alley. For the rehearsals the runtime has been about 3 hours, and for the gig it was two three hour sessions. I've found that if you start the two from a dead stop (just hitting 'record'), the slave more often than not will start milliseconds after the master, which is easily synced in post production. It seems if you have the two units on standby (pause and record) and then hit record from there, the slave seems to start in better sync with the master.

With regards to drifting, I haven't had a problem with that during the initial recordings at all. All 8 tracks lined up beautifully in Logic. But there did seem to be some out of sync stuff happening later. See, after putting all the files in Logic and mixing, I decided I wanted to redo some keyboards on one particular song cuz I choked during the live performance. So what I did was bounce down the initial tracks to an mp3 and import to iTunes and then into my iPod. I hooked up my keyboards to the R-44 and rerecorded the keyboard parts while playing the bounced iPod song as a guide. (I didn't use my Duet to record straight into Logic because it made the keyboards sound too different). After I finished, I downloaded the file and popped it into Logic. I lined it up with the other original gig files, but as the song went on the new keyboard tracks started lagging farther and farther behind. At the end of the song the lag was quite noticable. Perhaps it could have been me, but I *really* don't think I would have played those rerecorded parts in such a progressively slower fashion like that without noticing it. Anyways, I had to do a bit of trimming sections and move them forward to match up, no major trauma. But I wonder what was the cause of the progressive lagging? One thing that I am thinking about is that I did use the same sample/bit rate that I used during the initial recording (24/48), but I during the initial recording I used a Sandisk 16GB Class 4 card and during the rerecording I used a Panasonic 8GB Class 6 card. Could the different class speeds have something to do with it? Or maybe it was the bounced file, which went from 48 to 44.1 at some point? Hmmmmmm.

Anyways, just my observations.

[Ozpeter]

The mp3 playback on the iPod would certainly give rise to sync errors.

[Gutbucket]

When I run my twin R-09's, I manually sync the start by hitting play (from rec/pause mode) on both simultaneously.   I am amazed at how close they stay in sync, drift-wise.  They are so close that I cannot hear any drift over a 2 hour file when playing back from both units.  Don't have my DAW software loaded currently so I haven't looked to see how close they really are.  Just sayin'.

Thinking about it now, I suppose I should try swapping cards between recorders, then playing back the same files and see if they still stay in sync.  That should exaggerate any differences between the clocks rather than compensating for them.

[BayTaynt3d]

Sometimes "really close" is actually worse cause that's when phasing can start getting all crazy on you, heh.

[Gutbucket]

Truth. 

I should add that I'm not mixing the two sources so the possible (probable?) phase bugaboos aren't an issue. I'm just playing them back through separate speakers - they only mix in the air.

[BayTaynt3d]

Got the Oade unit in the mail two days ago. So far, so good. This thing is sweet.

Haven't used it in the field yet, but I made two battery test runs and got about 2:50 total each time on a set of RatShak rechargeable AAs. I formatted the card each time before starting the recording, and ran with all four phantom power channels on (mics were attached). That's not bad. Not sure I'll run with the DVD battery or not. A second set of AAs would be a lot smaller than hauling a DVD battery. That said, the DVD battery must last forever with this thing (haven't tested that, won't bother, heh).

Also, for the record, it seemlessly split at 2 gigs for both tests, and it also saved the file both times before shutting down while running low on AA battery power. I've been using a class 6 Transcend 8 gig card, and didn't have a problem running at 2xStereo 24/44.1.

The other thing that was interesting was I guess I have to choose "Adapter" for the power supply when using my DVD battery for some reason? Otherwise it just says "low batt" and turns off. I guess it just barely ins't putting out enough voltage or something? Seems odd considering the unit runs just fine when set to "Adapter" mode. Maybe it doesn't matter. Seems like if it did register though at 9v, then the battery meter would be active while running on DVD power. That doesn't really matter though cause the DVD batt has it's own meter. Anyway, I just wonder if I might need a new external battery? Anyone having this same experience?

