I’m using WaveLab Elements 9 only for converting my masters (before or after mastering), - from 48k to 44k. -
What’s the best (highest quality) method - to do this specific conversion?
Thanks
.
(How come no one is responding? Is this a stupid question or too complicated?)
.
It’s not a stupid question but it can be complicated and subjective.
Before WaveLab 9, sample rate conversion in WaveLab was handled by the Crystal Resampler which was not exactly known for the best sound, especially as better options came to be.
With WaveLab 9, the Crystal Resampler was replaced by the SoX Resampler. You can see a extensive comparison of sample rate conversions here:
http://src.infinitewave.ca
Many people use the SoX Resampler built in the WaveLab Master Section and are OK with the sound.
I prefer to use Weiss Saracon for sample rate conversion needs. I don’t use WaveLab. I use the excellent WaveLab feature of “Custom Montage Copy” to recreate a montage at a new sample rate though after I to the sample rate conversion in another app.
I generally speaking do all my processing at 96k. When I need to go down to 44.1k or 48k, I sample rate convert to the desired sample rate using Saracon, and then recreate the montage at the desired sample rate with all the markers/data/etc retained thanks to “Custom Montage Copy” found in WaveLab 9.
The only processing that is done after the sample rate conversion is the dithering of course. I load dither plugin in the montage master section to dither to either 24 or 16-bit as needed, after all other processing is done.
It sounds like a lot of extra work but it’s not really because of the ease of “Custom Montage Copy” and the shortcuts I’ve set up.
Other popular 3rd party sample rate conversion is iZotope RX, or Myriad.
In WaveLab Pro, you get some options of “Standard, High, Best” etc. but it seems that in WaveLab elements, you can only choose the new sample rate but not the resampling quality. I don’t use Elements often enough to know more about this though.
“NOTE
Using the Resampler plug-in in the quality mode High to change the sample rate results in the same quality as when using the Resample option in the Audio Editor. However, that is only the case if the sample rate in the Sample Rate dialog exists in the values of the Resampler Sample Rate pop-up menu. If you choose a custom sample rate, another algorithm is used, which results in a lower quality of what the Resampler can achieve.”
This appears in the operation manual of both Pro 9 & Elements 9.
The “note” is clear, (there is no need to repeat it).
My understanding is that the SoX algorithm is now used also in the Resample option in the Audio Editor (in the Sample Rates which are mentioned in the “note”, - no need to repeat that).
By the way, - Justin, - since you mentioned “Best”, - Is there better than “High” in Resampler on Pro 9? - Are you sure? - (Because in Elements 9 there isn’t).
Can anyone confirm if “Resample” (Audio Editor) is of identical quality in Pro 9 & Elements 9?
(Without repeating the “note”).
Thanks
.
Attached is a picture of the Resampler options in WaveLab Pro 9. I’m not sure what is used or possible for Elements.
+1 for Weiss Saracon
I use R8Brain (from Voxengo) - still hard to beat.
Paul
Thanks for the picture Justin. In Elements 9 there’s only “Standard” & “High”.
(I was quite shocked. I wasn’t aware of the fact that they put similar items - but in different qualities. - Maybe I just missed something).
.
I’ve made the following 48k to 44k conversions of the same source:
WaveLab Elements 9 - Audio Editor Resample
WaveLab Pro 9 - Plugin Resampler (“High”)
WaveLab Pro 9 - Plugin Resampler (“Very High”)
WaveLab Pro 9 - Plugin Resampler (“Best”)
iZotope RX 5 “CD” (company preset)
iZotope RX 5 “Steep” (own preset)
Then I compared with phase reverse, - and here are the results:
Those two are absolutely identical:
WaveLab Elements 9 - Audio Editor Resample
WaveLab Pro 9 - Plugin Resampler (“High”)
All the rest resulted in a difference between each other - of a white-noise which measures -80db.
.
Just wondering… do you sometimes need limiting or level adjustment since sample rate conversion may sometimes affect the level?
I set my output ceiling to -1dBFS and use a good true peak limiter like the new Limiter 6 by Tokyo Dawn to help minimize this. So, yes, the peak levels do increase but I leave enough headroom for it to not be an issue after sample rate conversions, as well as after most mp3/lossy encodings.
If I needed to adjust the level down overall it would be no problem though because I haven’t reduced the bit-depth yet at this point. I could just adjust the level (or add a final limiter again) before the final dither.
As much as I love WaveLab, I probably wouldn’t use Elements just for the sake of sample rate conversion and nothing else. If you need to do other work in WaveLab, having built-in SRC for some users is OK but Elements is a bit weak as a stand alone SRC.
If the only reason for using Elements is SRC, then I would probably look elsewhere, especially for “mastering grade” SRC.
I think even Foobar which seems to be free has SoX resample in it, Voxengo r8 Brain is said to be good as well, also free and Windows. For Mac, Myriad, RX, and of course Saracon (not cheap) are also good options. RX is of course Mac/PC.
Thanks for the clarification.
Thanks again for writing, Justin.
In my opinion, nowadays, they appear to be all more or less the same.
(Indeed, there were times that they weren’t).
.
In the meantime, I’ve made another comparison (with phase reverse), - of conversions from 48k to 44k.
The following two are also NOT identical, and show a difference of a very low level white-noise (-60 to -80db);
WaveLab Elements 9 - Audio Editor Resample
WaveLab Pro 9 - Audio Editor Resample
.
I find it a bit bizarre.
.
I’d still like to see a single-line text box command line entry for the complete available SRC parameters in preferences in Pro. If not a simple interface to do the same.
It’s because Pro uses Best for the resample function, and Elements uses High for the resample function, because Elements doesn’t have Best. Maybe it’s a little odd, but I don’t think Pro should be changed to use High for the resample function just to make them the same.
Makes sense.
But then, Elements should be made to be using “Best”.
In any case, - a difference of this sort should be specified, - otherwise, it’s misleading, and appears as if the function is similar, in both versions (Elements & Pro).
.
I’d agree with you here - not that they should be the same, but the difference should be known. In the comparison on Steinberg.net, both versions get the same green check mark at the ‘Integrated real-time Resampler’ item, suggesting there is no difference.