[ironmine]

How does my DSP chain look? Curious to hear if I have my components/plugins in the correct/logical order

Hi WoodyLuvr,

(1) The first plugin should be a high-quality upsampler so that all further calculations are done with a higher precision and less quality loss. Use the ratios 2X or 4X or 8X. It means it's better to upsample 44/16 to 88/24 or 176/24 or 352/24, but not to 96/24 or 192/24 or 384/24. Of course, 32 bits are even better than 24 bits.

(2) The last three plugins in the chain must be:
a) Ditherer. Dither the signal to 16 or 24 bits depending on your DAC input specifications.
b) Volume control adjustment (not to exceed 0 dB or preferably even less, in order to avoid digital clipping)
c) Volume and clipping monitor

2b and 2c sometimes can be combined in the same plugin.

All your other processing plugins should be put in between (1) and (2).

My Foobar has this component installed - http://www.yohng.com/software/foobarvst.html
which connects to VST-chainer called Console ART-Teknika - http://www.console.jp/en/

ART-Teknika basically lets you load and connect almost any VST plugins. You just place VST plugins there, move them around and connect them to each other with virtual wires any way you want. So, it's very graphic, visual and convenient. After you connected all your plugins and set them up you just save this preset in Console. So, you can have a bunch of VST-chains, each with it's one parameters, and you can open them "on the fly", while the music is playing, for instant comparison.

 

[WoodyLuvr]

@ironmine Thank you for the detailed advice... unfortunately much of it went right over my head! LOL. Though I believe I have the Advanced Limiter placed correctly, as per your advice, last in the chain for volume and clipping prevention.

Currently, the VST plugin adapter I use is foo_vst:

So besides missing a "resampler" first in the chain (are there any that you would recommend and do I really require it?) do I have the rest of my DSPs ordered correctly? FYI: all of my files are MP3 256-320 kbps with sampling set at 24 bit 44,100 Hz for my DragonFly.

Please give me an example of your DSP chain so I may better understand your advice.

Regarding your Crossfeed VST (112 dB Redline Monitor) how does it compare with Meier Crossfeed?

 

[ironmine]

@ironmine Though I believe I have the Advanced Limiter placed correctly, as per your advice, last in the chain for volume and clipping prevention.

I am not sure what exactly the Advanced Limiter does, as I could not find the description of it and it does not have any controls. I just hope it's not a compressor!

But, if you monitor the signal level and make sure that it stays below 0 dB or -1.0 dB, then you don't need any limiter, be it "advanced" or "stupid".

So besides missing a "resampler" first in the chain (are there any that you would recommend and do I really require it?) do I have the rest of my DSPs ordered correctly?

Yes, you are still missing a resampler, ditherer, volume adjustment and clip detector.

FYI: all of my files are MP3 256-320 kbps with sampling set at 24 bit 44,100 Hz for my DragonFly.

Why would you want to use such low-quality source material as compressed mp3 ??? It's not serious, man...

 

[WoodyLuvr]

I am not sure what exactly the Advanced Limiter does, as I could not find the description of it and it does not have any controls. I just hope it's not a compressor!

But, if you monitor the signal level and make sure that it stays below 0 dB or -1.0 dB, then you don't need any limiter, be it "advanced" or "stupid".

Apparently, it does nothing except detect and prevent clipping (especially from DSPs) and thus why it has been recommended to be set last in the DSP chain. Also, it has been recommended if you are using Replay Gain in Foobar and have processing set at "Apply Gain" only under Playback Preferences.

Yes, you are still missing a resampler, ditherer, volume adjustment and clip detector.

If you don't mind please do a quick screen capture of your DSP list in Foobar so I can visualize and research all the plugins/components you are using.

Why would you want to use such low-quality source material as compressed mp3 ??? It's not serious, man...

Unfortunately, my ears are honestly unable to hear the difference between 256 kbps MP3 and FLAC/ALAC/WAV so I have simply not gone that route. Plus, all my music is from Google Play which is limited to that.

@ironmine
Thus far though with my plugins is my DSP chain correctly ordered you think?

 

[pinnahertz]

Music without a crossfeed sounds like a bee buzzing in the ear.

This feeling is awful, weird and unnatural.

Anti-crossfeed guys may say what they will, but it's not pleasant.

It's not the way we hear sounds in the real world. If you listen to headphones without a crossfeed for a long time, yes, you may get used to it. That's what most headphone listeners have done, unfortunately - they have become accustomed to a wrong presentation of sound. That's why they resist.

Even with cross-feed, headphone listening isn't how we hear sounds in the real world. If you're after the "real world", the only thing that even sort of works is binaural. Otherwise, the "real world" isn't even a goal in any recording. Cross-feed simple presents a different perspective of a totally artificial world that may or may not be a subjective improvement.

