[Rob the Comic]

Yes i do use crossfeed. Never been a fan of crossfeed. Until i bought SPL Phonitor xe and tested their Matrix. It's been in use since.
It is very subtle, does not colour sound. And makes listening much more relaxed and enjoyable. Works for me.

Happy to read this as I have the XE arriving next week. I will jump through all the hoops and see what is what - thanks.

 

[71 dB]

I ruined my self-confidence here with my crossfeed posts, but Wednesday I start a course on mixing using Pro Tools. I hope after the course I am not considered a clueless fool anymore! At least I DON'T care anymore, because I have done the damn course! At least people can't say I haven't done the course!

 

[bigshot]

ProTools is like all programs. The basics are pretty easy, but there are levels upon levels of added features that you can learn a little at a time.

 

[71 dB]

ProTools is like all programs. The basics are pretty easy, but there are levels upon levels of added features that you can learn a little at a time.

Pro Tools doesn't seem seriously difficult. Ease of use is a good marketing point. My problem is asperger, which makes it difficult to learn "series of small steps" as automated actions. Learning logical connections between things (system thinking) is easy for me, but only a small part of learning a new software is about that. For example learning short cuts is hard for me, because they are often a bit "random." so there is not much logic behind. Why does "Command" + "=" toggle between Edit and Mix windows? Somebody just chose that shortcut for whatever reason and it's like memorazing the decimals of pi.

 

[ironmine]

Has anyone here mentioned https://www.dsoniq.com before? Sounds really great.

 

[FlavioWolff]

I made crossfeed for use with Eqalizer APO
If someone is interested and want to try. Delay is in samples for 44.1kHz

Thanks, I'm gonna try this out. Does it approach more of the CMOY, or the Meier method?
Is it anyhow different from HeSuVi's crossfeed (which also uses EqAPO)?
Thanks!

 

[71 dB]

Double crossfeed? Huh?

Generally it seems to be a bad idea to do crossfeed twice. Using typical crossfeeders the sound becomes too centered and dead. However, there seems to be ways to do double crossfeed successfully. I have been using an arrangement where I have my "wide crossfeeder" followed by a very simple crossfeeder:

Wide crossfeeder is a variation of the CMOY circuit where the crossfeed cut off frequency is lowered to 300 Hz from 800 Hz causing much bigger phase differencies at low frequencies. It emulates the situation where the speakers are on both sides of the head rather than at 30° angles on front. The sound appears wide and not "forward" the way normal CMOY does.

Simple crossfeed is laughable simple, but gives surprisingly nice results similar to Meier crossfeed: It is just a coil and resistor in series connecting left and right channels so that low frequencies "leak" between channels. I use a "UPOC" variation of this idea which uses a few more resistors to make the circuit work better with any output or headphone impedances.

The resulting sound using these two crossfeeders is not too narrow, because the first crossfeeder makes the sound wide and the use of two different crossfeed topologies (CMOY is "X" topology and Meier/simple coil crossfeeders are "H" topology) gives the flavor of both to the sound: "X" topology tries to emulate two speakers at fixed angles while "H" topology spreads continuously the soundstage from left to right. The miniature soundstage I hear using this double crossfeed arrangement is very "solid" and "even", but also small. The sound appears near my head. It feels like wearing a helmet protecting against excessive spatiality. The crossfeeders I use for this have fixed values for crossfeed, but it is possible by making both crossfeeders a bit "milder" the miniature soundstage could increase in size on the expense of the "solidness" and "evenness" of the sound.

Multiple crossfeed? Huh?

If double crossfeed can work, how about multiple crossfeed? What if we crossfeed the signal 10 or 20 times using extremely mild crossfeeders? It would be approaching the final result slowly and might be a method to have a fixed crossfeed that works well with all kind of material with excessive spatiality of varying degree. The biggest problem with this ideas is the massive complexity and amount of variables to optimize! What are the optimal crossfeed types and in what order?

 

[bigshot]

I want a helmet to protect me from excess space!

