*** This bug has been confirmed by popular vote. ***

can the sampling rate be made higher, even if you drop the resolution, to allow
the tablet to be used as a primitive oscilloscope?

Comment 3 Andre Klapper 2008-10-21 23:51:47 UTC

(In reply to comment #0)
> If I'm reading the TSC2301 datasheet correctly, it sounds like it supports
> stereo audio recording at up to a 48KHz sampling rate, is there any way to do
> this on the N800?

That's not really a question for a bugtracker - aren't there applications out
there that support this? Can I get a how-to to reproduce this issue?

> Does the audio jack have a stereo input, or only MIC level mono?

What is this about? N800? I don't see any input jack for the N810.

> Either way, is there any possibility of recording at higher than an 8KHz
> sampling rate, whether from the built-in microphone, or audio input, stereo or
> mono?

Which application did you use to try this?

Comment 4 Ryan Abel 2008-10-22 00:08:12 UTC

(In reply to comment #3)
> What is this about? N800? I don't see any input jack for the N810.
> 

Both the N800 and N810 have a 3rd line on the 3.5mm jack for mono input. They
also have built-in mics that should be capable of capturing better than 8KHz.

https://garage.maemo.org/projects/maemo-recorder/ is likely one thing being
discussed.

(In reply to comment #3)
> (In reply to comment #0)
> > If I'm reading the TSC2301 datasheet correctly, it sounds like it supports
> > stereo audio recording at up to a 48KHz sampling rate, is there any way to do
> > this on the N800?
>
> That's not really a question for a bugtracker - aren't there applications out
> there that support this? Can I get a how-to to reproduce this issue?

What? Are you asking if there's applications that support recording at sampling
rates higher than 8KHz?

Maybe I'm dense, but it seems a perfectly valid question for the bug tracker to
me.. where else would you suggest I ask?

How to reproduce:
Step 1. Run any audio recording software
Step 2. Try to record at a sampling rate higher than 8KHz
Step 3. Fail, for unknown reasons that no one will bother to explain.

> > Does the audio jack have a stereo input, or only MIC level mono?
>
> What is this about? N800? I don't see any input jack for the N810.

It's the port you plug earphones or a headset into, on both the N800 and N810.

> > Either way, is there any possibility of recording at higher than an 8KHz
> > sampling rate, whether from the built-in microphone, or audio input, stereo or
> > mono?   
>
> Which application did you use to try this?

It's been a while, but from memory:

gstreamer   
rec (from sox)
alsarecord  
maemo-recorder
esdrec
..and probably half a dozen others.

I'm pretty sure this is a well known issue.. and I'd love to know if it's
simply an artificial limitation.

If the hardware supports it, I just don't understand why they'd cripple it in
software.

I recall reading somewhere that recording above 8 khz sampling rate is
possible, but one would need to use the DSP.  Seems unlikely to me that a
modern Arm can't sample faster than 8 khz.  You know, my 1987 Mac II, with 16.7
MHz 68000 could sample at 22 khz. It wasn't limited by the CPU.  All i'm saying
is that i've no idea why.  Ideally, i'd like 44 khz 16 bit samples.

Comment 7 Andre Klapper 2008-10-22 02:15:36 UTC

(In reply to comment #5)
> What? Are you asking if there's applications that support recording at 
> sampling rates higher than 8KHz?
> 
> Maybe I'm dense, but it seems a perfectly valid question for the bug 
> tracker to me.. where else would you suggest I ask?

Well, that was exactly my question - if it's just an issue of the application
you used and that you didn't mention. Clarified, thanks. :)

> If the hardware supports it, I just don't understand why they'd cripple it in
> software.

I've asked the Nokia folks. They told me that "while the chipset might support
such features, the rest of the architecture might not."

So there's no chance to get this fixed for N8x0, but maybe for the next
hardware product (that we all don't know enough about currently).

In the end, this boils down to a hardware request which is basically out of
scope for this bugtracker.

> I've asked the Nokia folks. They told me that "while the chipset might support
> such features, the rest of the architecture might not."

Thanks for asking them, appreciate it! :)

It's still unclear though.. does "might not" actually mean *might* not, or do
they mean the physical architechture absolutely *does not* support 8KHz?

> So there's no chance to get this fixed for N8x0, but maybe for the next
> hardware product (that we all don't know enough about currently).
> 
> In the end, this boils down to a hardware request which is basically out of
> scope for this bugtracker.

If it is actually a hardware problem, sure.. but I'm really not convinced by
the "might not" answer from Nokia, that makes me think the people you talked to
don't actually know, or maybe it's a language thing.

Any chance you could get a crystal clear answer from Nokia?

Thanks

I should note I'd be thrilled to get higher than 8K for the internal mic or the
external input, if either one were physically possible.

Comment 10 Andre Klapper 2008-10-22 12:48:09 UTC

I'm very, very sure it means "does not". :-/

Comment 11 Andre Klapper 2008-10-22 22:23:13 UTC

So, just to clearly say this again:

This bug report is INVALID for 770/N800/N810 because of hardware limitations. 
I don't know about future hardware plans.

Discussion of *Nokia hardware* requests is out of scope for this Bugzilla
(*Maemo software*), and I feel quite bad closing this one with its 14 votes.
Sigh.
Let's cross fingers for the next hardware.

I realize that it may be pointless, but could we please get a bit more in the
way of details about why this proposal isn't possible?  44.1kHz @16bit may not
be possible, but you're saying that no sample rates faster than 8kHz is
possible?  Where possible I guess means can be done for reasonable periods of
time without dropping blocks.

I find it very hard to believe that this hardware can handle nothing better
than 8kHz.

Comment 13 Andre Klapper 2008-10-28 15:41:44 UTC

I've asked whether I'm allowed to list the exact hardware restriction reasons
here in public without breaking any NDA contracts, but haven't received an
answer yet...

The DSP does processing of inputs depending on the sampling rate. For diabo PE5
we'll update the dsppcmsrc to also offer 16KHz. Unfortunately there is not
setup for other rates and the code is not flexible enough as it is to support
other rates in pass-thru mode (no processing at all).

In freemantle there won't be such issue at all, as we move to pulseaudio with
audio processing on host side.

I know its a compromise, but thats all we can offer for now.

Comment 15 Quim Gil 2008-12-01 10:01:48 UTC

According to Stefan's answer, this feature request can be resolved.

Comment 16 Andre Klapper 2008-12-17 20:08:10 UTC

So, was today's 5.2008.43-7 == PE5? If so, this has been released.

Comment 17 Andre Klapper 2008-12-18 02:21:16 UTC

Ah. 5.2008.43-7 == PE4.
OK, keeping the 4.1+ target milestone for PE5.