ZOOM F8 Ext Auto Sync (via TC)

[Philip Perkins]

(I apologize to JW for starting a new F8 thread, but I have a specific technical question for F8 users and didn't want it buried in the thread with overall impressions and speculations about the recorder's future etc.  If this is not cool I will move this thread or delete and try again there.)

My question is for F8 users--have you used the F8 TC mode in which it uses incoming external TC to control its record (sample) clock?  Have you used this mode over a long take (like a concert set, 1+ hr) and did sync hold with other devices?  There have been other devices (that I've owned) that would do this (MOTU Traveler, Tascam HDP2 among them) and it worked very well, but it isn't something that SD or Zax (or Sax?  Aaton?) recorders do (this is intentional on SD's part at least, they have told me).  So....did this work?  Was the audio glitch free and in sync with whatever?  thx

[EmRR]

On 5/28/2016 at 2:51 PM, Philip Perkins said:

My question is for F8 users--have you used the F8 TC mode in which it uses incoming external TC to control its record (sample) clock?  Have you used this mode over a long take (like a concert set, 1+ hr) and did sync hold with other devices?  There have been other devices (that I've owned) that would do this (MOTU Traveler, Tascam HDP2 among them) and it worked very well, but it isn't something that SD or Zax (or Sax?  Aaton?) recorders do (this is intentional on SD's part at least, they have told me).  So....did this work?  Was the audio glitch free and in sync with whatever?  thx

 

Old post, but I'm wondering the same thing.  I've got an F8n on the way so I'll know soon enough and will file a report.

 

I'll be checking to see if it integrates with a MOTU AVB unit as an aggregate device under macOS when configured for multitrack interface mode with a DAW, and I'm assuming it'll need EXT AUDIO CLOCK SYNC along with LTC feed from the MOTU system.   It'd be a real bonus for my usage list if that works.   

 

I did find one music user who is syncing a pair of F8's for concert remotes, reports no problems, but I don't think he's syncing to video.  He's not tried running as an aggregate interface. 

[IronFilm]

On 5/29/2016 at 6:51 AM, Philip Perkins said:

 

My question is for F8 users--have you used the F8 TC mode in which it uses incoming external TC to control its record (sample) clock? 

It does use it to stay in sync with another F8, that I know for sure, as Zoom has confirmed this multiple times. Don't know about the details of if it would work as well with other devices as well other than another F8. 

[EmRR]

Thanks.  I've seen mentioned several places that MOTU and Zoom both have proprietary methods to derive/extract clock from time code, so I'm hopeful.  It won't be a deal breaker if it's a fail.  

[EmRR]

Unit in hand, have experimented a little.  I can say it works in aggregate device mode for a short run, can't yet speak to a long run yet.  I only had time to check it with 'drift correction' checked on the Zoom under AudioMIDI with it as secondary slave device, MOTU device as master clock, no time code connection, which leads to:

 

 a response from Zoom customer support which said:

Quote

 

This type of functionality (two F8n in 'multitrack interface mode' or an F8n and another coreaudio compliant device) may be possible, but is not officially supported.

Unfortunately, time code input and output are not available in audio interface mode.

 

 

I hadn't run an aggregate device in several operating systems, and didn't recall the 'drift correction' checkbox, which is apparently a renamed 'resample' option.  I had fed a mono source into each device and recorded for a few minutes.  When I look at the waveforms, there's a latency offset difference that is not corrected on the way in, and once aligned visually to waveform peaks, a phase torch display shows the extreme top frequencies waving back and forth a bit which I believe shows evidence of resampling, as implied by the checkbox.  The two files sound correctly mono with no apparent phasing artifacts when remixed together in mono.  Not wild about the idea of resampling, but then the Mac is doing it in the background all the time and it's never obvious/apparent.  

 

I sent a response suggesting it would be useful in a future update to have time code access in interface mode and/or a word clock option.  I don't really know what hurdles exist on that front.  

 

I'll report in more detail about varied test conditions as I can.

 

 

[EmRR]

On 5/28/2016 at 2:51 PM, Philip Perkins said:

My question is for F8 users--have you used the F8 TC mode in which it uses incoming external TC to control its record (sample) clock?  Have you used this mode over a long take (like a concert set, 1+ hr) and did sync hold with other devices?  There have been other devices (that I've owned) that would do this (MOTU Traveler, Tascam HDP2 among them) and it worked very well, but it isn't something that SD or Zax (or Sax?  Aaton?) recorders do (this is intentional on SD's part at least, they have told me).  So....did this work?  Was the audio glitch free and in sync with whatever?  

