[rickcr42]

I am still trying to figure out where this "200khz" comes from.

Original Linkwitz : 700hz

Chu Moy Variant : 700hz

Simpson Variant : 700hz

Meier Natural :660

Meire Natural bass: 700hz

Quote:

As with the natural crossfeed filter, the direct signal with the Linkwitz filter shows a signal loss at lower frequencies, (-1.0 dB at 60 Ohms, -0.35 dB at 50k Ohms). However, more important is that a mono-signal at frequencies below 700 Hz is increased by up to 1.3 dB at a 60-Ohm load and up to 1.9 dB (!) at 50k Ohms. The delay times for the Linkwitz design (figure 2) are fairly natural, as the crossfeed signal has similar filter frequencies and thereby should have similar delay times as the natural crossover filter.

John Conover : 684hz

So they all have the hinge at the midband and effect from that point downwards set by a capacitor.If too high in the range there would be a high frequncy blend that would blow the left/right imaging and effectively be a monoing device.Because the 3db point is set to these ranges there will obviosly be some effect in higher ranges but this diminishes at the standard for a single pole filter which is -6dB per octave meaning it will slope graudually (as intended) rather than be a rapid cutoff or dead stop band.
The low bass is most often already a mono signal,usually below 100hz or so, meaningless in this or any other context of blending because it is already a mono signal.

Quote:

Add to the problem of crossfeed is that it adds electronics in the signal path, just as equalization does, and this has to degrade the signal, however slightly

Strictly speaking no.All are totally passive devices using resistors/capacitor filters except for the Headroom and original J.Conover but even he offers a totally passive desin option (direct Headphone Drive).The Jan Meier active circuits are for the actual headphone amp and he chooses to buffer the input of the crossfeed to present it with a low impedance and a known impedance so the network response can be predicted but this is not essential unless the driving source has a high impedance such as some tube stages.The output if connected to a headphone amp will see the input impedance of the headphone amp as if a buffered output and since this is pretty much standardized at 100K ohms the network response can be predicted.changes in either input or output impedance (network driving source and output load ) will cause slight changes in network response but not extreme enough to ruin the effect.
The Meier Audio Corda Cross addresses the possible impedance mismatches by adding in passive bass and treble trim controls so the effect can be actually tailored to individual taste.I will not comment on the Headroom matrix because even though I know how it works it is their design and if they wanted the actual design details out would have posted them on their web page.

The only downside to most crossfeed devices is there is some negative gain (oxymoron

),some level attentuation as any passive network will have.The other problem is not all are totally out of circuit when bypassed and effectively are "blend" controls even when in the bypass position.Both are easily fixed by upping the volume control on the amp for the attenuation and installing a proper defeat switch for the blending.Other than that it comes down to personal choice and preference but the fact that recordings mixed for speaker playback are different than those recorded for headphone playback (rare !) is not in doubt.
I do find it interesting and something I have observed over time is usually the most vocally opposed to a headphone crossfeed network mostly never even listened to one and if they did only for a very brief time.Those that have extensive time with one and still question the validity usually complain not that it destroys the sound but that the effect is so unobtrusive as to be hard to detect-as it should be-so what is the point of having it ?

Quote:

No doubt the bigger and more filled-in soundstage is often desirable when inner detail is not important to the listener. There is no free lunch - either crossfeed or no crossfeed has its good points and it bad.

Detail is not effected,just the direction it comes from unless crappy parts are used

Quote:

He doesn't sell that anymore...

But he does still have step-by-step build instructions on his web page including a proper bypass position on the "tweak" page (suggested by me

).Total build time is about 15 minutes.build cost about ten bucks american so an easier/cheaper "toy" for experimenting would be hard to find.Probably would make sense before laying out cash for a commercial device to whip one up and see if YOU like the effect or not.

****Edit****

even those who do not know which end of the soldering iron is the hot end can usually find some local pimply faced technical high school attendee who will be more than happy to whip one up.If that is not an option maybe some old beezer family member with a HAM Radio liscense.They ALL know electronics as it is required to pass the test for a liscence.

