|
Post by Aristrotle on Jan 27, 2019 15:33:48 GMT
So I was exploring headphone measurements on DIY-AUDIO-HEAVEN and RTINGS. I payed most of my attention to DT1990's because I just got my own one today. then I noticed there is a prominent difference in the treble range: RTINGS' measurement shows two treble peak, one at 7khz and the other at 9khz side by side, but DIY-AUDIO-HEAVEN's measurement shows only one treble peak at somewhere lower than 8khz. To figure out which one is true, I listened to swept-frequency signals a few times. Then I hear a peak around 7.2khz and another peak above 10khz but not too far (the tone generator I used only generates signals lower than 10khz).
It turns out that DIY-AUDIO-HEAVEN's measurement sounds more true to me, but also I found RTINGS' explaination on how they measure headphones very trustful. Their K702 measurement coincide with my personal listening experience.
What's going on there?
------------------------------------------
When I was reading HD800s measurement the other day, another conflict between RTINGS and DIY-AUDIO-HEAVEN was found:
"The 10kHz peak even sticks out 10dB above the rest...When this level is reduced by EQ one must be careful not to reduce the 10kHz part too much. That part is responsible for the fine detail reproduction and instrument ‘placement/localization’. When that level is reduced to the same level as the mids the headphone sounds ‘bland’ and lifeless.", says the review. But RTINGS' plot shows a dip more than a peak there. (btw I tried to add an artificial 10khz peak by using software equalizer to my k702, its separation did improved a bit, along with more sharpness and detail, pretty interesting.)
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Jan 27, 2019 18:24:30 GMT
I recently had this discussion with a HATS measuring advocate.
My response was that I don't believe HATS are suitable for headphone measurements at all. It all seems logical to use Pinna and fake ear canal and then only correct for the ear canal afterwards. The reasoning is that this is closer to real life situation. That part is certainly true and if I had paid many thousands for a HATS + analyzer etc I would expect this to work well.
Trouble is... every headphone reacts differently on a HATS as shown by the 'soundstage' plot. It would seem like there is no single correction for headphones. Also why introduce a bandpass filter (ear canal) as well as 'floating' bandpass filters (the Concha/Pinna) and afterwards try to correct for this.
On top of that I have this other idea.
Consider someone measures a speaker in anechoic conditions at 1m or 3m on axis. This measures perfectly 'flat'. Then the soundwaves reaching the side of our head are still 'flat' and is similar to the speakers response. This is what we hear. Now the Pinna comes into play and changes the FR and at the ear it measures awful. The brain has calibrated itself to hear sounds as they are.
The same applies for headphones. The headphone speaker emits sound. Only a short travel later it arrives at the Pinna. It is changed by the Pinna (different then with speakers as the angle differs tremendously) and again is a mess at the eardrum.
Of course we actually 'hear' the speaker as flat and so does the driver of the headphone. Yes, the Pinna changes things when we measure headphones but this also happens when we would measure a speaker with a Pinna attached.
My experience is that when one EQ's headphones acc. to plots made with a HATS and get them to measure 'flat' on a HATS and we do this for 2 different headphones they should SOUND very similar as well as they measure the same and the HATS mimics our head. Well.. when you EQ them to those plots they still sound quite different. When I do the same with my measurements they still sound somewhat different but the tonal balance is much closer.
The conclusion I draw from this is that to arrive at a good EQ based on measurements the flat plate rig (such as mine) gives 'better' results compared to a HATS. Still there are some differences because the Pinna does 'something' of course.
So.. as headphone measurements are still in a kinf of infancy and every owner of a rig does have their own 'compensation' based on how they hear it and all HATS show different results with the same headphones it may just be a matter of finding which plots 'correllate' the most with your hearing and then use that.
|
|
|
Post by Aristrotle on Jan 28, 2019 5:55:32 GMT
Wow...Thanks. You explained a lot. So I checked RTINGS' raw frequency response measurement again after I woke up in the morning and see your reply, realized it could be a consequence of artificial pinna gain which doesn't exist on my ears. And I saw another interesting thing I forgot to mention last night. Here is the raw frequency response of the left driver they tested: The sub bass extention of this individual sample is beyond good, the curve slightly turned upward as frequency goes down. Yet I have only seen such curve on 64 Audio tia Fourte, a multi driver IEM, but DT1990 is a single driver (semi-)open back headphone! Also it's the only sample among many DT1990 measurements I've ever seen. It can't be some random testing error since the FR below 2khz was averaged from five tests on five real humans. I think manufacturers should really manage to find out what caused that and produce more of it.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Jan 28, 2019 10:24:24 GMT
I think most manufacturers know very well what they are doing and tune their headphones while keeping manufacturing costs as low as possible. Its a trade off.
Pads, baffle and driver design and above all seal is the root cause of bass roll-off Most open planar headphones all measure absolutely flat to below 10Hz.
|
|
|
Post by Aristrotle on Jan 28, 2019 11:19:57 GMT
Yeah. Planar headphones are famous for flat bass response. But in this specific sample the bass response are more than flat, it's tilting up! Which means it has more sub bass than high bass. Although it's subtle and made by chance. This is one of the reasons why I decided to buy DT1990 —— best bass extention among all open back dynamic headphones. 64 Audio tia Fourte has a more significant tilting and made by design. Here's FR of tia Fourte I just seeked out:
It's fuzzy because it was shot from a video record.
