JSmith
Major Contributor
Digital attenuation is just a multiplication of all the data samples by a fixed coefficient.What if digital attenuation can bring loss of resolution because of data compression?
JSmith
Digital attenuation is just a multiplication of all the data samples by a fixed coefficient.What if digital attenuation can bring loss of resolution because of data compression?
Digital attenuation is just a multiplication of all the data samples by a fixed coefficient.
JSmith
PCM data is composed by integer numbers. If you multiply for a fraction you can have rounding errors.
24 bits is enough resolution to basically make it a non issue.
@Kal Rubinson says it can be done. There’s is DSP control somewhere in settings with parametric EQ and what not, I’d start looking there.
The level of the lowest bit of a 24 bit source is about 144 dB below maximum level. This then is the level of the rounding errors (a bit crude). So you can attenuate up to 24 dB (144 - 120) before rounding errors reach audible levels at 120 dB below maximum signal.Fair enough, but, that is if you are getting an 16bit source and up-sampling to 24bit to avoid the rounding issues. But, what if you are dealing with an already 24bit (or 32bit) source?
... plus dithering (at least for 16 bit sources).Digital attenuation is just a multiplication of all the data samples by a fixed coefficient.
Digital attenuation can also hide flaws in a DAC. There are DACs where THD rises close to 0 DBFS which may be audible but not when running attenuated.Hmmm, but in that case you would be modifying the source, no? What if digital attenuation can bring loss of resolution because of data compression?
How are you determining how much to attenuate?
I think you are doing it right. I had the same experience with 2 different amps which I could not tell apart level matched and blind. Lots of people read about level matched blind testing but experiencing it for yourself is powerful.OK, so, I've got two RCA passive attenuators (basically boxes with 20k pots) and one RCA A/B passive switcher. I hocked two different DACs to my PC and used JRiver to play on both at the same time the same files. I first made one big 440Hz tone file and played on both and used my multimeter on V(AC) to equalize both (both played on about 1.8V). Plugged both to the A/B box and the box to my L50 headphone amp. Switched back and forward, change DACs...
Conclusion: I just don't have enough "golden" ear to hear a difference ... They sound so similar that is kind of splitting hairs to find anything to pay attention to. And almost any combination of two DACs sound extremely similar. I feel that I am doing something wrong...
Do you guys have any specific music recommendation that let the differences more obvious?
I think you are doing it right. I had the same experience with 2 different amps which I could not tell apart level matched and blind. Lots of people read about level matched blind testing but experiencing it for yourself is powerful.
I "can" hear "some" difference but let's be honest, it is close to nothing!