Re: [casper] pfb_coeff_gen_calc shifted sinc arguments

[Nathan West]

I’m very happy to hear that. I was working with the ATA taps which looked 
strange and lead me down this path. We probably saw the same thing and wound up 
at the same result— I’m just late to the party!

Cheers,
Nathan

From: casper@lists.berkeley.edu  on behalf of Jack 
Hickish 
Sent: Thursday, May 30, 2024 6:18:47 AM
To: casper 
Subject: Re: [casper] pfb_coeff_gen_calc shifted sinc arguments

You don't often get email from jackhick...@gmail.com. Learn why this is 
important<https://aka.ms/LearnAboutSenderIdentification>
and was pointed out to me by the late Frankin Antonio (lest I take credit 
for a level of detail-orientation I don't possess)

On Thu, 30 May 2024, 01:46 Andrew Martens, 
mailto:and...@sarao.ac.za>> wrote:
Hi Nathan

This bug (if we are talking about the same one) was fixed in a commit 2 years 
ago but has not made it into the main repo yet.

It was pointed out to me by Jack Hickish.

For more info see

https://github.com/ska-sa/mlib_devel/commit/702d962842b53f76ec661b27b1156679ccb71929

The fixed version is available at

https://github.com/ska-sa/mlib_devel/blob/m2021a/casper_library/pfb_coeff_gen_calc.m

Cheers
Andrew

On Thu, May 30, 2024 at 3:11 AM Nathan West 
mailto:nat...@quadrature.dev>> wrote:
Hi all,

I was comparing the coarse channelization response and taps from a CASPER 
instrument to other window design routines and noticed a small discrepancy 
comparing to pfb_coeff_gen_calc.

The argument to the sinc to generate an ideal lowpass filter is typically 
-wc..wc. Different routines get the range in a few ways, but sticking with the 
matlab version in 
https://github.com/casper-astro/mlib_devel/blob/a557a844f8421f9860876b0216dd6758508d8f2e/casper_library/pfb_coeff_gen_calc.m#L23C25-L23C43
 it would typically be something like [0:alltaps-1]/alltaps - (alltaps-1)/2​ to 
give a range of -(alltaps-1)/2 : (alltaps-1)/2​ inclusive. That would then be 
scaled by the cutoff which is included in this routine as 1/(2^PFBSize).

Including the cutoff factor, to sample -wc:wc would normally be
[0:alltaps-1]/(TotalTaps*2^PFBSize) - (TotalTaps*2^PFBSize)/2​

This is very slightly shifted version used in the casper routine by a delay of 
1/(2*2^PFBSize)​. The casper shifted version is guaranteed to produce a leading 
0 since the equivalent TotalTaps*2^(PFBSize-1) will always be an integer but 
produces slightly different taps than standard routines. Anyway, I was mostly 
hoping someone might help me understand the design choice a little more or if 
it's actually a bug.

Thanks,
Nathan
[https://af-south-1.api.getmessagecontrol.com/onprem_image_fetch?dep=DKsFNr%2FdE4ZSNMVbhQo8iA%3D%3D7EjproNhHQf5Q4T%2BaJSmW%2F%2FIQ4KqD8jQHyvFW4CcXivSTWxlLy37%2BIbPAIasHwgPFVyQY8ALjBuYePq6cDIUiGfoFIPPCwhAnA3%2F0Be0AxMFLK6GVMUpA1zbPE254sU69mzmXLrqQr43W70Fckyz4vy28vBYwFz3OUtpi%2BiWCjwAoPpUhcA2o3z1xctZEiWC6DLQsKimqBO%2FPhThx09MJhTwejjqUW6YEIMugFLcmCTU%2FIDAzLLiaB5tlopGU2r7Occbp%2FeUNwG6vSml3w%2BXofzswcvfj7HFMK1n%2FKXPy7I5JBHxYpbDQ7NDQvhXJQXTCruzFPFyfIeTMZaG7ud0IM5qLAo2YaMbPwp2f1g8AnWta7yRw%2FiUNUeFXsGPuQ8t%2FyNpoqb3Yk9Afy%2FLbRC%2BtGE6FmRe38NYm1Zd3HJ0z7SJFxvwhHVN1uiKMHJtpFpxjbV4z1clPu3W0qgRGC74Iw%3D%3D]<https://github.com/casper-astro/mlib_devel/blob/a557a844f8421f9860876b0216dd6758508d8f2e/casper_library/pfb_coeff_gen_calc.m#L23C25-L23C43>
mlib_devel/casper_library/pfb_coeff_gen_calc.m at 
a557a844f8421f9860876b0216dd6758508d8f2e · 
casper-astro/mlib_devel<https://github.com/casper-astro/mlib_devel/blob/a557a844f8421f9860876b0216dd6758508d8f2e/casper_library/pfb_coeff_gen_calc.m#L23C25-L23C43>
Contribute to casper-astro/mlib_devel development by creating an account on 
GitHub.
github.com<http://github.com>