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Mar 15, 2019 20:49:06 GMT
I find the raised treble of the dt1990 really beneficial if I’m honest, whether it’s fake detail or just simple raised treble. My ears don’t go up anything as high as they used to so having a boost up there helps me quite a lot. The other bonus is that recently, I made a film using a device called a gimbal, which keeps the camera steady. If I hadn’t monitored the sound with the DT1990, I would probably have missed the fact that the gimbal was giving out quite a high ‘squeal’ from its motors. It was low enough to be reproduced on a TV soundbar, but I couldn’t hear it until I put the 1990’s on. Plain as daylight so I had to put in a notch filter to get rid of it. I only found out about the ‘squeal’ because my daughter could hear it planly on TV. The Dt1990 picked it out for me. The nice thing about these headphones is that I can wear them all day because they’re so comfortable and they really do get into the mix so that you can plainly hear any problems. They’re also great for just listening. Kind of like a dt990 on steroids. They’re selling for a more realistic price now as well these days. I stopped using the Sony mdr7506 altogether for monitoring once I heard these.
|
|
|
Post by Aristrotle on Mar 16, 2019 6:37:25 GMT
I do appreciate its fine treble, not the mid-treble peak but even and well extended (and lifted of course) high-treble, which delivers high resolution. Back when I first got my dt1990, I didn't heard what people say "hyper details" compared to my k702 until I surfed to one of my favorite piano pieces, "Morning Love" played by Carlos Cipa. Things had been going normally before I started to hear a series of very high frequency ringing from about 1:00. At first I thought it was some harmonic distortions caused by the headphone drivers as I never heard it before. But the ringing didn't occur at the beginning part which is also the quieter part of the track, even when the player hit the same keys. Then I switched back to k702 and listened carefully. Appeared that the ringing was there, but barely noticeable.
I also tried some Nightwish's songs that I thought k702 wasn't good enough to reproduce. Sadly I still heard a mess, especially muddy bass. Few days after believing me not being able to hear good metal music out of poor amp power, crosstalk or whatever reason, I ran into an Epica's album where I heard rapid, rich, punchy bass, clean, clear vocals and ordered, spacious image, joyful sound in short.
I think dt1990 really has a revealing and unforgiving character. Sometimes it ruins your listening experience, other times it pleases you more than ever before. Whether the recording/mixing is good or bad, dt1990 shows it to you.
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Mar 16, 2019 11:18:46 GMT
Absolutely Aristrotle. They are a bit more like lasers by comparison. Mind you, I listened to an HD800 and that was even more treble laden. I’ve used a DT990 for a long time along with Senn HD650 and various others, but as soon as the dt1990 was on my head, I wanted a pair. Like a massively improved dt990 and a lovely package with the box and pads. The bass is very well controlled which surprised me. It kind of punches nicely without the flabbiness of the dt770. What also surprises me is the frequency response graphs showing such a high peak in the treble and yet, on the head, I don’t feel that they’re sharp at all. Well etched and actually, less sharp than the hd800 I think. Maybe my hearing has dropped away up there a bit, but to me, they sound almost ‘warm’. Improved mids compared to the dt990 and a kind of ‘dryness’ to the sound with a clarity and speed of bass that is unusual for a dynamic headphone. The HD800 I think, is more ‘spaced out’ but for me, the bass weight and clarity of the Beyer won me over. It goes surprisingly low and has great impact. You wouldn’t think that they go that low at first either. The sound is kind of unassuming. For the difference in price, I was happier with the Beyers as well if I’m honest. I’d been toying around for a long time over the hd800 but when I finally got to hear one, I felt that the sound signature just didn’t suit my tastes at all. To me, it was a bit too edgy and thin, although the bass was good. The other thing that I find is that the dt1990 doesn’t seem touchy about listening volume. It stays well focused at low volume and just grows with volume. With many headphones, I’ve always found that they snap into focus at certain volumes (different for all headphones) but these seem particularly well focused at both ends of the volume scale. Given that the dt1990 is now retailing at around £399, I felt much happier about it; especially the build quality as well. Now they don’t come off my head. I like their sound a lot. For editing movies I mainly used Sony mdr7506’s because they give a ton of top end which helps to identify mic hiss. Plus a raised mid bass which points out lower end unwanted rumbles and hums with ease, but the Beyers, as you say, really do bring them out. It is indeed a kind of blessing and a curse at the same time. With bad recordings, you will definitely know it on these!! Beyer did a really good job with these. I should have done a review on them when I got them really because they are terrific headphones. Until I heard these, I stuck to my guns regarding an upper price limit for a headphone. It was £300 but these have caused me to review it now to £400.
|
|
|
Post by Aristrotle on Mar 18, 2019 13:47:46 GMT
The other thing that I find is that the dt1990 doesn’t seem touchy about listening volume. It stays well focused at low volume and just grows with volume. With many headphones, I’ve always found that they snap into focus at certain volumes (different for all headphones) but these seem particularly well focused at both ends of the volume scale. It might have something to do with " Equal-loudness contour", which indicates how much sound pressure level it actually needs to achieve the same loudness to human ears at different frequencies. "Phon" is the unit of the loudness we subjectively feel, while "dB SPL" is measured objectively. As you can see, the scales in the bass range are denser than scales in mid range, which means the loudness of bass increases faster than mid does as volume goes up.