I am sure that 10 or 20 years from now, headphone listening without some sort of crossfeed would be regarded as an oddity, a thing of the past. It's like watching TV now in black & white mode.

Well let's see now. We've had stereo recordings available to the consumer for 65 years (starting with tape in 1952) and stereo headphones for 59 years (Koss, 1958) with the current escalation of popularity beginning in 1979 with the Walkman (38 years). Cross-feed has existed in one form or another for at least 39 years (the Apt-Holman preamp had variable cross-feed), and could at any time have been achieved with minimal, though not insignificant cost of manufacture. Yet it's still extremely rare today. I'm not sure another 10-20 years is going to change anything.

Every single feature that is standard today was introduced as a major change and easily percieved audible improvement at an acceptable cost, like the jump from mono to stereo, tape noise reduction, analog to digital audio, stereo to multi-channel, etc. Unsuccessful would be things like Loudness compensation which showed up in the 1960s, but since it never really worked as intended, and eventually vanished. It's been replaced by DSP-based volume-aware loudness compensation, but those options aren't understood and so ignored by most consumers. And so far, higher than CD audio has not succeeded in dominating the music market, which is interesting, because it really is a low-cost upgrade with imperceptible improvement.

Where does Cross-feed fit into this? It's a clearly audible change, it's low cost, the concept is anything but new and has had numerous market tests with free solutions available now. Why isn't it already on every single iOS and Android device, which between the two covers the vast majority of all devices in the world that headphone listening sources? Lack of market awareness? If it costs nothing, but provides clear benefit, it becomes a marketing edge. If it were presented as an advantage to consumers and included on a lot of product, the market would become aware...but that's not done.

Looking to the more esoteric devices in the headphone world, things like DAC/Amp combos, or headphone amps, we find very, very few include cross-feed. Within Headphone.com's offering of 32 headphone amps, around 10% have cross-feed. Their most expensive offering, the Sennheiser HDVD 800, has this sentence in its description: "It provides balanced sound, maximum precision and impressive spacial accuracy." Spatial accuracy? That MUST be cross-feed, right? Nope, none at all. One manufacturer who did include it on many past models, Headroom, has discontinued all of those products, with their current offering lacking cross-feed.

So the question to answer is: Why is cross-feed not standard today? It can't be ignorance, at least, not in the headphone amp market. But it hasn't sustained it's presence even there, with today's total number of products including fewer than ever with cross-feed. The extrapolated statistics have the presence of cross-feed in headphone amps slowly approaching zero in 10 years or less. Could that be because it's happening elsewhere, like in software? Possibly, but not likely. Headphone amps are very targeted to headphones, and would be the perfect place for a cross-feed option. So would digital music players, but it's not there either...much.

I feel compelled to say at this point that I'm not really a cross-feed hater, nor am I opposed to it. It's a tool that is good when appropriately and properly used. I have it, I use it, and I try it on a lot of material, but usually choose to keep it off. I would, however, rather have a cross-feed option than not. Going forward, that's not looking like something that's going to happen.

 

[71 dB]

Interesting. So you are correcting for hard-panned below 1kHz, but not correcting hard-panned above (I know it's a gradual transition). If I may...why not? Wouldn't hard-panned sounds above 1kHz be bothersome to you as well?

ILD problems above 1 kHz are pretty insignificant. Some recordings do have such a "harsh" upper end, that I find a little treble crossfeed beneficial in "softening" the sound, but generally the main issue is lower frequencies. 1 kHz here is a symbolic border between "ILD problems/no ILD problems" areas. As you said, it's gradual transition.

They are not comparable, though. Both effects (overly wide separation) and HF response changes that are related to headphone position my be detrimental, but not in the same way. There's no means of comparison without statistical analysis of subjective testing, and I guess we won't go there.
How would you know, though? Don't you reposition headphones for best (most pleasing) response? Each time you put them on? Some of us do that, and some of us do care.

I don't care and hardly even hear the difference. Moving you head while listening to speakers means similar changes to sound. I try to tackle relevant issues of sound reproduction.

 

[jgazal]

This is going to be a wild speculative post, but since nobody risked to give an explanation, I will try.

