frodeni
100+ Head-Fier
- Joined
- Aug 27, 2015
- Posts
- 141
- Likes
- 28
Next up is imaging.
Again, this is used to describe a sonic trait, which is the positioning of sounds within the soundstage.
The soundstage is the room in which these sounds can possibly be placed.
If not using imaging to describe this, another term must be deviced, to describe the sonic trait.
As usual, things are not cut in stone. That is why this in the forums in the first place.
As usual, I start of with what is out there, on the topic.
http://www.stereophile.com/content/sounds-audio-glossary-glossary-r-s#w6O96HtboD7HDUGd.99
http://www.stereophile.com/content/sounds-audio-glossary-glossary-i-m#KsZDLsEi4Q1Z1KoO.99
http://www.head-fi.org/t/220770/describing-sound-a-glossary
https://en.wikipedia.org/wiki/Stereo_imaging
[size=17.03px]Width[/size]
Placement by width of a guitar, will depend upon the recording. For a single mic recording, it should be pin pointed. All string action should be in the same position by width. The guitar case, and its sounds, also should fall into the same space in width.
For an acoustic recording, using two mics, this is not the case. The string play on the neck of the guitar, should be reproduced to the right. Any action of fingers playing them, should be slightly to the right as of width. The main hitting of the strings, the right handed play, should be slight to the left as by width. They might also move by width, as the neck play is not fixed by the width axis.
The body of the guitar, the case, is bigger in size. It should thus extend by width. The string play should be more pinpointed, and postitioned on top of the wider guitar case.
[size=17.03px]Dept[/size]
A guitar recording usually has little dept, as the guitar has little physical dept. But the guitar could easily be a part of a bigger setting.
For multi mic recordings, a single mic is used for the guitar. All it sounds should thus be at one single point by dept.
For an acoustic recording, using just two mics, the placement could also be by dept. But in this case, the reproduction is not pin pointed. Not at high magnification.
If a guitar plays with a philharmonic orchestra, It may be in the back or the front. Depending on how small it is in the picture, or its magnification, the extension by width will differ by the dept of the magnification.
[size=17.03px]Width[/size]
A philharmonic, recorded by two mics, will usually extend the entire width. Since magnification for each instrument is low in magnification, they are more pin pointed and extends little by width. There is often groups of similar instruments, and this group can be heard by width.
[size=17.03px]Dept[/size]
In a philharmonic, different instruments are placed in rows. Some in front, some in the back. The positioning by dept, both by position and extent, is imaging.
Thus:
Imaging is the positioning of sounds in width and dept. (as usual, it probably can do with a lot of work)
I would suggesting to break it down to accuracy, and extension.
Some gear, only manges to pinpoint parts of the sound of an instrument, while other parts are drifting, or impossible to pinpoint.
Sometimes, the position seems to drift, often shifting by tone.
A philharmonic is a great example. Playing a tune like "I dovregubbens hall" by Grieg, accuracy might start out pretty fine, but at the end of that insane piece of art, as everything is at the climax, pinpointing anything might be tricky with a lot of gear. Both for width and dept. Some recordings starts out really carefully with that tune, and the ability to pinpoint at low volume, all the instruments, often sets gear apart.
For some reason, that guitar often has it elements, the right and left hand play, and the box, placed all wrong. Speakers in particular, as they often have separate planes separated by height for different speaker elements.
Headphones are not great at dept. Nor should they be, as the soundstage is shallow by dept. But for accuracy by width, some are pint point accurate. Or so I am told. My HD800 only do 16/44.1 music, and strings for the philharmonics wonder a bit in the soundstage for a number of recordings, and are hard to accurately pinpoint. Recordings differ a lot, and for some, everything seems to be spot on for imaging.
Accuracy are often times frequency dependent. The lows might not have much accuracy at all. Base might just be a rumble. Rigs plagued with digital noise often times struggle with imaging cymbals or perk in general.
Once you get the hand of it, you will often times hear mics picking up other instruments or singers, as they are meant to do. Singers will thus sometimes wander in the soundstage. They do for "Chess", to name one.
I have given a lot of examples above, others could fill inn more details as well.
Also, you need to stick with what you hear, and not what you assume to be correct. Some compression algorithms, removes non essential sounds first, like room acoustics. Like echo effects. They also remove the finer details of instrument details. This will improve the pinpointing of these instruments, as a lot of noise surrounding them, simply is removed. Their main traits remain. Less is rendered, thus it is easier to image. Less confusion.
So, imaging may be more accurate for lossy compressed playback, than the lossless original. This is why, sonic traits needs to be broken into quantifiable pieces. Just looking at imaging alone, is bloody confusing, giving plenty of false positives.
This is plain hard to do by ear, and given the current stage of tech, setting up an objective test for this, will be extremely hard. In particular on the headphone and speaker part. Hopefully, there will be thrown a Nyquist like solution at it sometime in the future, resulting in a brilliant way to test this.
For now, I have none.
Again, this is used to describe a sonic trait, which is the positioning of sounds within the soundstage.
The soundstage is the room in which these sounds can possibly be placed.
If not using imaging to describe this, another term must be deviced, to describe the sonic trait.
As usual, things are not cut in stone. That is why this in the forums in the first place.
As usual, I start of with what is out there, on the topic.
What is out there?
