Playlist not registering SPP as Slave

[dsetto]

Hello. My Tempest, in Playlist mode, does not seem to be registering SPP (Song Position Pointer) information when it's slaved to Pro Tools 9 (Mac OS 10.6). (It's via 5-pin din, from the Digidesign 003.) My Tempest is brand new. I assume it has OS 1.4 because the manual was labeled as OS 1.4. (I haven't updated the OS from how it shipped.)

Regardless of the PT measure I play from, Tempest pretty much always begins with Playlist step 1; beat 1/measure 1. Once in a while Tempest starts in a different place in its Playlist. (I couldn't replicate this on command, though.)

I've looked over the manual & searched google, this & the other DSI forum to find an answer. From what I can tell, there used to be an SPP issue, but it seems it was addressed.

Any suggestions appreciated.

Thanks,

dsetto

[dsetto]

More info:
Tempest OS: 1.4.0

I tried assessing SPP behavior with Logic Pro 9, but Tempest's playlist didn't correspond with Logic's timeline either. (In Logic, I checked "allow to send SPP while playing" in the midi clock section of Midi Sync Preferences.)

I tried to play around some more to see if I could detect patterns. In Tempest, I took a Demo Project, and made all 16 beats in the playlist 4 measures long. In Pro Tools (& Logic), I put markers every 4 measures, numbered accordingly to the beats.

What works:
If I play from the DAW's measure 1, beat 1 (the beginning of the song), then, the Tempest playlist plays through the beats as expected. I.e., beat 2 begins at measure 5, beat 3 begins at measure 9, etc….

However, if I press stop on the DAW transport, and then I place the DAW locator at the start of what would be the 1st measure of a new Tempest beat (other than beat 1), the expected beat does not play. I also tried placing the DAW locator 2 measures before a new beat is supposed to trigger to allow some room to "find itself".

What was peculiar is:
After stopping the DAW transport from random places, if I placed the DAW locator at measure 3, a bunch of the times, the expected results would occur. Tempest Beat 2 would begin at measure 5. I can't recall if this was consistently reliable, or just more often than not.

Here's the other peculiar, reoccurring thing:
Often, when I placed the DAW transport 2 measures prior to what would correspond to a beat higher up in the playlist, such as beats 5-15, when the DAW transport got to the position that would be the new upcoming beat, Tempest would begin Beat 2 there.

--
Obviously, I should look into updating my Tempest OS to the most current one. I assume it should be the most current Beta that's being assessed. But, something tells me that's not going to make a difference, as according to the forums, SPP matters were addressed earlier on.

--
I think the next troubleshooting step to take is to use a MIDI interface other than the Digidesign 003 I've been using.

Help is surely appreciated.

Thanks.

[dsetto]

I connected a MOTU Express 128. I set it up in Apple's Audio MIDI Setup. I disabled all MIDI connections other than the midi out to Tempest. I selected Tempest receives MIDI Beat Clock.

In Pro Tools, I disabled Loop Play, and I set the Session length to 999 measures. (To make sure Pro Tools isn't looping within the DAW.)

The results were as before.

Then, I initialized the Tempest Project. Loaded up new sounds. Created 16 new beats on pads 1-16. Created a playlist with beats 1-16 as Playlist steps 1-16. Set up markers in a new Pro Tools session corresponding to the playlist steps.

Here's what I've observed:
1. When beginning at PT measure 1, beat 1, Tempest plays through playlist as expected.
2. When beginning 2 measures before steps 1-9, Tempest registers the SPP and plays the expected Beat during the corresponding Tempest Playlist step that is in accordance with the PT Session location.

However, the following behavior is not as expected.
3. When I set the PT locator on the PT timeline that should correspond to 2 measures before the beginning of a new Tempest Bear for beats 10-16, rather than triggering the expected Tempest playlist step/beat, Tempest plays Playlist Step 2 (which is Beat 2), and then, proceeds to subsequent Playlist steps from there.