 

[FlavioWolff]

Double crossfeed? Huh?

Generally it seems to be a bad idea to do crossfeed twice. Using typical crossfeeders the sound becomes too centered and dead. However, there seems to be ways to do double crossfeed successfully. I have been using an arrangement where I have my "wide crossfeeder" followed by a very simple crossfeeder:

Wide crossfeeder is a variation of the CMOY circuit where the crossfeed cut off frequency is lowered to 300 Hz from 800 Hz causing much bigger phase differencies at low frequencies. It emulates the situation where the speakers are on both sides of the head rather than at 30° angles on front. The sound appears wide and not "forward" the way normal CMOY does.

Simple crossfeed is laughable simple, but gives surprisingly nice results similar to Meier crossfeed: It is just a coil and resistor in series connecting left and right channels so that low frequencies "leak" between channels. I use a "UPOC" variation of this idea which uses a few more resistors to make the circuit work better with any output or headphone impedances.

The resulting sound using these two crossfeeders is not too narrow, because the first crossfeeder makes the sound wide and the use of two different crossfeed topologies (CMOY is "X" topology and Meier/simple coil crossfeeders are "H" topology) gives the flavor of both to the sound: "X" topology tries to emulate two speakers at fixed angles while "H" topology spreads continuously the soundstage from left to right. The miniature soundstage I hear using this double crossfeed arrangement is very "solid" and "even", but also small. The sound appears near my head. It feels like wearing a helmet protecting against excessive spatiality. The crossfeeders I use for this have fixed values for crossfeed, but it is possible by making both crossfeeders a bit "milder" the miniature soundstage could increase in size on the expense of the "solidness" and "evenness" of the sound.

Multiple crossfeed? Huh?

If double crossfeed can work, how about multiple crossfeed? What if we crossfeed the signal 10 or 20 times using extremely mild crossfeeders? It would be approaching the final result slowly and might be a method to have a fixed crossfeed that works well with all kind of material with excessive spatiality of varying degree. The biggest problem with this ideas is the massive complexity and amount of variables to optimize! What are the optimal crossfeed types and in what order?

Very intriguing. Have you ever thought about researching this in the university?

 

[71 dB]

Very intriguing. Have you ever thought about researching this in the university?

Well, I haven't worked at the university for 15 years...

 

[starepiernikowe]

Hello

I have been using crossfeed something like 14 years now.
In the beginning I really had problems with it, because all "types" of crossfeed offered by different plugins sounded to me strange.
I really didn't like the stereo image that was presented as two speakers "in the headphones".
I mostly used BS2B in foobar2000 because it didn't make my head hurt.
But I wanted something that will feel natural, and after few years of looking and testing I have finally found it.
I use this little plugin for foobar2000 since the latest version came for f2k 1.1

I used to have it at default settings wit mixer level changed to 15% - that widens the sound stage.
Now with my new buy Shure SRH1840 I use it with all settings set to default "mixer level 25%".
If you use f2k I highly recommend this plugin.

To me, what it does; it takes the music from inside of my head to "on my face and sides".
Without feeling they delay and the speaker placement, the sound is on me, on my face.
What this plugin does is very unique around all other crossfeed plugins.
Try it out, write what you think.
http://www.naivesoftware.com/


Test sample:
https://drive.google.com/drive/folders/1ZT3dEF69iCGcvJh1PfooEEG2yk3Ch8jw?usp=sharing


Paweł

 

[sonitus mirus]

It is probably the Jan Meier natural crossfeed. I like that with headphones and I have had a few headphone dac/amps that included this DSP option. Even my current RME ADI-2 DAC FS includes 3 crossfeed options, with Meier being one of them. I greatly prefer listening to music with speakers, and Meier gets my headphones closer to stereo speaker sound for me, though still not particularly similar.

 

[FlavioWolff]

Hello,
I'm currently experimenting HeSuVi's Crossfeed implementation via EqualizerAPO.
Which settings should I input to have a Jan Meier-like approach?

