as I've mentioned, the cables on my main system are difficult to change, screw connections behind cabinets etc, + is an active bass set-up. switching back and forth just isn't going to happen !!
when I get time I may see if I can find some cable that make a difference on one of my secondary systems, but those first 4 words are a killer at the moment.
back to marking assignments !!
when I get time I may see if I can find some cable that make a difference on one of my secondary systems, but those first 4 words are a killer at the moment.
back to marking assignments !!
Andy Graddon said:Its such an odd result.. its seems to indicate you might have been recognising a difference but the wrong way around.
that might indicate you were listening for the wrong cues, why is anybody's guess, but maybe when listening sighted you had an expectation of a particular difference, but they were actually the other way around. dunno !!!
anyway, bet its going to bug you for a while
It does annoy me a bit. I suspect having got the first cable wrong the overall "negative" result was inevitable - ie when I heard a change I assumed it was the other cable. Nine wrong cables in a row ...
fredex said:
The distinction is blurry, at least to me..It hurts to think that much...
fredex said:
Why would it seem obvious???
The sillyness in the entire premise all the audio testing I've seen to date strikes me funny. (the alternative is to cry, but luckily I am not an "audio" guy..
Using an adaptive instrumentation system to examine absolute changes is entirely absurd. I cannot believe it is still used.
Use differential techniques to distinguish alterations.
Here's a simple speaker wire test..this will work only if the resistances are comparable...btw..
1. Use a mono signal.
2. With two identical speaker wires in place, sit in the hotseat, adjust the balance to produce the image dead center. Use an object as a visual reference to center...
3. Swap ONE CABLE...ONLY ONE.
4. Listen without adjustments of any kind to the system.
You are now using the human localization capability to discern any alterations in ITD and IID. And need I remind you how incredibly good humans are at doing this??
I await.
Cheers, John
Edit...LAZY?? Use two zip cords. After initial centering, split the zip the entire length to increase the inductance. Wanna go further? Use two ribbon sets, and then split one...this'll go from 15 nH per foot to about 300 nH per foot..
Too late to edit the previous post...
Meant to say split one of the zips.
Here's a graph of a #12 parallel wire inductance. As is seen, the inductance of the pair will go from a typical of about 150 to 160 nH per foot, to a value of about 600 nH per foot at five inch conductor spacing.
For a 10 foot run, that translates into a difference between 1.5 uH to about 6 uH.
Cheers, John
Meant to say split one of the zips.
Here's a graph of a #12 parallel wire inductance. As is seen, the inductance of the pair will go from a typical of about 150 to 160 nH per foot, to a value of about 600 nH per foot at five inch conductor spacing.
For a 10 foot run, that translates into a difference between 1.5 uH to about 6 uH.
Cheers, John
Attachments
jneutron said:
How incredibly good? The word "useless" comes to mind. I sit in an open office area with about ten phones. My team of four people are constantly wrong in guessing which phone is ringing. By up to 180 degrees. And not consistently wrong either. We may be okay at localisation outdoors, but in a room I couldn't think of a worse test procedure.
tnargs said:
For you perhaps, but certainly not for the bulk of the human race.
I trust you only listen to monophonic playback? Or you have only one ear.
tnargs said:
A worse test procedure than what? Are you changing the cables on the phones?😉
The entire reason for a controlled experiment or test is to CONTROL the experiment. Bringing up silly contrived examples of chaos certainly does nothing for the discussion.
Within a room, the speakers are placed symmetrically, the listener is in a specified location, there are no confounding noises, no confounding first arrival reflections.
Within an office situation, the noisemakers (phones) are designed for omnidirectional emission, they are typically on desks, there are significant numbers of reflections.. none controlled. Many factors make your example a poor one within the context of the discussion.
Cheers, John
KSTR said:
Any perceived "height" difference in imaging is a pyschoacoustic phenomenon relating to proximity effect, not to actual height placement in a 2 channel soundfield.
You can't "mix" a sound higher, meaning you can't "pan" it higher. You can pan it right to left, or anywhere in between, but you can't go "up" or "down.
I've had this debate a number of times, and at quite great length.
It's impossible. If it were, recording engineers would be dancing in the streets, as the available increase in headroom for mixing would be a godsend.
Before you debate me, try a simple test with a pair of bookshelf speakers. Place them with the tweeter at ear height, listen for the "height" of a feature instrument or vocal, judge it's distance from the floor.
