[FlavioWolff]

No need to upsample anything to 48KHz. The main thing is to downsample everything to 48KHz. But on Windows, if you use Direct Sound, then you don't have to worry about resampling, it will be done through Direct Sound.

I see! But my dac still sees 44.1khz when playing music through OOYH

 

[FlavioWolff]

Guys, I’ve seen a lot of different suggestions and it all became a little confusing.
What’s the recommended target curve to eq our headphones in order to enjoy OOYH the best? Or, at least, the best curve to start?

Adding a bit to this:

Regarding EQ (target curve), there is apparently no consensus here.

This is what I gathered so far by reading most of this thread:

- edwardsean and phoenixdogfan recommend the Sonarworks Reference 3 curve

- Zenvota suggests that we do some "inverting EQ" to compensate from double HRTF (by reducing the 2-6khz area, IIRC. Didn't understand if he EQed to any other curve before EQing that region)

- Jakko advocates the Harman Curve or the Diffuse Field Curve, saying that absolute flat could be bad because it removes the natural 3khz peak from the ear canal's ressonance (but could work for wonky/very bright measurements).

So, while zenvota suggests reducing the 2-6khz region gain, jakko suggests using curves that boost 3khz. Despite me being a total noob, they seem to be taking opposite directions here.

- Darin, when answering an email I sent him asking about headphone EQ and OOYH, said:

"Hi Flavio,

In theory, a neutral response in the headphones is desirable. That way the sound you hear will be a representation of the sound of the speakers in the room when they were measured.

But I would EQ the headphones so they sound good to you and your preferences.
Also spending on the headphones, some people find some of the presets to be a little bright. You can EQ your headphones to taste if the sound from that particular preset sounds good to you.

It's hardly an exact science when it comes to EQing headphones."

Then I asked this:

"Thanks again. Reading the OOYH thread, I found a post of yours saying that you checked the effects with headphones eqed to “flat”. Do you recall what flat target that was? A flat line? The Harman target?
I’m really curious and fond to achieve the best reproduction of the rooms.
Thanks!"

And the final answer was:

"Hi Flavio,
The target was a flat line vs. Harman Target.
-Darin"
______________________________________________________

In my particular case (AKG K371 which follows the Harman Curve very closely), most presets are too bassy (almost unbearable) and also too bright when uneq'd or eq'd to Harman (2013 or 2018)
I have then created a Diffuse Field and a total flat EQ using AutoEQ (based on oratory1990's measurements), and total flat sounds excellent with almost all of the OOYH presets. Great tight bass and balanced mids and highs (Cello preset is a total bliss - I can picture myself in a very deep retangular room with speakers far far beyond me!); diffuse field sounded a bit bass shy and too bright with most, but still much better than Harman (or no-eq).

Can we discuss a bit more about this topic? I think this is absolutely essential.

(sorry for my English)

 

[edwardsean]

Adding a bit to this:

Regarding EQ (target curve), there is apparently no consensus here.

This is what I gathered so far by reading most of this thread:

- edwardsean and phoenixdogfan recommend the Sonarworks Reference 3 curve

- Zenvota suggests that we do some "inverting EQ" to compensate from double HRTF (by reducing the 2-6khz area, IIRC. Didn't understand if he EQed to any other curve before EQing that region)

- Jakko advocates the Harman Curve or the Diffuse Field Curve, saying that absolute flat could be bad because it removes the natural 3khz peak from the ear canal's ressonance (but could work for wonky/very brights measurement).

So, while zenvota suggests reducing the 2-6khz region gain, jakko suggests using curves that boost 3khz. Despite me being a total noob, they seem to be taking opposite directions here.

- Darin, when answering an email I sent him asking about headphone EQ and OOYH, said:

"Hi Flavio,

In theory, a neutral response in the headphones is desirable. That way the sound you hear will be a representation of the sound of the speakers in the room when they were measured.

But I would EQ the headphones so they sound good to you and your preferences.
Also spending on the headphones, some people find some of the presets to be a little bright. You can EQ your headphones to taste if the sound from that particular preset sounds good to you.

It's hardly an exact science when it comes to EQing headphones."

Then I asked this:

"Thanks again. Reading the OOYH thread, I found a post of yours saying that you checked the effects with headphones eqed to “flat”. Do you recall what flat target that was? A flat line? The Harman target?
I’m really curious and fond to achieve the best reproduction of the rooms.
Thanks!"

