RAAL-requisite SR1a Ribbon Headphone Review

[JohnYang1997]

(I think I followed that whole thread, and also posted a good deal in it as it was developing and I certainly don't remember that being concluded, so it's not particularly reasonable to ask him to read that whole thread to support your point.)

That's what I learned after anticipating that thread. There wasn't a clear conclusion and it's a long thread. But we basically ended up agreeing that buying 5128c isn't the wise choice. And GRAS would be the better solution. And here's what happened.
Oh and that was my conclusion over that thread. I didn't mean Amir made that conclusion.

 

[DuxServit]

This reminds me of the 1977 Stax SR-Sigma.

 

[LTig]

Does the RAAL really consume 100 W or more or does it just require higher voltage levels?
@amirm: could you measure the impedance over frequency, please?

 

[GDK]

Looks awfully heavy and bulky.

I would have thought that issue would be addressed by the not at all asinine decision to use carbon fibre for the frame. Plus, they go really fast.

 

[DualTriode]

@amirm,

I want to see test gear porn. Please show us photos of the hammer and power supplies.

Whisper in our ear, tell us if the condenser microphones are pre-polarized or externally polarized?

Thanks DT

 

[Robbo99999]

That's what I learned after anticipating that thread. There wasn't a clear conclusion and it's a long thread. But we basically ended up agreeing that buying 5128c isn't the wise choice. And GRAS would be the better solution. And here's what happened.
Oh and that was my conclusion over that thread. I didn't mean Amir made that conclusion.

Ah, there lies the difference.

 

[RayDunzl]

20hz to 1k takes more than half the graph i think it could be tweaked a bit?

Consider a piano keyboard...

Do you think it is laid out incorrectly for the way it sounds?

The keys are laid out linearly, but the frequencies produced rise exponetially.

Each higher note on the piano is (roughly) the 12th root of 2 higher in frequency.

Note frequency x 1.05946309436 = next higher note in a 12 note octave (western music standard).

If you lay a keyboard next to a log graph, and compare frequencies:

1k to 4k is all scrunched together compared to 100 to 400, but the ear doesn't perceive those notes that way. Each just covers a two octave range.

 

[solderdude]

Does the RAAL really consume 100 W or more or does it just require higher voltage levels?
@amirm: could you measure the impedance over frequency, please?

Most of the power is dissipated in the 'interface box' which basically is a huge resistor + capacitor network = about 7 Ohm.

The headphone itself is 0.02 Ohm (yes basically a short) and does not require a lot of voltage.
The headphone cable adds more resistance and dissipates more power than the headphone itself.
The efficiency is 85db/mW so 122 dB/V including the cable. It thus needs less than 1V (5A) = 5W to play very loud and the largest portion of this is dissipated in the cable. (never mind any damping factor)

Jason (Schiit) liked the SR1A so much that he constructed an amp that can drive the 0.2 Ohm load directly so bypassing the resistor/EQ box.

 

[Vini darko]

Most of the power is dissipated in the 'interface box' which basically is a huge resistor + capacitor network = about 7 Ohm.

The headphone itself is 0.02 Ohm (yes basically a short) and does not require a lot of voltage.
The headphone cable adds more resistance and dissipates more power than the headphone itself.
The efficiency is 85db/mW so 122 dB/V including the cable. It thus needs less than 1V (5A) = 5W to play very loud and the largest portion of this is dissipated in the cable. (never mind any damping factor)

Jason (Schiit) liked the SR1A so much that he constructed an amp that can drive the 0.2 Ohm load directly so bypassing the resistor/EQ box.

Wow full crazy town

 

[amirm]

Care to explain please?
This from Amirm I see as no negative point

Answering this and related questions, I have not yet made a decision which fixture to go with. The evaluation is not complete on the GRAS 45CA until I show the results to you all.

Some things are clear at this point though:

1. B&K 5128 costs more than 2X the GRAS 45CA. This is a huge consideration for me by itself and combined with other factors.

2. Getting a good fit on GRAS 45CA is a lot easier than 5128. In some sense, having a more realistic head on the 5128 limits placement much more than 45CA. And its pinna is stickier making small adjustments harder. I probably could figure it out but it has been a breath with the 45CA so far (relatively speaking).

3. The Harman target curve is available for the GRAS and is definitely correct and predictive of preference. I can EQ with confidence and see the listening results track the target curve (within some margin of error). The target we computed for the BK 5128 does not appear to be as correct especially in higher frequencies. We are at the mercy of Harman creating the 5128 target curve which may never happen as there is not a lot of value in it for them. Without one we will be shooting in the dark.

4. GRAS 45CA is considered authoritative by manufacturers and industry.

5. This is work in progress but it seems that BK 5128 does tease out more information about high frequency response. The 45CA's high resolution pinna dampens this area good bit. Without #3 though, it is hard to interpret what should be there though.

6. Both companies have been very supportive.

7. On Jude's report, he only tested a couple of IEMs with 5128 and has yet to provide any other measurements despite promising such months ago. Nor has he explained why despite his huge investment in GRAS 45CA, it is all yesterday's news for him. He is in the BK advertising for the 5128 which taints what could otherwise be a neutral position.

8. The reliability of measurements for headphones remains fuzzy the higher the frequencies. So no matter which way you go, you can't get ironclad data.

 

[The Jniac]

Wow. $3500 USD plus the cost of a decent speaker amp, and it loses to HE-500 for which I paid $500 CAD. Still, if they could get the bass to be flat to about 30Hz it could reach the dizzying heights of "good". I disagree with using the Harman curve as a target because, as far as I have been able to find, it is based on listener preference rather than true objective neutrality. My personal ideal response is: the response that produces the same response at the eardrum as that produced by a loudspeaker that has a perfectly flat response at the listening position.