By the way, this is the first time I use Midi Beat Clock, and attempt to synchronize a DAW and a drum machine. (A long time ago I had experience synchronizing two DAWs via MTC without a hitch.)


When you sync Tempest to a DAW via Midi Beat Clock (without an additional hardware master clock) does SPP work for Tempest Playlist steps over 9?

Any thoughts?

[dsetto]

After trying different Playlist variations I've determined the following.

For DAW, Pro Tools 9 or Logic Pro 9, with MOTU Midi Express 128 sending Midi Beat Clock via 5-pin din to Tempest.

Tempest gets expected DAW's Midi Clock SPP if I begin DAW anywhere before DAW's measure 9. If I play anywhere on or after DAW's measure 9, Tempest consistently begins at Playlist Step 2, and plays from there. But that's not in coordination with DAW's timeline.

This brings to mind that Tempest's max beat length is 8 measures. Yet, I don't know if that's related. I would think I should pursue Pro Tools further, but as the identical behavior happens in Logic with two different MIDI interfaces, I wonder if there's something I am supposed to configure in Tempest.

So, for now, for me, Tempest's Playlist only responds to SPP within the first 8 measures of a DAW's timeline. I was hoping it would respond at least to a song's length.

For now, I'll assume user error, including not having updated the OS from the received 1.4.0.

[dsetto]

I updated Main OS to 1.4.0.15b and the Voice OS to 1.5. I then initialized Project, loaded sounds, made new Beats and tested the Playlist while slaved to a DAW via 5-pin Din.

When in Playlist mode, Tempest doesn't play the correct bar # (or, step) when the DAW is started on or after measure 9.

The same issue occurs when I synced Tempest to a hardware sequencer via MIDI cable.

Other than this SPP-Playlist issue, Tempest's MIDI Clock sync to the DAW is excellent from my short-duration tests.

Has anyone ever been able to have a slaved Tempest's Playlist correctly register a location when starting the master sequencer at measure 9 or greater?

Thank you.

[Steven Morris]

I updated Main OS to 1.4.0.15b and the Voice OS to 1.5. I then initialized Project, loaded sounds, made new Beats and tested the Playlist while slaved to a DAW via 5-pin Din.

When in Playlist mode, Tempest doesn't play the correct bar # (or, step) when the DAW is started on or after measure 9.

The same issue occurs when I synced Tempest to a hardware sequencer via MIDI cable.

Other than this SPP-Playlist issue, Tempest's MIDI Clock sync to the DAW is excellent from my short-duration tests.

Has anyone ever been able to have a slaved Tempest's Playlist correctly register a location when starting the master sequencer at measure 9 or greater?

Thank you.

There were a few instances where I have successfully used the Playlist mode on some of my work in DAWs (I did a cover of Metroid 2's SR-388... I just checked my Tempest video and can confirm I used the playlist mode successfully on that one. That was in April/May of 2015). I always thought it was because I was using Reaper and that I have some quirky things about my projects (like the tracks start on negative bar numbers and things like that). At any rate, this has been one of the reasons why I don't use my Tempest more in what I do. I haven't had time to test the new OS's Playlist mode with any of my newer projects but have been meaning to and am really curious about this. FWIW, I usually use a USB cable and completely different hardware than what you're using.

I'd love to get to the bottom of this. Thank you for taking the time to post all of your troubleshooting notes. I won't have the time to do such in-depth tests for another week or two, but I am actually planning on running some tests. I'll be sure to post my findings in this thread .

I've been meaning to talk to Ace Waters about this (He's a YouTuber who makes great VGM covers and uses the Tempest for his drum tracks). I'll see if he has any insight. He's also a Reaper user.

[bozo]

Well i dont know if SPP working or not, but I do know the Tempest will not start properly in when using playlist and slaved to another clock.
So "song mode" (play list) is useless unless its the master clock.

[dsetto]

Thank you for responding. It's great to read about your experiences.