Stereophile
http://www.stereophile.com/content/sounds-audio-glossary-glossary-r-s#w6O96HtboD7HDUGd.99
http://www.stereophile.com/content/sounds-audio-glossary-glossary-i-m#KsZDLsEi4Q1Z1KoO.99
Head-Fi article
http://www.head-fi.org/t/220770/describing-sound-a-glossary
Wikipedia
https://en.wikipedia.org/wiki/Stereo_imaging
Breakdown
There is more agreement on this term. Agreement is a great thing.The guitar
A run through of the reproduction of a guitar, as of imaging. Played right handed.[size=17.03px]Width[/size]
Placement by width of a guitar, will depend upon the recording. For a single mic recording, it should be pin pointed. All string action should be in the same position by width. The guitar case, and its sounds, also should fall into the same space in width.
For an acoustic recording, using two mics, this is not the case. The string play on the neck of the guitar, should be reproduced to the right. Any action of fingers playing them, should be slightly to the right as of width. The main hitting of the strings, the right handed play, should be slight to the left as by width. They might also move by width, as the neck play is not fixed by the width axis.
The body of the guitar, the case, is bigger in size. It should thus extend by width. The string play should be more pinpointed, and postitioned on top of the wider guitar case.
[size=17.03px]Dept[/size]
A guitar recording usually has little dept, as the guitar has little physical dept. But the guitar could easily be a part of a bigger setting.
For multi mic recordings, a single mic is used for the guitar. All it sounds should thus be at one single point by dept.
For an acoustic recording, using just two mics, the placement could also be by dept. But in this case, the reproduction is not pin pointed. Not at high magnification.
If a guitar plays with a philharmonic orchestra, It may be in the back or the front. Depending on how small it is in the picture, or its magnification, the extension by width will differ by the dept of the magnification.
The philharmonic
A quick run through for a philharmonic[size=17.03px]Width[/size]
A philharmonic, recorded by two mics, will usually extend the entire width. Since magnification for each instrument is low in magnification, they are more pin pointed and extends little by width. There is often groups of similar instruments, and this group can be heard by width.
[size=17.03px]Dept[/size]
In a philharmonic, different instruments are placed in rows. Some in front, some in the back. The positioning by dept, both by position and extent, is imaging.
Wrap-up
Imaging as an experience is made up of a lot of sonic traits, but imaging as a sonic trait term, is best suited to define positioning of sounds only.
Thus:
Imaging is the positioning of sounds in width and dept. (as usual, it probably can do with a lot of work)
Quantifying
The whole point of such a definition is that it is quantifiable, easy to recognize, and easy to describe.I would suggesting to break it down to accuracy, and extension.
Accuracy
Accuracy is simply the ability to pinpont the location of a sound. And keep it there. This should apply for all frequences as well.Some gear, only manges to pinpoint parts of the sound of an instrument, while other parts are drifting, or impossible to pinpoint.
Sometimes, the position seems to drift, often shifting by tone.
A philharmonic is a great example. Playing a tune like "I dovregubbens hall" by Grieg, accuracy might start out pretty fine, but at the end of that insane piece of art, as everything is at the climax, pinpointing anything might be tricky with a lot of gear. Both for width and dept. Some recordings starts out really carefully with that tune, and the ability to pinpoint at low volume, all the instruments, often sets gear apart.
For some reason, that guitar often has it elements, the right and left hand play, and the box, placed all wrong. Speakers in particular, as they often have separate planes separated by height for different speaker elements.
Headphones are not great at dept. Nor should they be, as the soundstage is shallow by dept. But for accuracy by width, some are pint point accurate. Or so I am told. My HD800 only do 16/44.1 music, and strings for the philharmonics wonder a bit in the soundstage for a number of recordings, and are hard to accurately pinpoint. Recordings differ a lot, and for some, everything seems to be spot on for imaging.
Accuracy are often times frequency dependent. The lows might not have much accuracy at all. Base might just be a rumble. Rigs plagued with digital noise often times struggle with imaging cymbals or perk in general.
Extension
Some sounds should have extension. This is different from accuracy. The reproduction should extend by space. Like the box of a guitar. This is typically achieved by great accuracy. Like that guitar box, that extends around the strings being played by the right hand.Physics
The physics is explained in the soundstage thread.Subjective testing
Imaging is tricky to subjectively test. The reason being the listening material in particular. Get to know a recording well, and stick with it. You will need to know the music by heart.Once you get the hand of it, you will often times hear mics picking up other instruments or singers, as they are meant to do. Singers will thus sometimes wander in the soundstage. They do for "Chess", to name one.
I have given a lot of examples above, others could fill inn more details as well.
Also, you need to stick with what you hear, and not what you assume to be correct. Some compression algorithms, removes non essential sounds first, like room acoustics. Like echo effects. They also remove the finer details of instrument details. This will improve the pinpointing of these instruments, as a lot of noise surrounding them, simply is removed. Their main traits remain. Less is rendered, thus it is easier to image. Less confusion.
So, imaging may be more accurate for lossy compressed playback, than the lossless original. This is why, sonic traits needs to be broken into quantifiable pieces. Just looking at imaging alone, is bloody confusing, giving plenty of false positives.
Objective testing
Channel separation could be a pre-speaker test.This is plain hard to do by ear, and given the current stage of tech, setting up an objective test for this, will be extremely hard. In particular on the headphone and speaker part. Hopefully, there will be thrown a Nyquist like solution at it sometime in the future, resulting in a brilliant way to test this.
For now, I have none.