[Ozpeter]

The voltage you set for external battery is the voltage that the battery goes down to when it is flat - the "terminal voltage" I think the manual calls it.  So if you set it to 9 volts with a 9 volt battery (which when fully charged might slightly exceed that, but not for long) then the R-44 will consider it flat the moment it heads down to that voltage.  You can only work out the correct figure by trial and error, or measurement of your external battery voltage when it is almost flat.  In fact, given that the R-44 asks for 9 volts minimum, using a 9 volt nominal battery might underpower it when the battery discharges a bit.  Again, simply something you'd need to try.

[BayTaynt3d]

The voltage you set for external battery is the voltage that the battery goes down to when it is flat - the "terminal voltage" I think the manual calls it.  So if you set it to 9 volts with a 9 volt battery (which when fully charged might slightly exceed that, but not for long) then the R-44 will consider it flat the moment it heads down to that voltage.  You can only work out the correct figure by trial and error, or measurement of your external battery voltage when it is almost flat.  In fact, given that the R-44 asks for 9 volts minimum, using a 9 volt nominal battery might underpower it when the battery discharges a bit.  Again, simply something you'd need to try.

That's what I figured, and what I was a little worried about (that a 9v battery is just barely good enough). The batts powered the R4 no problem for hours and hours though, so even though this is a different unit, seems so similar that I'd be surprised if these batts didn't work just as good for the R44. Interesting...

[RobC]

I ran my R-44 with a 9v 5400mAh Lithium Ion battery ("walmart style DVD battery) with no problems at a festival 2 weeks ago.
I did have to set to "Adapter" for it to work.  I got a good amount of recording time off of a fully charged battery.   I did have 4 fully charged AAs inside it too,  that way if the external battery does get too low it will switch to the internal batteries.

I also ran it with a 12v SLA and still had to set it to "Adapter"  when i had it set to 12v i got the low battery warning right away.

[Ozpeter]

About R-44 markers:-

They seem to be somewhat non-standard when it comes to viewing them in DAW software.  They way they work seems to be that every R-44 file contains markers 0 to 99, all at zero time, or possibly at some negative times.  When you press the "mark" button, the first mark is given a real time.  However, Adobe Audition doesn't see them as normal wave file cues (well, it does, but in rather garbled fashion).  However, Wavosaur (tiny free audio editor) does see them.

Meanwhile, the next version of Reaper ($50 very fully featured DAW) can show cues in wave files, and optionally create timeline markers from those cues.

I've found that if I open an R-44 file in Wavosaur, then create new markers alongside the original ones (the work of a moment, and you can place them better than you did at the concert on the fly) and save, Reaper's next version will then see them.  You only need to see them in one file from a multitrack set of course.  And you can get them into Reaper's timeline by shortening the left edge of the file by a tiny amount, so the stack of unused ones doesn't show, then use the command to copy the rest onto the timeline.

Saves having to hunt through a long concert working out where each song started.

Alternative strategy, if you have pre-recording enabled, is to press stop in applause between songs, and immediately press record.  That will create a new set of files with only a few milliseconds of lost material, as the pre-record buffer starts filling as soon as you pressed stop.  So in your DAW you see each song as a separate set of files.  But check at home whether yours is as gapless as mine when used that way.

Meanwhile I've emailed Edirol about the flakiness of their markers and I believe they are looking into it.  Let's hope for a firmware upgrade to fix this issue in due course.

[BayTaynt3d]

FWIW, the markers in the R4 and R4 Pro were the same way, I just don't bother with them at all as I prefer to do my tracking in CDWave (99% of the time I don't stick with my deck the whole show to track in real time, plus you probably have to fine tune them anyway, you don't have the modify the original file, I can save the cue sheet for later reference, and guarenteed no SBEs). But it'd be sweet if the markers showed in my NLE, so don't get me wrong, I'd be stoked if they did.

And by the way, for all you new R44 owners out there, I've touted this before, but I'll say it again, which is that the laptop locking hole is not to be ignored or forgotten. Even if you only lock the deck to your stand, the odds just got that much lower that someone is going to walk away with the rig -- and more so if you can get the deck, string through the stand, and lock it to something. Love that feature. Obviously not needed for every show, but the piece of mind for other shows rules.