 

[third_eye]

I use the the lowest crossfeed setting on my Prehead; although the soundstage does shrink very slightly, the overall sound to my untrained ears is more coherent and enjoyable for long-term listening. In fact, at this point I'm not sure if I would even want an amp without the crossfeed option.

 

[ken36]

Outstanding read. thanks.

 

[Scrith]

I use the Crossfeed plugin in Foobar2000, which works great. There are several other crossfeed-like plugins for Foobar, though I have only tried one or two of them (and prefer the regular Crossfeed one).

Adding crossfeed at the digital level is probably going to sound better than an analog hardware crossfeed device/circuit, and provides you with some flexibility (most of the plugins can be configured to customize the various aspects of the crossfeed effect).

I've found that not using Crossfeed while listening to headphones gives me a minor headache after 10-30 minutes.

 

[gpalmer]

Quote:

Originally Posted by rickcr42 But he does still have step-by-step build instructions on his web page including a proper bypass position on the "tweak" page (suggested by me ).

Nice catch Rick, last time I hit his site I didn't see that up there but it's been a while!

 

[White Fox]

Quote:

Originally Posted by Scrith Adding crossfeed at the digital level is probably going to sound better than an analog hardware crossfeed device/circuit, and provides you with some flexibility (most of the plugins can be configured to customize the various aspects of the crossfeed effect).

Yeah, someone come out with a CDP/DAC with customizable crossfeed. In the future, physical media will probably be a thing of the past anyway - at least that's what some want you to believe. For now, I don't like involving my computer for listening to music, or making CD burns with crossfeed for my entire collection. But I might try making a couple of burns to try this Foobar plug-in. Still need to try Headroom crossfeeds, though.

Those who would describe crossfeed as making the sound 'muddy' or 'congested' should try all of the crossfeeds available or specify which crossfeed design(s) you're talking about.

 

[gpalmer]

Quote:

Originally Posted by rickcr42 I do find it interesting and something I have observed over time is usually the most vocally opposed to a headphone crossfeed network mostly never even listened to one and if they did only for a very brief time.Those that have extensive time with one and still question the validity usually complain not that it destroys the sound but that the effect is so unobtrusive as to be hard to detect-as it should be-so what is the point of having it ?

I have used both the Meier and HeadAmp designs (and some software). The degradation in the signal, particularly the frequency range bothered me far more than the lack of crossfeed. I think Hirsch nailed it on this one, the brain does a much better job than the electronics I've heard...

 

[Sovkiller]

Quote:

Originally Posted by gpalmer I have used both the Meier and HeadAmp designs (and some software). The degradation in the signal, particularly the frequency range bothered me far more than the lack of crossfeed. I think Hirsch nailed it on this one, the brain does a much better job than the electronics I've heard...

The brain simply can't fix that problem, that is why some of those fellows developed the crossfeed theory, and devices...You may like it or not, and that is different, you can get used to the lack of it, or not, but you will never hear a totally lifelike music reproduction (unless a binaural recording) without it.....

It is very simple to explain, the crossfeed simulate the listening in the real world, giving some of the sound of the right channel to the left, and vice, like a bleeding done in studios, there is no way the brain could fix that mixing as there is no physical connection between both channels, while listening with headphone, unless there is hole in the head that connects both ears, or you manage to connect both earpieces, in some mechanical way, like a tube or the like to take the signal from one side to the opposite, or an extremelly open headphones, in which you literally could hear one channel from the other....

You can live with it, or without it, the choice is your as the devices exist, but there is absolutelly no way that the brain could fix that problem....that is physics...not voodoo magic...

It is also true that the crossfeed offer some changes in the spectrum of freqs, that the Meier one, has into consideration and compensate in part, there is no perfect crossfeed to the date, and all of them introduce anomalies in the signal, but no more than the anomalies due to the lack of it, IMO....

 

[gpalmer]

Quote:

Originally Posted by Sovkiller The brain simply can't fix that problem, that is why some of those fellows developed the crossfeed theory, and devices...