And the final answer was:

"Hi Flavio,
The target was a flat line vs. Harman Target.
-Darin"
______________________________________________________

In my particular case (AKG K371 which follows the Harman Curve very closely), most presets are too bassy (almost unbearable) and also too bright when uneq'd or eq'd to Harman (2013 or 2018)
I have then created a Diffuse Field and a total flat EQ using AutoEQ (based on oratory1990's measurements), and total flat sounds excellent with almost all of the OOYH presets. Great tight bass and balanced mids and highs (Cello preset is a total bliss - I can picture myself in a very deep retangular room with speakers far far beyond me!); diffuse field sounded a bit bass shy and too bright with most, but still much better than Harman (or no-eq).

Can we discuss a bit more about this topic? I think this is absolutely essential.

Flavio, I think it is awesome that you're taking such a comprehensive approach. I do think a flatter EQ is desirable so that you're not adding correction on top of correction, if that makes sense. I think part of why you are getting such a wide range of responses, is because there are so many variables at play here. A theoretically "flat" curve helps to provide at least some basis to play around with other factors.

In the end, I guess it goes without saying that, you do have to explore, adjust, and trust your ears. I know that it is a bit of useless advice. However, as I've often said, in these discussions, "if it sounds good it is good." I don't think it's efficient to get too bogged down with all complications of theoretical psychoacoustics. Not that it's not complicated, but because adjusting for each individual set of ears, perceptions, systems, tastes, etc., is more complicated than we can account for at present. So, you have to go with a fair amount of tuning by ear.

I use sonarworks for my HD800 and Audeze reveal for my LCDi4s. Both of these are custom solutions for their respective phones, but I also have another EQ in the chain that I adjust manually.

As an aside, the Cello setting is my bliss as well!

All the best to you.

 

[FlavioWolff]

Flavio, I think it is awesome that you're taking such a comprehensive approach. I do think a flatter EQ is desirable so that you're not adding correction on top of correction, if that makes sense. I think part of why you are getting such a wide range of responses, is because there are so many variables at play here. A theoretically "flat" curve helps to provide at least some basis to play around with other factors.

In the end, I guess it goes without saying that, you do have to explore, adjust, and trust your ears. I know that it is a bit of useless advice. However, as I've often said, in these discussions, "if it sounds good it is good." I don't think it's efficient to get too bogged down with all complications of theoretical psychoacoustics. Not that it's not complicated, but because adjusting for each individual set of ears, perceptions, systems, tastes, etc., is more complicated than we can account for at present. So, you have to go with a fair amount of tuning by ear.

I use sonarworks for my HD800 and Audeze reveal for my LCDi4s. Both of these are custom solutions for their respective phones, but I also have another EQ in the chain that I adjust manually.

As an aside, the Cello setting is my bliss as well!

All the best to you.

Thank you for your insights.
I understand what you mean! I'm aware that there will never be perfection, and that every pair of ears will hear things differently. What I'm seeking is a bit of a consensus on what are the best practices regarding EQ when using OOYH, more about achieving the original recorded sound (albeit impossible) than about individual preferences (which could be tweaked afterwards).
Flatter (your answer), for example, is a good guideline, but some of the advices found here are not towards "flatter" at all, if I understood them correctly. That's what causing confusion! What should be a starting point before adjusting to our preferences and HRTF particularities?

In the end, it's all about enjoyment. But the mind plays a great role in perceived sound quality, especially when one is aware that he/she is following recommended practices.

 

[edwardsean]

Thank you for your insights.
I understand what you mean! I'm aware that there will never be perfection, and that every pair of ears will hear things differently. What I'm seeking is a bit of a consensus on what are the best practices regarding EQ when using OOYH, more about achieving the original recorded sound (albeit impossible) than about individual preferences (which could be tweaked afterwards).
Flatter (your answer), for example, is a good guideline, but some of the advices found here are not towards "flatter" at all, if I understood them correctly. That's what causing confusion! What should be a starting point before adjusting to our preferences and HRTF particularities?

In the end, it's all about enjoyment. But the mind plays a great role in perceived sound quality, especially when one is aware that he/she is following recommended practices.

Yes, I think I understand what you're getting at. I think it's a worthwhile endeavor. I would be interested if there is a consensus as well.

However, in everything I mentioned, my focus wasn't really about personal preferences. I was also talking about the best route to get optimum sound quality. I emphasize the need for manual EQ editing, and using your ears throughout from beginning to end, not just as a matter to suit to taste. I think it is necessary as a guide and correction for proper sound, even if you do find a consensus.

I don't mean to say that SQ is all subjective, and that everyone hears differently so go with whatever sounds good to you. I think there is an objective scientific answer to what you are looking for. I don't think that's quite possible within the limitations of OOYH. To find the actual right EQ curve, you really need to measure your own ears with the actual system in the actual room.

This is some really complicated stuff mathematically, and you need to get a lot right for it to work. Here, with OOYH we are dealing with approximations. We are using systems, rooms, HRTFs, etc, with frequency responses that have no objective relation to your ears, equipment, and the curve you select. We are trying to gain consistent results with the way a curve responds with the arbitrary relations of an IR of one system/room to another, and to the donor HRTF to your HRTF and system.

