New Record Day - Binaural Shootout (Revel, Klipsch, Spatial, Q Acoustics)

[ctrl]

Okay, admittedly I find binaural speaker comparisons cool - if the original source is also available. What I particularly like is that Ron makes the recordings in a still "normal" listening room (the room has been slightly optimized).

The problem with binaural recordings is that you don't have "standardized" headphones to judge the absolute sound of the recording. However, if the original source is also present, then you can perceive sound/timbre coloration by comparison with the original - and man are these audible in the recordings.

 

[ctrl]

Have put together a few samples from Ron's recording as an attachment.

First you always hear the original source, then the binaural recording.

The original source sounds "louder" than the binaural recordings, although all recordings were normalized to -1dB.
I know too little about recording technology to be able to fairly compensate for the "dynamic losses" due to the listening distance.

One speaker sounds like an old gramophone, with others this effect is also slightly perceptible

 

[amirm]

It is a good attempt but I am not getting much out of the demo. Spatialization is above my head rather than in front of me as one would want to hear from speakers. I hear the difference between the small and large speakers but that is it. Are you getting something else out of it?

 

[Vini darko]

I enjoyed this video by Ron. Was a fun way to spend an hour. It is intresting how much is lost in translation between the source file and demo through the same headphone. He must have put alot of work into setting it all up.

 

[richard12511]

Interesting video, and thanks for adding the original track. Ron is a member here.

@New Record Day , were all of the speakers in the youtube video level matched? Some of the speakers in the test sounded louder to my ears.

 

[ctrl]

Are you getting something else out of it?

In the small sample comparison (unfortunately had to convert back to mp3 because of the upload size) I hear clear differences in timbre. Particularly striking for me is that some speakers have a strong to slight "gramophone sound".

The dynamic differences are also clearly audible, but I'm not sure if Ron has done a careful level matching.

 

[richard12511]

For the piano tracks in the video:

Revel F226Be
Klipsch Heresy
Q Acoustics Concept 500
Spatial Sapphire
Spatial X5

I'm gonna spend some time going back and forth with those. I would actually advise against watching the whole video(just use the timestamped links), as the commentary is unfortunately biased in between the songs, so much so that it will no doubt bias your preference.

Still, comparing against the original track is probably the only true way to really tell, though it's testing for correctness rather than preference. @ctrl your clips are a little too short for me. They end before my brain has time to start making any sort of decision. Would you mind splitting them up into 5 files? Each file containing the source and 1 speaker?

 

[ctrl]

@ctrl your clips are a little too short for me. They end before my brain has time to start making any sort of decision. Would you mind splitting them up into 5 files? Each file containing the source and 1 speaker?

@amirm will definitely award me the Headless Panther for this.

 

[richard12511]

In the small sample comparison (unfortunately had to convert back to mp3 because of the upload size) I hear clear differences in timbre. Particularly striking for me is that some speakers have a strong to slight "gramophone sound".

The dynamic differences are also clearly audible, but I'm not sure if Ron has done a careful level matching.

I hear now what you're talking about.

Having gone through the comparison a few times now, I'd say

2 has it the worst, but 3 and 4 also clearly have it. 1 sounds the closest to the original, with 5 being the second best. Are these in order of the video?

If so, that's odd for me, as while watching the video, the Heresy and the X5 were the only 2 I really didn't like. X5 and Sapphire had the best spatial quality to my ears, but something about the X5's tonality ruined it for me. It was like one of the instruments was a bit too loud and it was messing with the balance. With your clips, the X5(assuming it's in order) doesn't sound bad.

 

[amirm]

@amirm will definitely award me the Headless Panther for this.

I will send you my hosting bill for next month....

 

[ctrl]

Are these in order of the video?

In post#2 with the short samples, the order is in the file name:
1. original_revel_original_klipsch_original_qacoustics
2. original_M3_original_X5

 

[richard12511]

@amirm will definitely award me the Headless Panther for this.

Awesome thanks. Going back and forth it's now clear that the X5 is louder(confirmed with an spl meter), which unfortunately biases the result.

 

[ctrl]

If so, that's odd for me, as while watching the video, the Heresy and the X5 were the only 2 I really didn't like. X5 and Sapphire had the best spatial quality to my ears, but something about the X5's tonality ruined it for me. It was like one of the instruments was a bit too loud and it was messing with the balance. With your clips, the X5(assuming it's in order) doesn't sound bad.

When adjusting the volume on the short samples by turning the volume control, the Revel and the QAcoustics (slightly less good) would be okay for me, the Spatial X5 doesn't come close tonally to the original. With the M3, I miss the midrange, it sounds "thin". The Klipsch sounds as hollow as an empty poured out can of Pepsi.

