I just like the character with a long hat sitting on a block next to Joe and Jason. Great to see DMS stepping in front of the camera in these discussions. Great discussion as usual or as US people say “as always”.
Loving this subject!! There are so many variables that change the way sound waves reach your ear, from the instrument, the processing, and mixing, the micing, the amp, dac, source, etc. I never like to strive for a specific sound, I just go by how I feel when I hear things haha.
This was a fun one! I like this content so much and have learned quite a bit watching the abyss channel, DMS, and especially from Resolve/Andrew from the headphone show. That guy goes hard with the measurements and even provides his subs with eq profiles specifically for cans like the LCDX ect... I'd still love to see the 1266 Phi TC measured for frequency response and plotted against the latest harman target curve... I've read that the suspension system of the TC makes it difficult to get a proper measurement/seal with the typical measurement rigs. Ideally one day we would have the ability to measure resolution/detail as well as the punch/slam/dynamics of bass which doesn't show up in the frequency response. Most people confuse the bass measurement level in a curve with punch and slam.
Good discussion. Personally I like to use Harmon target as my EQ basis, then adjust as needed. The nice thing is that with the AutoEQ project they compensate to match that curve, and you can compare different headphones (mainly) on some technical performance and can help out a lot of headphones just to at least reach that "neutral" level. But even when you can EQ to it, the physical characteristics of the drivers/headphones make them sound vastly different. For me, the Focal target is pretty closed off and muffled sounding. If you EQ to harmon and (especially) compensate for their weird 3-6K dip, it becomes so much more enjoyable to listen to and clean where you can hear the technical performance and ignore the usual tonal balance that separates a lot of products. Its always a lot of experimentation though when it comes to preference, and frequency response is only a small part of the picture.
Thank you for this video. I couldn’t agree more. The sound of a headphone should be what sounds best for you and not an arbitrary graph or review. I remember when I bought my 6xx headphones and how excited I was to be in the ‘audiophile’ club. But when they arrived and I started listening to them how disappointed I was with the sound. The seemed flat and muddy to me and where not enjoyable. I thought there must be a problem with mine and literally put them back in the box and did not touch them for a year or more. Then one day I was talking to a friend who had more experience with headphones and related gear and he told me the, first it was ok for me to not like them ‘out of the box’ and it did not mean there was a problem with me. It was just that the way I heard the music through them was the way I liked my music. No judgment and no reproach. Just my preference. He suggested I take a few of my favorite tracks and EQ them until they sounded good to me. So I did. And it has made all the difference and I now love them. Also I learned that it’s OK to adjust a headphone for subjective listening and not always going for a curve that others liked (not that there is a problem with other people liking a different sound) and just enjoy my music how I like it to sound.
These are interesting videos your company makes. Douglas, your guest in the black shirt, really knows what he's talking about. You should feature him in every video.
I agree that flat is a difficult term to define exactly. Suppose instead that we go with the goal of making our headphone present to us exactly what the mastering engineer intended. I have an idea about how to do this: 1. First take a bunch of measurements to obtain your own HRTF curve to a reasonable degree of accuracy. 2. Use a spectrum analyzer to obtain an averaged frequency plot from a given song's source file 3. Apply the inverted HRTF curve to the song using an equalizer 4. Play this through the headphones and record this output with a transducer 5. Apply the spectrum analyzer to this recording to obtain its averaged frequency response plot 6. Subtract this plot from the source file response plot to obtain an error curve 7. The inverted error curve is the EQ correction which should give you something that approximates what the mastering engineer intended
This podcast sort of reassures my habit of tweaking with my EQ for different headphones. I was always berated by my brother whenever I start teaking the EQ, saying I'm destorying the music by teaking the settings and I'm a damn amature . But some equipments are just don't sound enjoyable without a slight tweak, and it sort of puzzles me why can't I tweak my equipment to let myself like the set of headphones a lot more? Well, at least I learned today that there's nothing wrong with what I'm doing and I shouldn't feel ashamed of it either.
There's certainly nothing wrong with tweaking eq on your equipment. The issues in some cases becomes that you might end up on a habit of making every headphone sound the same as you tweak them to the same sound signature you are used to (some headphones are really good for certain scenarios compared to other). This becomes a problem if you start doing doing EQ soon after you got a headphone because if you allow yourself to listen to the new headphone for a while you might just find something you didn't even know you were missing, the problem is just that you end up with people who become fanatics against EQ use, that becomes wrong as well.
The Harman target is trying to solve this "What is flat" question. They're just trying to make every studio use this target, so if someone wants to hear what flat sounds like, you just point to one single headphone. But I also think that S. Olive's methodology is not quite right. They just EQ a single headphone to different kinds of headphones' response, which is quite limiting I think. Their purpose is good, but the methodology is not great.
