Roland Cloud

[Gamma-UT]

Is this an exact 5080 emulation (exact meaning bit accurate)? The hardware 5080 can output SPDIF and Adat and its base waveforms and SRX expansion cards are 44K

"I like the analogue warmth of oldskool sample-rate conversion."

[EvilDragon]

I doubt it, probably FX come from Fantom just like in JV-1080 plugin, so that would be a point of departure.

"I like the analogue warmth of oldskool sample-rate conversion."

Fail! If you use digital output, there's no SRC happening

[Gamma-UT]

Fail! If you use digital output, there's no SRC happening

With 48kHz outputs from a 44kHz signal chain? What wizardry is this?

[outerspacecat]

One thing that makes the XV-5080 perform much better than the JV-1080 is MIDI timing. I guess that Roland just somehow ports the internal OS of the hardware into the software world, because the timing of the virtual JV-1080 or the SoundCanvas is as bad as with the hardware.

Hm, I see nothing weird with timing to received MIDI of either JV-1080 or SC plugins?

EDIT: Those heavy drift seems to be in Logic Pro X only
The timing of the SoundCanvas is that bad, that you don't even have to zoom in.
Attached a simple 2 bar MIDI region (green) with 4-to-the-floor kick drum (Standard 1 default kit) at 100 BPM rendered to audio (orange), hard cut at the beats (no zero crossings or relative grid games going one). 25 ms drift. Nothing more to say. JV-1080 isn't hat horrible of course but together with the Korg legacy sample based virtual instruments pretty much on the loose end. Not that bad wit simple sounds, drums as extreme example.
I wish that they would eliminate these old hardware MIDI or processing flaws in the software version.

[outerspacecat]

One thing that makes the XV-5080 perform much better than the JV-1080 is MIDI timing. I guess that Roland just somehow ports the internal OS of the hardware into the software world, because the timing of the virtual JV-1080 or the SoundCanvas is as bad as with the hardware. For tight drums completely unsuitable by today's standards. The XV-5080 timing doesn't drift that much that's why I think the XV-5080 has also its right to exist beside the JV-1080 (nostalgia, anyway).

Never had anything like that in any of their plugins. It's probably something in your daw.

No, this has nothing to do with the DAW. See extreme example in my post above. Same is true for the Korg M1 plug-in for example.

[Funkybot's Evil Twin]

No, this has nothing to do with the DAW. See extreme example in my post above. Same is true for the Korg M1 plug-in for example.

I don't claim to have the right answer here or be more knowledgeable on this, but how on Earth could it be early in some instances like you're seeing unless it had something to do with how the plugin interacts with the DAW? Like they're getting MIDI signals from the DAW, playing them back out of time, then telling the DAW "hey, this is late a few ms so shift this back to compensate for the latency." Like they went out of their way to simulate jitter but then went a step further and tried to compensate for it. I dunno...it's all very odd and not making sense to me based on your screen capture.

[outerspacecat]

No, this has nothing to do with the DAW. See extreme example in my post above. Same is true for the Korg M1 plug-in for example.

I don't claim to have the right answer here or be more knowledgeable on this, but how on Earth could it be early in some instances like you're seeing unless it had something to do with how the plugin interacts with the DAW? Like they're getting MIDI signals from the DAW, playing them back out of time, then telling the DAW "hey, this is late a few ms so shift this back to compensate for the latency." Like they went out of their way to simulate jitter but then went a step further and tried to compensate for it. I dunno...it's all very odd and not making sense to me based on your screen capture.

Yes, probably something like that. Internal buffering (this is true for hardware also), transfer to the DAW, PDC. There are a few plug-ins with that kind of behavior, not that bad of course, but any ms drift is pretty annoying at least for my understanding of metric accuracy.
viewtopic.php?f=98&t=535951&p=7598672#p7598672
viewtopic.php?f=1&t=467127&p=7600329#p7600329
To be honest, I wonder if anyone in here ever uses all this stuff for real world production as obviously nobody notices or cares about a 17 ms shift in Avenger 1.5 for example (this is obviously a bug and is getting fixed). FIXED IN 1.5.4!
And to anticipate it again, no, it's not my computer or my DAW. I have this behavior (at least on different macOS versions) in practically all major DAWs.
I would be happy if more users would complain to the respective companies (here Roland Cloud) or developers, so that there are improvements in this regard. The possibility to change the buffer size (Synthmaster) or latency compensation (Avenger) is exactly what should be integrated in such plug-ins.