Solderdude had remade some figures for us to see the differences of these contours easier (further explained here).
The chart on the right suggests that if a recording sounds flat at 80 phon, you need to decrease bass to make it sounds flat at higher volume. And conversely, you could get a tonal balance of higher loudness at lower volume with fine-tuned bass boost, which is what dt1990 does.
BTW, what does it mean by "space out"?
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Mar 18, 2019 21:35:12 GMT
‘Spaced out’ = sound stage. The sounds are more separated than in the dt1990, although the imaging on the dt1990 isn’t bad at alleither.
I made a binaural recording of a walk that I did last week for a video I was making and on the DT1990, the placing of sounds was very good. On an hd800s it seemed almost wider than it really was, but extremely ‘airy’
I also have a Fostex TH900 but to be honest, I think that I prefer the sound of the DT1990 to that one as well.
|
|
|
Post by Aristrotle on Mar 19, 2019 12:20:45 GMT
Oh, I see.. Asked because some dictionary told me it stands for "getting high on drugs". XD
Speaking of soundstage, I remember a recording engineer was once saying like even K240s perform better than HD800 for classical recording mastering use, because everying on HD800 got laid back due to its large but irregular soundstage, therefore you would mess up positioning. (I don't know if it's true cuz I've never worked in a studio. And I personally think you should use loudspeakers in that case.)
DT1990 could have more air with A pads on. Maybe "A" stands for "airy" and "B" stands for "bassy". XD*2 To me, chamber music and solo instrument sound most attractive on A pads, as for noisier musics I always eager to swap to B pads after an hour or two. Although theoretically those pads only affect bass and airness is more about treble, I did find that airness could get lost while boosting bass with software EQ and no physical change being applied to headphones. Haven't found a basis yet. What's worse, I've read some texts against what I said yesterday today: In the early 1970s hi-fi manufacturers provided a “loudness” button which put in a bass and treble boost in order to flatten the Fletcher– Munson curves, and so provide a simple compensation for the reduction in hearing sensitivity at low levels. The action of this control was wrong in two important respects:
- Firstly, it directly used the equal loudness contours to perform the compensation, rather than the difference between the curves at two different absolute sound pressure levels, which would be more accurate. The latter approach has been used in professional products to allow nightclubs to achieve the equivalent effect of a louder replay level.
- Secondly, the curves are a measure of the equal loudness for sine waves at a similar level. Real music on the other hand consists of many different frequencies at many different amplitudes and does not directly follow these curves as its level changes. We shall see later how we can analyze the loudness of complex sounds. In fact because the response of the ear is dependent on both absolute sound pressure level and frequency it cannot be compensated for simply by using treble and bass boost.[1]
1. David Howard, Jamie Angus-Acoustics and Psychoacoustics, Fourth Edition-Focal Press (2009), Page 94.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 19, 2019 19:28:28 GMT
A= Analytic, B= Balanced
|
|
|
Post by Aristrotle on Mar 20, 2019 0:33:36 GMT
I knew, my friend. I knew it from the instruction manual. I was just trying to make a joke of them. Looks like it isn't ridiculous enough to make people think it's a joke but rather confusing. I'll try harder next time What I concerned is, there are quite some reviews saying A pads are more neutral and balanced while B pads boost bass. Also some hifi enthusiasts said A pads got the traditional beyer sound that they'd admired. But for me, while A pads do wow me with airy and fantastic treble, I couldn't live with them too long. I could really use a bit of boosted bass to balance boosted treble. Glad we have choices here.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 20, 2019 12:29:25 GMT
I didn't know it came with 2 different pads when I measured it. My plots were made with the B pads (fitted stock) as were Rtings measurements.
If I'd known I would have measured both pads. Too bad Rtings didn't measure both pads, would heave been nice to see the difference. They mentioned the amount of bass is lower in the A pads (and thus the treble stands out even more).
Good to have choices indeed. Excellent headphones but the treble peak bothered me the short time I listened to it. For monitoring it may be welcome, maybe even for mastering as one dials back that lifted part of the FR a bit. This is the area where most headphones have peaks anyway.
|
|
|
Post by Aristrotle on Mar 20, 2019 15:43:07 GMT
They did actually. But only presented in a video shortly in which they compared dt1990 with hifiman sundara (at around 6:24). You can find the video link at the head of their reviewing page of dt1990.
I don't know why they didn't post it with measurements from B pads as well. Maybe they thought the sounds with A pads are not ideal. Or maybe their website wasn't ready for one headphone with two sets of sound.
|
|