Consider the following crosstalk cancellation with speakers and externalization with headphones:

13 Is the 3D realism of BACCH™ 3D Sound the same with all types of stereo recordings?
(...)
All other stereophonic recordings fall on a spectrum ranging from recordings that highly preserve natural ILD and ITD cues (these include most well-made recordings of “acoustic music” such as most classical and jazz music recordings) to recordings that contain artificially constructed sounds with extreme and unnatural ILD and ITD cues (such as the pan-potted sounds on recordings from the early days of stereo). For stereo recordings that are at or near the first end of this spectrum, BACCH™ 3D Sound offers the same uncanny 3D realism as for binaural recordings18. At the other end of the spectrum, the sound image would be an artificial one and the presence of extreme ILD and ITD values would, not surprisingly, lead to often spectacular sound images perceived to be located in extreme right or left stage, very near the ears of the listener or even sometimes inside of his head (whereas with standard stereo the same extreme recording would yield a mostly flat image restricted to a portion of the vertical plane between the two loudspeakers).
(...)

By the way, I recently created a PRIR for stereo sources that simulates perfect crosstalk cancelation. To create it, I measured just the center speaker, and fed both the left and right channel to that speaker, but the left ear only hears the left channel because I muted the mic for the right ear when it played the sweep tones for the left channel, and the right ear only hears the right channel because I muted the mic for the left ear when it played the sweep tones for the right channel. The result is a 180-degree sound field, (...).

Binaural recordings sound amazing with this PRIR and head tracking.

Using the first PRIR, central sounds seem to be in front of you, and they move properly as you turn your head. However, far-left and far-right sounds stay about where they were. That is, they sound about the same as they did without a PRIR, and they don't move as you turn your head. In other words, far-left sounds stay stuck to your left ear, and far-right sounds stay stuck to your right ear. It's possible to shift the far-left and far-right sounds towards the front by using the Realiser's mix block, which can add a bit of the left signal to the front speaker for the right ear, and a bit of the right signal to the front speaker for the left ear.
(...).

Why is this happening?

In the second example, why the hard panned sounds are stuck to the side and do not derotate according to the headtracking?

First, let’s see how the convolution, interpolation and headtracking work.

An standard 2 channel personal impulse response - PRIR comprises 12 impulses:

1. looking center + left speaker playing + left ear measuring;
2. looking center + left speaker playing + right ear measuring;
3. looking center + right speaker playing + left ear measuring;
4. looking center + right speaker playing + right ear measuring;
5. looking left + left speaker playing + left ear measuring;
6. looking left + left speaker + right ear measuring;
7. looking left + right speaker + left ear measuring;
8. looking left + right speaker + right ear measuring;
9. looking right + left speaker + left ear measuring;
10. looking right + left speaker + right ear measuring;
11. looking right + right speaker + left ear measuring;
12. looking right + right speaker + right ear measuring.

ipsolateral impulses (1, 4, 5, 8, 9, 12)
contralateral impulses (2, 3, 6, 7, 10, 11).

If you have only three looking angles (i.e. -30, 0, +30), how does the convolution engine set the frequency levels, time arrivals and phase delays for any angle between -30 degrees and 0 degrees and between 0 degrees to +30 degrees?

Frequency levels, time arrivals and phase delays for any angle are calculated by an interpolation algorithm.

Then head-tracking helps to externalize sounds.

The standard PRIR convolution, interpolation and headtracking allow to exactly emulate, with headphones, how the measured room/speakers sounds.

Since speakers were located at +/-30 degrees and contralateral impulses describes an important proportion of the acoustic crosstalk, the expression “how the measured room/speakers sounds” above means mostly flat image restricted to a portion of the vertical plane between the two virtual loudspeakers. Even if the content has hard panned sounds. Virtual speakers and hard panned sounds within will derotate according to the head-tracking.

So if you are fond of crossfeed for headphones that emulates what would be to listen to standard stereo recordings, in the old-fashioned way (i.e. with two speakers in a room and corrupting acoustic crosstalk), just use a PRIR convolution without interpolation and headtracking.

So far, so good.

But Erik wanted to test with headphones how Professor Choueiri’s crosstalk cancellation would sound in his room.

What has he done?

Erik’s crossfeed free PRIR comprises also 12 registries, but now both speakers are at 0 degrees and contralateral impulses registries are left blank (2, 3, 6, 7, 10, 11).

Since only ipsolateral impulses are computed in Erik’s crossfeed free PRIR, acoustic crosstalk info is lost and no crossfeed is added.

But there is another catch in Erik’s crossfeed free PRIR.

In standard PRIR’s, speakers are measured at +/-30 degrees.

One the one hand, ipsolateral impulses capture frequency levels, time arrivals and phase delays introduced by room reflections (RIR component of his PRIR) and the filtering effect of his pinna plus the scattering effect of the ipsolateral side of his face (two of three HRTF components of his PRIR) varying according to the looking angles.

On the other hand, contralateral impulses captures room reflections (RIR component of his PRIR) and acoustic crosstalk (note that acoustic crosstalk is a mix of room reflections and a third HRTF component). Such third HRTF component is the effect of his head shadowing directional frequencies.