These are the available settings:

Frequency Cutoff: The frequency that is central to the gain changes set below.
Direct Shelf Boost: The gain increase of the higher frequencies for the adjacent ear.
Cross Attenuation: The frequency-independent gain decrease of the audio signal to the opposite ear.
Cross Delay: The frequency-independent delay of the audio signal to the opposite ear.
Cross Shelf Decay: The gain decrease of the higher frequencies for the opposite ear
Constant 6 dB/Oct Decay: The value above will be ignored and a constant 6 dB/Oct decay is used instead for the higher frequencies at the opposite ear.

They default as:

Frequency Cutoff: 700 Hz
Direct Shelf Boost: 30 dB/10
Cross Attenuation: 90dB/10
Cross Delay: 18 Samples
Cross Shelf Decay: 100 dB/10
Constant 6dB/Oct Decay: unchecked by default.

Thank you!

 

[starepiernikowe]

It is probably the Jan Meier natural crossfeed. I like that with headphones and I have had a few headphone dac/amps that included this DSP option. Even my current RME ADI-2 DAC FS includes 3 crossfeed options, with Meier being one of them. I greatly prefer listening to music with speakers, and Meier gets my headphones closer to stereo speaker sound for me, though still not particularly similar.

Naive from what I know is not based on Meier, to my ears Meier is to much in front - at least the f2k software version is.
Naive started as a winamp crossfeed in 2002.
https://hydrogenaud.io/index.php?topic=4140.msg42151#msg42151

 

[71 dB]

Hello,
I'm currently experimenting HeSuVi's Crossfeed implementation via EqualizerAPO.
Which settings should I input to have a Jan Meier-like approach?

These are the available settings:

Frequency Cutoff: The frequency that is central to the gain changes set below.
Direct Shelf Boost: The gain increase of the higher frequencies for the adjacent ear.
Cross Attenuation: The frequency-independent gain decrease of the audio signal to the opposite ear.
Cross Delay: The frequency-independent delay of the audio signal to the opposite ear.
Cross Shelf Decay: The gain decrease of the higher frequencies for the opposite ear
Constant 6 dB/Oct Decay: The value above will be ignored and a constant 6 dB/Oct decay is used instead for the higher frequencies at the opposite ear.

They default as:

Frequency Cutoff: 700 Hz
Direct Shelf Boost: 30 dB/10
Cross Attenuation: 90dB/10
Cross Delay: 18 Samples
Cross Shelf Decay: 100 dB/10
Constant 6dB/Oct Decay: unchecked by default.

Thank you!

HeSuVi's Crossfeed seems to be using "X-topology" approach while Jan Meier uses "H-topology" approach. This means that the resulting soundstages are different in nature. "H-topology" approach gives edgier and more "surround"-like soundstage while "X-topology" approach is more like listening to stereo speakers.

People seem to want to have "their personal settings" for crossfeed, but actually each recording calls for it's own setting depending on the spatial nature of the recording. Some recordings don't need crossfeed at all, while some other recordings need massive crossfeeding. Here are what I would recommend for three basic settings weak, moderate and strong:

WEAK:
Frequency Cutoff: 800 Hz
Direct Shelf Boost: 17 dB/10
Cross Attenuation: 100 dB/10
Cross Delay: 11 Samples (@ 44.1 kHz)
Cross Shelf Decay: 200 dB/10
Constant 6dB/Oct Decay: unchecked by default.

MODERATE:
Frequency Cutoff: 800 Hz
Direct Shelf Boost: 26 dB/10
Cross Attenuation: 60 dB/10
Cross Delay: 11 Samples (@ 44.1 kHz)
Cross Shelf Decay: 200 dB/10
Constant 6dB/Oct Decay: unchecked by default.

STRONG:
Frequency Cutoff: 800 Hz
Direct Shelf Boost: 38 dB/10
Cross Attenuation: 20 dB/10
Cross Delay: 11 Samples (@ 44.1 kHz)
Cross Shelf Decay: 200 dB/10
Constant 6dB/Oct Decay: unchecked by default.