Take the same speaker, flip it upside down but adjust the height so that the tweeter is at the same listening height, now judge the height distance of the aforementioned feature instrument or vocal from the floor.
It will be the same.
There is no "height" plane in L/R two channel. You would need 4 channels, L/R plus U/D (up and down).
There is too much misinformation that is repeated about what "tricks" recording engineers can do in a studio.
Cheers
Movies mixes do it all the time by means of HRTF DSP. By your argument we can only perceive height by tilting our single two 'microphones' laterally located either side of the head.
rdf said:
Agreed - you can adjust the sound's reflection off the "floor" to make it progressively quieter and later, while reducing the primary sound volume slightly as it goes higher.
I have heard test CDs with sounds mixed in a way which makes them move vertically upwards in increments. Spooky!
Never heard it in music tho.
@tnargs etc
Because our location abilities for police/ambulance siren tones is so poor, some of them here now mix in bursts of white noise. I presume it is the higher frequencies which help. Phone ringing bounces around an office - you need anechoic wallpaper!
rdf said:
Movie mixes? You mean in 5.1? If you are talking about Q Sound type effects, that's artificially done with phase and distortion (I forget the methods, but I used to be up to speed on it years ago).
See the part in my post where I mentioned it can't be done in 2 channel.
Tilting microphones will change proximity to the sound source, FR response variations due to comb filtering effects because of the change of direct vs reflected sounds, and the difference in FR due to the polar pickup pattern of the microphones.
It will change the FR character and proximity (ratio of direct and reflected sounds, early and late reflections), but not the "height" of the image.
Take two microphones in an X/Y stereo configuration, and record someone snapping their fingers at various heights. Have them snap their fingers 1' off the floor, 4' off the floor, then 8' off the floor.
You will not be looking 8' in the air while listening to the playback of the snapping that was recorded at that height.
It doesn't work that way. If it did, like I said earlier, you'd hear about all these recording techniques to place images in vertical soundfield (in 2 channel audio).
It just doesn't exist. You can't pan a sound right or left, when you have only one speaker, you can't pan a sound up or down when you don't have 2 channels that are up and down.
Try the bookshelf speaker experiment I suggested earlier. It illustrates it perfectly.
Or better yet, try it with a pair of single driver speakers.
With an ambulance you are introducing doppler effects, a moving(in proximity/amplitude) sound source, changing reflections due to buildings etc..., long echoes, a sound source of varying pitch, and the fact that the "attack" of the sound is not a sharp transient (sounds that are sharp and percussive, are much easier to localize because the initial transient has a very fast [amplitude] rise and decay time, which leaves a lot of room for the early and late reflections to more clearly heard).
Cheers
Daygloworange said:
Yes, it mimics the same phase and distortion mechanisms generated by acoustic shadowing of the head and reflections/delays/equalizations of the pinna. The point is it can be done and those effects are still audible, if less effective, in 2-channel playback of 5.1 soundtracks. It also appears to be common with modern pop recordings.
That it mimics phase and distortion of human hearing is merely rhetoric. It makes for good marketing. How humans hear and loclize sounds, is very hard to plot and graph.
While novel, Q Sound type effects (in 2 channel) always sound very "out of phase" to me.
Those kinds of combining of L/R signals in both speakers with out of phase tricks to artficially create crosstalk separation always sound weird.
The image is always "weighted" to one side, and leaves a "sonic vacuum" in the opposing ear, and any shifting of your head makes for strange effects. Not what I call very natural.
Unless you resort to these synthesized sound effects, you will not get an image up in height in 2 channel audio.
Cheers
While novel, Q Sound type effects (in 2 channel) always sound very "out of phase" to me.
Those kinds of combining of L/R signals in both speakers with out of phase tricks to artficially create crosstalk separation always sound weird.
The image is always "weighted" to one side, and leaves a "sonic vacuum" in the opposing ear, and any shifting of your head makes for strange effects. Not what I call very natural.
Unless you resort to these synthesized sound effects, you will not get an image up in height in 2 channel audio.
Cheers
@ Alan Hope,
the test result mustn´t annoy you.
You have unerringly choosen the most difficult test setup.
While identification itself is a possible choice the DUTs are usually presented together with references to recalibrate your perception.
Did i already mention that some training under the exact test conditions could help? 🙂
@ ravon,
the statistical method used in the classical ABX-Setup is a onetailed test and dismiss therefore high listener failure rates.