In Erik’s PRIR both speakers are measured at 0 degrees, with +/-30 looking angles.

On the one hand, ipsolateral impulses capture frequency levels, time arrivals and phase delays introduced by all the room reflections (RIR component of his PRIR), the filtering effect of his pinna, the scattering effect of the ipsolateral side of his face and the effect of his head shadowing directional frequencies varying in degree according to the looking angles.

On the other hand, contralateral impulses, that in standard PRIR’s would capture the effects of room reflections and acoustic crosstalk, are left in blank, as I said before.

As he describes, he gets externalization with binaural recordings.

That is because natural ILD an ITD are already in the content. In others words, the effect of your head shadowing directional frequencies (or as, I said earlier, the acoustic relationships between both hemispheres that are separated by the sagital plane were preserved in the content).

If Erik plays back with his crossfeed free PRIR the old stereo songs that contains hard panned sounds, why the hard panned sounds don’t derotate?

Hard panned sounds don't have ITD at all, because there is no reference for the original time in the opposite channel. Every difference in time will be deemed just as the hard panned instrument started to play later (obviously it does not affect artistic coherence with other sounds as ITD are in the range of microseconds). They don’t have ILD either because their full level is on one single channel.

Since Erik’s PRIR has ipsolateral impulses measured with speakers at 0 degree that, as I said, capture the effect of his head shadowing directional frequencies, the convolution engine can introduce, in the playback step, his personal ILD and ITD to center sounds while rotating his head.

That happens because when he turns his head:

a) looking left (maximum looking angle is still 30 degrees), center sounds become more shadowed by left ipsolateral impulses and at the same time center sounds become less shadowed and more scattered by right ipsolateral impulses;

b) looking right (maximum looking angle is still 30 degrees), center sounds become less shadowed and more scattered by left ipsolateral impulses and at the same time center sounds become more shadowed by right ipsolateral impulses .

And why that does not work with hard panned sounds? Why in such strong stereo separation recordings, hard panned sounds are stuck to the side when using Erik’s crossfeed free PRIR?

Because when he turns his head looking left (maximum looking angle is still 30 degrees), hard panned sounds in the left channel become more shadowed by left ipsolateral impulses, but absolutely nothing happens with such hard panned sounds at the right headphone driver, since no correlated sounds are feed into the convolution of right ipsolateral impulses.

Note that the absence of crossfeed seems to make reflections more perceptible (so better room acoustics would yield no echo?):

By the way, after making the PRIR, the "window" setting should be reduced to 200ms to prevent the right ear from hearing a faint echo of the left ear's signal, and vice versa.

So it is clear that stereo recordings with hard panned sounds won’t result a good spatial rendering with Erik’s PRIR.

What does Erik do to circumvent such error?

The Realiser A8 mix block allows mixing the left channel into the right channel and vice versa with rough steps (consider 0 as mute and 1 as full scale, the step is 0.1; don’t know how to express that logarithmically in dBFS).

If @Erik Garci mix at least 0.1 of each channel one another then he is able “to shift the far-left and far-right sounds towards the front”.

Do the hard panned sounds also start to derotate according to the headtracking?

Yes.

Why?

Because only using ipsolateral impulses Erik is to feed some level of left channel hard panned sounds into the convolution of right ipsolateral impulses.

Then why the original far-left and far-right sounds shift directly towards the front? Why the far-left and far-right sounds shift directly to the front and not gradually, going from "stuck to the sides" to rotating as if they were outside the +/-30 degrees looking angles?

Because the ITD and ILD are derived from and is therefore limited by the maximum time difference captured by ipsolateral left and right looking angles.

Could the 0.1 mix ratio be higher than his natural ILD?

Yes.

That’s why, as far as I understood, Erik said:

That would be much easier than manually muting the microphones during measurements, and just about any PRIR could be used.

Allowing fractional values would be even better, such as 0.5 (-6 dB) or 0.1 (-20 dB).

What if the Realiser could mix both channels in increments such as 1dBFS?

We are waiting for Smyth Research response.

Does anybody have an alternative explanation that is not expressed as programming codes or equations?

 

[71 dB]

Music without a crossfeed sounds like a bee buzzing in the ear.

This feeling is awful, weird and unnatural.

Anti-crossfeed guys may say what they will, but it's not pleasant.

It's not the way we hear sounds in the real world. If you listen to headphones without a crossfeed for a long time, yes, you may get used to it. That's what most headphone listeners have done, unfortunately - they have become accustomed to a wrong presentation of sound. That's why they resist.