Of course it can Sov, have you heard of hallucinations? The brain can interpret the sensory perceptions and manipulate them any way it wants. It's like having a very powerful computer manipulating streams of data...

 

[sacd lover]

I have crossfeed on a used singlepower tube amp I just bought. After listening to the wide open soundstage of the normal singlepower tube amps the crossfeed sounds to closed in... something like what Hirsch described. I like using the amp much better without the crossfeed. The normal presentation has more transparency and openess. Fortunately the amp has a normal jack and a crossfeed jack.

 

[Whitehall]

I'm new to serious headphony but I'm using a vintage receiver while listening to FM, CD, and, best of all, SACD. The need for something like crossfeed to approximate binaurial recording is VERY apparent to me.

The crossfeed filter circuits are interesting in that I could use them in the tape loop with no additional active components (I think).

On the other hand Dolby Headphone seems much more sophisticated in theory - I haven't a good chance to do a listening test.

It would be best implemented in a CD/DVD-A player since it is purely digital domain.

Two downsides seem to be a limit on resolution to 96kHz/24bits and no DSD implementation. One could use an analogue signal and A/D converters to feed the processing chip (is a dedicated chip available?) then back to analogue to drive you cans. Seems like a royal pain.

Of course, if one only listened to CD tracks on the computer, there are software packages with Dolby Headphone included but that ties you to a sound card and no SACD or high resolution stereo DVD-A at 192/24 (competitive with SACD to my ear).

I do like the idea of cutting one's own CDs using DH on the computer - I'm going to try that.

I'm sure someone out there can add, correct, or expand the above.

 

[rickcr42]

Quote:

The crossfeed filter circuits are interesting in that I could use them in the tape loop with no additional active components (I think).

Yes

Quote:

On the other hand Dolby Headphone seems much more sophisticated in theory - I haven't a good chance to do a listening test.

totally different on many levels but the short and sweet is

Dolby headphone-headphone surround decoder/ambience simulator and actually adds a new signal

Crossfeed Network-moves the position of the image forward and compensates for having a lump of head in between the left and righ earpices thus simulating what happens with a left and a right loudspeaker in a listening room where there being no "barrier" between the two allows for the left and right channels to mix

Quote:

It would be best implemented in a CD/DVD-A player since it is purely digital domain.

Best implemented as a passive device for universal use with ANY source input

Quote:

Two downsides seem to be a limit on resolution to 96kHz/24bits and no DSD implementation. One could use an analogue signal and A/D converters to feed the processing chip (is a dedicated chip available?) then back to analogue to drive you cans. Seems like a royal pain.

See above.Only digital has a bit rate and resolution limit.Analog is whatever goes in comes out and being passive you avoid adding active circuitry self noise,ADC/DAC.power supply noise,AC line abberations,etc

Quote:

Of course, if one only listened to CD tracks on the computer, there are software packages with Dolby Headphone included but that ties you to a sound card and no SACD or high resolution stereo DVD-A at 192/24 (competitive with SACD to my ear).

Many of which are processor resource intensive and if you multitask or use another processor heavy program can and will have lock ups.Passive also avoids this

Quote:

I do like the idea of cutting one's own CDs using DH on the computer - I'm going to try that.

Be aware those CDs will become useless for any but headphone playback.This matrix does not translate back to loudspeaker use and why "flip the switch and insert matrix" really the best way to go if you want your CDs to be "dual use" speaker/headphone compatible

Quote:

I'm sure someone out there can add, correct, or expand the above.

nope

 

[Whitehall]

Thanks for the contribution, RICKCR42.

I suspect the question of crossfeed vs Dolby Headphone (DH) boils down to which enhances the headphone experience the most - or at least which do I prefer - even after the debits are considered (cost, noise, convenience, availablity).

Plus creating one DH disk allows one to listen to headphone-specific music on any machine with a player. One could also store it in an iPod.

Since I wrote my post last night, I was able to do some more research.

