I'm only asking because I couldn't here much difference, while usually I can. Would it be possible/necessary to offer high resolution versions of the current sound banks?
Thanks and sorry if you consider this a stupid question
How does Rhino deal with 24bit 96kHz?
-
- KVRist
- 30 posts since 30 Mar, 2002
-
- KVRAF
- 3388 posts since 29 May, 2001 from New York, NY
Hi,
It's true the built-in waveforms started their life as 16 bits, and are antialiased for 44.1 KHz. Would it be possible to provide better quality ones ? Maybe (I'd also need to change the audio engine), but not straightforward - especially since my own system can only play 16 bits, 48 KHz. Would it be necessary ? Actually, I don't know. You guys tell me. Comments, anyone ?
'Tick[/b]
It's true the built-in waveforms started their life as 16 bits, and are antialiased for 44.1 KHz. Would it be possible to provide better quality ones ? Maybe (I'd also need to change the audio engine), but not straightforward - especially since my own system can only play 16 bits, 48 KHz. Would it be necessary ? Actually, I don't know. You guys tell me. Comments, anyone ?
'Tick[/b]
-
- KVRian
- 598 posts since 12 Mar, 2002 from An American in Japan (I moved)
I think the sound quality of rhino is second to NONE...
It sounds great. Mu problem is I really am not into programming, so I'm hoping some KvRians are busy making patches
I would like to be able to load my on wav files, 16 bit or not..
It sounds great. Mu problem is I really am not into programming, so I'm hoping some KvRians are busy making patches
I would like to be able to load my on wav files, 16 bit or not..
peace,
david
david
-
- KVRAF
- 3388 posts since 29 May, 2001 from New York, NY
iDavid,
Well, loading your own files is on my huge list of "todo" things. But I don't want Rhino to be compared to a sampler...
If you have some wavs that you think would be useful, please send them my way, and I'll put them in a set of additional waveforms. So everybody can benefit from them, and more kvr'ians get inspired to do patches
'Tick
Well, loading your own files is on my huge list of "todo" things. But I don't want Rhino to be compared to a sampler...
If you have some wavs that you think would be useful, please send them my way, and I'll put them in a set of additional waveforms. So everybody can benefit from them, and more kvr'ians get inspired to do patches
'Tick
-
- KVRian
- 598 posts since 12 Mar, 2002 from An American in Japan (I moved)
Can't save that I do, unless you want some StickBass wavs..
I really want to import vocal wavs. I know Absynth 2 is going to be able to do that, but I ever have gotten along so well with Absynth.
I've gotten plenty to paly with at the moment, so I'll wait on rhino
I really want to import vocal wavs. I know Absynth 2 is going to be able to do that, but I ever have gotten along so well with Absynth.
I've gotten plenty to paly with at the moment, so I'll wait on rhino
peace,
david
david
-
- KVRist
- Topic Starter
- 30 posts since 30 Mar, 2002
Hello Big T,
I'm thinking that the bitdepth won't matter much, because I assume that Rhino has 32bit internal precision and that it sends this to the host at 32bit float, is this right? If so, then I assume that all of Rhino's effects are working at 32bit and not 16bit, is this also right?. Yeah, so it seems like a higher bit depth won't make that much of a difference, aside from the noise floor (which is not noticeable at the moment).
The samplerate matters though, doen't you think so? What does Rhino do when the host is using 88.2kHz or 96kHz? I wouldn't want Rhino or the host to upsample from 44.1 to 48, 96 or 192, but there are three ways to handle this that I can think of:
1. Rhino could have two sets of waves, 44.1 and 48, and could intelligently choose the best match for the host samplerate then just allow the host to upsample in multiples of 2.
2. Same as above, but Rhino could use it's own internal sample doubler at the earliest point in Rhino's signal chain.
3. Rhino could just have seperate banks for each samplerate and intelligently choose the best match.
I think 2 above would be the best approach, but my knowledge of such things is limited. I'm sure you could come up with a better way.