I am sure that 10 or 20 years from now, headphone listening without some sort of crossfeed would be regarded as an oddity, a thing of the past. It's like watching TV now in black & white mode.

Totally agree with this. The dominant problem of no crossfeed is this "bee buzzing", but there is more.

- Bass with too much ILD sounds fake. Crossfeed gives it "physicality", realism.
- Sound image is "broken" due to spatial distortion. Crossfeed is able to fix this. Spatial distortion masks real spatial information. Crossfeed reduces/removes spatial distortion so that real spatial information is revealed. Instruments are located at pinpoint accuracy instead of spreading all over the place (requires good recording thou).
- Listening fatique which crossfeed is famous for reducing/removing.
- Without crossfeed the music is located mostly above shoulders near ears, which sounds annoying.

Crossfeed makes the sound "jump" in front of me at a distance dictated by the spatial information of the recording, but varying from 1' (30 cm) (crappy spatiality) to perhaps 5' (1.5 m) (excellent spatiality). The latter is almost like listening to speakers in terms of soundstage depth. If I sit in front of my speakers and imagine the sound coming from them I can almost fool myself that I am listening to speakers, but that's only with the recordings with excellent spatiality (typically a multichannel SACD of classical music). When I began crossfeed, for some time I had hard time believing it's possible to achieve improvement of this scale using some simple electric circuits. It seemed magic, but after thinking these things I did understand why this it possible. Spatial distortion is very detrimental because our hearing is so sensitive to it. Reducing it even a little helps a lot. Our spatial hearing can be fooled as long as we have reasonable spatial cues that make sense and that's why the spatial information doesn't need to be perfect after crossfeed. As long as it makes sense our hearing is able to "get the picture" and we are fine.

The problem of crossfeed is not how it "messes up" the sound (that's a misunderstanding of the whole concept). The problem is that each recording contains different amount of spatial distortion so for optimal result one must be able to adjust the crossfeed level. If you don't crossfeed enough, you don't really experience the real benefits fully and if you crossfeed too much, you get a bit dull and "monophonic" sound. It's the optimal level where crossfeed shines. If you have just one fixed level, say -8 dB, only those recordings with optimum crossfeed level between -9 dB and -7 dB work optimally with it (1 dB accuracy seems to be enough imo) and that's probably 50 % at most of all your music, perhaps only 25 %.

You are spot on saying heaphones listeners have accustomed to a wrong presentation of sound. That's exactly how it is. Pinnahertz even admitted to this saying he wants to hear The Who with the same hard panned spatiality he used to hear in the 70's. The day I discovered crossfeed I admitted myself having done headphone listening wrong for years. It's crazy to assume we do things the right way from the beginning. At least I don't listen to headphones without crossfeed all my life.

 

[71 dB]

71 dB, thanks for your explanations and for the photo.

I am not really knowledgeable in electronics, I cannot understand this diagram. If it were presented as a bunch of inter-connected VST plugins, I would understand it better

No problem. I have a degree in electric engineering, so these diagrams aren't difficult to me, but of course they are if you don't know electronics. The schematics has more components than electrically needed, because I have multiplied some resistors to increase power handling capability. This is a headphones adapter, my amp feeds it like a speaker so we are talking about watts of power at peaks, a small fraction of it sent to the headphone. So using 0.6 W metal film resistors I can increase power handling to say 2.4 W by using four 47 Ω resistors in paraller/serial configuration instead of just one. Three 10 Ω resistors in series is just one 30 Ω resistor with 1.8 W of power handling. The power handling requirements are carefully analysed. However, the power handling capabilities are overkill so my typical listening levels don't make the resistors warm so much to be noticable by finger.

 

[gregorio]

[1] Let's say you have 12 dB of ILD at bass on your recording. [2] It seems improbable that the producers of the record intented that, kickdrums a feet from your head. ... Kickdrums are expected to be located to some distance meaning ILD at bass is limited to a few decibels. ... [2a] How does kickdrum sound if you listen to it that close? .... [3] When you listen to speakers, all of this is automatically taken care of by acoustic crossfeed.

1. That would be unusual.
2. You are just making that up, you have no idea. In fact in some genres the kick drum is processed very dry and designed/intended to sound close relative to most other elements of the mix.
2a. A real kick drum typically sounds COMPLETELY different, at any distance. In virtually all pop and rock of the last 40 years or so, the kick drum is heavily processed and sounds nothing like a real kick drum. You are again falling into the audiophile trap of making a comparison with reality, when there is no reality in the first place!
3. No, it's not! Firstly, a kick drum is typically panned centrally or near centrally, so there is very rarely a large ILD. Secondly, the reason a dry processed kick drum doesn't sound like it's in your head when reproduced with speakers is because of listening room reflections and has little or nothing to do with ILD! Thirdly, when I'm mixing a track with a very dry kick and when I listen with cans, if that kick sounds too present relative to the other elements in the mix, I'll typically add some amount of an "ambience" type reverb. This will likely have little or no impact when played on speakers as the effect will likely be overwhelmed by room reflections but it will move the kick back into a more desirable relative position when listening with cans. HP crossfeeding will damage that processing.

