Propellerhead Software

Go Back   Propellerhead Forum > Feature Suggestion Forum

 
 
Thread Tools Display Modes
  #1  
Old 2010-12-21, 08:35
eXode's Avatar
eXode eXode is online now
 
Join Date: Dec 2003
Posts: 3,486
Exclamation Oversampling in Reason devices



Incorporate an oversampling option to the Reason synths and devices. This would be the smartest way to improve the existing synths and devices withouth having to break backwards compatibility.

In my mind, the oversampling option could be added to the "Mix Channel" device that gets added with the devices in the rack, that way you could set the entire chain that utilizes that particular mix channel to X amount of oversampling. We could have the following oversampling options: Normal, 2x, 4x and 8x (maybe even up to 16x?).

This would also leave people with less powerful systems the freedom to set Reason to standard quality but then render their output in high quality. Win Win for everyone!

This is an issue that we have only recently started to touch upon. I have made various requests in the past on how to improve the quality of Reasons synths, especially the older ones (Subtractor and Malström). I want to suggest a realtively simple but effecient way to update all the synths and devices for this age and day.

Clarification:

It struck me that not everone might now what oversampling is.

Well, I'll be happy to explain (what little I know anyways). Oversampling basically means that you render the audio in i.e. a synth at X times higher than the actual playback rate. This is a trick that have been used for years in various software and VA synths as means to get rid of or atleast partially remove unwanted artifacts such as aliasing and to increase overall fidelity.

In computer games a similar technique known as full screen anti aliasing, or FSAA for short is quite common. It means that you render the produced image in X times the screen resolution and then downsample it back to the actual resolution as means to remove jagged edges in a 3D environment.

Last edited by eXode; 2011-09-29 at 20:27.
  #2  
Old 2010-12-21, 10:58
WeaponX323's Avatar
WeaponX323 WeaponX323 is offline
 
Join Date: Nov 2004
Posts: 265
+1

I've lately been running my projects at higher sample rates to "force" oversampling. It yields an improved upper frequency response when processing sounds, to my ears. Eats a lot of CPU this way, though. Efficient oversampling would definitely be appreciated by me.
  #3  
Old 2010-12-21, 12:58
tiftof's Avatar
tiftof tiftof is offline
 
Join Date: Sep 2007
Posts: 2,649
+1000
Running a project on higher sample rates really kills my computer. Still it's pretty cool how it really opens up a mix.
  #4  
Old 2010-12-21, 13:55
Anomecron's Avatar
Anomecron Anomecron is offline
 
Join Date: Jan 2004
Posts: 2,904
I with you on this one - all the way Exode.
  #5  
Old 2010-12-23, 10:42
propsiman propsiman is offline
 
Join Date: Oct 2007
Posts: 107
+1
good idea
  #6  
Old 2011-06-20, 15:57
eXode's Avatar
eXode eXode is online now
 
Join Date: Dec 2003
Posts: 3,486
This one needs a bump! For me this is on my personal Top 5 list together with MIDI-Out of most important features.
  #7  
Old 2011-06-20, 18:55
selig's Avatar
selig selig is online now
 
Join Date: Jun 2003
Posts: 10,505
Quote:
Originally Posted by tiftof View Post
+1000
Running a project on higher sample rates really kills my computer. Still it's pretty cool how it really opens up a mix.
Which is why, if they do adopt this feature, it should be at the 'device by device' level. If you did the whole song file (or all the devices) you'd have worse problems than just raising the sample rate (because you'll also be doing the 'real-time' conversion back down, which is not know for it's quality, speed, or CPU efficiency). :-)

In the mean time, only switch to the higher sample rate when rendering either the entire file or just your synth tracks - then your CPU won't have to actually try to PLAY the file and you can render files as large as needed. :-)
__________________
Giles Reaves, aka 'selig'
Audio Illusionist, Musical Technologist
Selig Audio, LLC
USA

Last edited by selig; 2011-06-20 at 18:58.
  #8  
Old 2011-06-20, 19:37
jkheal jkheal is offline
 
Join Date: Jul 2008
Posts: 1,576
I'm proabably going a little off-topic here, but I've always wondered why it is recommended to work at a higher bit depth and sample rate (i.e. 24bit, 48kHz), and yet all of the audio samples that come with Reason are 16bit, 44.1kHz.

Am I wrong, or do the audio files that come with Reason have to be resampled (and redithered) if you're working at a higher resolution? And then if your final product is 16/44.1, do they not have to be resampled and redithered again. Does this introduce degredation in qualtiy?
__________________
Jon Heal
---
Reason 7.0.1 • A computer • A few boxes that do something or other • Virtually no talent and/or skill

Ye Olde FishCAM
  #9  
Old 2011-06-20, 20:44
selig's Avatar
selig selig is online now
 
Join Date: Jun 2003
Posts: 10,505
Quote:
Originally Posted by jkheal View Post
I'm proabably going a little off-topic here, but I've always wondered why it is recommended to work at a higher bit depth and sample rate (i.e. 24bit, 48kHz), and yet all of the audio samples that come with Reason are 16bit, 44.1kHz.

Am I wrong, or do the audio files that come with Reason have to be resampled (and redithered) if you're working at a higher resolution? And then if your final product is 16/44.1, do they not have to be resampled and redithered again. Does this introduce degredation in qualtiy?
ALL audio must be dithered when doing any changes - the R/R environment is 32 bit float (64 bit summing in Record), so the minute you even change the level by 0.1 dB you are creating a 'bigger' number, bit depth wise. The only condition where you will begin and end with a 'clean' 16 bit sample is if you load it in an audio track and export it without doing anything (but what would be the point in that?).

As to sample rate, consider this: 44.1 kHz only applies if you play the sample at it's original pitch - shift the pitch (to cover different pitch ranges), pitch bend it, add virbrato, etc, and you ARE changing the sample rate! So even a 44.1 kHz sample played in a song with other 44.1 kHz audio will only be "44.1 kHz" if you do nothing to the sample. This is called "interpolation" and is available as a 'high quality' option on the NN-XT. Not all samplers interpolate pitch equally…

This is similar to how a 16 bit sample will only stay 16 bit if you do NOTHING to it. And it's highly unlikely to not to anything to a sample!
__________________
Giles Reaves, aka 'selig'
Audio Illusionist, Musical Technologist
Selig Audio, LLC
USA
  #10  
Old 2011-06-20, 21:01
jkheal jkheal is offline
 
Join Date: Jul 2008
Posts: 1,576
Thanks for the tutorial, Giles. Very helpful. So then I guess if you pitched a sample down 50%, Reason would have to interpolate and redither from a 20.5kHz sample, yes? You'd lose a lot of detail, I reckon.

I think I'll only pitch up, or at least severly limit pitching down ... unless I'm trying to make something sound weird.
__________________
Jon Heal
---
Reason 7.0.1 • A computer • A few boxes that do something or other • Virtually no talent and/or skill

Ye Olde FishCAM
 

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
MIDI Implementation Charts andrthad Phead User Forum (read only) 3 2008-12-08 17:42
using reason devices to process other signals DrumAddic Phead User Forum (read only) 3 2006-06-15 22:40
Strange Pops in My NN-19. DJKS Phead User Forum (read only) 38 2004-09-19 17:30
Reason / DP / OMS Test Results Within... arothman General Forum (read only) 2 2003-03-05 03:26


All times are GMT +2. The time now is 20:43.