[71 dB]

You convolve once, then it's a low load fixed filter that is easy to run. It's been done, though not specifically an impulse response for HRTF. Audyssey's AMP app (now discontinued...nuts!) ran headphone EQ which started by essentially taking impulse response in their lab, developing the "tuning" filter, then users downloaded it and ran it on their iOS device. It worked fine even on devices several generations previous to current. In fact, the apps issues had nothing to do with running the DSP filter, that was actually the easy part.

I'm not sure why you're arguing about using the "right HRTF", when any generalized HRTF beats any 'normal' cross-feed. It's clearly impractical to measure every users HRTF, so it must be a generalization, perhaps a choice of 2 or 3, like small-medium-large, or whatever.

It depends on how well you want things done. Convolution with real HRTF > approximative IIR or FIR of a HRTF > normal cross-feed. It's also a question of diminishing returns. Normal cross-feed takes you to 80 %, approximative filters to 90 % and convolution with real HRTF to 95 % (numbers illustrate the principle). If 90 % is enough, why isn't 80 % as well?

All I'm saying is the technology is in our hands now, and there's a market gap. The missing critical information is market perceived need, which is probably why it hasn't been done. None of the details of the DSP algorithm are road blocks at all. It sounds like "Mr. Cross-Feed" is pushing back on the idea of a $2 DSP cross-feed app. Huh.

Huh.

 

[castleofargh]

You convolve once, then it's a low load fixed filter that is easy to run. It's been done, though not specifically an impulse response for HRTF. Audyssey's AMP app (now discontinued...nuts!) ran headphone EQ which started by essentially taking impulse response in their lab, developing the "tuning" filter, then users downloaded it and ran it on their iOS device. It worked fine even on devices several generations previous to current. In fact, the apps issues had nothing to do with running the DSP filter, that was actually the easy part.

I'm not sure why you're arguing about using the "right HRTF", when any generalized HRTF beats any 'normal' cross-feed. It's clearly impractical to measure every users HRTF, so it must be a generalization, perhaps a choice of 2 or 3, like small-medium-large, or whatever.

All I'm saying is the technology is in our hands now, and there's a market gap. The missing critical information is market perceived need, which is probably why it hasn't been done. None of the details of the DSP algorithm are road blocks at all. It sounds like "Mr. Cross-Feed" is pushing back on the idea of a $2 DSP cross-feed app. Huh.

talking about that, there is a Waves NX app for cellphones. I didn't think about it because it's originally made for a little blue tooth head tracking toy that's faaaar from being bug free, but you can still use it without the tracker and get their simulation for the head for free. they ask for 2 measurements of the head to set up their target. it's more intuitive than asking for delay, attenuation, where to roll off...
if you take the room setting down, what's left is xfeed

 

[pinnahertz]

It depends on how well you want things done. Convolution with real HRTF > approximative IIR or FIR of a HRTF > normal cross-feed. It's also a question of diminishing returns. Normal cross-feed takes you to 80 %, approximative filters to 90 % and convolution with real HRTF to 95 % (numbers illustrate the principle). If 90 % is enough, why isn't 80 % as well?



Huh.

We've traded positions.

 

[Zapp_Fan]

It seems to me the two above quotes are somewhat at odds with each other. Which is it, we mix for speakers and headphones, or we mix for speakers primarily then do a cursory check for headphones? (It's the latter, BTW).

I'm kinda wishing there were fewer statements and assumptions about what audio pros do made by people who don't seem to actually know.

Sorry for the slow uptake here, been quite busy. This is a fair comment, and it bears mentioning that when people say "check on headphones" they usually mean for frequency response, not stereo image.

Well, this Ryoki Ikeda is certainly not as bad as Merzbow. Anyone ever hear Merzbow? I'm not going to put a link down, but if you are curious I will warn you: turn down your volume! I found the existence of Merzbow after browsing the DR database one day and organizing results by album with least dynamic range. Merzbow and a couple other "noise artists" earned a prestigious 0 DR. Yep, that's 0 db of range. Nada. Nill. Nichts. It'll drill your flippin' brain out. Japanese electronic music seems to be heavily influenced by noise.