As described earlier a situation in which a listeners detects a difference but gives constantly a wrong answer would not be detected by a onetailed statistical test.
A twotailed test is more suitable to this situation. That´s the reason why in Alan Hope´s statistical test the result is significant.
Due to the different methodoloy used in a ABX one can argue that listeners detecting differences but giving wrong answers are more unlikely and therefore a onetailed test could be justified.
But that position is imho not scientifically proved and therefore i´d argue it´s always better to use twotailed statistical tests.
Les Leventhal wrote a paper about this topic related to audio tests back in 1994:
Leventhal, Les `Statistical significant poor performance in listening tests´
JAES Volume 42 Issue 7/8 pp. 585-587; July 1994
@ Daygloworange,
if you reread KSTR´s post, you´ll see, that he exactly expressed the same point of view as you, but maybe not in the same wording. 🙂
But, and on that he is absolutely right, as stereo is an illusion, there are some ways to trick our perception into the feeling that sounds were coming from different directions than the horizontal plane spanned by the two loudspeakers in front of us.
And it is reported that some listeners percept a _phantom source_ located direct in front of their head as if it were elongated by up to 30 degree vertical.
That´s what KSTR´s post was about.
the test result mustn´t annoy you.
You have unerringly choosen the most difficult test setup.
While identification itself is a possible choice the DUTs are usually presented together with references to recalibrate your perception.
Did i already mention that some training under the exact test conditions could help? 🙂
@ ravon,
the statistical method used in the classical ABX-Setup is a onetailed test and dismiss therefore high listener failure rates.
As described earlier a situation in which a listeners detects a difference but gives constantly a wrong answer would not be detected by a onetailed statistical test.
A twotailed test is more suitable to this situation. That´s the reason why in Alan Hope´s statistical test the result is significant.
Due to the different methodoloy used in a ABX one can argue that listeners detecting differences but giving wrong answers are more unlikely and therefore a onetailed test could be justified.
But that position is imho not scientifically proved and therefore i´d argue it´s always better to use twotailed statistical tests.
Les Leventhal wrote a paper about this topic related to audio tests back in 1994:
Leventhal, Les `Statistical significant poor performance in listening tests´
JAES Volume 42 Issue 7/8 pp. 585-587; July 1994
@ Daygloworange,
if you reread KSTR´s post, you´ll see, that he exactly expressed the same point of view as you, but maybe not in the same wording. 🙂
But, and on that he is absolutely right, as stereo is an illusion, there are some ways to trick our perception into the feeling that sounds were coming from different directions than the horizontal plane spanned by the two loudspeakers in front of us.
And it is reported that some listeners percept a _phantom source_ located direct in front of their head as if it were elongated by up to 30 degree vertical.
That´s what KSTR´s post was about.
Daygloworange said:
As flatly incorrect as saying "how the ear perceives distortion is hard to plot, therefore lossy compression techniques are pure marketing rhetoric". And while I agree reproducing height without such manipulation is open to question, you're wrong to discount reports of height on DSP-manipulated material on the basis of what sounds like a poor job reproducing Q-sound. It certainly is nowhere like the ~220 degree wrap-around I've heard on Roger Waters' "Amused to Death" for example.
The greater point in all this however is discounting the DSP between your ears is a sure-fire route to wrong answers. The reproduction system starts at the instrument and ends at the illusion your brain creates. You can't discount what happens past the eardrum.
I'm not a fan of DSP based effects holographic effects.
I did not discount what DSP does, or anything else. Stop trying to set up strawmen. I simply said that DSP doesn't sound natural to me.
My whole point was discussing that you can't capture height placement in 2 channel audio with microphones with a standard L/R record/playback system.
I could care less about DSP effects....
Cheers
I did not discount what DSP does, or anything else. Stop trying to set up strawmen. I simply said that DSP doesn't sound natural to me.
My whole point was discussing that you can't capture height placement in 2 channel audio with microphones with a standard L/R record/playback system.
I could care less about DSP effects....
Cheers
You're implying reports of height in home reprodcution are erroneous based on questionable limiting assumptions regarding source material, or discussing a scenario of limited relevance. Personal feelings about DSP don't enter into it, the quoted statements are incorrect or deceptive by being incomplete. Height can be mixed into a recording with modern equipment, reports of height in home reproduction are not unreasonable if qualified with source material, it's not a proximity effect though the latter might augment it. Strawman?