EDIT: Avenger latency report to DAW fixed in 1.5.4

[sinemotor]

I also have this MIDI timing issue. The Sound Canvas was the worst. I couldn't have a normal, perfectly on the grid sound with it. I also have this problem with the D-50 but only when I already have many tracks in the project. Maybe it is related to the Audio Unit with MacOS ?

[outerspacecat]

I also have this MIDI timing issue. The Sound Canvas was the worst. I couldn't have a normal, perfectly on the grid sound with it. Maybe it is related to the Audio Unit with MacOS ?

No, that's concept-related and has nothing to do with the DAW, no need for investigation here. Just ask the support for improvement in this regard.
The hardware of the SoundCanvas was really bad as well (hence my thought that they somehow compiled the hardware OS to the plug-in world), especially if you played 16 MIDI tracks at once. No zoom function needed (and available) in the 90s, your ears were enough to hear that loose timing.

[PAK]

Whuuuut? The timing of the SoundCanvas is that bad, that you don't even have to zoom in.
Attached a simple 2 bar MIDI region (green) with 4-to-the-floor kick drum (Standard 1 default kit) at 100 BPM rendered to audio (orange), hard cut at the beats (no zero crossings or relative grid games going one). 25 ms drift. Nothing more to say. JV-1080 isn't hat horrible of course but together with the Korg legacy sample based virtual instruments pretty much on the loose end. Not that bad wit simple sounds, drums as extreme example.

I don't use it, but checked Sound Canvas (same BPM and kit you used) and Cubase (with PDC) was a consistent 3ms (dropped to 2.5ms @ 96kHz).

[outerspacecat]

Whuuuut? The timing of the SoundCanvas is that bad, that you don't even have to zoom in.
Attached a simple 2 bar MIDI region (green) with 4-to-the-floor kick drum (Standard 1 default kit) at 100 BPM rendered to audio (orange), hard cut at the beats (no zero crossings or relative grid games going one). 25 ms drift. Nothing more to say. JV-1080 isn't hat horrible of course but together with the Korg legacy sample based virtual instruments pretty much on the loose end. Not that bad wit simple sounds, drums as extreme example.

I don't use it, but checked Sound Canvas (same BPM and kit you used) and Cubase (with PDC) was a consistent 3ms (dropped to 2.5ms @ 96kHz).

Did you test with an instrument part or a drum part as well? I would highly be interested what your numbers are triggering a simple kick drum. Also don't use 120 BPM (there is a possible alignment depending on the ms). Cubase, I guess your on PC then?

[jksound]

very bad working roland cloud manager application

[PAK]

Did you test with an instrument part or a drum part as well? I would highly be interested what your numbers are triggering a simple kick drum.

That was the numbers triggering the kick on the kit you mentioned at 100bpm.

Cubase, I guess your on PC then?

Yep, didn't check on Mac OS.. would have to install Sound Canvas there though

[outerspacecat]

Did you test with an instrument part or a drum part as well? I would highly be interested what your numbers are triggering a simple kick drum.

That was the numbers triggering the kick on the kit you mentioned at 100bpm.

Cubase, I guess your on PC then?

Yep, didn't check on Mac OS.. would have to install Sound Canvas there though

Thanks for the effort. I hope it isn't a macOS related cause in the end. But as mentioned other plug-ins behave similar (VST2, VST3, AU even AAX) and I'm not the only one recognizing these issues. Korg legacy would be the best example. However, I'm going to check the same settings in Cubase again, right now (macOS).

[outerspacecat]

Did you test with an instrument part or a drum part as well? I would highly be interested what your numbers are triggering a simple kick drum.

That was the numbers triggering the kick on the kit you mentioned at 100bpm.

Cubase, I guess your on PC then?

Yep, didn't check on Mac OS.. would have to install Sound Canvas there though

Thanks for the effort. I hope it isn't a macOS related cause in the end. But as mentioned other plug-ins behave similar (VST2, VST3, AU even AAX) and I'm not the only one recognizing these issues. Korg legacy would be the best example. However, I'm going to check the same settings in Cubase again, right now (macOS).

Yes, indeed significantly better in Cubase 10.5 as VST3 instrument. 3 ms is much better than 25 ms. However the problem with "too early rendered notes" is still there which should be improved. I also can't get why there has to be a drift for such a simple plug-in as the sound canvas at all.