[1] Yes, but our brain still has to make sense of it. Artificial becomes easier if you have cues from the reality such as reasonable ILD. [2] My ears don't like large ILD below 1 kHz and for some reason scientific knowledge of human hearing does give explanations why (kickdrums don't fit into ear canals). [3] But that's just me. I don't know how your ears work.

1. Again, you are referencing to reality, when there is no reality in the first place! The reason our brain can "make sense" of all the spatial information anarchy on a recording is because it was created/applied by a human being who also has a brain. And again, the mix would always be checked on cans and if what was applied for speaker reproduction doesn't make sense, then something can often be done to have it make better sense, assuming of course that "making sense" was the original goal, which it sometimes isn't!

2. Careful here! You are stating a preference and then attempting to support that preference with "scientific knowledge" but that scientific knowledge is inapplicable. What has a kick drum not fitting into your ear canal got to do with anything? Again, we are not talking about a real kick drum, we are talking about an imaginary, made-up kick drum which is nothing like a real kick drum! Additionally, you do realise that most manufactured/manipulated kick drums have a very significant component typically somewhere between 1kHz and 2kHz, the initial attack or click of the kick as opposed to the boomy resonance?

3. Exactly, it's just you (and a few others). You don't know how my ears work, you don't know how yours work, you don't know how the engineers and artists ears work or what they were trying to achieve and you don't know what a kick drum sound is!

[1] I see it differently. To me acoustic crossfeed tames the excessive ILD levels with speakers creating ILD values that make sense. [1a] Crossfeed with headphones does similar (but not identical) thing. [1b] A hard-panned instrument becomes a spatially panned instrument with reasonable combination of ILD and ITD information. [2] That works great for me. ... [3] I don't get why acoustic crossfeed is fine, but electric crossfeed isn't.

1. You can see it however differently you choose but you seeing it differently does NOT change the actual facts. The actual facts being that acoustic crossfeed is NOT the only or even necessarily the most important factor at play with speakers, unless you're talking about speakers in an anechoic chamber.
1a. No, crossfeed does a very different thing, it does NOT take into account the "acoustics" part of "acoustic crossfeed"!
1b. No, it does NOT become a "spatially panned instrument"! It will probably become more appropriately panned relative to it's intended pan position with speakers but it does not address the "spatial" part of your "spatially panned instrument" statement. Indeed, simply crossfeeding with an ITD may actually damage/destroy that "spatial" part of "spatially panned instrument" more than not crossfeeding! You are ignoring the fact that in practise we don't just have ILD panning on an instrument, that instrument will also almost certainly have some sort of STEREO reverb (or stereo delay based effect) and the timing and levels of the reflections created by that reverb are likely to be damaged by a simple crossfeed and ITD of 250uS!

2. And that's fine. With some recordings it works better for me too but with others it doesn't and I don't know which recordings have had adjustments applied to help with non-crossfed HPs or have deliberately not been adjusted because the artists wanted it to sound that way when played back on HPs and that's why I don't use crossfeed.

3. That's obvious. Maybe you are somewhat insensitive/ignorant of the "spatial" part of "spatially panned instruments" and are therefore not bothered that you are often damaging/destroying it? Or, maybe I'm overly sensitive because I spend so much time working specifically on it? I'm not sure, which is why I state that my preference is not to use crossfeed and that using it can reduce fidelity BUT, I would never say that everyone will definitely benefit from not using crossfeed and that it would be idiotic to use it!

Let's use your vehicle analogy again: Let's say studio playback is like a MotoGP race bike, playback on home speakers is not the same, more like a consumer superbike. Playback on HPs is a like a car, a completely different vehicle. Now we could do something to that car to make it more like a motorbike. If we hacked-off two of it's wheels, it would have two wheels just like a motorbike but of course it's not really anything like motorbike, what we've really got is a nonsense, un-drivable vehicle and I personally would rather have just a normal car with all it's relative faults. ... This analogy is poor on numerous levels, one of which is in practice such a two-wheeled car might sometimes give us a better experience than a normal car but it serves the purpose of illustrating that just fixing one aspect/element doesn't necessarily get us something better.