Very interesting! My preferred FR will vary with the kind of music I listen to as well. So the headphone's job is to be able to withstand a little bit of tuning without sounding "off". Imho.
I feel like "flat" and "preference" are two very different notions that are talked about using the same word, "flat". And that is confusing. Mesuring flat compared to a preference curve doesn't mean being flat, because that is so subjective. I think that the term "flat" should be kept for a technical/theoratical context where it makes sens to be used. If not avoided at all.
dBHL or dBSPL thats the question. I think that HL really needs more implementation And i dont think musicians master for flatness but for versatility. Mix is for how you wanna sound, mastering is to sound good. mixing is the art and mastering the science.
Assuming gear and source and EQ is all perfect, throw in humans and there will be guaranteed differences. How our brain perceives sound and what life experiences (eg accidents) change how we hear, or how we are feeling at any given moment when we hear also changes how we process sound. Not including externals like trained/experienced listeners, personal beliefs/preferences, blah blah blah.
If I was Making music I would make 2 versions of the same album making music 1 capture the whole room No compression and if I can I will increase the Dynamics to max The 1st 1 would be for audio files Called the audio file version second version Called the consumer version works Sounds compressed very consumer like In stuff that consumers a like I would be using the abyss AB1266 Phi TC For mixing mastering and my pure enjoy my headphone The brings my excitement in my life
Also factor in that with speakers, measurements are taken at 1 meter from the baffle. Ok so in that case the listener would have to be listening to the speakers from 1 meter to get that response in their ears... then factor in our ears not getting frequencies the same way as mics do... thats a step away from realism at least as to what the engineer wanted you to hear. Then add that the instruments were not recorded 1m away from the mic for us to have to listen 1m away like what binaural is conceptually doing which is hearing the event from where it was recorded etc etc
Speakers and mics have to be flat (at least in lab conditions) to reproduce sound in the most uncolored way possible. The rest of the "processing" is done by your ears in much of the same way it would happen in a live venue. Speakers in good rooms don't deviate much from flat except for a slight roll-off in the highs. Really it's not as complicated as the video makes it seem. It's more complicated when you want to carry that knowledge to tuning headphone drivers. The interaction there is more complex than what you get with speakers. In fact for speakers, the data shows that in blind tests people overwhelmingly prefer "flat" speakers as opposed to colored wonky sounding ones. Some similar research has been done by Harman for headphones so that gives you an idea why headphone targets aren't just pulled out of thin air. Abyss doesn't even have a target because they haven't published one and they've admitted multiple times that they do not use measurements when developing their headphones. So they're essentially working completely in the dark.
BTW, the world's "best headphone", the HE-1 is very closely tuned to the Harman target (which seems to indicate that Sennheiser reached a conclusion similar to Harman in their research) and almost nobody would say it's a boring headphone.
what is flat? well not dms or whatever his camera is on, why is dms bouncing up and down all through this? is the camera on somthing that is vibrating? also while im at it, not a fan of these 2:1 ratio vids, same on dms channel, i guess you think its cool, but its not. other then that keep up the great work !
Read the Harman research and make an actual attempt to explain it before you criticize it lol. What's more likely to be closer to good sound: a "generic" target that is based on most people's preferences or whatever random boutique company thinks good sound is? Well, one of those probably involved millions of dollars in research while the other is pure conjecture. For people wearing the engineer badge you sure don't think like one.
The problem is that the number of subjects in Harman research is only 268 people in Canada including 12 trained listeners. 268 is enough to make statitical inference but not enough to draw a decisive conclusion, especially when studying public's preference. I think Olive underestimates the effect of change of culture in his paper. In Asia, especially in China, Harman Target is mostly considered bass heavy and dark treble because Chinese people never listen to the test music Olive used in his research and the music listened by Chinese people includes lots of their traditional Chinese instruments where a bit more treble will bring more enjoyment and heavy bass seldom exists in those instruments. So if he draw 268 people from China and use different set of music, the Harman Target may vary a lot compared to the 2018 Target. In fact, Harman Target is not meant to be a perfect 'neutral' or 'flat' target, the research is to find the best Target response that preferred by the general market Harman competed in and they need to compete with consumer headphone companies, that's why the bass is boosted from the equivalence of in-room response of flat speakers in Harman reference room, the latter which I believe is much more close to the actual 'flat‘. Though I do think a bass boost is needed for headphone response due to the essential acoustic differences from speaker listening. Also Harman Target explains very little about the Spatial effect of headphones, which is also very important in headphone listening. So it is better to take Harman as a loose standard rather than a strict Target.