 

[amirm]

@amirm: could you measure the impedance over frequency, please?

You mean electrical impedance and phase? If so, how would I interface to it and what difference would it make? You would be feeding a signal to an amplifier which then goes through their impedance matching box. Since you never see the headphone itself, its impedance is immaterial.

Since its impedance as stated is near zero, it would be hard to measure it as it would short out any source I use to drive it directly.

 

[amirm]

How loud do you think is the SR1A ?

I have not bothered to calibrate my measurement gear so can't give you SPL ratings. Subjectively I can tell you that if the HD-650 is 10 on max loudness, the SR1a starts to get distorted in highs around 5. It can get as loud as 7 and then it won't get any louder and/or will clip. Comfortable long term listening would be around 3 to 4 so it is definitely usable. At 7 I can tolerate its level but it is loud.

And I should clarify that at 10 the HD-650 is literally resonating my earlobes. So that is not a usable level either other than for quick testing.

 

[DualTriode]

Wow full crazy town

Okay plot an impedance curve of that. (a box full of resistors)

That looks like not so clever brute force equalization to bring the FR closer to the idealized Harmon transfer function.

It looks to me to be begging the next generation to be active equalization.

Perhaps two generations from now they will be powered ear monitors on your head.

Makes my neck feel a bit achy just looking at it.

Thanks DT

 

[solderdude]

click on the link 'about 7 Ohm' in post #72 and there is the impedance plot of the box.
The box isn't about the Harman curve but to bring down frequencies above 400Hz a bit to get it to be audible 'flat'.
The lack of lower bass is inherent to the open baffle design. It is actually quite impressive they can get the frequency down this low without using pads and only using the baffle.

 

[mathman]

Wow, they look super comfy... oO

 

[Soniclife]

My personal ideal response is: the response that produces the same response at the eardrum as that produced by a loudspeaker that has a perfectly flat response at the listening position.

Do you mean a perfectly flat speaker measured under anechoic conditions, at the listening position, or a speaker that measures perfectly flat at the listening position?

 

[JohnYang1997]

Do you mean a perfectly flat speaker measured under anechoic conditions, at the listening position, or a speaker that measures perfectly flat at the listening position?

He meant what the ear drum hear at the listening position with speakers which have flat anechoic response.

 

[Degru]

An interesting thing you can do with headphones this 'open' is some people keep their subwoofers on while using these. I've tried it and it's okay. I probably wouldn't trust things to be completely flat or easily measurable with a mix of a subwoofer or two + headphones, but it can make it more fun.

I had some Stax L-300 headphones before this which are pretty similar (unfortunately not at the same time to directly compare) and they were my 2nd favorite headphones for mixing/recording (Raal's #1) They both lack bass though. I found I needed to eq the high end more with the L300's for them to sound flat, whereas the Raals were closer from the start. My ears touched the inside of the L300's which was annoying (you can buy thicker pads though)

I had Focal Elears and Hifiman Sundaras/400i's and Monoprice M565c's at the same time as the Raals though when I got them out of curiosity. (have since sold all but Raals)
For 'fun' listening: I liked the Sundara's the best and they had the most bass. If all you do is listen to music for fun - Planars are definitely fine! The Sundara's clamped on my head though so that was annoying. The high end was confusing for mixing (same with the other dynamic/planars)
As far as closed back: M565's I liked more than Oppo PM3's or Audeze sine (sine sounded flatter, but small for ears) Focal Elears/HD800's/AKG Q701's or K701's/etc I liked when I had them but I preferred the bass of Planars.

Are the Stax/Raals worth it ... maybe if you use headphones a lot or do work in audio. I'd always prefer Stax (even the cheapest model) or Raal as for me I prefer them for mixing/audio type work. I used to have multiple sets of headphones to check mixes/speakers ... but now I now just have the Raals (though I'd be fine with the $600 stax set)

One thing that is great about the raals for professional-type use is if you're stuck using them for hours on end editing things or something: you won't get super sweaty ears like regular headphones. As far as the price: for me it was cheaper to buy a used pair of raals and use a switcher for my speaker amp than buying a high end stax setup + spendy ESL headphone amp you can only use with electrostatic headphones)

You can fix the L300's bass extension by performing blue tack mod and sealing the pad mount mechanism.

 

[Degru]

this is simular to AKG K1000. I look all the years for a heaphone as AKG K 1000 but with no clipping bass. there is none see. I hope there come more as this for better price.
How loud do you think is the SR1A ?. when it have 90 db, with 100 watt then it have 87 db with 50 Watt. for me are 80 db good enough and 7 db for bass boost. do you think the headphone can reach that without clipping ?

the AKG K1000 i buy in the middle of 90 as far i remember. it give a much better room feeling and not so much in the head fealing as other headphones. and it give better transients as speakers in nearfield. On my K1000 was problem with lots bass distortions. it do mettalic distort on some songs and low bass very much when overall loudness is more than (i guess 74 db). but i need 80 db to hear good. I have not the headphone amp for this. i have osciloscope and i see no clipping in the headphone out of my equeipment. with headphone out it reach around 80 db, loud enough and it is wear good. headphones that press on my ears i can not wear more than 20 minutes. this i can wear hours. only tAhe bass clipping problem.

Have a look at Mysphere 3.2, which is the direct successor to k1k pretty much. You need a source with 60ohm or so impedance (or resistors in line) to damp it properly however.