 

I have run one experimental pass.  F8n chasing time code from a MOTU 16A, and set for 'external audio clock sync'.   Dual mono fed to the 16A and the F8n.  This works pretty well.  I recorded an hour and ten minutes at 88.2kHz using 30ND LTC, set for ‘External Auto Record’ with ‘External Audio Clock Sync’ on.    The F8n started recording when I hit record in the DAW.  I found initial offset appeared to be low, in the order of a few samples, much better than all other sync attempts.   After correcting the initial offset on the imported file in the DAW, by the end of the recording the offset appeared to be 1.6mS, fine for any mono source, probably fine across a typical multitrack recording, but fails the dual mono remix to mono test with the usual phasing artifacts in mono, or side to side shift with files hard panned opposite.  Phase scope looks very weird, with non-linearities all the way down into low frequencies.  Now if this was only available in interface multitrack mode…..

 

Other tests:

 

F8n sending LTC to the MOTU 16A, significant latency difference and drift within minutes.  The whole system had been warmed up several hours.  The imported WAV file from the F8n did drop into the correct initial time spot.

 

Free sync, no connection between devices, pretty much the same as F8n sending LTC after manual lining up of initial waveform.  

 

Aggregate device with drift correction checked, 2nd test, I found a moment where there was an instantaneous shift in phase linearity, so I wouldn't trust an initial correction to maintain for any amount of time.  This happened about 6 minutes into the experiment, then timing maintained for the next 14 minutes.

 

Aggregate device without drift correction checked, significant latency offset and timing drift within minutes.

 

[Mark H]

On 1/21/2019 at 7:11 PM, IronFilm said:

It does use it to stay in sync with another F8, that I know for sure, as Zoom has confirmed this multiple times. Don't know about the details of if it would work as well with other devices as well other than another F8. 

Just wanted to add a comment here – as a new member of the group. I found this group after a bit of Internet searching because I’m having immense problems syncing an F8n and a new F8n pro.

 

Certainly zoom implies that you can synchronise the two units as standalone recorders, but it seems that whatever I try, it doesn’t work (PS just doing audio so no video involved).

 

I’ve tried the continuous timecode out of one F8n into External (audio sync) of the other, and basically it doesn’t work and just drifts – typically it will be 400 samples out after 40 minutes or so, but sometimes it can be 1700 samples out after 10 minutes. I’ve tried different cables, I’ve tried terminating the BNC cable, just nothing works. Timecode is getting between the two units, just any sort of word clock seems to be non-existent.

 

I also wanted to be able to sync the two units together while acting as interfaces to my windows PC – I can get all 16 channels into the PC, and monitor via a different soundcard, but the sets of 8 files from each f8 drift in sync from each other (of course they are in sync within each device’s set of files).

 

I’m open to suggestions, but just about to return the new F8n pro as I really wanted to be able to record more than eight tracks on location occasionally, and it would be great to use them to record up to 16 at home while using them as interfaces.

 

Looking for any suggestions – and also, to the moderator of this site, if this should be a new post, please let me know.

[Tobi A]

Sounds like you need to sync word clock not just TC. Wordclock is on its own connector, not sure if the Zoom has that?

[Mark H]

1 hour ago, Tobi A said:

Sounds like you need to sync word clock not just TC. Wordclock is on its own connector, not sure if the Zoom has that?

I am unsure now too! It doesn’t have a separate bnc wordclock connection, but it does list options for external tc input, which include:

 

- External

- External (audio clock sync)

 

I can use timecode to start recording from one f8n / replicate timecode on the receiving device with / without having “Audio clock sync” selected, but the recordings don’t synchronise any better with “Audio clock sync” selected.

 

These are the specific mentions of this in the manual.

 

So either I’m doing something wrong or I have misunderstood and it will never work. I have owned wordclock devices in the past, so I understand what it is. In the absence of a reply from Zoom, I suspect that this feature does not actually do what I would call Wordclock sync. 😞 

 

[The Documentary Sound Guy]

Audio clock sync does sound like a wordclock sync.  Maybe the issue is that the sending recorder isn't sending wordclock?  Is there an equivalent "audio clock sync" option for the TC output?