--
You received this message because you are subscribed to the Google Groups 
"casper@lists.berkeley.edu<mailto:casper@lists.berkeley.edu>" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to 
casper+unsubscr...@lists.berkeley.edu<mailto:casper+unsubscr...@lists.berkeley.edu>.
To view this discussion on the web visit 
https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/MW4P221MB10470D49AF6C96FC1CEA3828BBF32%40MW4P221MB1047.NAMP221.PROD.OUTLOOK.COM<https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/MW4P221MB10470D49AF6C96FC1CEA3828BBF32%40MW4P221MB1047.NAMP221.PROD.OUTLOOK.COM?utm_medium=email&utm_source=footer>.

[https://af-south-1.api.getmessagecontrol.com/onprem_image_fetch?dep=MaxgAqX1iOqPHUgkPTZiOA%3D%3DbA4cHSNwB78IBiGDRi8saQgBEooxBYRF2tlSkPm%2F3gX6%2Fg6Ex%2BP6bd%2BchOt%2FXm5eK0CwujoQiJWSHuB%2BhD9Y%2BlTuJfrcULr0bRUs05HkwXEiup2HU7eNf3GqPzp5LWIz%2Fk4KtjZ976i3f2aiJVmn0sMeaUsbfvzwIR%2B56bbW5h5ntVLMDKrQK8Ha9FoT2RPm5WBLIk4nOdDMLUvYoKNjkq4c%2Fcurvln1%2FbZNz2c2xSDpOMjYl7yXgKJgxv8cy33OkF8gNLpAkb2dEh50c5x6vJ4fJOMP4AWL1Y9XImgVUQQFz%2B7RMpf3mhFkiXZ0mLNJv9T8OUARhTETArBS1aakB2e83hLTqRcOHRn6h6Gp4QA%3D]



Disclaimer

The information cont

Re: [casper] pfb_coeff_gen_calc shifted sinc arguments

[Jack Hickish]

and was pointed out to me by the late Frankin Antonio (lest I take
credit for a level of detail-orientation I don't possess)

On Thu, 30 May 2024, 01:46 Andrew Martens,  wrote:

