Clipping & Underpowered Speakers

[antcollinet]

Tony, he may be crossing over to the magic of letting out the smoke..........I suggest a LOT of caution.

The question was (turning it around a little) "will disconnecting the woofers increase the stress on the tweeters"

My answer assumed the amp volume/gain setting remains the same - in which case the tweeters are unaffected.

Sure - if he uses woofers disconnected to allow him to max out the amp volume, then magic smoke may ensue depending on the output capability of the amp and any resulting clipping, compared with tweeter rating.

 

[killdozzer]

@danadam @tonycollinet Thank you very much! You did some work.

Yes, I did think of some sort of upscaling. Point being that it's needed. Without it, it would be a loss. The process you're describing almost looks like a trade off. You take away volume and substitute it with information. If I understood correctly, this would imply that 24bit DAC always upscales 16bit. Do they all work in this fashion or is this a tech spec one should look for?

 

[killdozzer]

@danadam @tonycollinet

One other thing, this almost looks like an ideal loudness. At least before it starts taking away from the signal.

 

[antcollinet]

@danadam @tonycollinet Thank you very much! You did some work.

Yes, I did think of some sort of upscaling. Point being that it's needed. Without it, it would be a loss. The process you're describing almost looks like a trade off. You take away volume and substitute it with information. If I understood correctly, this would imply that 24bit DAC always upscales 16bit. Do they all work in this fashion or is this a tech spec one should look for?

I'm not an expert on the internals of dacs - but I think they have to. You can see from the two sinewave in my chart. If they didn't upscale the 16 bit values of +/- around 32K, for a dac output that expects values +/- around 8 million at full scale, then you'd get a tiny ouput voltage from the dac.

I think the point is the 24 bits can fully contain all the information in the 16 bit signal even when it is upscaled by the 48dB. It then allows you to scale that amount down by any arbitrary amount down to -48dB and STILL retain all the information.

 

[antcollinet]

@danadam @tonycollinet

One other thing, this almost looks like an ideal loudness. At least before it starts taking away from the signal.

Not sure what you mean by that.

 

[danadam]

The process you're describing almost looks like a trade off. You take away volume and substitute it with information.

Not sure I understand. To me it doesn't look like trade off, but rather like the only solution that makes sense.

If I understood correctly, this would imply that 24bit DAC always upscales 16bit. Do they all work in this fashion or is this a tech spec one should look for?

I don't know how common it is, but some DACs don't even accept 16-bit as an input, so the upscaling is done already on the OS/driver side. And for those that do accept 16-bit, as I wrote above, I can't think of a reason why they would do anything other than the upscaling.

According to ALSA on my machines:

• ADI-2 Pro FS accepts only 32-bit data,
• AudioQuest DragonFly Red accepts only 24-bit data,
• Fiio K3 accepts both 32-bit and 16-bit data

debian]$ cat /proc/asound/Pro57760455/stream0
RME ADI-2 Pro (57760455) at usb-0000:00:13.2-3, high speed : USB Audio

Playback:
  Status: Stop
  Interface 1
    Altset 1
    Format: S32_LE
    Channels: 2
    Endpoint: 0x02 (2 OUT) (ASYNC)
    Rates: 44100, 48000, 88200, 96000, 176400, 192000, 352800, 384000, 705600, 768000
    Data packet interval: 125 us
    Bits: 32
    Channel map: FL FR
    Sync Endpoint: 0x82 (2 IN)
    Sync EP Interface: 1
    Sync EP Altset: 1
    Implicit Feedback Mode: No

debian]$ cat /proc/asound/v10/stream0
AudioQuest AudioQuest DragonFly Red v1.0 at usb-0000:00:12.1-3, full speed : USB Audio

Playback:
  Status: Stop
  Interface 2
    Altset 1
    Format: S24_3LE
    Channels: 2
    Endpoint: 0x03 (3 OUT) (ASYNC)
    Rates: 44100, 48000, 88200, 96000
    Bits: 24
    Channel map: FL FR
    Sync Endpoint: 0x83 (3 IN)
    Sync EP Interface: 2
    Sync EP Altset: 1
    Implicit Feedback Mode: No

ubuntu]$ cat /proc/asound/K3/stream0
GuangZhou FiiO Electronics Co.,Ltd FiiO K3 at usb-0000:00:14.0-8, high speed : USB Audio

Playback:
  Status: Stop
  Interface 1
    Altset 1
    Format: S32_LE
    Channels: 2
    Endpoint: 1 OUT (ASYNC)
    Rates: 44100, 48000, 88200, 96000, 176400, 192000, 352800, 384000
    Data packet interval: 125 us
  Interface 1
    Altset 2
    Format: S16_LE
    Channels: 2
    Endpoint: 1 OUT (ASYNC)
    Rates: 44100, 48000, 88200, 96000, 176400, 192000, 352800, 384000
    Data packet interval: 125 us

 

[killdozzer]

Not sure what you mean by that.

Maybe I worded it poorly English not being my first language (or me just speaking it poorly). If you can attenuate the signal without changing it, it looks like ideal attenuation. This is what otherwise 'loudness' button is supposed to do, that's why I say it looks like ideal 'loudness' (by that I mean function on an amp). My Yamaha has a dial that promises exactly this - a nonlinear attenuation that takes care of following a desired curve.

Not sure I understand. To me it doesn't look like trade off, but rather like the only solution that makes sense.

Same as above, I'm probably to blame. One coming in place of the other I tried to describe as trade off. You give volume in order for signal to remain the same. A poor choice of term.

Yes, solution makes sense to me as well. If you don't need high SPL, you can use that part of info to preserve the signal. I'm all for that. I just didn't know a DAC can work in this way. It's very interesting.

I remember some audiophiles having fits about up-scaling feeling it to be "false", "inserted/created", "not part of the original signal" etc. Not that I care, but I supposed this (usually high paying) part of the market will be catered with some "boutique / audiophile" models of DAC that don't upscale. Otherwise, I don't know how they get through their day.

 

[killdozzer]

• ADI-2 Pro FS accepts only 32-bit data,
• AudioQuest DragonFly Red accepts only 24-bit data,

Only one more question, if you find the time; how is this manifested? Merely as an example - imagine you have a CD Transport, you play red-book CD and you connect your Transport into one of these DACs, would it simply not play or immediately up-scale it?

 

[danadam]

Only one more question, if you find the time; how is this manifested? Merely as an example - imagine you have a CD Transport, you play red-book CD and you connect your Transport into one of these DACs, would it simply not play or immediately up-scale it?

My examples were about USB connection, CD Transport would be rather connected over S/PDIF, I think. In that case I only know what wikipedia says:

S/PDIF is meant to be used for transmitting 20-bit audio data streams plus other related information. To transmit sources with less than 20 bits of sample accuracy, the superfluous bits will be set to zero. S/PDIF can also transport 24-bit samples by way of four extra bits; however, not all equipment supports this, and these extra bits may be ignored.

 

[killdozzer]

My examples were about USB connection, CD Transport would be rather connected over S/PDIF, I think. In that case I only know what wikipedia says:

OK, here's an example for USB connection; you have EAC of a Red Book CD ripped on your PC. You send that signal to one of those DACs that don't accept lower than 24-bit. Result is you don't hear anything or it gets up-scaled?