@@tomtang8708 They don't claim it's a strict target. At the very least a headphone's tuning should acknowledge the existence of an ear gain peak. They've actually tested it in multiple countries so far with similar results from country to country.
@@DEALUX First al all, any of free field, diffuse field or Harman has gain peak and even quite similar patterns. The key thing is the fine-tuning of the curve and bass/treble preference. Second, Please don’t exaggerate their findings. In terms of country of listeners, the number of subjects are even smaller and not conclusive at all. I have the report and paper by my side, and here is direct quoting.’’72 Americans, 72 German, 20 Chinese, 48 Canadians and 37 various, (a bass/treble preference graph is presented);more research is needed in terms of country of listener as a factor of preferred bass and treble levels.’’
@@tomtang8708 It's more research than any random headphone company has ever done. And the target is still based on objective measurements tweaked to sound more pleasing (i.e. more bassy) to the average listener. I mean it's no surprise that people love the HD600 and 650 and those follow the target pretty closely in the mids and highs. Even though our anatomy is different our preferences still tend to converge in meaningful ways, which is what the Harman research tried to show. It makes sense. If individual HRTFs were hugely different there would be no such a thing as hi-res or high quality recordings because the actual quality of the mix would depend on the subjective hearing of the mixing/mastering engineer so much that what he finds natural and balanced would sound odd to the average listener.
Cool insight and appearance from DMS. Thanks
DMS brings a lot to the table, so to speak. One of the best videos from Abyss yet.
I just like the character with a long hat sitting on a block next to Joe and Jason. Great to see DMS stepping in front of the camera in these discussions. Great discussion as usual or as US people say “as always”.
This has been a fascinating conversation to listen to, I enjoyed it a lot👍
Thank you, guys!
yes its always a pleasure to listen when wise people are talking
Love you guys. You guys are the best. So much info.
Loving this subject!! There are so many variables that change the way sound waves reach your ear, from the instrument, the processing, and mixing, the micing, the amp, dac, source, etc. I never like to strive for a specific sound, I just go by how I feel when I hear things haha.
That purple diana is just.. sexy.
This was a fun one! I like this content so much and have learned quite a bit watching the abyss channel, DMS, and especially from Resolve/Andrew from the headphone show. That guy goes hard with the measurements and even provides his subs with eq profiles specifically for cans like the LCDX ect... I'd still love to see the 1266 Phi TC measured for frequency response and plotted against the latest harman target curve... I've read that the suspension system of the TC makes it difficult to get a proper measurement/seal with the typical measurement rigs. Ideally one day we would have the ability to measure resolution/detail as well as the punch/slam/dynamics of bass which doesn't show up in the frequency response. Most people confuse the bass measurement level in a curve with punch and slam.
Good discussion. Personally I like to use Harmon target as my EQ basis, then adjust as needed. The nice thing is that with the AutoEQ project they compensate to match that curve, and you can compare different headphones (mainly) on some technical performance and can help out a lot of headphones just to at least reach that "neutral" level. But even when you can EQ to it, the physical characteristics of the drivers/headphones make them sound vastly different.
For me, the Focal target is pretty closed off and muffled sounding. If you EQ to harmon and (especially) compensate for their weird 3-6K dip, it becomes so much more enjoyable to listen to and clean where you can hear the technical performance and ignore the usual tonal balance that separates a lot of products. Its always a lot of experimentation though when it comes to preference, and frequency response is only a small part of the picture.
Thank you for this video.
I couldn’t agree more. The sound of a headphone should be what sounds best for you and not an arbitrary graph or review.
I remember when I bought my 6xx headphones and how excited I was to be in the ‘audiophile’ club.
But when they arrived and I started listening to them how disappointed I was with the sound. The seemed flat and muddy to me and where not enjoyable.
I thought there must be a problem with mine and literally put them back in the box and did not touch them for a year or more.
Then one day I was talking to a friend who had more experience with headphones and related gear and he told me the, first it was ok for me to not like them ‘out of the box’ and it did not mean there was a problem with me. It was just that the way I heard the music through them was the way I liked my music. No judgment and no reproach. Just my preference.
He suggested I take a few of my favorite tracks and EQ them until they sounded good to me. So I did. And it has made all the difference and I now love them.
Also I learned that it’s OK to adjust a headphone for subjective listening and not always going for a curve that others liked (not that there is a problem with other people liking a different sound) and just enjoy my music how I like it to sound.
These are interesting videos your company makes. Douglas, your guest in the black shirt, really knows what he's talking about. You should feature him in every video.
