Results 1 to 2 of 2
07-03-2013, 11:34 PM #1
- Join Date
- May 2013
Offline Izotope > Audirvana Upsampling
Moving the thread as it seems more approrpriate here.
Been experimenting with offline upsampling with Sample Manager and Audirvana upsampling of Redbook to 192khz and found upsampling with Sample Manager to be significantly better sounding than upsampling with Audirvana. The soundstage is much wider and deeper and overall sounds much more real and 3D. The only problem is the large file size.
Sample Manager doesn't allow advanced setting for Izotope so I have no idea what the setting is. My process is saving the aiff files to 32bit first, and then upsampling the 32bit files to 192khz. My Audirvana setting is：
Filter max length: 1,500,000
Cutoff freq: 1.03
Anyone know the setting for Sample Manager Izotope upsampling? If possible, would like to avoid the large file size. Or is the drop in sound quality a result of the extra real-time processing when Audirvana upsamples on the fly?
07-04-2013, 01:47 AM #2
A typical album upsampled to 24/196 AIFF is 4gb to 5gb in size, so I can see where HD storage, although cheap, could be a problem if you have a large music library. I don't hear much about anyone using 32-bit depth. That would make the files even larger. You could use Apple Lossless to cut that down by about 40%, but I use AIFF to keep my music library cross-platform compatible, so it will play on PC, too.
As you your specific query, although I don't know for sure, I would assume that the five iZotope presets in Sample Manager are the same as those in A+, i.e. the "best" quality setting would be steepness 150db, filter 500,000, cutoff 1.00 Nyquist, A/A 200, pre-ringing 1.00. The ideal solution for you would be to purchase and download the Audiofile Engineering "Action Pack" add-on ($19.95), which will upgrade your iZotope to the "advanced" version, which, I assume, will give you the same, user-cusomizable control of the five slider settings as in A+.
Interestingly, I did some A/B listening tests with my pre-upsampled 24/96 files played with A+ 184.108.40.206 (forced upsampling off, obviously) vs. playing the straight redbook files using the A+ iZotope SRC settings set to 24/96, and adjusting the sliders similar to your settings described above. I am hearing the opposite of what you are hearing. I thought there was better instrument and vocal separation, and a more "live" sound to the performance doing the upsampling in A+ real-time. The presampled files sounded good, but flattened, with the vocals and instruments muddled together a bit more. I reiterate that I am only upsampling to 24/96. 192kHz might produce a completely different result, and YMMV.
If you lock in your upsamplng in advance, you will be stuck with it that way forever. If you keep your redbook files in native format, you have the flexibility of adjusting the playback upsampling however you want with A+, as well as fitting 4x as much music on your HD. Also, A+ does preload the tracks and performs the upsampling before playing the track, so as long as your computer has enough RAM and processing power, this should not be an issue.
Let me make few other points. The basic iZotope upsampling in Sample Manager does not do 176kHz, which is a major drawback, as that is the preferred mathematically-correct 4x sample rate from redbook. Not sure about the "Action Pack SRC"upgrade--I don't have it.
Upsampling redbook to 192kHz (or 96kHz for that matter) is not ideal, requiring the software to interpolate the square waves. I hope I am stating this correctly, I am not a scientist, I only know enough about this stuff to make me dangerous.
If any other forum members are doing the same experiments with Sample Manager vs. real-time upsampling, I'd be interested in hearing their opinions, too.OWC 2TB RAID FW HDD > MacBook Pro i5 8/120GB SSD > optimized Mavericks 10.9.1 on SanDisk SDHC card > Audirvana Plus > Kimber USB > NAD M51 > Kimber Hero > NAD C 275BEE > B&W CDM7SE