Render in Place Bug

This might have been pointed out before but I noticed today when working with a Rack VST and rendering it in place… if there’s another take of midi, other than than the take you are mixing down… the CC changes such as modulation/expression interfere with the selected take, and the resulted rendered file isn’t correct.


Michael

Render in place never produced exact results in Cubase. Some plugins just don’t render properly. It is better to use the old Export function. Or even record audio directly to another track in real time, to really get what you hear.

hmm - I prefer to know what the bugs are and be aware of them rather than just not use functions. It performs well for me most of the time.

Hi all, i did an experiment on this after seeing it again. what i found wasnt good. :astonished:

http://keyzs.zapto.org/dl/net_picxs/render_ip.jpg

There are timing issues with RIP as the rendered audio start times do not line up. It is also discovered that different VSTi render differently. My suspicion is perhaps something to do with ASIO Guard. In other tests, i lowered ASIO Guard & then disabled it totally and got different results. Hence i left it as that. At the moment, ASIO Guard is set back to NORMAL

Using sonicstate’s export method and place audio into project is working perfectly.


not really sure where do we go from here… perhaps others may wish to chime in. Many thanks. Cheers!!!

Wow, totally agree. I’ve rendered and heard no obvious errors but am now wondering what I should be listening for, timing errors?

I thought that I could trust this function now in 8.5.15

You can trust it in general, for quick production work. Select clips, render section, and you get something out of it. If that something is good enough to keep working on, you keep it. It’s a take. May be different or not, if it sounds good, it is good.
But if you want exactly the same as you were hearing before, or even exactly the same to byte, then it may not be the same. Or it may be, depends.
So, in production phase, it may be just fine, to just bounce some clips down, and work further on them. But I wouldn’t trust it in mixing phase, for creating stems or any kind of sensitive operation.
I mostly noticed problems with automation - some plugins just don’t render properly. I noticed even VST Amp Rack rendered completely wrong long sustained release notes of guitar, creating sustained feedback where there was none in realtime. But for me mostly problems are with automation, it just isn’t rendered correctly sometimes, too often, so take care.

Can you guys post a detailed repro* and put it in the “issues” subforum?

https://www.steinberg.net/forums/viewforum.php?f=230

Sorry, no spare energy here to go through all the issue rules atm, and make report only to be ignored as usually. Maybe somebody else will want to bite into that, surely there are some enthusiasts.
When such errors surface during work, I don’t make notes what settings I had. Reporting a bug is not my priority, not my job or commitment atm, so I don’t attempt to reproduce it and document it. I just find a way arround it and keep working.
If I was one of beta testers, then it would be different. I applied for beta testing or “Steinberg Solvers” program a month ago through that questionaree, and nobody even contacted me, not even a note or thanx. Quite embarrasing, you put yourself out and then silence from the other side, I feel like a fool.
Posting repros by users is an extra endeavour which needs to be recognized from the side of Steinberg. Make me a beta tester, and I will report all bugs properly with full repro and all.

For what its worth, the issue report has been posted however at the moment it is not showing up. Hopefully it will in a couple of hours then.

cheers…

having exactly the same problems.

is the ticket open already ?

Yup… read all about it :wink:

https://www.steinberg.net/forums/viewtopic.php?f=230&t=95710

Thanks for posting this. Looks like the first RIP was quite accurate with True Piano, but the subsequent RIP with True Piano was not?

Was it the same piano sample used in both cases? I could envision where different instruments could have different attacks and give similar results, even with RIP working properly. Or even the same instrument, given any possible variation/randomness in the sample played back.

Maybe a sharp attack instrument, like a clave or drumstick would be helpful to look at …

Thanks so much for posting!

Hi Alexis, not at all… glad to be a part of this.

Actually the first track immediately under the MIDI track was exported and “re-inported” using the Export function.

The actual True Pianos RIP is further down. Do zoom in on the attached pict and you’ll see its actually way off. Yup no change is piano settings. Same MIDI track data all the way just the VSTi change only.

You’ll also notice that Kontakt is also way off.

Not sure if its visible properly, but on Cubase if you zoom in to the MAX MAX, the Korg M1 is very slightly out but this time, it either before or after the count - problem is the shift is like 1 or 2 counts plus / minus.

hope this helps all… cheers!!!

EDIT: I forgot to mention that all RIP and the single EXPORT wave was normalised to -6db just for visual purposes. looking at the image again, i think i left out this process for the Halion RIP wave.

I can not reproduce this, RIP and mix downs produce the same.
No timing issues or velocity issues, they even nullify.

Hi guys… thank you all for participating and testing this out…

i would like to apologise for the methodology being unclear. i had some time to redo this and attached is a screen shot of the results. i will be posting an update to the original issue report as soon as the post comes back online as verified by a moderator. All the updated steps and methods will be posted there too. Will have the link as soon as it is available.


http://keyzs.zapto.org/dl/net_picxs/render_ip2.jpg


hope it helps and many thanks again.

This thread, https://www.steinberg.net/forums/viewtopic.php?f=230&t=75106 , which addresses this RIP issue (with repros and pictures) also said that the normal export audio mixdown was also inaccurate in the same way, adding one sample to the end of the render, based on the bpm.

If you haven’t visited that topic, it may be worth it. Any thoughts? For me, even if true it would not affect how I work I believe (I don’t glue repeats in multiples, e.g.), but still it would be nice to know about.

Thanks -

Hi Alexis, thanks again for the above post. Looks like things may be heating up with this wired RIP issue… it could uncover a bigger problem lurking behind as the normal export audio mix down is also being questioned. At the moment we may not be able to make heads or tails out of this.

Its good to hear that it doesn’t affect your workflow and i think it should not affect mine either.

Personally i use RIP (all dry) to convert MIDI to audio, then disable and hide the original Instrument track. If i need to rework things like velocity, expression or CC changes, i just delete the rendered .WAV track and rework the MIDI again. When all is done, RIP and see if the .WAV fits into the mix… something like that… :slight_smile:

good link and many thanks again. Cheers!!!