Sample rate problems Studio one

[lawrenceyoungenid]

Hello everyone. I’ve completely torn my hair out. This problem is so frustrating that this is my first time ever asking for help and advice on a message board. Here’s what I’m using. Studio one 5, an Avid Eleven Rack as my recording interface, this is connected to my recording computer which is an hp z602 with 96 gigs of ram and 2 e5 Xeon processors. This is connected to a waves sound grid server. Somehow when I was recording a song, Studio one was set to 48k sample rate but my interface and the sound grid weren’t. I believe they were at 44. So whenever I tried to mix the song down it sounded like the chipmunks. I recognized it was probably a sample rate issue and I tried to change it in studio one. Now it sounds like the chipmunks in studio one also. I found I can get it to play back at the right tempo and pitch if I open studio one And the interface and sound grid in 48k And then change the interface and sound grid to 44 but leave studio one alone. This is fine until I try to mixdowns the song at which time studio one recognizes the mismatch and changes it back to chipmunks. What’s the fix? I’m going crazy.

[lawrenceyoungenid]

Ok I found out to fix it in case any one ever has this problem. I have a friend who has Adobe Audition and he opened the files in it and went to edit - interpret sample rate - and he put in 440000 as the rate. This fixed it. Now I can sleep lol

[CS70]

No studio one expert but you could just resample it when importing?

I mean, why insist on using a track at a different sample rate than the rest? It’s a recipe for headaches.

Note that software resampling can be very good, but not all DAW implementations are equally well made.

Check for example https://src.infinitewave.ca/ to see what’s what with yours.

[James Perrett]

Ok I found out to fix it in case any one ever has this problem. I have a friend who has Adobe Audition and he opened the files in it and went to edit - interpret sample rate - and he put in 440000 as the rate. This fixed it. Now I can sleep lol

I've had to do this with certain files too. Usually it is down to the interface being set to one sample rate when the track is enabled for record and then the sample rate changes before the record button is actually pressed. So the file is recorded at the new sample rate but the header still contains the old sample rate.

Even worse - I have a few DAT tapes where the sample rate changes in the middle of the tape. Computer files can't handle sample rate changes part way through the file.

[Hugh Robjohns]

Yes, it's a problem with the file descriptor -- claiming to be at a sample rate it's not actually recorded at. There are utilities around that can help you re-write the file headers with the relevant information -- as you've found. It shouldn't happen, of course... but it's not hard to accidentally mis-configure a system that can then result in mis-flagged files!

[manwilde]

Hello. This has happened to me this week, and I´ve solved the problem using the Adobe Audition trick, so thank you very much indeed for pointing that... but now the files don´t align to the click they were recorded to. They sound fine by themselves, but I can´t for the life of me find a spot were they lock to the grid all along the full song, and the beats land on "in-beteween" measures ("17.2.263", for instance).

I know the band played in tempo (I wasn´t present at the recording, but I´m mixing it) because on the other three songs for that session there are not important discrepancies in perfomance vs. click track, and these are accomplished jazz musicians playing pop-songwriter tunes.

The band is planning to overdub more stuff over this tracks, also playing to a click, so any idea on how to fix this issue would be greatly appreciated!.

[CS70]

If you forget the grid (and the original tempo) for a moment, are the files they still musical? Do they align to each other? Or do you hear timing problems?

[manwilde]

Yes, aside from the grid issue, all sounds fine...

[manwilde]

Near the end of the song tempo is off by 1.6 BPM according to Reaper...

[CS70]

Near the end of the song tempo is off by 1.6 BPM according to Reaper...

Hehe that means little without knowing the tempo and time signature of the song

But regardless: don't know Reaper, but most DAWs have the sample rate as a project property - so is it correct there?

[manwilde]

Yes, all settings are correct, what surely happened is what Hugh described: the tracks were armed for recording prior changing the sample rate of the session, so the headers for the wav files were saying "ok, this are 44.100Khz files" when actually the recording was made at 48k.
The three subsequent songs didn´t suffer this issue, as the correct sample rate (48k) was already set before opening each session... Thanks for the suggestion anyway!.