[1] That is exactly why I need crossfeed to transform "Decca tree information" to something my ears and head understands. [2] If they use say a Jecklin Disk then the spatial information is already compatible with my spatial hearing (spatial distortion free) and I most probably listen to the recording crossfeed off.

1. No, you have that backwards! The Decca Tree mic array was invented BECAUSE it translates so well to what our ears/brains understand! It's entirely possible that a Decca Tree recording would sound subjectively better in HPs than speakers, which is why the use of outrigger mics along with a Decca Tree eventually became the popular/preferred setup.

2. You can't use a Jecklin Disk with a Decca Tree array, a Jecklin Disk only works with a specific type of mic arrangement: Two small diameter mics placed in a close stereo setup, like an ORTF pair for example. And, as with the more modern usage of a Decca Tree arrangement, there are probably very few classical/orchestral recordings in the last 30 years or more which only use such a simple mic setup, virtually always there would be a number of other mics mixed in.

[1] I don't get why we shouldn't make spatial information less distorted. [2] If a recording was clearly made for speakers (having acoustic crossfeed) and [2a] clearly sounds horrible with headphones without crossfeed then to me it's a nobrainer to use crossfeed, but that's just me… …it makes sense to me but apparently not to everybody.

1. Because applying crossfeed to HPs doesn't make the spatial information less distorted, it can make some aspects of the spatial information MORE distorted!
2. And how do you know if a recording "was CLEARLY made for speakers"? It probably has been primarily made for speakers but that doesn't mean that some aspects of the mix have not been modified so that it still works relatively well on HPs, un-crossfed HPs!
2a. "Horrible" is a purely subjective term and exactly what makes a piece/type of music horrible to some is what makes it great to others! It's often difficult to judge what is just horrible and therefore bad and what is intentionally horrible and therefore good if we're after fidelity but still potentially bad relative to a particular individual's preferences. That makes sense to me but apparently not to many audiophiles, for whom if something sounds better (relative to their preferences) then it is good period for everyone and they then simply redefine the word "fidelity" to equal their perceived "better".

G

 

[Erik Garci]

This is going to be a wild speculative post, but since nobody risked to give an explanation, I will try.

Great explanation. It gets even more interesting when you consider the Hafler PRIR. Basically you hear the sum from the center-front speaker (L+R to both ears), and you hear the differences from the center-back speaker (L-R to left ear, and R-L to right ear).

I made a Hafler PRIR where the center-back speaker was actually measured in front, but I turned my head in the opposite direction. I looked right instead of left and looked left instead of right. This way, the center-back speaker has the same spectral balance as the center-front speaker, and head-tracking helps me distinguish which sounds are from the front versus the back. Maybe Smyth can add a Hafler mode that works for any PRIR that has a center speaker or a closely-spaced pair.

My next plan is to add crosstalk-free height channels.

Technical note: The A8's mix block cannot produce differences on its own because it does not allow negative coefficients, thus the negative signals must be generated by another device (such as one with balanced outputs) and fed to the A8 on other channels.

 

[gregorio]

(1) The first plugin should be a high-quality upsampler so that all further calculations are done with a higher precision and less quality loss. Use the ratios 2X or 4X or 8X. It means it's better to upsample 44/16 to 88/24 or 176/24 or 352/24, but not to 96/24 or 192/24 or 384/24. Of course, 32 bits are even better than 24 bits.

(2) The last three plugins in the chain must be:
a) Ditherer. Dither the signal to 16 or 24 bits depending on your DAC input specifications.
b) Volume control adjustment (not to exceed 0 dB or preferably even less, in order to avoid digital clipping)
c) Volume and clipping monitor

1. No, that's not how processing works! You will NOT get either higher precision or less quality loss. Precision is dictated by the plugin's internal bit depth which is either 32 or 64 bit float. So whether you feed a plugin with 16 bits of data or 16 bits plus 8 additional zeros, the precision and result are identical. Increasing the sample rate is likewise pointless, there's no data above the Nyquist point to process and you don't magically get from upsampling. The only time a higher sampling rate could make a difference is in the case of some non-linear processes, in which case any decent plugin will internally upsample anyway. Actually, upsampling could (theoretically) make matters worse, some plugins operate at a single internal sample rate. Say for example a plugin operates at 96kHz, you upsample to say 176kHz, the plugin down samples to 96, performs it's algorithm and when complete upsamples the result back to 176kHz again, to match the incoming sample rate. What have you gained by upsampling in the first place? You'd have been better if you'd left it at 44.1kHz! You've fallen into the old audiophile trap of confusing the bit depth of the audio file with the bit depth of the processing environment! The bit depth (and precision) has nothing to do with the bit depth of the file but the bit depth of the internal processing of the plugins and the bit depth of the data connections between those plugins.

