Hey guys, need an advice from you.

I was thinking about sampling EW QL Spaces IR Library, it has very good IRs... Somebody already posted the whole thing on the sister site, but every sample has a strange gap from 65 to 85 ms (don't quite understand how they ended up with that), and it is audible later in the reverb, so not very usable unfortunately.

I am subscribed to the EW cloud, so i have Spaces included, but the plugin often makes strange glitches, clicks and pops (like when a drive is overloaded) on my pretty capable PC (i7 5820 64gb ram win10). I sampled a few IRs, and tried them in Altiverb, and with these Altiverb is working perfectly. It's about the same CPU usage, but no clicks and pops. Also, these sampled IRs to my ears sound identical to the originals in Spaces.

Now, i would not like to go through whole process, which is A LOT of work, just to realize they are actually not the same. So, what I did...

I downloaded "SPIKE 44K.WAV" file from this page: https://www.audioease.com/altiverb/sampling.php

To sample a true stereo IR, with that mono spike, i made two stereo files, one with spike on the L channel (R is silent), and other with the spike on the R channel.

Then I put both in the track with Spaces as an insert, with these settings:
DRY: 0
WET: default (preset dependable)
PREDELAY: 0
INPUT: default (0db)
Some true stereo IR/preset, right...

I exported separately both L and R spikes with verb and ended up with two stereo files of spike with stereo verb. Then I imported both stereo files in Altiverb 7 with the "without post-processing" option. Altiverb recognized all four channels, L-L, L-R, R-L and R-R, which I then saved as an Altiverb preset/format. The result for me was the identical reverb response comparing to the same IR in Spaces. I was doing extensive A/B blind comparisons. Now, my questions are:

1. The only difference was the levels as i had to turn up the "Out" knob on Altiverb to match the sampled IR's reverb level with Spaces'. I guess the problem is too low level of the SPIKE audio, so maybe i should turn that up before sampling? But i don't know how much actually. Is there any standard?

2. The audio files of the SPIKE with reverb were very low level, and turning them up in the daw, like +20db up, revealed some strange noise. Near the end of the reverb tail, while it fades out, there is some high freq chopped noise. Not sure why is that and what it is ... maybe sounds familiar to you? Exporting with higher level SPIKE ended with no such noise...or less. But i would really love to know what that is. I am guessing it affects the final verb tail too.

3. Does this process sound right to you? Is it possible to make identical IR by doing this? And that IR to sound the same in a different convo engines?

I would gladly share the sampled library with you guys once I make these.

EDIT: I tried to null the two audios, same piano sample with Altiverb and Spaces, sampled vs original IR, and while they didn't null completely, the master output droped to -54db. The "difference" have freqs all over the spctrum, it pretty much sounds the same but much quieter. Why they don't null?

 

Last edited: Oct 2, 2017

I don't know much about reverb, but I would try the sweep sine files for maximum quality. Specially seeing that the SPIKE wav file is for sampling hardware (gear).

Anyway, you should be able to achieve good quality with both sample wavs.

 

Yes, and first I tried that which resulted in IR nothing like the original. That is when I saw the rest of the linked page where it says that for digital gear (with digital connections) one should use SPIKE.

 

Last edited: Oct 2, 2017

You're sampling a digital plugin reverb which samples a hardware analog gear.
It's kind of an intermediate territory, but if the spike works better for you go with it.
Anyway, I'm into the whole IR thing so even I don't have your EWQL plugin if I have time I'll try to sample other quality reverb plugin.
If you find a solution I'll like you to share here if you're so kind.

Keep it ))