Zap VLDTMF causes continuous DTMF tone - Bug?

I have a repeatable problem on two separate asterisk installations.

Placing a call from a Zap channel, to a SIP connected provider (voicepulse) to a PSTN phone. If the receiver of the call presses a keypad number during the conversation, a continuous tone is fed back to the caller. Both parties must hang up, the conversation is over.

Placing a call from a SIP connected handset (not a zap channel) does not result in the same problem.

Increasing the debugging, I see:

[May 10 10:16:43] DEBUG[4238] rtp.c: - RTP 2833 Event: 00000001 (len = 4)
[May 10 10:16:43] DEBUG[4238] rtp.c: Sending dtmf: 49 (1), at xx.xx.xx.xx
[May 10 10:16:43] DEBUG[4238] chan_zap.c: Started VLDTMF digit ‘1’
[May 10 10:16:43] DEBUG[4238] rtp.c: - RTP 2833 Event: 00000001 (len = 4)
[May 10 10:16:43] DEBUG[4238] rtp.c: - RTP 2833 Event: 00000001 (len = 4)
[May 10 10:16:43] DEBUG[4238] rtp.c: - RTP 2833 Event: 00000001 (len = 4)
[May 10 10:16:43] DEBUG[4238] rtp.c: Sending dtmf: 49 (1), at xx.xx.xx.xx
[May 10 10:16:43] DEBUG[4238] rtp.c: - RTP 2833 Event: 00000001 (len = 4)
[May 10 10:16:43] DEBUG[4238] rtp.c: - RTP 2833 Event: 00000001 (len = 4)

This problem did not exist prior to upgrading to:

zaptel 1.4.10.1 (from zaptel 1.4.0)

First installation:
Prior config:
** No problem here
Debian GNU/Linux 3.1
libpri-1.4.0
zaptel-1.4.0
asterisk-1.4.0

First installation:
Current config:
** Problem is here
Debian GNU/Linux 3.1
libpri-1.4.4
zaptel-1.4.10.1
asterisk-1.4.19.1

Second installation:
Current config:
** Problem is here
Debian GNU/Linux 4.0
libpri-1.4.3
zaptel-1.4.7.1
asterisk-1.4.10.1

I have been running Asterisk for several years. I have extensive (20+ years) UNIX internals experience. I can help debug this problem.

Can anyone provide some feedback?

Should I submit a bug report to Digium?

Any ideas?

Thanks!