2. Dither should be the last step here, not the first! If you are changing the volume in your processing environment, which is presumably a 32 or 64bit environment, then the result of that volume change is obviously a 32 or 64bit word length, which you are then truncating to 16 or 24bit because you applied dither before the volume change. However, if you're outputting 24bit to your DAC there is no point in dithering, as truncation artefacts would be way below audibility, even at 16bit truncation artefacts would probably be inaudible.

G

 

[pinnahertz]

ILD problems above 1 kHz are pretty insignificant.

Sorry...huh? ILD above 1kHz is very significant, and if you cross-feed to correct below 1kHz and ignore above, you'll skew the perceived position of the LF fundamental with the HF harmonic or attack content. You'll break the image into separate sources! I'm sure you can see that in a complete analysis of HRTF. And of course, mentioned this before...I see no attempt at ITD "correction", ignoring half the spectrum.

I've experimented with ILD-based and ILD/ITD/HRTF-based cross-feed. Care to guess which works better? I know you can't whip that last one up with resistors and caps, but since there's a DSP or three pretty much in everything, it's not a problem, it's software. And already done.

Still not a preference of mine in either case.

Some recordings do have such a "harsh" upper end, that I find a little treble crossfeed beneficial in "softening" the sound, but generally the main issue is lower frequencies. 1 kHz here is a symbolic border between "ILD problems/no ILD problems" areas. As you said, it's gradual transition.

The "harshness" issue isn't fixed by cross-feed, though, either with ILD or ITD alone or in combination. You have to work in the full HRTF. And like I said, even then I don't think it works universally.

I don't care and hardly even hear the difference. Moving you head while listening to speakers means similar changes to sound. I try to tackle relevant issues of sound reproduction.

Interesting opportunity opens here. If our positions were reversed on the above, would I be accused as being spatially deaf?

I hope that's over.

 

[71 dB]

Even with cross-feed, headphone listening isn't how we hear sounds in the real world. If you're after the "real world", the only thing that even sort of works is binaural. Otherwise, the "real world" isn't even a goal in any recording. Cross-feed simple presents a different perspective of a totally artificial world that may or may not be a subjective improvement.

True, only binaural can reach very high level of real world feel, but crossfeed modified the sound to be if not "real world" at least more natural and pleasant. The problem with binaural is that you need your own or at least very similar HRTF-signature for it to work really well. Crossfeed simulates HRTF at so coarse level that the result works for anybody. "Real world" is even today a challenge in sound reproduction. Spatial distortion free sound is not.

Their most expensive offering, the Sennheiser HDVD 800, has this sentence in its description: "It provides balanced sound, maximum precision and impressive spacial accuracy." Spatial accuracy? That MUST be cross-feed, right? Nope, none at all.

That's silly marketing talk. Without crossfeed Sennheiser HDVD 800 doesn't offer spatial accuracy unless the recording happens to be one of the 2 % free of spatial distortion. Rich morons can spend their money on overprised products like that. I have my 50 bucks DIY headphone adapter with 6 level crossfeed and enjoy real spatial accurary on almost all recordings (some recordings are bad beyond hope of fix). That's the benefit of having a degree on electric engineering and acoustics.

Sennheiser Orpheus has 2 level crossfeed. For the price it's kind of the minimum..

 

[71 dB]

This board is too active for me! I try to keep up with the pace...

1. That would be unusual.
2. You are just making that up, you have no idea. In fact in some genres the kick drum is processed very dry and designed/intended to sound close relative to most other elements of the mix.
2a. A real kick drum typically sounds COMPLETELY different, at any distance. In virtually all pop and rock of the last 40 years or so, the kick drum is heavily processed and sounds nothing like a real kick drum. You are again falling into the audiophile trap of making a comparison with reality, when there is no reality in the first place!
3. No, it's not! Firstly, a kick drum is typically panned centrally or near centrally, so there is very rarely a large ILD. Secondly, the reason a dry processed kick drum doesn't sound like it's in your head when reproduced with speakers is because of listening room reflections and has little or nothing to do with ILD! Thirdly, when I'm mixing a track with a very dry kick and when I listen with cans, if that kick sounds too present relative to the other elements in the mix, I'll typically add some amount of an "ambience" type reverb. This will likely have little or no impact when played on speakers as the effect will likely be overwhelmed by room reflections but it will move the kick back into a more desirable relative position when listening with cans. HP crossfeeding will damage that processing.
G

1. Perhaps. Even ILD larger than 6 dB starts to be too much imo.
2. Yeah, but to be in the middle. Dry and mono.
2a. Of course, but that doesn't change how spatial hearing works.
3. Yes, fortunately. Just explaining spatial distortion to you.