Higher sampling rate causing delays to change/sound different? WTF

This forum is for discussing Rack Extensions. Devs are all welcome to show off their goods.
User avatar
Flavolous
Posts: 427
Joined: 17 Aug 2016
Location: ZW | GB

Post 12 Sep 2019

Hey guys, I've had this problem for a while when using Neutron or Spectre. I predominantly work in 44.1 to save on computing power but nowadays I'm exporting my sessions at 96k for mastering. The timing for Neutron as a particle emitter changes when I have it set up as a delay in the send fx. This seems to happen with Spectre as well (but its a different kind of problem resulting from the sampling rate change though). Suddenly I get lots of high frequency resonance dancing around which would not have been there when the session was at 44.1k. What is the cause of this, these may be seperate issues between the devices but figured I might get more understanding of one or the other issue. My workaround for this is to bounce at the strip at 44.1k and then mute then send, but I kind just want to resolve this without having to resort to workarounds now.

Here are some sound examples:


Its basically a kong synthesised drum. I bounced the send channels. The first version of each take is at 44.1k (which is the sound I wanted) the 2nd version of each take is the 96k version (not the result I'm looking for). The first set of sounds are Neutron, the 2nd set of sounds are Spectre

User avatar
antic604
Posts: 2239
Joined: 23 Oct 2017
Location: Warsaw, Poland

Post 13 Sep 2019

It's not a Reason phenomenon exclusively. For example my tracks in Ableton or Bitwig also sounded different at different sampling rates, because there was a lot of modultion and other time-based effects that somehow got unsynced ever so slightly, but it made an audible difference.

I'm not saying this is what should be happening, but some devices probably aren't coded and tested enough and such things pop up.

And sometimes perhaps it just can't be avoided altogether? If for example Neutron's or Spectre's controls work in number of "samples" rather than "miliseconds" or "bars", then higher sampling rate will obviously have impact. And that would definitely make sense for a spectral FX like those. Again, not looking for excuses, just trying to come up with explanation.
Bitwig 3 // Reason 10 // Cubase Pro 10 @ Lenovo Y730 15'' i7-8750h
my music - https://soundcloud.com/antic604
other stuff - "Psytrance in Bitwig Studio" tutorial

TritoneAddiction
Posts: 1826
Joined: 29 Aug 2015
Location: Sweden

Post 13 Sep 2019

In my experience most synths (and samples) sound a bit different when I switch between sample rates. I don't know the science behind it. I just know what I hear.
To avoid any problems I always work in the same sample rate as when I bounce the track.

I know some people say we shouldn't be able to hear any difference. That may be true once the song is bounced and you convert the bounced track to another sample rate.
But in Reason, when I'm working on a track there can be a significant difference between how the sample rates sound. Again I have no idea how these things work. All I know is, on my computer, when I'm in Reason I hear a difference.

RealReasonHead
Posts: 25
Joined: 21 Jun 2019

Post 13 Sep 2019

Some devices are buggy in my experience but any non-linear processing that introduces aliasing will sound different at higher sample rates. But the two examples you've provided seem to be of the buggy type.

With that being said, there's no reason to export at a different sample rate than the one you've worked at. If the mastering engineer operates at a higher sample rate for fidelity reasons (debatable, in my opinion, but fair enough in the mastering stage) he can always upsample the 44.1 file to the sample rate he wants with basically no downside.

User avatar
Flavolous
Posts: 427
Joined: 17 Aug 2016
Location: ZW | GB

Post Yesterday

RealReasonHead wrote:
13 Sep 2019
no reason to export at a different sample rate than the one you've worked at
Definite noticeable quality difference. The high end in particular, the depth of the sound. Some people cant hear it at all but when exporting a mixdown to master it will be 96khz as long as I can help it. During production it doesn't matter, yes!

Anyway thank you so much for the input! May have to do a bit of reading up on audio engineering and coding but will flag this with the two developers!

RealReasonHead
Posts: 25
Joined: 21 Jun 2019

Post Yesterday

Flavolous wrote:
Yesterday
Definite noticeable quality difference. The high end in particular, the depth of the sound. Some people cant hear it at all but when exporting a mixdown to master it will be 96khz as long as I can help it. During production it doesn't matter, yes!
This is probably due to distortion plugins or badly programmed plugins. Aliasing is less of an issue with higher sample rates but I don't see the reason why one would mix at a lower sample rate and adjust the mix to sound good and then switch to a higher sample rate to have the sound being altered again. I've noticed a very very noticeable difference when using Scream 4 but when mixing I adjust to it. The actual objective difference in quality (not just a subjective difference in sound/timbre that you would adjust to if you had mixed at a different sample rate to begin with) I think is not that big and even negligible if you're not using a huge amount of distortion.

User avatar
selig
Moderator
Posts: 7908
Joined: 15 Jan 2015

Post Today

Hey Flave, As others mentioned, most folks work at the target sample rate to avoid surprises.
IF you want a brighter mix with more depth, work it until you love it. Any subtle gains in export can be offset by any negatives - not all "changes" are for the better…
Also as other said, any advantages of mastering at higher rates can be achieved by upsampling before mastering, something a lot of MEs already do.

Best solution if you cannot mix at the target sample rate is to export at both the working sample rate and the higher rate, and trust your ME to choose the best option to get the desired results. :)
Selig Audio, LLC

  • Information
  • Who is online

    Users browsing this forum: No registered users and 0 guests