 

[Kvalsvoll]

Good work, this is entertaining, whether it is useful could be left to the viewer to judge, or is there more to it.

Does it represent each speaker in a way that gives them a fair presentation - both compared to the original, and compared to the other speakers.

Thanks to @ctrl for posting those sound samples, made it possible to do this efficiently. I started watching, then realized original was not in the video, so over to the other room, find it on tidal, play there, then go back to the first room to listen more, realize the difference is too huge to make much sense out of it. But the samples makes it easy to do a quick compare, adjust volumes to get a better match, ignore parts to isolate individual properties.

Comparing 1. and the last ones 4. and 5., I found the 1. speaker, the Revel, to be the only one with a neutral midrange that did not colour the piano significantly, this one was lower in volume, and overall tonal balance is off due to colourations/increased level in the bass range and perceived roll-off in the high frequencies. Missing a lot of infromation/detail at higher frequencies. soundstage moves forward (deeper, farther away), no good rendering of instruments left. The other (4. and 5.) were coloured in the midrange, better bass balance, soundstage more forward similar to original, but reduced clarity.

The question is then, is this representative of the sound that could be experienced in the room when the recording was made. Could this be done just as good or even better, using 2 mics in normal stereo configuration.

 

[ctrl]

The question is then, is this representative of the sound that could be experienced in the room when the recording was made. Could this be done just as good or even better, using 2 mics in normal stereo configuration.

This depends heavily on your individual HRTF. If your individual HRTF corresponds well to the average HRTF of the dummy head and your headphones provide a neutral reproduction (neutral in the sense of Sean Olive's findings), the result should be very close to the reality in the recording room.

When recording with two microphones, the influence of the HRTF is missing, which is a disadvantage.
However, nowadays it is easily possible to mathematically convolve the recordings made (in post-processing or directly during recording) with an HRTF - this can be an average HRTF or your individual HRTF.

Whether this provides as good or better results than with a dummy head, I cannot assess.

Does it represent each speaker in a way that gives them a fair presentation - both compared to the original, and compared to the other speakers.

As long as the binaural speaker recording is only ever compared to the original source, the comparison is pretty fair - this applies regardless of the headphones used, since only differences between the original and the recording are taken into account. So, one can rate how much the loudspeaker deviates from a "neutral reproduction".
But one can't make an absolute classification, except for blatantly deviating speakers, if the headphone deviates too much from a neutral reproduction.

I have already linked here how well this works - the Neumann KU 100 dummy head is also used there. With neutral-sounding loudspeakers, you can hardly tell the difference between the original and the binaural recording - of course, the recording room also plays a role here.


The direct comparison of the loudspeakers with each other is problematic, since it strongly depends on how well your headphones "harmonize" with the dummy head recording. If that is the case, absolute sound ratings and direct comparisons are certainly not completely off the mark.

 

[Kvalsvoll]

This depends heavily on your individual HRTF. If your individual HRTF corresponds well to the average HRTF of the dummy head and your headphones provide a neutral reproduction (neutral in the sense of Sean Olive's findings), the result should be very close to the reality in the recording room.

When recording with two microphones, the influence of the HRTF is missing, which is a disadvantage.
However, nowadays it is easily possible to mathematically convolve the recordings made (in post-processing or directly during recording) with an HRTF - this can be an average HRTF or your individual HRTF.

Whether this provides as good or better results than with a dummy head, I cannot assess.



As long as the binaural speaker recording is only ever compared to the original source, the comparison is pretty fair - this applies regardless of the headphones used, since only differences between the original and the recording are taken into account. So, one can rate how much the loudspeaker deviates from a "neutral reproduction".
But one can't make an absolute classification, except for blatantly deviating speakers, if the headphone deviates too much from a neutral reproduction.

I have already linked here how well this works - the Neumann KU 100 dummy head is also used there. With neutral-sounding loudspeakers, you can hardly tell the difference between the original and the binaural recording - of course, the recording room also plays a role here.


The direct comparison of the loudspeakers with each other is problematic, since it strongly depends on how well your headphones "harmonize" with the dummy head recording. If that is the case, absolute sound ratings and direct comparisons are certainly not completely off the mark.

This is not so simple. How the binaural recording represents the experience heard in-room depends on more factors than you can see in a simple frequency response. The response of this type of recording is very sensitive to sound field directivity.

Then you have people like me, who insist on listening to this on speakers.