“There’s certainly a community that wants some really weird stuff.” I raised my eyebrows at that, but yeah, there’s no accounting to taste.
I agree that flat is a difficult term to define exactly. Suppose instead that we go with the goal of making our headphone present to us exactly what the mastering engineer intended. I have an idea about how to do this:
1. First take a bunch of measurements to obtain your own HRTF curve to a reasonable degree of accuracy.
2. Use a spectrum analyzer to obtain an averaged frequency plot from a given song's source file
3. Apply the inverted HRTF curve to the song using an equalizer
4. Play this through the headphones and record this output with a transducer
5. Apply the spectrum analyzer to this recording to obtain its averaged frequency response plot
6. Subtract this plot from the source file response plot to obtain an error curve
7. The inverted error curve is the EQ correction which should give you something that approximates what the mastering engineer intended
Great thoughts !
DMS seems to be sitting on a very bumpy chair 2:20
This podcast sort of reassures my habit of tweaking with my EQ for different headphones.
I was always berated by my brother whenever I start teaking the EQ, saying I'm destorying the music by teaking the settings and I'm a damn amature .
But some equipments are just don't sound enjoyable without a slight tweak, and it sort of puzzles me why can't I tweak my equipment to let myself like the set of headphones a lot more?
Well, at least I learned today that there's nothing wrong with what I'm doing and I shouldn't feel ashamed of it either.
There's certainly nothing wrong with tweaking eq on your equipment. The issues in some cases becomes that you might end up on a habit of making every headphone sound the same as you tweak them to the same sound signature you are used to (some headphones are really good for certain scenarios compared to other). This becomes a problem if you start doing doing EQ soon after you got a headphone because if you allow yourself to listen to the new headphone for a while you might just find something you didn't even know you were missing, the problem is just that you end up with people who become fanatics against EQ use, that becomes wrong as well.
The Harman target is trying to solve this "What is flat" question. They're just trying to make every studio use this target, so if someone wants to hear what flat sounds like, you just point to one single headphone.
But I also think that S. Olive's methodology is not quite right. They just EQ a single headphone to different kinds of headphones' response, which is quite limiting I think.
Their purpose is good, but the methodology is not great.
ugh never ask a flat earther everything is flat with those guys
Very interesting! My preferred FR will vary with the kind of music I listen to as well. So the headphone's job is to be able to withstand a little bit of tuning without sounding "off". Imho.
the earth is FLAT. lol...
man, so cool magenta !
Here for DMS
I feel like "flat" and "preference" are two very different notions that are talked about using the same word, "flat". And that is confusing.
Mesuring flat compared to a preference curve doesn't mean being flat, because that is so subjective.
I think that the term "flat" should be kept for a technical/theoratical context where it makes sens to be used. If not avoided at all.
I just wanna hear the creators intent
dBHL or dBSPL thats the question. I think that HL really needs more implementation
And i dont think musicians master for flatness but for versatility. Mix is for how you wanna sound, mastering is to sound good.
mixing is the art and mastering the science.
I agree that Harman sounds really well on 6xx series. And I think it's the biggest... bias probably is the right word in the research.
great discussion
klark teknik dn370 is the device I use for EQ'ing my headphones.... I have my turntable and DAC fed into it.
Assuming gear and source and EQ is all perfect, throw in humans and there will be guaranteed differences. How our brain perceives sound and what life experiences (eg accidents) change how we hear, or how we are feeling at any given moment when we hear also changes how we process sound. Not including externals like trained/experienced listeners, personal beliefs/preferences, blah blah blah.
If I was Making music I would make 2 versions of the same album making music 1 capture the whole room No compression and if I can I will increase the Dynamics to max The 1st 1 would be for audio files Called the audio file version second version Called the consumer version works Sounds compressed very consumer like In stuff that consumers a like I would be using the abyss AB1266 Phi TC For mixing mastering and my pure enjoy my headphone The brings my excitement in my life
Also factor in that with speakers, measurements are taken at 1 meter from the baffle. Ok so in that case the listener would have to be listening to the speakers from 1 meter to get that response in their ears... then factor in our ears not getting frequencies the same way as mics do... thats a step away from realism at least as to what the engineer wanted you to hear. Then add that the instruments were not recorded 1m away from the mic for us to have to listen 1m away like what binaural is conceptually doing which is hearing the event from where it was recorded etc etc
you guys said most of it. I really love this video.