Ha, if you think Merzbow is bad, try some Prurient... I don't listen to him, not sure why anyone does...

So why is cross-feed not universal at this late date? Could it be it's not universally desired or accepted? I don't know, but I'm in the "not universally desired" camp.

I'm relatively sure that it's because manufacturers consider crossfeed to have only niche appeal and their customers aren't asking for it. The vast bulk of the audio market is comprised of people who have no critical listening experience. Headphones you'd consider mediocre or even horrible make most people quite happy. For most consumers, a crossfeed feature would be akin to putting a spoiler on a bicycle. I don't say this to deingrate the average listener in any way - it's just that most people wouldn't feel a need for crossfeed, or necessarily even appreciate the difference.

Those who are more engaged with sound quality will be assumed to be capable enough to implement crossfeed themselves. And, this thread somewhat proves that, no?

You convolve once, then it's a low load fixed filter that is easy to run.

I think for real convolution this is not strictly true, but the general point is true, which is that you could probably convolve with an HRTF IR on most smartphones today.

I wanted to bring up another question, which is related, but not strictly about crossfeed.

The site RTINGS.com has automated headphone testing and rating, and their sound quality ratings depend on whether a headphone is open or closed. As I understand it, open headphones get an automatic bonus for the "critical listening" score. They measure openness in part by measuring acoustic crosstalk within the room from ear to ear - for headphones!

This struck me as odd, and I assume they're just using openness as a proxy for quality, and don't believe that audio leaking from one ear to the other is an objective measure of sound quality. Rather, acoustic crosstalk on headphones is a measure of openness, which is a heuristic/proxy for overall sound quality. That's my guess anyway...

Anyone know for sure what this metric is getting at?

 

[Sc00p]

I'm not sure if this is all in my head. I have never really used any kind of cross feed. Mainly because any amps i chose didn't have this kind of feature, and i didn't seek it out as i have always read a lot of negativity surrounding it. Anyways, my current amp has it and i thought i would try it out.
Before trying the cross feed function: I prefer music through headphones rather than speakers. But for a long time, i have felt on many tracks that the right ear sounded too dominant, many songs sounding unbalanced or skewed. I started to wonder if it was my hearing, but if i change to mono there is no inbalance. This is for multiple amps, dacs, headphones so i knew it wasn't my gear.
I put up with it because, even though i don't hear it with speakers, my other preferences lean heavily towards headphones.
So, last night i decided to have a long listening session with the 3d crossfeed on my amp. I am inexperienced, but this seems to have fixed my problem. Everything sounds so much more balanced (not mono, it is still very much stereo), more natural.

Presuming this isn't in my head. What could be a possible explanation? without crossfeed eg, a singer who is imaged in the middle, is in the middle. But if the singer is imaged to come from both sides, it quite often feels to be louder from my right. Could it be something to do with timings? With crossfeed enabled. This annoying effect is gone for me.
Am i describing an effect when others say "without crossfeed is unnatural sounding" ?

 

[Zapp_Fan]

Presuming this isn't in my head. What could be a possible explanation? without crossfeed eg, a singer who is imaged in the middle, is in the middle. But if the singer is imaged to come from both sides, it quite often feels to be louder from my right. Could it be something to do with timings? With crossfeed enabled. This annoying effect is gone for me.
Am i describing an effect when others say "without crossfeed is unnatural sounding" ?

A couple possibilities - one is that your headphones have a driver matching problem, i.e. different frequency response on each ear, which can cause pretty noticeable imabalances like what you describe. If you have nice headphones this is unlikely, but you can easily check by listening to a frequency sweep and then seeing if the sound seems to move from side to side.

Another is that the song is just mixed that way, with something harder left/right than sounds natural. The crossfeed inherently diminishes that.

Another possibility is your ears have uneven frequency response, it's not unusual to have that type of hearing loss over time.

 

[71 dB]

So, last night i decided to have a long listening session with the 3d crossfeed on my amp. I am inexperienced, but this seems to have fixed my problem. Everything sounds so much more balanced (not mono, it is still very much stereo), more natural.