> Hi Nathan
>
> This bug (if we are talking about the same one) was fixed in a commit 2
> years ago but has not made it into the main repo yet.
>
> It was pointed out to me by Jack Hickish.
>
> For more info see
>
>
> https://github.com/ska-sa/mlib_devel/commit/702d962842b53f76ec661b27b1156679ccb71929
>
> The fixed version is available at
>
>
> https://github.com/ska-sa/mlib_devel/blob/m2021a/casper_library/pfb_coeff_gen_calc.m
>
> Cheers
> Andrew
>
> On Thu, May 30, 2024 at 3:11 AM Nathan West  wrote:
>
>> Hi all,
>>
>> I was comparing the coarse channelization response and taps from a CASPER
>> instrument to other window design routines and noticed a small discrepancy
>> comparing to pfb_coeff_gen_calc.
>>
>> The argument to the sinc to generate an ideal lowpass filter is typically
>> -wc..wc. Different routines get the range in a few ways, but sticking with
>> the matlab version in
>> https://github.com/casper-astro/mlib_devel/blob/a557a844f8421f9860876b0216dd6758508d8f2e/casper_library/pfb_coeff_gen_calc.m#L23C25-L23C43
>>  it
>> would typically be something like [0:alltaps-1]/alltaps - (alltaps-1)/2​
>> to give a range of -(alltaps-1)/2 : (alltaps-1)/2​ inclusive. That would
>> then be scaled by the cutoff which is included in this routine as
>> 1/(2^PFBSize).
>>
>> Including the cutoff factor, to sample -wc:wc would normally be
>> [0:alltaps-1]/(TotalTaps*2^PFBSize) - (TotalTaps*2^PFBSize)/2​
>>
>> This is very slightly shifted version used in the casper routine by a
>> delay of 1/(2*2^PFBSize)​. The casper shifted version is guaranteed to
>> produce a leading 0 since the equivalent TotalTaps*2^(PFBSize-1) will
>> always be an integer but produces slightly different taps than standard
>> routines. Anyway, I was mostly hoping someone might help me understand the
>> design choice a little more or if it's actually a bug.
>>
>> Thanks,
>> Nathan
>>
>> 
>> mlib_devel/casper_library/pfb_coeff_gen_calc.m at
>> a557a844f8421f9860876b0216dd6758508d8f2e · casper-astro/mlib_devel
>> 
>> Contribute to casper-astro/mlib_devel development by creating an account
>> on GitHub.
>> github.com
>> **
>> **
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "casper@lists.berkeley.edu" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to casper+unsubscr...@lists.berkeley.edu.
>> To view this discussion on the web visit
>> https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/MW4P221MB10470D49AF6C96FC1CEA3828BBF32%40MW4P221MB1047.NAMP221.PROD.OUTLOOK.COM
>> 
>> .
>>
>
>
>
>
> *Disclaimer*
>
> The information contained in this communication from the sender is
> confidential. It is intended solely for use by the recipient and others
> authorized to receive it. If you are not the recipient, you are hereby
> notified that any disclosure, copying, distribution or taking action in
> relation of the contents of this information is strictly prohibited and may
> be unlawful.
>
> --
> You received this message because you are subscribed to the Google Groups "
> casper@lists.berkeley.edu" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to casper+unsubscr...@lists.berkeley.edu.
> To view this discussion on the web visit
> https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/CADEwHTeL4C6SxvLKmaM42Vg8F83ZQmFQGXMYJ0568F9qwr5iwA%40mail.gmail.com
> 
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"casper@lists.berkeley.edu" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to casper+unsubscr...@lists.berkeley.edu.
To view this discussion on the web visit 
https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/CAG1GKSkZrw-wvFBph%2B8SS1O7bGntAk97dSv7ObMJjXcjeF7JLA%40mail.gmail.com.

Re: [casper] pfb_coeff_gen_calc shifted sinc arguments

[Nathan West]

Hi Andrew,

Yes, that version corrects the issue I found. Thanks for confirming.

Regards,
Nathan

From: casper@lists.berkeley.edu  on behalf of Andrew 
Martens 
Sent: Thursday, May 30, 2024 12:46 AM
To: casper@lists.berkeley.edu 
Subject: Re: [casper] pfb_coeff_gen_calc shifted sinc arguments

You don't often get email from and...@sarao.ac.za. Learn why this is 
important<https://aka.ms/LearnAboutSenderIdentification>
Hi Nathan

This bug (if we are talking about the same one) was fixed in a commit 2 years 
ago but has not made it into the main repo yet.

It was pointed out to me by Jack Hickish.