Speakers and mics have to be flat (at least in lab conditions) to reproduce sound in the most uncolored way possible. The rest of the "processing" is done by your ears in much of the same way it would happen in a live venue. Speakers in good rooms don't deviate much from flat except for a slight roll-off in the highs. Really it's not as complicated as the video makes it seem. It's more complicated when you want to carry that knowledge to tuning headphone drivers. The interaction there is more complex than what you get with speakers. In fact for speakers, the data shows that in blind tests people overwhelmingly prefer "flat" speakers as opposed to colored wonky sounding ones. Some similar research has been done by Harman for headphones so that gives you an idea why headphone targets aren't just pulled out of thin air.
Abyss doesn't even have a target because they haven't published one and they've admitted multiple times that they do not use measurements when developing their headphones. So they're essentially working completely in the dark.
@@DEALUX Your tone shows a definite bias.
Great video.
DMS is bouncing like his really nervous it kind of make me nervous 😅
What is Abyss flat ?
One word. Cocaine. If you know you know.
why is DMS fidgeting around so much? It's making me nervous watching him.
Extra coffee kind of day
-D
BTW, the world's "best headphone", the HE-1 is very closely tuned to the Harman target (which seems to indicate that Sennheiser reached a conclusion similar to Harman in their research) and almost nobody would say it's a boring headphone.
The 1266 costs 5k and it has better detail what a bargain
Depends on your definition of detail. People still overwhelmingly praise the HE-1 despite being stupidly overpriced. And $5K is still overpriced.
@@DEALUX but the he1 is a lot more overpriced
@@shadouxg7150 Why ?
@@jsharp9735 the 1266 tc is more detailed, and it's a Plannar so u can EQ to whatever target u want, the he 1 is way way more expensive
Earth is flat.. 😂 😂 😂
your GF.
what is flat? well not dms or whatever his camera is on, why is dms bouncing up and down all through this? is the camera on somthing that is vibrating? also while im at it, not a fan of these 2:1 ratio vids, same on dms channel, i guess you think its cool, but its not. other then that keep up the great work !
Read the Harman research and make an actual attempt to explain it before you criticize it lol. What's more likely to be closer to good sound: a "generic" target that is based on most people's preferences or whatever random boutique company thinks good sound is? Well, one of those probably involved millions of dollars in research while the other is pure conjecture. For people wearing the engineer badge you sure don't think like one.
The problem is that the number of subjects in Harman research is only 268 people in Canada including 12 trained listeners. 268 is enough to make statitical inference but not enough to draw a decisive conclusion, especially when studying public's preference. I think Olive underestimates the effect of change of culture in his paper. In Asia, especially in China, Harman Target is mostly considered bass heavy and dark treble because Chinese people never listen to the test music Olive used in his research and the music listened by Chinese people includes lots of their traditional Chinese instruments where a bit more treble will bring more enjoyment and heavy bass seldom exists in those instruments. So if he draw 268 people from China and use different set of music, the Harman Target may vary a lot compared to the 2018 Target. In fact, Harman Target is not meant to be a perfect 'neutral' or 'flat' target, the research is to find the best Target response that preferred by the general market Harman competed in and they need to compete with consumer headphone companies, that's why the bass is boosted from the equivalence of in-room response of flat speakers in Harman reference room, the latter which I believe is much more close to the actual 'flat‘. Though I do think a bass boost is needed for headphone response due to the essential acoustic differences from speaker listening.
Also Harman Target explains very little about the Spatial effect of headphones, which is also very important in headphone listening. So it is better to take Harman as a loose standard rather than a strict Target.
@@tomtang8708 They don't claim it's a strict target. At the very least a headphone's tuning should acknowledge the existence of an ear gain peak.
They've actually tested it in multiple countries so far with similar results from country to country.
@@DEALUX
First al all, any of free field, diffuse field or Harman has gain peak and even quite similar patterns. The key thing is the fine-tuning of the curve and bass/treble preference.
Second, Please don’t exaggerate their findings. In terms of country of listeners, the number of subjects are even smaller and not conclusive at all. I have the report and paper by my side, and here is direct quoting.’’72 Americans, 72 German, 20 Chinese, 48 Canadians and 37 various, (a bass/treble preference graph is presented);more research is needed in terms of country of listener as a factor of preferred bass and treble levels.’’
@@tomtang8708 It's more research than any random headphone company has ever done. And the target is still based on objective measurements tweaked to sound more pleasing (i.e. more bassy) to the average listener. I mean it's no surprise that people love the HD600 and 650 and those follow the target pretty closely in the mids and highs. Even though our anatomy is different our preferences still tend to converge in meaningful ways, which is what the Harman research tried to show.
It makes sense. If individual HRTFs were hugely different there would be no such a thing as hi-res or high quality recordings because the actual quality of the mix would depend on the subjective hearing of the mixing/mastering engineer so much that what he finds natural and balanced would sound odd to the average listener.