The only proper solution is to correct EQ by a personal measurement. Barring that, I tend to think that listening throughout the process of selecting presets, and making final adjustments by ear may be more important than whether you start with one theoretical curve or another.

Still, you are asking for a best practice, and that makes a lot of sense. I've chosen the flat curve as my starting point, and I would also be curious to hear other thoughts.

 

[jincuteguy]

So it's 2020 and does anyone still use this Out Of Your Head software for gaming with your High End headphones and DAC / Amp?

 

[Zenvota]

So it's 2020 and does anyone still use this Out Of Your Head software for gaming with your High End headphones and DAC / Amp?

I'm still using it for music and occasionally movies(just due to my own circumstances), havent been playing games lately. but it does work fantastic for gaming, especially paired with bass shakers.

 

[manukmanohar]

I'm still using it for music and occasionally movies(just due to my own circumstances), havent been playing games lately. but it does work fantastic for gaming, especially paired with bass shakers.

Would be interested to know what hp/iems you are using them with, and what preset worked for you?

 

[Zenvota]

Would be interested to know what hp/iems you are using them with, and what preset worked for you?

desktop system in my sig, primarily open planars with large angled velour pads, current signal non negative feedback amplifiers, and signal/power conditioning. I havent found dynamic/hybrid iems to work as well(i recently got a planar set but i havent tried them yet). They seem to lack the response of the planars and dont render the natural decay and reflections as well. Really benefits from low distortion, fast impulse, free of ringing/resonance, to get a really natural sounding rendering. I primarily use the Genelecs preset for music due to the type of music i listen to and the studio monitor sound as im sitting at a desk with no tactile transducers. But i have several presets and they all work well, typically only needing a little channel balance or treble rolloff to sound great. The more reverberant large speakers large room presets work much better with tactile transducers and lights off, feeling those low frequencies on most of your body compared to just your head makes a huge difference in externalization. I just do the headphone correction manually with eq apo and id imagine im doing some hrtf correction as well. If im not listening to metal a wide array of presets sound perfectly natural for other music.

Let's see, I have the
Acoustic Zen Crescendos
Magico Q3(boomy)
Genelecs(most lofi metal)
Egyptian Theater(big theater sound for old movies)
Focal Scala Utopian(rich)
Ribbons(clinical)
Cello Stradivari Premiere(very reverberant)
AZen Adagio Plus(bit more forgiving of crap recordings)
Volti Vittoras(good for crunchy guitars)
PBN Audio Sammy(good for bass guitars)
JMLab Nova Utopia Be(bigger brighter sounding room)
Marten Coltrane Memento V2(bigger bassier room)

Using specific equipment, correcting the headphones/hrtf, tactile transducers, and cleaning up external noise made a tremendous difference in the perceived naturalness and externalization of the rendering, "phantom like imaging"

 

[jincuteguy]

desktop system in my sig, primarily open planars with large angled velour pads, current signal non negative feedback amplifiers, and signal/power conditioning. I havent found dynamic/hybrid iems to work as well(i recently got a planar set but i havent tried them yet). They seem to lack the response of the planars and dont render the natural decay and reflections as well. Really benefits from low distortion, fast impulse, free of ringing/resonance, to get a really natural sounding rendering. I primarily use the Genelecs preset for music due to the type of music i listen to and the studio monitor sound as im sitting at a desk with no tactile transducers. But i have several presets and they all work well, typically only needing a little channel balance or treble rolloff to sound great. The more reverberant large speakers large room presets work much better with tactile transducers and lights off, feeling those low frequencies on most of your body compared to just your head makes a huge difference in externalization. I just do the headphone correction manually with eq apo and id imagine im doing some hrtf correction as well. If im not listening to metal a wide array of presets sound perfectly natural for other music.

Let's see, I have the
Acoustic Zen Crescendos
Magico Q3(boomy)
Genelecs(most lofi metal)
Egyptian Theater(big theater sound for old movies)
Focal Scala Utopian(rich)
Ribbons(clinical)
Cello Stradivari Premiere(very reverberant)
AZen Adagio Plus(bit more forgiving of crap recordings)
Volti Vittoras(good for crunchy guitars)
PBN Audio Sammy(good for bass guitars)
JMLab Nova Utopia Be(bigger brighter sounding room)
Marten Coltrane Memento V2(bigger bassier room)

Using specific equipment, correcting the headphones/hrtf, tactile transducers, and cleaning up external noise made a tremendous difference in the perceived naturalness and externalization of the rendering, "phantom like imaging"

Damn that's a lot of presets, i think you have to pay $150 for 1 preset, then every extra preset costs $25 more.
So which Preset to you that you think is the best for gaming? like in terms of surround sound cues. I don't care for music. I have Hifiman Arya headphones.