For more info see

https://github.com/ska-sa/mlib_devel/commit/702d962842b53f76ec661b27b1156679ccb71929

The fixed version is available at

https://github.com/ska-sa/mlib_devel/blob/m2021a/casper_library/pfb_coeff_gen_calc.m

Cheers
Andrew

On Thu, May 30, 2024 at 3:11 AM Nathan West 
mailto:nat...@quadrature.dev>> wrote:
Hi all,

I was comparing the coarse channelization response and taps from a CASPER 
instrument to other window design routines and noticed a small discrepancy 
comparing to pfb_coeff_gen_calc.

The argument to the sinc to generate an ideal lowpass filter is typically 
-wc..wc. Different routines get the range in a few ways, but sticking with the 
matlab version in 
https://github.com/casper-astro/mlib_devel/blob/a557a844f8421f9860876b0216dd6758508d8f2e/casper_library/pfb_coeff_gen_calc.m#L23C25-L23C43
 it would typically be something like [0:alltaps-1]/alltaps - (alltaps-1)/2​ to 
give a range of -(alltaps-1)/2 : (alltaps-1)/2​ inclusive. That would then be 
scaled by the cutoff which is included in this routine as 1/(2^PFBSize).

Including the cutoff factor, to sample -wc:wc would normally be
[0:alltaps-1]/(TotalTaps*2^PFBSize) - (TotalTaps*2^PFBSize)/2​

This is very slightly shifted version used in the casper routine by a delay of 
1/(2*2^PFBSize)​. The casper shifted version is guaranteed to produce a leading 
0 since the equivalent TotalTaps*2^(PFBSize-1) will always be an integer but 
produces slightly different taps than standard routines. Anyway, I was mostly 
hoping someone might help me understand the design choice a little more or if 
it's actually a bug.

Thanks,
Nathan
[https://af-south-1.api.getmessagecontrol.com/onprem_image_fetch?dep=DKsFNr%2FdE4ZSNMVbhQo8iA%3D%3D7EjproNhHQf5Q4T%2BaJSmW%2F%2FIQ4KqD8jQHyvFW4CcXivSTWxlLy37%2BIbPAIasHwgPFVyQY8ALjBuYePq6cDIUiGfoFIPPCwhAnA3%2F0Be0AxMFLK6GVMUpA1zbPE254sU69mzmXLrqQr43W70Fckyz4vy28vBYwFz3OUtpi%2BiWCjwAoPpUhcA2o3z1xctZEiWC6DLQsKimqBO%2FPhThx09MJhTwejjqUW6YEIMugFLcmCTU%2FIDAzLLiaB5tlopGU2r7Occbp%2FeUNwG6vSml3w%2BXofzswcvfj7HFMK1n%2FKXPy7I5JBHxYpbDQ7NDQvhXJQXTCruzFPFyfIeTMZaG7ud0IM5qLAo2YaMbPwp2f1g8AnWta7yRw%2FiUNUeFXsGPuQ8t%2FyNpoqb3Yk9Afy%2FLbRC%2BtGE6FmRe38NYm1Zd3HJ0z7SJFxvwhHVN1uiKMHJtpFpxjbV4z1clPu3W0qgRGC74Iw%3D%3D]<https://github.com/casper-astro/mlib_devel/blob/a557a844f8421f9860876b0216dd6758508d8f2e/casper_library/pfb_coeff_gen_calc.m#L23C25-L23C43>
mlib_devel/casper_library/pfb_coeff_gen_calc.m at 
a557a844f8421f9860876b0216dd6758508d8f2e · 
casper-astro/mlib_devel<https://github.com/casper-astro/mlib_devel/blob/a557a844f8421f9860876b0216dd6758508d8f2e/casper_library/pfb_coeff_gen_calc.m#L23C25-L23C43>
Contribute to casper-astro/mlib_devel development by creating an account on 
GitHub.
github.com<http://github.com>




--
You received this message because you are subscribed to the Google Groups 
"casper@lists.berkeley.edu<mailto:casper@lists.berkeley.edu>" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to 
casper+unsubscr...@lists.berkeley.edu<mailto:casper+unsubscr...@lists.berkeley.edu>.
To view this discussion on the web visit 
https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/MW4P221MB10470D49AF6C96FC1CEA3828BBF32%40MW4P221MB1047.NAMP221.PROD.OUTLOOK.COM<https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/MW4P221MB10470D49AF6C96FC1CEA3828BBF32%40MW4P221MB1047.NAMP221.PROD.OUTLOOK.COM?utm_medium=email&utm_source=footer>.

