Sampling Rate Bug

[Marc Hildebrant]

Group,

I have been experimenting with various sampling rates during recording. The last rate I tried was 192 KHz.
When I changed the sampling rate to a lower value, the program crashes. In the picture, I tried to change the rate to 48.

The exact same commands work in DC 8.5

Marc

[Craig Maier]

Hi,

Ok - we will look into it. As an aside, can it change to 44.1 kHz and are you using the Make Waves signal generator for the test?

Craig

[Craig Maier]

Hi Again,

I just tried it with the Make Waves generator without any problem. I made a sine wave signal, of 1 kHz - 6dB, 16 bits, 192 kHz. That went fine and played. Then, I brought up the sample rate converter and changed it to 48 kHz (same bit depth) and it did it without crashing. Did you create the original 192 kHz file with a sound card? Some sound cards have issues with producing proper 192 kHz sample rates from what I have heard.

Craig

[Craig Maier]

Hi Again,

I just repeated the test, but this time using a sound card on this Dell Laptop. It has a Realtek High Definition Audio chip-set as its sound-card. I recorded at 192 kHz, stereo, 16 bits (my own voice). It played fine. Then, I converted it down to 48 kHz (CD Quality setting) and it did that and not crash and the result played fine too.

Craig

[Marc Hildebrant]

Craig,

All the tests were for recording, not making waves. It's a realtek on the motherboard. The tests that I did do confirm that the sample rate went up to 192 KHz.

Strange, because the same settings\sound card works fine with DC 8.5 on the same computer.

Marc

[Craig Maier]

Marc,

Yeah, I see. I ran the same test with a Realtek and it worked ok. This computer (and its chipset) are similar to yours (but maybe not identical). I can not explain the bug, especially if DC8.5 reacts differently on your machine compared to Forensics10. Not sure what to do about it right now. Maybe Rick will be able to reproduce it on one of his systems. BTW - is it completely reproduce able? If you re-boot, does it yield a different result? Does re-initialize on play have any effect? Is it 16 or 24 bit (I tested only 16 bit).

Craig

[Craig Maier]

Hi,

As an aside, I just measured the frequency response of the Realtek chip-set on this computer. Everything above 48 kHz sampling rate is emulated. Thus, 96 kHz and 192 kHz are not true sample rates of the incoming analog signal (at least on this Dell laptop). The frequency response of the chipset at 48 kHz is the same as it is at 96 kHz and 192 kHz. The top end in all cases is around 23 kHz or so.

Craig

[Marc Hildebrant]

Craig,

Well, I can't explain it but only state the facts with "software" bugs.

Today, the DC 10 recording bug did not occur. I could change the sample settings with no problem. I have no idea why it's O.K. today.

Also, with regard to the Realtek chips on my system. This computer came from Tracer as one of the "Audio Center" types. I check all the sample rates from 48 KHz to 192 KHz and they are correct.
The 192 KHz was flat up to just before 100 KHz using a sine wave input. The chip set works correct.

Marc

[Craig Maier]

Hi,

The Realtek chipset on this laptop must be of lower quality. I suspect that they are 'emulating' the sample rates above 48 kHz on this machine which would explain the limitation in top-end bandwidth. Realtek probably offers a wide array of products in their lineup; Dell probably does not use the best set for their laptop line.

Craig