Presuming this isn't in my head. What could be a possible explanation? without crossfeed eg, a singer who is imaged in the middle, is in the middle. But if the singer is imaged to come from both sides, it quite often feels to be louder from my right. Could it be something to do with timings? With crossfeed enabled. This annoying effect is gone for me.
Am i describing an effect when others say "without crossfeed is unnatural sounding" ?

The idea of crossfeed is to make the sound more natural for headphones reducing excessive stereo separation. When you listen to loudspeakers, there acoustic crossfeed because left ear hears rigth speakers and vice versa. So, if you find crossfeed natural, you are simply hearing it as intented.

 

[SilverEars]

Perhaps that's a better solution for movie watching with headphones? Try crossfeed? Is there a good plug-in for windows to try out for movies, videos? For videos and movie watching I prefer speakers as it sounds more natural.

 

[bigshot]

I think what you're reacting to is how 5.1 gets folded down into 2 channel for headphones. Dialogue works much better when it's isolated in that center channel speaker than when it is halfway between your ears with cans.

 

[71 dB]

Perhaps that's a better solution for movie watching with headphones? Try crossfeed? Is there a good plug-in for windows to try out for movies, videos? For videos and movie watching I prefer speakers as it sounds more natural.

Multichannel movie sound has a lot of stuff going on in rear channels. When those are downmixed to stereo, there's a lot of stereo separation, because rear channels are encoded to stereo out of phase. When I watch movies I usually use strong crossfeed for that reason and the result is good imo.

 

[Malfunkt]

Perhaps that's a better solution for movie watching with headphones? Try crossfeed? Is there a good plug-in for windows to try out for movies, videos? For videos and movie watching I prefer speakers as it sounds more natural.

For movie watching, even from Netflix with multi-channels, definitely check out Dolby Atmos for Headphones. Search for it in the Windows store. It really is excellent. It will work on stereo sources too, but its best for multi-channel.

 

[pinnahertz]

Multichannel movie sound has a lot of stuff going on in rear channels. When those are downmixed to stereo, there's a lot of stereo separation, because rear channels are encoded to stereo out of phase. When I watch movies I usually use strong crossfeed for that reason and the result is good imo.

There are several downmix algorithms, a common one does result in Ls and Rs mixed out of phase, because that downmix is compatible with ProLogic decoding. There are other algorithms, such as Dolby Headphone, that begin with the raw 5.1/7.1 or even 2.0 ProLogic (Dobly Stereo, if a legacy track), and result in virtual surround in headphones by processing the surround channel(s) with appropriate spatial cues.

Hardly cross-feed, though. Basic cross-feed of an LtRt mix or downmix will not localize surround properly at all, and will, if applied strongly enough, upset the intended mix balance of LCRS. A lot more has to be done to get that right.

 

[71 dB]

There are several downmix algorithms, a common one does result in Ls and Rs mixed out of phase, because that downmix is compatible with ProLogic decoding. There are other algorithms, such as Dolby Headphone, that begin with the raw 5.1/7.1 or even 2.0 ProLogic (Dobly Stereo, if a legacy track), and result in virtual surround in headphones by processing the surround channel(s) with appropriate spatial cues.

Hardly cross-feed, though. Basic cross-feed of an LtRt mix or downmix will not localize surround properly at all, and will, if applied strongly enough, upset the intended mix balance of LCRS. A lot more has to be done to get that right.

Cross-feed of a Lt/Rt mix or downmix is the best I can do (on DVDs and Blu-rays) with the stuff I have. Fortunately for me it works really well despite your efforts.

 

[pinnahertz]

Fortunately for me it works really well despite your efforts.

My efforts? To do what, exactly? All I did was explain the reality of that downmix situation.

 

[71 dB]

My efforts? To do what, exactly? All I did was explain the reality of that downmix situation.

Crossfeed is like sonic umbrella: It doesn't stop the rain, but it keeps you from getting wet. My effort is to keep people dry. Your effort is to have only sunny days. My way is to buy an umbrella, your way is to move to a paradise island where it never rains.

Also, I don't speak to anyone with Dolby Headphone. I have never heard it, but I'm sure they are better off than I. I am speaking to people who don't use any kind of separation reduction system and don't even know about them or what to know more.