For example this one:
Analysis of speaker cables
But, I'm sure that your remember having raised the "human head movement problem" in a a lot of other threads as well.
Is I recall, 1 sample of red book CD is 10 degrees. Or was it 20 - significant. Cant find the info source...
//
1 sample at used FS. One could manipulate the clock phase but it would require different clocks per channel....
//
The ability of humans to detect interchannel time differences as low as (at least) 10us was long established, therefore quite surprising that the first cd-players only used one da-converter for reproduction.
The MAA (minimum audible angle) of 1 degree corresponds to roughly 10us, but is frequency dependent and sensitivity is best in the frontal median plane.
As usually some distribution exists among humans it is quite probable that some will be able to discern even smaller ITDs.
The MAA (minimum audible angle) of 1 degree corresponds to roughly 10us, but is frequency dependent and sensitivity is best in the frontal median plane.
As usually some distribution exists among humans it is quite probable that some will be able to discern even smaller ITDs.
So your saying moving a one speaker 2mm closer will shift the image noticable?
Hard to believe.
More thoughts. How does ITD relate to 2 speakers and phantom image. The ITD happenes from both speakers but we hear it as one sound in the middle. So moving one speaker forward doesn't really change the ITDs of either speaker just the ITD between speakers, and this may not be as audible. And it seems that cable dependent ITD is the same as one speaker Moving as not as audible as an actual ITD chage in a mono signal.
I linked to this earlier too Stereo Perception, Sound Localization & Auditory Cues Although we use ITD all the time to locate sound sources it explains how for stereo reproduction using speakers it is largely irrelevant and why panning is used instead. The phrase "inter channel delay" has crept into the conversation at least a couple of times as if it's the same as ITD, I don't think it is.
Last edited:
NO!!! It is absolutely NOT!!!!!!
Love using the caps, make my argument stronger, right?
I have detailed what cable "geometry" is needed to reduce ITD variation despite variations in speaker impedance vs frequency vs acceleration.
I have detailed how to spot a sensitivity between cables and speaker impedance.
I have presented a design of cables which alleviates the issue.
My "design" requires quadrupling the number of bog standard zip cables, using #18 or #16 zip cables in parallel to drop the RFZ.
I do not sell cables.
I do not sell anything.
My "solution" does not require expensive cables. And, it requires only a specification of L, C, and R.
....And, I do not clutch at straws.
Jn
Last edited:
In point of fact, my foray into ITD was started because I was unable to get a centered proper lateralization with my headphones. It turned out my computer soundcard had one DAC and two S/H. There was a half rate delay between ears I could not compensate using the pan control.
Jn
If we agree, that this problem doesn't exist, I'm fine with that.
Indeed the "ITD between speakers" is the foundation of the socalled "time-arrival-stereophony" and there were quite a few studies examing the effects. Afair, I've linked some in the past but could do again it if there is some interest.
Humans are different and so it is not surprising that some are more sensitive to missing time difference information than others.
For example JJ is on the records that he can't stand listening to records done with pure pan-potting.
It is still a mistery why most humans perceive a virtual sound source in the median frontal plane if both speakers distribute the same signal. And our listening sense has to cope with the fact, that very similar signals are arriving at different times due to the two speaker arrangement, but it is still a pattern anyway. And our listening sense seems to be quite good in detecting differences in patterns, so if in this situation the time pattern changes, it can be detected.
As said before, it is frequency dependent (due to physiological reasons) and usually the detection ability vaires from human to human (due to training, physiological differences and habits).
Last edited:
With 44.1 Khz, each new sample comes every 23usec.
With 2 channels from one DAC, samples come in between, which is 12,5 usec inter channel delay.
However almost all modern DACs upsample to 192 Khz, meaning that the interchannel delay is only 2.6usec.
In both cases I don’t think it can be audible and if so, I strongly suspect other causes like unequal FR of the reconstruction filters in combination with small level differences.
Looking at low cost DAC test, slightly unequal FR and gain is often the case.
Hans
With 2 channels from one DAC, samples come in between, which is 12,5 usec inter channel delay.
However almost all modern DACs upsample to 192 Khz, meaning that the interchannel delay is only 2.6usec.
In both cases I don’t think it can be audible and if so, I strongly suspect other causes like unequal FR of the reconstruction filters in combination with small level differences.
Looking at low cost DAC test, slightly unequal FR and gain is often the case.
Hans
- Status
- Not open for further replies.