I'm thinking that the bitdepth won't matter much, because I assume that Rhino has 32bit internal precision and that it sends this to the host at 32bit float, is this right? If so, then I assume that all of Rhino's effects are working at 32bit and not 16bit, is this also right?. Yeah, so it seems like a higher bit depth won't make that much of a difference, aside from the noise floor (which is not noticeable at the moment).
The samplerate matters though, doen't you think so? What does Rhino do when the host is using 88.2kHz or 96kHz? I wouldn't want Rhino or the host to upsample from 44.1 to 48, 96 or 192, but there are three ways to handle this that I can think of:
1. Rhino could have two sets of waves, 44.1 and 48, and could intelligently choose the best match for the host samplerate then just allow the host to upsample in multiples of 2.
2. Same as above, but Rhino could use it's own internal sample doubler at the earliest point in Rhino's signal chain.
3. Rhino could just have seperate banks for each samplerate and intelligently choose the best match.
I think 2 above would be the best approach, but my knowledge of such things is limited. I'm sure you could come up with a better way.
-
- KVRAF
- 3388 posts since 29 May, 2001 from New York, NY
Rhino is just using whatever tempo the host requests. Then, whatever the sample rate, it attempts to block all frequencies above 22050 Hz. Therefore, higher sample rates give you better attenuation of all the aliased frequencies above 22050 Hz.
In my opinion, this was the best option - when using a sample rate of 96 Khz, you don't really care of all the frequencies between 22050 and 48 KHz - nobody can hear them anyway !. But you do care of the improved definition in the 0-22050Hz range.... No ? Do I have it all wrong ? I'm wondering what the other synths out there do ?
'Tick
In my opinion, this was the best option - when using a sample rate of 96 Khz, you don't really care of all the frequencies between 22050 and 48 KHz - nobody can hear them anyway !. But you do care of the improved definition in the 0-22050Hz range.... No ? Do I have it all wrong ? I'm wondering what the other synths out there do ?
'Tick
-
- KVRist
- Topic Starter
- 30 posts since 30 Mar, 2002
Big T, that is exactly what I care about. But are you saying that there is actually no benefit at all to using Rhino at higher samplerates? It is still doubling the samples when the host is using 88.2/96kHz right? So does Rhino use something similar to method 2 that I outlined above, with the addition of removing all frequencies above 22050Hz? If not, then it seems there could still be benefit from implementing method 2 or something similar or better. But still, I guess before I can ask any smart questions, what actually happens with the samplerate in Rhino when the host is set to 96kHz? Does Rhino upsample internally before all of Rhino's effects and so on? Or, is it the host that upsamples the output from Rhino? Does Rhino require any upsampling when used at these higher resolutions?But you do care of the improved definition in the 0-22050Hz range.... No ? Do I have it all wrong ? I'm wondering what the other synths out there do ?
Sorry for all of the questions.
-
- KVRAF
- 3388 posts since 29 May, 2001 from New York, NY
Yes, when the host requests to work at 96 KHz, Rhino upsamples the initial waveforms at this sample rate (and processes everything at 96 KHz, including the effects, from this point on).
'Tick
'Tick
-
- KVRist
- Topic Starter
- 30 posts since 30 Mar, 2002
Okay, I was mistaken. Rhino sounds dramtically different at 96kHz; in some ways good and some bad. I created a comparison between 44.1 and 96kHz, along with some pictures of what the waveforms look like:
www.freewebs.com/agnishvatta/InstantScore.zip
You'll have to copy and paste it into the address bar of your browser. I downsampled the 96kHz file to 44.1 so that you could here it.
www.freewebs.com/agnishvatta/InstantScore.zip
You'll have to copy and paste it into the address bar of your browser. I downsampled the 96kHz file to 44.1 so that you could here it.
-
- KVRAF
- 3388 posts since 29 May, 2001 from New York, NY
Agnishvatta,
I think you found a monster bug... I need to do more investingations, but I think envelopes in non-sync mode have different times depending on the sample rate. This is quite obvious in your "instant score" sample.
I need to check this DC offset thing for the other preset too.
'Tick
I think you found a monster bug... I need to do more investingations, but I think envelopes in non-sync mode have different times depending on the sample rate. This is quite obvious in your "instant score" sample.
I need to check this DC offset thing for the other preset too.
'Tick