This particular topic will likely produce some of the most inconsistent responses, with nearly all of them having some validity. The same system with two differing files at the same levels will produce two differing results.
First, if the DAC is using digital attenuation to address amplitude, the ability to attenuate without loss will depend on the bit depth of the file. CD's offer 16 bit of data and LPCM files are often at 24 bit (but then we can also digress into the LSB/MSB factors as well). But lets go with the lowest factors for this example. Using red book standard, a 32-bit volume control can fully attenuate a digital signal without affecting its dynamic range. This being the most common format and the typical file format most music is published in gives it a pretty good starting point. If we move into 24 bit files, this dramatically reduces down to 44db of attenuation before perceptible loss. This can occur during high resolution playback and will be very system dependent on how much impact will occur.
This by no means addresses the interaction between the analog stages of the pairing. It can be measured, but often must be done by ear to determine how well they may interact.
You could even achieve differing results with a speaker change. If one is significantly more sensitive and requires more attenuation, its resulting sound may differ more so than the speaker change as the digital dither could have greater impact at the differing amplitude levels.
So I don't believe any answer provided can established a preferred route. Just a demonstration that differing configuration may have varied results.
First, if the DAC is using digital attenuation to address amplitude, the ability to attenuate without loss will depend on the bit depth of the file. CD's offer 16 bit of data and LPCM files are often at 24 bit (but then we can also digress into the LSB/MSB factors as well). But lets go with the lowest factors for this example. Using red book standard, a 32-bit volume control can fully attenuate a digital signal without affecting its dynamic range. This being the most common format and the typical file format most music is published in gives it a pretty good starting point. If we move into 24 bit files, this dramatically reduces down to 44db of attenuation before perceptible loss. This can occur during high resolution playback and will be very system dependent on how much impact will occur.
This by no means addresses the interaction between the analog stages of the pairing. It can be measured, but often must be done by ear to determine how well they may interact.
You could even achieve differing results with a speaker change. If one is significantly more sensitive and requires more attenuation, its resulting sound may differ more so than the speaker change as the digital dither could have greater impact at the differing amplitude levels.
So I don't believe any answer provided can established a preferred route. Just a demonstration that differing configuration may have varied results.