The effect of cables are another question.
rdf said:
I remember many years back I had a CD by Dean Peer entitled "Ucross" During one of the songs ---{refered to as Tone Poems IIRC}--- the sound emanated from my speakers ---{which were the original Carver Amazings with a 60" line-source ribbon and four 12" woofers}--- at a level that seemed as if he was seated while playing this particular song. Then all of a sudden either Dean actually stood up from a seated postion while he was being recorded or else the engineer applied some technic to the recording because the sound's point of emanation moved up approx 12" or so! Everyone who I ever played this CD for commented how they had never so noticably heard sound move up before! It was a perfect example of a soundstage having HEIGHT as well as depth and width.
On another completely different note I had a CD that was either "Water Sign" or "Bamboo Forest" by Steve Reid ---{the ex-Rippingtons percussionist}--- that presented a completely normal soundstage presentation until all of a sudden in one song some music would emanate directly from my left side ---{as if the musician was standing or a speaker had been placed at a point that was exactly to the left side of my couch}--- it was a very eerie event and actually startled me the first time it occured.
I've never heard anything like that before or after without using surround sound or one of the 5.1 movie setups with it's additional amps and speakers. Yet these both occured using only a stereo setup with 2 speakers...
Reports? People report placing pictures of themselves in freezers improve audio playback.
Make a valid argument by proving me wrong with fact. I'd
love to learn the mechanics of capturing "height" in a recording, and how I missed that during my 20+ years of recording experience.
I've written enough about "how" it is not possible to "encode" height differential and "decode" it with normal record playback techniques.
It's simple physics. A microphone diaphragm goes in and out due to air pressure fluctuations of a sound source. A speaker driver does the inverse of what the microphone did.
A microphone only captures the pressure fluctutations at the diaphragm in an in and out movment of the diaphragm, it can't recognize where in the hemispere the pressure comes from.
Inversely, a driver can only push forward, it can't push air off to one side or the other, or UP or DOWN.
I don't know how much clearer I can explain this....
I have given you a number of simple examples and experiments, none of which you have tried (I'm sure, nor will you).
Cheers
Make a valid argument by proving me wrong with fact. I'd
love to learn the mechanics of capturing "height" in a recording, and how I missed that during my 20+ years of recording experience.
I've written enough about "how" it is not possible to "encode" height differential and "decode" it with normal record playback techniques.
It's simple physics. A microphone diaphragm goes in and out due to air pressure fluctuations of a sound source. A speaker driver does the inverse of what the microphone did.
A microphone only captures the pressure fluctutations at the diaphragm in an in and out movment of the diaphragm, it can't recognize where in the hemispere the pressure comes from.
Inversely, a driver can only push forward, it can't push air off to one side or the other, or UP or DOWN.
I don't know how much clearer I can explain this....
I have given you a number of simple examples and experiments, none of which you have tried (I'm sure, nor will you).
Cheers
First, chill out. We haven't even discussed physical/psychological reasons this could happen.
Get irritated at those who ignore the myriad of reasons cables sound the same and keep claiming their ears are trained and infallible.
Second, I would like to propose a theory on how we can hear height in recorded sounds played back on two speakers at the same elevation.
When sound originates from a "central source" and is picked up by two microphones it is a complex waveform consisting of direct and reflected energy-floor, ceiling and sides. Now if the source elevation is changed, the vertical direct/reflected ratio (time and tonal balance) will change and comb filtering will alter the spectral balance of the sound. Microphones will measure this and ear/brains will hear it. Also, floors are shiny and ceilings are usually rough and low density(might be backwards for studios). By altering direct/ceiling/floor ratios there will be a difference in sound.
The sound is then reproduced on speakers:
ALL speakers have nonuniform vertical directivity, so they will reproduce white noise at different linear distortion levels depending on the vertical angle, most likely with differences between a high measurement and a low measurement. (tweeter/fr driver is near top edge of enclosure, edge diffraction shows its ugly head)
1.Now if you know the sound is changing elevation your brain will explain this spectral imbalance as a height change, which is purely an auditory illusion.
2.The ear/brain combo can detect vertical changes, we do it every day. So if we are listening to music that has a certain tonal balance and the room has a certain "response" and all of the sudden the music has a shift in frequency our brain may recognize this shift from our daily lives as similar to another height change we experienced...
Not definitive but if someone wants to create and perform a test to prove it, it would not be that difficult.
Third, lets try and get back to cables.
- Status
- Not open for further replies.