My findings thus far with the setups I've described above:

For me, Tempest does sync tightly to my two DAW's Midi Beat Clock when I begin the DAW from measure 1, beat 1. I am impressed at the "tightness" of this sync without an external device that syncs to the DAW's audio track and sends out Midi Clock.

-
Steven, you inspired me to check out USB sync. To my delight, Tempest Playlist did properly register Pro Tool 9's Song Position Pointer when I started Pro Tools on or after measure 9, including before.
Unfortunately, by ear, I found the sync between PT & Tempest to be off in the 1ms range when I would begin Pro Tools on or after measure 9.

From what I've read online, I figure the tightest sync between a DAW and Tempest will be with an external device that syncs to a DAW's audio track, and sends off MIDI Clock. Also I assume Tempest can not receive isolate SPP info from USB simultaneously with receiving isolated MIDI Beat Clock from the 5-pin din port. (Which is entirely reasonable.) So, because of both these things, and what I value, proper SPP registering via USB is not for me, if it's not the tightest sync possible between Tempest and a DAW.


 

[dsetto]

I am intrigued to hear about other folks' experiences with Tempest syncing to a DAW via the 5-pin Din port, and if Playlist registers properly when starting the DAW at or after measure 9. Thank you.

[dsetto]

SPP test.

Bottom line: what was not working for me in 1.4.3 re: SPP is now working in 1.4.5.1.


T=Tempest
PT=Pro Tools
SPP=Song Position Pointer
MBC=Midi Beat Clock

With PT 12.4 vanilla on 10.8.5 on early MBP '08.

- it works. For me.

- What works? SPP, in my one-off test.


Tempest Beat and Playlist prep.
- From an initialized project, I made a new sound collection into a beat.
- I recorded 10 different 2-bar patterns. 120bpm, with T as master, not synced to PT.
- I made a playlist where it was simply T Beats 1-10, each 2 bars.

Configuration:
- Then, USB connection. Set tempest as midi clock slave via USB. PT sending MBC.

PT Setup
- In PT I set up 21 markers, 2 bars apart, starting at m1. So, last marker is at m41. (m=measure). I labeled each marker Beat 1 to Beat 10, repeated that, with final marker Beat 1.
- Set T in Playlist mode

T-PT SPP Test
- Dropping PT playback cursor anywhere within these 41 measures led to Tempest playing the correct corresponding location in its playlist. * This was easily confirmed by matching Tempest' lit pad to the labeled Memory Location in PT.
- *This was not working in beta 1.4.3.
- - in other words, starting PT from a PT memory location that read Beat 9, led Tempest to play beat 9.
- This same correct behavior occurred for:
- - A wide, wide variety of different tempos. (I did not try a tempo change)
- - Dropping PT playback cursor both in grid mode (set to bar, beat) and also in slip mode.
- - looping as small as 1/2 note durations (my PT wouldn't loop any smaller)
- - even when looping straddled a Tempest beat change
- - quick succession of dropping PT playback cursor, to try to trip up T.

When I discovered the SPP bug in beta Main 1.4.3, the PT-T correspondence was off for what I believe were 9 or greater measures in PT.


I am thrilled. This is a core function for getting best of both worlds out of Tempest and DAW.



I am curious about:
- how this works in other DAWS
- What things have I not considered, that I would use, that I could test to ensure SPP works for those situations
- Via 5-pin DIN
- T as master. (Requires non-PT DAW)
- Shorter loop range (for non-core use reasons)
- That this continues to work.

[dsetto]

I should have written I "programmed" 10 2-bar Beats at the Tempest, prior to involving Pro Tools. (Not "recorded".)

[John the Savage]

SPP test.

Bottom line: what was not working for me in 1.4.3 re: SPP is now working in 1.4.5.1.

Hi dsetto,

I promise I'm not saying this to be argumentative, but I'm not sure what the purpose of the above tests are, using previous incarnations of the firmware?

The reason why SPP is working for you with Beta OS 1.4.5.1, where it was not otherwise working in all those other versions, is because Beta OS 1.4.5.1 is the first and only operating system wherein this functionality was actually addressed and fixed.

