[Csnd] Problem with the 'tvconv' opcode DFT mode

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

[Csnd] Problem with the 'tvconv' opcode DFT mode

Nathan Holmes
Hello,

When using the new tvconv opcode, any partition size above 2 gives odd results. With sizes 4-32, there are high frequencies introduced into the output which remind me of aliasing from downsampling; and as partition size continues to increase, the output devolves into a light crackling.

I attached some of the output I get from the example file doc/manual/examples/tvconv.csd, which uses a partition size of 256.

In contrast, when the partition size is 1 (and thus, using the direct convolution mode instead of the DFT mode), the opcode seems to work as I'd expect. Interestingly, a size of 2 (which I believe uses DFT mode) seems to work pretty well too. It's just above 2 that things get weird.

Or is this all expected behavior?

I'm using Windows 7 x64 with Csound 6.09.1

Neat opcode, in any case!

Nathan
Csound mailing list [hidden email] https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

tvconv-output.wav (1M) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd] Problem with the 'tvconv' opcode DFT mode

Oeyvind Brandtsegg-3
Hi Nathan, 
There was a bug, but it has recently been fixed in source. It seems the fix narrowly missed this binary release.
There is also a related opcode 'liveconv' that allows dynamic updates to the IR and using longer IRs. Both tvconv and liveconv was developed in the context of a research project on crossadaptive modulation http://crossadaptive.hf.ntnu.no/
best
Oeyvind

2017-07-13 7:45 GMT+02:00 Nathan Holmes <[hidden email]>:
Hello,

When using the new tvconv opcode, any partition size above 2 gives odd results. With sizes 4-32, there are high frequencies introduced into the output which remind me of aliasing from downsampling; and as partition size continues to increase, the output devolves into a light crackling.

I attached some of the output I get from the example file doc/manual/examples/tvconv.csd, which uses a partition size of 256.

In contrast, when the partition size is 1 (and thus, using the direct convolution mode instead of the DFT mode), the opcode seems to work as I'd expect. Interestingly, a size of 2 (which I believe uses DFT mode) seems to work pretty well too. It's just above 2 that things get weird.

Or is this all expected behavior?

I'm using Windows 7 x64 with Csound 6.09.1

Neat opcode, in any case!

Nathan
Csound mailing list [hidden email] https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here



--
Csound mailing list [hidden email] https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd] Problem with the 'tvconv' opcode DFT mode

Victor Lazzarini
In reply to this post by Nathan Holmes
It is probably the case that the bug fix did not make it into the Windows version. The issue was exactly as you described.

Unfortunately, this means that the fix will need to wait until the next release (unless Mike re-does the
package).

best regards

Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy
Maynooth University
Ireland

> On 13 Jul 2017, at 06:45, Nathan Holmes <[hidden email]> wrote:
>
> Hello,
>
> When using the new tvconv opcode, any partition size above 2 gives odd results. With sizes 4-32, there are high frequencies introduced into the output which remind me of aliasing from downsampling; and as partition size continues to increase, the output devolves into a light crackling.
>
> I attached some of the output I get from the example file doc/manual/examples/tvconv.csd, which uses a partition size of 256.
>
> In contrast, when the partition size is 1 (and thus, using the direct convolution mode instead of the DFT mode), the opcode seems to work as I'd expect. Interestingly, a size of 2 (which I believe uses DFT mode) seems to work pretty well too. It's just above 2 that things get weird.
>
> Or is this all expected behavior?
>
> I'm using Windows 7 x64 with Csound 6.09.1
>
> Neat opcode, in any case!
>
> Nathan
> Csound mailing list [hidden email]<mailto:[hidden email]> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here
> <tvconv-output.wav>

Csound mailing list
[hidden email]
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd] Problem with the 'tvconv' opcode DFT mode

Nathan Holmes
Oh, that's good news! Thanks everyone.

I really like tvconv, and I'll definitely be doing some experimentation with liveconv in the future too. In fact, some months back, I found myself wanting to modify IR tables in realtime using the older convolution opcodes, but with not-so-great results; so I was excited when I saw that Oeyvind was developing an opcode to do just that. Great work on these, they're much appreciated!

I'll do some reading on the cross-adaptive processing site, it seems fascinating. :)

-Nathan

On Thu, Jul 13, 2017 at 2:49 AM, Victor Lazzarini <[hidden email]> wrote:
It is probably the case that the bug fix did not make it into the Windows version. The issue was exactly as you described.

Unfortunately, this means that the fix will need to wait until the next release (unless Mike re-does the
package).

best regards

Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy
Maynooth University
Ireland

> On 13 Jul 2017, at 06:45, Nathan Holmes <[hidden email]> wrote:
>
> Hello,
>
> When using the new tvconv opcode, any partition size above 2 gives odd results. With sizes 4-32, there are high frequencies introduced into the output which remind me of aliasing from downsampling; and as partition size continues to increase, the output devolves into a light crackling.
>
> I attached some of the output I get from the example file doc/manual/examples/tvconv.csd, which uses a partition size of 256.
>
> In contrast, when the partition size is 1 (and thus, using the direct convolution mode instead of the DFT mode), the opcode seems to work as I'd expect. Interestingly, a size of 2 (which I believe uses DFT mode) seems to work pretty well too. It's just above 2 that things get weird.
>
> Or is this all expected behavior?
>
> I'm using Windows 7 x64 with Csound 6.09.1
>
> Neat opcode, in any case!
>
> Nathan
> Csound mailing list [hidden email]<mailto:[hidden email]> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here
> <tvconv-output.wav>

Csound mailing list [hidden email] https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here
Loading...