These recordings are interesting, because they include speakers that are actually different. I listened briefly once more, this time with replaygain, and these are my observations - comparing original - speaker #1 - speaker #5:

- #1 is tonally tilted with too much bass and lack of highs.
- #1 is very colored in bass range - some notes too loud, others missing.
- #1 is dynamically compressed, sounds soft. like transients are muted.
- #1 soundstage slightly forward, missing essential reverb from the recording, reduced clarity.
- #1 tonal balance of piano is excellent.
- #5 is like a copy of the original, just with some coloration in the mid on the piano and slightly less clarity and added room reverb from the listening room.

The faults in #5 can be fixed with eq - if they are real. #1 has faults that is impossible to correct once the sound has left the speaker.

All those differences I observed can be measured and analyzed. Use the original as reference, and you will see changes in frequency response that correlates well. Dump the samples into masvis, and see that the dynamic range is actually reduced in speaker #1.

 

[richard12511]

This depends heavily on your individual HRTF. If your individual HRTF corresponds well to the average HRTF of the dummy head and your headphones provide a neutral reproduction (neutral in the sense of Sean Olive's findings), the result should be very close to the reality in the recording room.

When recording with two microphones, the influence of the HRTF is missing, which is a disadvantage.
However, nowadays it is easily possible to mathematically convolve the recordings made (in post-processing or directly during recording) with an HRTF - this can be an average HRTF or your individual HRTF.

Whether this provides as good or better results than with a dummy head, I cannot assess.



As long as the binaural speaker recording is only ever compared to the original source, the comparison is pretty fair - this applies regardless of the headphones used, since only differences between the original and the recording are taken into account. So, one can rate how much the loudspeaker deviates from a "neutral reproduction".
But one can't make an absolute classification, except for blatantly deviating speakers, if the headphone deviates too much from a neutral reproduction.

I have already linked here how well this works - the Neumann KU 100 dummy head is also used there. With neutral-sounding loudspeakers, you can hardly tell the difference between the original and the binaural recording - of course, the recording room also plays a role here.


The direct comparison of the loudspeakers with each other is problematic, since it strongly depends on how well your headphones "harmonize" with the dummy head recording. If that is the case, absolute sound ratings and direct comparisons are certainly not completely off the mark.

How would this be affected by listening to this on speakers? That's what I've been using for all of these comparisons.

 

[richard12511]

This is not so simple. How the binaural recording represents the experience heard in-room depends on more factors than you can see in a simple frequency response. The response of this type of recording is very sensitive to sound field directivity.

Then you have people like me, who insist on listening to this on speakers.

These recordings are interesting, because they include speakers that are actually different. I listened briefly once more, this time with replaygain, and these are my observations - comparing original - speaker #1 - speaker #5:

- #1 is tonally tilted with too much bass and lack of highs.
- #1 is very colored in bass range - some notes too loud, others missing.
- #1 is dynamically compressed, sounds soft. like transients are muted.
- #1 soundstage slightly forward, missing essential reverb from the recording, reduced clarity.
- #1 tonal balance of piano is excellent.
- #5 is like a copy of the original, just with some coloration in the mid on the piano and slightly less clarity and added room reverb from the listening room.

The faults in #5 can be fixed with eq - if they are real. #1 has faults that is impossible to correct once the sound has left the speaker.

All those differences I observed can be measured and analyzed. Use the original as reference, and you will see changes in frequency response that correlates well. Dump the samples into masvis, and see that the dynamic range is actually reduced in speaker #1.

Important to note that speaker 5 was noticeably louder than speaker 1. Not sure how to measure by how much, but it was clear from the spl meter I have here that it was consistently 3-6dB louder.

 

[Kvalsvoll]

Important to note that speaker 5 was noticeably louder than speaker 1. Not sure how to measure by how much, but it was clear from the spl meter I have here that it was consistently 3-6dB louder.

Loaded all samples into foobar, and applied replaygain to the files. That fixed the loudness difference issue, but since tonal balance is faaaar off on #1, it is not possible to get this to sound equally loud no matter how you do it.

Just listened to the Bass Room album by Nenad Vasilic. This is actually very good music. The track used in the samples is very kind on the woofers, other tracks are different, some contain subsonics, some are very demanding. Good stuff, just remember to turn it up, it need to breathe to deliver that raw, physical sense of a real bass in your room.

 

[Kvalsvoll]

How would this be affected by listening to this on speakers? That's what I've been using for all of these comparisons.

Frequency response in mid centered around 1K typically, may be off. Depends on the directivity of the soundfield in the room when the recording was made.

Headphones does not work. As @amirm said in a previous post here.