Admittedly, I only skimmed through the above thread, so perhaps I've missed the point.  Anyway...

Cheers!

[dsetto]

Hey, hello.

My intended purpose for posting yesterday’s test was to get confirmation from others that the “SPP greater than 9 bars” issue is resolved. Although I intended to bring closure to the issue started in this thread and characterized below, I now realize my test yesterday was performed via USB. The issue I had was via DIN MIDI. I’ll retest via DIN MIDI. I am doing this publicly because this issue is important to me, and I am hoping to get independent confirmation from others. Although, I posted hastily, I think it’s good both get this going, and to assure it works via USB as well.

 
—

Whoops. I will describe a simplified test that I just performed on my Tempest that conveys the issue I've observed....

Hi dsetto--
Thank you. You've just found a new bug:

If slaving a playlist from MIDI clock received over DIN MIDI and starting at location 9.1.1 or higher, Tempest starts playing from the wrong position. Syncing from MIDI Clock over USB works fine.

Thanks much!
Roger Linn

from the “Tempest BETA OS - Main 1.4.5.1 and Voice 1.5” thread …

Main OS 1.4.4bb
   1   Fixed Bug: SPP messages greater than 9 bars cause unexpected behavior

[John the Savage]

Hey, hello.

My intended purpose for posting yesterday’s test was to get confirmation from others that the “SPP greater than 9 bars” issue is resolved. Although I intended to bring closure to the issue started in this thread and characterized below, I now realize my test yesterday was performed via USB. The issue I had was via DIN MIDI. I’ll retest via DIN MIDI. I am doing this publicly because this issue is important to me, and I am hoping to get independent confirmation from others. Although, I posted hastily, I think it’s good both get this going, and to assure it works via USB as well.

 
—

Whoops. I will describe a simplified test that I just performed on my Tempest that conveys the issue I've observed....

Hi dsetto--
Thank you. You've just found a new bug:

If slaving a playlist from MIDI clock received over DIN MIDI and starting at location 9.1.1 or higher, Tempest starts playing from the wrong position. Syncing from MIDI Clock over USB works fine.

Thanks much!
Roger Linn

from the “Tempest BETA OS - Main 1.4.5.1 and Voice 1.5” thread …

Main OS 1.4.4bb
   1   Fixed Bug: SPP messages greater than 9 bars cause unexpected behavior

Yes, but what I'm saying is, you don't need to test all those previous iterations of the firmware; those tests have already been done and thoroughly documented.  The SPP handling has been broken for years, we know that; indeed it has been broken since the very inception of the Playlist functionality.  So there's no value in going back and testing those older operating systems.  As you've copied from the change-log above, you can see that Beta OS 1.4.4bb addressed the "greater than 9 bars" issue, and Beta OS 1.4.5.1 further addresses two more issues...

Now, chances are there are still issues with this functionality that we've missed; and if you find one, by all means, report it... But if you want to be helpful, it's only the most recent beta that needs to be tested.  Finding and proving bugs in older versions is a moot point, and well, obviously there are bugs in the older operating systems.

Anyway, I'm just trying to spare you a long and tedious process that is otherwise unnecessary.

Cheers!

[dsetto]

The only test I did yesterday was on 1.4.1.5. ... I did the others in July 2016, when I got my Tempest, seeking forum help prior to reporting my results to the “Tempest BETA OS - Main 1.4.5.1 and Voice 1.5” thread.

[John the Savage]

The only test I did yesterday was on 1.4.1.5. ... I did the others in July 2016, when I got my Tempest, seeking forum help prior to reporting my results to the “Tempest BETA OS - Main 1.4.5.1 and Voice 1.5” thread.

Right... Sorry, I didn't remember this thread, and when I skimmed through it yesterday, I only noticed that the last few posts were all from the same day, and didn't realize the others were from months past (sheepish grin).  I'm asleep at the wheel here apparently.  Anyway, as you were (smirk)...

Cheers!