[Csnd] reverb (maybe reverbsc) in a 192 KHz sample rate project
| Date | 2023-07-07 08:26 |
| From | Michael Mossey |
| Subject | [Csnd] reverb (maybe reverbsc) in a 192 KHz sample rate project |
I like to run Pianoteq at 192 KHz in my DAW (Reaper) for more accurate and beautiful piano sound. For good reverb, I have Altiverb 7 but it doesn't run higher than 96 KHz. I have something called Sparkverb, an algorithmic reverb, and it runs at 192 KHz but doesn't sound that great. I'm wondering if reverbsc can run at 192 KHz or maybe not internally but in a 192 KHz project sample rate. Mike |
| Date | 2023-07-07 09:49 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] [EXTERNAL] [Csnd] reverb (maybe reverbsc) in a 192 KHz sample rate project |
|
The manual does not say anything against running it at different sampling rates, so I'd say it's good.
For piano sounds you might want to reduce the optional parameter ipitchm, which adds a bit of modulation to the delays.
Prof. Victor Lazzarini
Maynooth University
Ireland
On 7 Jul 2023, at 08:26, Michael Mossey <michaelmossey@gmail.com> wrote:
|
| Date | 2023-07-08 01:12 |
| From | Aaron Krister Johnson |
| Subject | Re: [Csnd] [EXTERNAL] [Csnd] reverb (maybe reverbsc) in a 192 KHz sample rate project |
Having looked at the code, IIRC, there are integer delay-line lengths that might not behave well too far from 44.1K...? I don't recall if the opcode source code auto-scales the sr. Whether or not the manual's note on `israte` means it will work and sound fine, or "will work, but is considered a 'special effect' " is an open question for you:israte (optional, defaults to the orchestra sample rate) -- assume a sample rate of israte. This is normally set to sr, but a different setting can be useful for special effects. On Fri, Jul 7, 2023 at 1:49 AM Victor Lazzarini <Victor.Lazzarini@mu.ie> wrote:
|
| Date | 2023-07-08 11:11 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] [EXTERNAL] [Csnd] reverb (maybe reverbsc) in a 192 KHz sample rate project |
|
I had a peek at the code and it takes note of the global SR (if israte is 0), so unless there is s bug in the code, it allows any sampling rate between the min and max allowed. These are set to 5KHz and 1MHz.
Setting israte is only needed if the said "special effect" is required.
It should work if the code is good.
Prof. Victor Lazzarini
Maynooth University
Ireland
On 8 Jul 2023, at 01:12, Aaron Krister Johnson <akjmicro@gmail.com> wrote:
|
| Date | 2023-07-08 12:30 |
| From | Oeyvind Brandtsegg |
| Subject | Re: [Csnd] [EXTERNAL] [Csnd] reverb (maybe reverbsc) in a 192 KHz sample rate project |
Hi, Just my 2cents: I think it will work all ok at higher sample rates. But the reverb algorithm, although fairly good, is not extremely refined, so I'm not sure it is worth the high sample rate. It could be refined significantly by adding allpass filters in the feedback loops, which is also likely what Costello does in many of the Valhalla algorithms. To the question of the sample rate adjustment, my humble opinion is that the sample rate does not matter all that much in terms of the internal resonances of this reverb. The important thing is that the delay lines have inharmonic (prime) relationships *between them*. One could argue that the relationship to the sr also matters in avoiding resonances, but since the delay times are also modulated/varying, such sample-rate exact multiples would happen seldomly. Of course the delay times will change with a change of sr, so it is good that the opcode is linked to the global sr, to keep the apparent "size" of the reverberation the same if the sr changes. all best Øyvind lør. 8. juli 2023 kl. 12:11 skrev Victor Lazzarini <Victor.Lazzarini@mu.ie>:
|
| Date | 2023-07-09 05:14 |
| From | Michael Mossey |
| Subject | Re: [Csnd] [EXTERNAL] [Csnd] reverb (maybe reverbsc) in a 192 KHz sample rate project |
Okay, thanks everyone. I'll try to get it working. Just to be clear, I'm not trying to improve the reverb by running at 192 KHz, but want that sample rate in my project so I can run PIanoteq at 192 KHz. On Sat, Jul 8, 2023 at 4:30 AM Oeyvind Brandtsegg <obrandts@gmail.com> wrote:
|