Regarding the timing issue, I would also think it´s a performance related thing, but my colleague swears that when the recording took place everyone was right on the click, just as on the other three songs recorded after that one. And when you listen to those three over the click, they don´t drift that much. I mean, it´s not quantised music, but they are on the click, whereas on the first one you can really tell the mistmatch already from the second verse on, so much so that they surely would have stopped the take on the spot and do another one... Plus, they all go in time with each other. Just not with the click.

I´m lost here, I don´t know what to do. Something to do with clocking, maybe?. I don´t know...

[manwilde]

Oh, I see now that I misunderstood your comment about the song signature and tempo. Sorry, it´s been a long day. The song is 70bpm and was recorded to a 4/4 click track. Whatever the bpm offset might be, the music doesn´t land on fixed spots of the grid, regardless of any grid subdivision (and I´ve tried for quite a long time...)

[CS70]

No problem.

How long is the track (in minutes, secs and msecs) when it plays ok (but off grid)?

[manwilde]

Ending tail aside, last beat lands at 4:40:05...

[manwilde]

-sorry, it´s 4:36.29, I had moved the clip from bar 0...

[CS70]

Hm... I had an idea, but with that length, if the idea was right I would expect the beat at the end to be off around 2.4, not 1.6 as you wrote.. so obviously my idea is not right.

[manwilde]

Tomorrow I´m going to the studio and mess around with the original session, which sounds pitched up because of this wrong sampling header problem. I took it home on a pen drive the day after the recording to fix it and do a basic mix. Maybe using the same machine where it was recorded something will behave differently. Unfortunately I don´t have Adobe Audition available at the studio, but anyway...
Thanks a lot for trying to help!.

[manwilde]

I´ve found an online song length calculator based on bpm and signature, but I didn´t get to any new conclusion... Also, finding tempo by sections using both Reaper and Audition gives different values depending on the section: 68.2 bpm, 67.8, 69... no wonder I can´t find a match on the grid!.
I also considered the chance of the musicians listening to the click with added latency, but buffer size was set at 128 samples, that gives something like 6 ms. roundtrip, not much, and was set the same for all the other songs were there was no problem with timing...
Thanks again!

[James Perrett]

It might be worth checking your clocking arrangements - I've had similar issues with a clocking loop where everything was trying to sync to everything else. The clock was roughly right but not quite at the right frequency.

[manwilde]

Back from the studio. Something was wrong with the clocking, it seems. Don´t wanna bore you with the details, also because Sonaworks Systemwide was involved in the mess, if not directly causing the trouble at least complicating it further.
I´ve reverted to the previous clocking arrangement that had worked with no fuss for over a year and uninstalled Sonarworks for now, since we´re rearranging the furniture and the monitors´ position and a new calibration profile must be created.

I´ll be watching the whole thing attentively until I make sure everything works as expected. We had changed the clocking setup two weeks ago and we did some tests and a paid recording session four days before this one, and all was well, so I still don´t know what went wrong...

Thanks OP, CS70, James, Hugh and everyone, really.

[manwilde]

Oh, by the way, the customer said the music sounded fine to him and he could lay down tracks over it without using a click. Phew!

[Hugh Robjohns]

Intriguing!

Clocking problems can be very frustrating and not always obvious.

It's easy if things just stop working and it all goes quiet, but some loop problems can cause things to speed up or slow down and, because the whole system does that it's not at all apparent until the file is played on an independent system.

So, it really is worthwhile drawing out the whole system and the digital imterconnections, and which device is the master clock. You can then identify how everything else is synchronised to it, and what switch settings or software configurations are required... And then you can make sure everything is set appropriately. It's all too easy to overlook a setting somewhere!

I also find it helpful, when in doubt, to use a 3kHz tone via an analogue input and record and play it back through the system. Clocking problems tend to be revealed as audible clicks. Regular clicks suggest something is free-running and not properly synchronised. Random clicks suggests an interconnection is inadequate in some way.

[manwilde]

Thanks for the tip, Hugh, will try that next day I´m in the studio.

[Eddy Deegan]

Was the original session recorded at the same tempo all the way through or were there BPM changes at certain points? If the latter, that's something that would definitely need to be accounted for when trying to match the audio to a new timing grid.

[manwilde]

Of course, but no, it was 70bpm all the way through. Now, don't ask me about the time signature, it's nothing esotheric but I'm still trying to figure it out. Anyway, as I said, they recorded to a 4/4 click...