mark... its as if they ran out of register bit space.. limited length. Higher clipping level is one New feature of the DAC-3.
in order to increase the top end, they dropped off the low level end. Looks better in measurement also.
Something along those lines ??
THx-RNMarsh
in order to increase the top end, they dropped off the low level end. Looks better in measurement also.
Something along those lines ??
THx-RNMarsh
Last edited:
That is what i think of when reverb tails are mentioned.... the recorded room reverb decay time...... but the music decaying at its end is from the room RT60 decay.
-RNM
Any news on the dual voice coil feedback woofer? I mentioned it here https://www.diyaudio.com/forums/sub...eaker-project-circa-1981-a-8.html#post5729022
The tails are probably produced by a variety of reverbs and delays layered on top of each other when pop music is mixed. In live recordings with 2-mics it is the room. But they tails don't exist in isolation in the recordings, so hard to measure them using commercial CD sources. One could certainly try to make one's own CD with some impulses or pulsed noise and then measure the decay spectrum coming out of different dacs or the same dac with different configuration settings. I say one could try it because I don't know a priori if dacs will respond to contrived test signals the same way they do to music recordings, since I don't have an exact theory of what is going on in the dacs to produce the effect, other than I can make the effect happen more or make it happen less by changing things that have jitter variations in common. No harm to try using test signals though, it might work quite well, just have to see. That being said, I have more work already lined up that I prioritize higher than measuring something that is clearly audible and for which the measurements are very unlikely to be useful for fixing the problem. The empirical evidence I have now is that I should continue to work on jitter reduction.
Last edited:
Quite true. Didn't try to measure those things, but distance was a few feet in front of the speakers (maybe 6' or 8', maybe a bit less, don't remember). Angle would have been hard to estimate because is was small. I would probably need two long rods hinged together at one end to use as a pointer for the two lateral extents of the perceived source. After that I could measure the angle I set the contraption to. Didn't bother with anything like that however. Not high on my priority list.
Register bit space??
They attenuate by 3.5 dB to give the intersample over headroom. Such low level detail lost by going from 129 dB to 125.5 dB dynamic range 🙄.
Yes, really. Not the low level ones.
I'm not the only one to notice the issue either. I noticed the effect when comparing Allo Katana, my modded dac, and DAC-3 through the same power amp and speakers. Not long after that, maybe 2-3 weeks Abraxalito posted a link in his dac thread about the problem. It's a long rambling article, but you could probably search for DAC3 or DAC-3 or maybe Benchmark and get to the interesting part. The article is about a rather expensive not very accurate dac that reproduces reverb tails pretty close to the master tapes that the CD recordings were made from. The guy bought some of master tapes and even the tape machines they were recorded on in an effort to find out what CDs should actually sound like on a fully functional good dac. Audio Fur and the Border Patrol DAC | Part-Time Audiophile
Just to correct a couple misconceptions posted here the last few days by those that should know better. First, second order distortion does not grow linearly it follows a power law as all weakly non-linear transfer functions do. Basic second and third order intercept stuff.
Second you do not need complimentary devices to cancel second order distortion, you only need a fully differential signal path. Again two identical signal paths with a weak second order distortion will have the second order distortion cancel when you look at the difference.
This is not the place to learn the basics.
Second you do not need complimentary devices to cancel second order distortion, you only need a fully differential signal path. Again two identical signal paths with a weak second order distortion will have the second order distortion cancel when you look at the difference.
This is not the place to learn the basics.
Just to be clear - My comment is based on that what you say you hear is the last part of the decay of room reverb. If it's something else you are talking about the my comment is not valid.
//
Many years ago i did a test on an amplifier and I changed the bias as I listened.
Using a THD meter, I did some experiments to see how much THD I could just detect.
Using a book shelf 2-way speaker of medium-good quality, I could not reliably detect a delta/ difference between no distortion and about 0.1%. At .05% it was impossible.
Today, I have to revise that number downward...... Using high quality headphones this time of much lower distortion than the speakers, I could easily hear a difference with .04% added (2H and 3H). In fact with music, i didnt like it at all.
This time I was listening to a headphone amplifier.... It uses Ic opamps with a pair of compl bipolar buffers added on. With high Z load the thd is well below -100dbv. But when I loaded it with 50 Ohms (my headphones are 52 Ohms), the THD went up to .04%
So now I have to learn what the new lower limit to just detectable difference is.
THx-RNMarsh
Using a THD meter, I did some experiments to see how much THD I could just detect.
Using a book shelf 2-way speaker of medium-good quality, I could not reliably detect a delta/ difference between no distortion and about 0.1%. At .05% it was impossible.
Today, I have to revise that number downward...... Using high quality headphones this time of much lower distortion than the speakers, I could easily hear a difference with .04% added (2H and 3H). In fact with music, i didnt like it at all.
This time I was listening to a headphone amplifier.... It uses Ic opamps with a pair of compl bipolar buffers added on. With high Z load the thd is well below -100dbv. But when I loaded it with 50 Ohms (my headphones are 52 Ohms), the THD went up to .04%
So now I have to learn what the new lower limit to just detectable difference is.
THx-RNMarsh
It's difficult to try to tease out exactly what you and John want in a design Richard.
0.04% distortion is not difficult to achieve - in fact I'd go as far as to say that's a pretty pedestrian performance figure if we are talking modern amplifiers (lets leave the esoteric stuff out of this for the sake of argument).
So if a headphone amp, or a power amp for that matter is producing, worst case, distortion levels that are orders of magnitude below this (THD, IMD whatever), how can anyone legitimately be able to claim they can hear differences?
0.04% distortion is not difficult to achieve - in fact I'd go as far as to say that's a pretty pedestrian performance figure if we are talking modern amplifiers (lets leave the esoteric stuff out of this for the sake of argument).
So if a headphone amp, or a power amp for that matter is producing, worst case, distortion levels that are orders of magnitude below this (THD, IMD whatever), how can anyone legitimately be able to claim they can hear differences?
I recall listening to some rock music and figuring out that the lead singer was recorded inside a separate booth by the way his voice echoed off the walls, vs the rest of the ensemble.
This was on a modified $60 clone of an older arcam usb dac through a decent fet head-amp, and $50 headphones.
This was on a modified $60 clone of an older arcam usb dac through a decent fet head-amp, and $50 headphones.
Just do an RT60 measurement - either recorded in a hall, or with an impulse source into your reproduction chain.
The more you try and describe what you mean by "reverb tails" the more vague it's becoming, to me. Above, it sounds like you mean a deliberately recorded artefact, and not a natural phenomenon of a room.
Very confusing.
What sort of jitter do you think is audible?
Same here with a Bob Marley LP, about 5yr. later I saw a documentary on the recording the band was in a warehouse each member in their own isolated recording space.
Why don’t you write to them?
I’d be mightily unhappy with that. But I’m wondering if it’s not something to do with they way they filter or some other issue - agree with you this sounds like a DSP issue.
- Status
- Not open for further replies.