[https://af-south-1.api.getmessagecontrol.com/onprem_image_fetch?dep=MaxgAqX1iOqPHUgkPTZiOA%3D%3DbA4cHSNwB78IBiGDRi8saQgBEooxBYRF2tlSkPm%2F3gX6%2Fg6Ex%2BP6bd%2BchOt%2FXm5eK0CwujoQiJWSHuB%2BhD9Y%2BlTuJfrcULr0bRUs05HkwXEiup2HU7eNf3GqPzp5LWIz%2Fk4KtjZ976i3f2aiJVmn0sMeaUsbfvzwIR%2B56bbW5h5ntVLMDKrQK8Ha9FoT2RPm5WBLIk4nOdDMLUvYoKNjkq4c%2Fcurvln1%2FbZNz2c2xSDpOMjYl7yXgKJgxv8cy33OkF8gNLpAkb2dEh50c5x6vJ4fJOMP4AWL1Y9XImgVUQQFz%2B7RMpf3mhFkiXZ0mLNJv9T8OUARhTETArBS1aakB2e83hLTqRcOHRn6h6Gp4QA%3D]



Disclaimer

The information contained in this communication from the sender is 
confidential. It is intended solely for use by the recipient and others 
authorized to receive it. If you are not the recipient, you are hereby notified 
that any disclosure, copying, distribution or taking action in relation of the 
contents of this information is strictly prohi

Re: [casper] pfb_coeff_gen_calc shifted sinc arguments

[Andrew Martens]

Hi Nathan

This bug (if we are talking about the same one) was fixed in a commit 2
years ago but has not made it into the main repo yet.

It was pointed out to me by Jack Hickish.

For more info see

https://github.com/ska-sa/mlib_devel/commit/702d962842b53f76ec661b27b1156679ccb71929

The fixed version is available at

https://github.com/ska-sa/mlib_devel/blob/m2021a/casper_library/pfb_coeff_gen_calc.m

Cheers
Andrew

On Thu, May 30, 2024 at 3:11 AM Nathan West  wrote:

> Hi all,I was comparing the coarse channelization response and taps from a
> CASPER instrument to other window design routines and noticed a small
> discrepancy comparing to pfb_coeff_gen_calc.The argument to the sinc to
> generate an ideal lowpass filter is typically -wc..wc.
> Different ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 
> ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌
>
> 
> Hi all,
>
> I was comparing the coarse channelization response and taps from a CASPER
> instrument to other window design routines and noticed a small discrepancy
> comparing to pfb_coeff_gen_calc.
>
> The argument to the sinc to generate an ideal lowpass filter is typically
> -wc..wc. Different routines get the range in a few ways, but sticking with
> the matlab version in
> https://github.com/casper-astro/mlib_devel/blob/a557a844f8421f9860876b0216dd6758508d8f2e/casper_library/pfb_coeff_gen_calc.m#L23C25-L23C43
> 
>  it
> would typically be something like [0:alltaps-1]/alltaps - (alltaps-1)/2​
> to give a range of -(alltaps-1)/2 : (alltaps-1)/2​ inclusive. That would
> then be scaled by the cutoff which is included in this routine as
> 1/(2^PFBSize).
>
> Including the cutoff factor, to sample -wc:wc would normally be
> [0:alltaps-1]/(TotalTaps*2^PFBSize) - (TotalTaps*2^PFBSize)/2​
>
> This is very slightly shifted version used in the casper routine by a
> delay of 1/(2*2^PFBSize)​. The casper shifted version is guaranteed to
> produce a leading 0 since the equivalent TotalTaps*2^(PFBSize-1) will
> always be an integer but produces slightly different taps than standard
> routines. Anyway, I