 

[Zenvota]

Damn that's a lot of presets, i think you have to pay $150 for 1 preset, then every extra preset costs $25 more.
So which Preset to you that you think is the best for gaming? like in terms of surround sound cues. I don't care for music. I have Hifiman Arya headphones.

I tended to pick up chunks of presets during of of the bi-annual half off sales.

Ah, well, Genelecs or AZen Adagio Pluses if you're not using bass shakers, but if you have shakers then alot of em sound really cool... the JMLabs are very spacious and bright so that could help with localization. The coltrane memento v2 is very spacious but bassy if you're looking for a more immersive tactile experience. The ribbons could also be good for localization.

gotta eq those Aryas though, just start with flattening em out / rolling off 2-10khz

 

[Takeanidea]

The OOYH software is a generic version of the Smyth Realiser. Most of the presets are copied from a file sharing system that Smyth setup for owners of the A8. It's clever that they've been made available through software, but they don't compare to the Realiser because the Realiser uses your own ears as a source rather than someone else's.

 

[edwardsean]

The OOYH software is a generic version of the Smyth Realiser. Most of the presets are copied from a file sharing system that Smyth setup for owners of the A8. It's clever that they've been made available through software, but they don't compare to the Realiser because the Realiser uses your own ears as a source rather than someone else's.

True. At the same time the Realiser also has the IR exchange and James Smyth said that one of his favorite IR captures was made from someone else's HRTF measurements.

 

[Zenvota]

The OOYH software is a generic version of the Smyth Realiser. Most of the presets are copied from a file sharing system that Smyth setup for owners of the A8. It's clever that they've been made available through software, but they don't compare to the Realiser because the Realiser uses your own ears as a source rather than someone else's.

Theyre all made with an A8. And yes personalized measurements will surely provide the best quality. But some basic hrtf correction can be made that will help the preset sound more natural / personalized.



Additionally these various programs and devices give us multiple options for different scenarios. The Realisers can be used without a pc, Impulcifer allows for all sorts of room correction, etc. And all of them require some level of user input. And Ill keep banging the equipment drum as it makes a world of difference.

@edwardsean did I see you posting on the uptone forum recently? :]

 

[edwardsean]

Theyre all made with an A8. And yes personalized measurements will surely provide the best quality. But some basic hrtf correction can be made that will help the preset sound more natural / personalized.



Additionally these various programs and devices give us multiple options for different scenarios. The Realisers can be used without a pc, Impulcifer allows for all sorts of room correction, etc. And all of them require some level of user input. And Ill keep banging the equipment drum as it makes a world of difference.

@edwardsean did I see you posting on the uptone forum recently? :]

Hi, yeah, I recently got a DAVE and that sent me on a deep dive of getting it the best signal. This has meant assembling a whole chain from server to reclocker and master clock, and, sigh, multiple high quality power supplies. So, Uptone has been in the mix.

I'm still devoted to OOYH. As my chain has gotten better and better, OOYH has scaled right along.

If you recall I was upscaling from 48KHz to DSD512–after OOYH processing. Now with Dave, I've switched to PCM and I'm upscaling from 48KHz to PCM768KHz. We discussed how we clearly heard an improvement to OOYH, but I saw a post calling for an end to such "ridiculous" assertions. I understand so much better now why this works.

Back then I didn't know the difference between upsampling and upscaling. I thought they referred to the same thing. I've come to realize that upscaling, as done by MScaler and HQP, are not just multiplying the sample rate. It's not just about increasing the frequency but how you're filling all that additional data.They are using extremely computationally heavy algorithms to "tap" the base (44.1/48KHz) digital waveform up to a million+ times. They then take this information and–intelligently–interpolate missing information to reconstruct the original analog waveform.

When this process is done post OOYH, it is not only attempting to reconstruct the original 2-channel data but OOYH's convolution of that source data. This is to say, it is not only digitally restoring the sounds and reverberations of the studio master but the sounds and reverberations of the system and room that OOYH captured. So, now I've added signal regeneration and reclocking downstream of the upscaling and each stage not only improves the audio data but OOYH's effect. The degradation of digital reproduction and digital convolution are being corrected simultaneously. Everything sounds more real, both the music and the effect of listening to it on speakers, while using headphones.

So, the sonic artifacts that made me feel like OOYH is off from actual speaker listening have steadily vanished. There is still the physics problem as no waves are actuating pressure in the room, but aside from that, OOYH–upsampled, upscaled, regenerated, and reclocked–is spectacularly convincing.

This again underscores the point that sometimes you know better to hear better, and other times you hear better to know better.

What's been new with you?