[Tobi A]

Look like they call timecode sync audio clock sync in thier manual. 

[Mark H]

19 hours ago, The Documentary Sound Guy said:

Audio clock sync does sound like a wordclock sync.  Maybe the issue is that the sending recorder isn't sending wordclock?  Is there an equivalent "audio clock sync" option for the TC output?

Good suggestion – I don’t think so, but I will double-check.

[The Documentary Sound Guy]

It's worth pointing out that the screenshot of the manual that you quoted doesn't describe a scenario where a Zoom recorder is providing the word clock; it only talks about an "external timecode" source (and it's worth noting that a "timecode" source isn't the same thing as a "word clock" source.

I actually think the terminology in the manual is mixed up; I think it's unlikely the Zoom provides word clock sync (as others have said, that's a separate connector and a different kind of signal than timecode).

I suspect the best you can do is turn on "Ext continuous" in addition to "Audio clock sync", which should at least jam the timecode continuously.  It won't prevent drift within a take, but it will at least keep the timecode updated so that takes start in sync.  This is the way the majority of recorders work these days; word clock sync is actually pretty rare.  It won't prevent the 1,700 sample drift in 10 minutes that you saw, but it's the best you can do without a more specialized recorder.

If you need to do long takes consistently or phase coherence across the recorders is important, you need to use recorders that are designed for word clock sync, not just timecode sync.  I believe the Sound Devices 7-series is still the most commonly available and economic option (even used) ... beyond that only the flagship Sound Devices models support it, plus some of the older Tascam recorders.

[Throwback]

I went down the route of clarifying word clock and the various Zoom F8 models about a year ago: there is much misinformation on-line suggesting that the F8 variations have it (which stems from an over-enthusiastic Zoom rep), but, eventually, Zoom confirmed what is unsurprising, and that the F8 models do not have word clock and, therefore, sample-accurate sync is not possible. I agree that a pair of 788s is a good route to go if you want that option at a relatively affordable price.

 

Cheers,

 

Roland

[The Immoral Mr Teas]

On 12/31/2024 at 6:01 PM, Throwback said:

... much misinformation ... (which stems from an over-enthusiastic Zoom rep),

but, eventually, Zoom confirmed what is unsurprising, and that the F8 models do not have word clock and, therefore, sample-accurate sync is not possible.

 

Yeah, I had trouble understanding how they had backward-compatibly changed timecode to allow their 'stamp' to keep reclocking but its good to know (as an F8 owner) not to bother thinking about it any more!

 

Thanks Roland! (Did you, btw, write up on any of this on your site or was it just personal time spent looking into the issue)?

 

Jez

[Throwback]

1 hour ago, The Immoral Mr Teas said:

 

Yeah, I had trouble understanding how they had backward-compatibly changed timecode to allow their 'stamp' to keep reclocking but its good to know (as an F8 owner) not to bother thinking about it any more!

 

Thanks Roland! (Did you, btw, write up on any of this on your site or was it just personal time spent looking into the issue)?

 

Jez

 

Hi Jez, No I didn't write anything up in detail on my website, though do say 'Amongst other limitations are the lack of word clock as, more latterly, I would like to be able to link recorders for higher channel counts (there is a lot disinformation out there, but, after pursuing this thoroughly, not least with Zoom, the F8/F8n/F8n Pro lacks the ability to provide the clocking for a slave unit and get sample-accurate sync)...'  (https://drbadphil.com/a-few-thoughts-on-field-recorders). It was painful trying to get a coherent answer from Zoom and it took a lot of emails to work through their unnecessary obfuscation of the fact that it has timecode and a reasonably accurate clock and no more. I also came across someone who knew exactly what they were doing and had tested the very evident drift. At that point I bought my first, rather nice and mint, 788!

 

Cheers,

 

Roland

[EmRR]

Somewhere here i confirmed that Zoom had told me there was no word clock option. TC will give you the same start point then all bets are off after that.  The closest thing you’ll get with 2 units as interfaces will be with macOS aggregate device with one unit resampling (drift correction) as it comes in.  Temperature makes a huge difference in comparative clock compatibility.  I free-sync frequently on live music jobs, sometimes there’s very little drift sometimes it’s huge.