Common sound cards are able to digitize analog inputs in two channels at 96/24. Cards with 6 channels of A/D conversion are rare or impractically expensive but would give DH more to work with. Consider $100 for such a card. With it one could take a stereo analog output of a SACD/DVD-A player and make a WAV file(s) that could then be DH-itized into a headphone only file. A CD drive could rip 44/16 WAV files off CDs directly.

With that, another program (discWelder Bronze ~$90) could burn the DH-ized WAV file (at 96/24) onto a DVD-A. Likewise, most CD burners come with software to transfer the DHized CD-quality file to a CD.

What I haven't located yet is a program to convert straight WAVs to DH WAVs. Some DVD programs will do this is real time off a DVD (PowerDVD, WinDVD, ~$50) but I haven't found a one that will create and save a converted file. Also, would it help to upsample the DH file (96/24) to 192/24 and how would one do that?

This way I would have dedicated headphone disks (DVD-A or CD) at a $1 a pop for the medium plus my time. One DVD-A would hold 5 or 6 CDs.

My least-cost path to trying this is finding that missing program - one to convert a WAV file into a DH WAV then practicing on CDs (using existing equipment on my computer). That will give me a better idea of what Dolby Headphone is capable of doing for me and how much I like it. There are other competing protocols out there too - Spatializer is one.

Those DVD player programs also might work for real time DH conversion of FM or LP sources too (with the sound card.)

Rick or anyone have direct experience with DH or could identify the software/hardware I would need?

 

[halcyon]

As long as we are replicating a number of acoustic sources with two different emitters, we need to make some decisions:

1) do we record, mix and master for loudspeakers?

2) do we record, mix and master for headphones?

In the first case listening to headphones can sound unnatural and un-intended, because the signals were not designed to be listened separately, with no overlapping of the left/right signals. In this scenario cross-feed can help, although it has it's issues (related to phase differences, no simulation of hrtf and perceptual change in bass amplitude*, unless accounted for by the crossfeed design).

In the second case (binaural recordings) crossfeed isn't really useful and can be downright detrimental.

In some older cases (like Beatles odd "stereo" recordings) crossfeed can definitely help you a lot and the case is more like 1) than 2). It will sound very odd with no crossfeed, compared to loudspeakers, on which it was designed to be listened to.

* Cross-feed impacts the perception of bass loudness, even when the bass signal itself is not diminished in the crossfed signal. This can be compensated for in the crossfeed circuitry design so that bass impact is not diminished.

 

[sxr71]

Quote:

Originally Posted by Hirsch You've already got a crossfeed processor that's far more sophisticated than any currently on the market: your brain. That's where sensory signals are integrated to form the sonic "image". Headphones are an unusual way to receive stereo sound, and the brain is not used to integrating the signal. This can lead to the "blob" effect, where there is a concentration of sound at the center and earpieces, with gaps in the image. Crossfeed fills these gaps, and can present a stronger initial image by replicating the type of sound that you're more used to hearing and processing (both channels audible to both ears). However, there's a downside. You can learn to integrate the channels in headphone audio. If you listen to headphones a lot, you become more sophisticated at processing the signal, and the auditory image become smooth across the "stage". Add crossfeed then, and you get a big blob in the center of the head. Ugh. I tend to regard crossfeed as an impediment to learning how to process sound from headphones. It's a shortcut to getting a good sonic image from headphones, but in its presence, you never undergoe the perceptual learning needed to create a good image from headphones without it. It does work, but IMO it's a crutch, and should only be used sparingly, or you become dependent on it for headphone listening.

I've tried only one implementation - Xin's. I found that it firstly made the image really tiny in that it seemed to be the all concentrated in the 7" distance between my ears. Of course one could argue that the "image" created by headphones is unnaturally large.


The second thing was that it really cut down the highs and made cymbals and such sound quite dull. Maybe you need headphones equalized for binaural recordings like the ER-4B to make it sound right. This is just a guess and I don't really know for sure. Maybe crossfeed + some sort of equalization will make it sound right to me.