[wslee]

I use a FIIO BTR1 for my Beyerdynamic DT770 32Ohm. Sound quality is good, but I need more power. Is the Radsone EarStudio more powered? Is there a difference between the power of the 3.5mm se and the 2.5mm balanced output?

FIIO BTR1 has AK4376, while ES100 has AK4375a.

Both DACs support the same output power, and H/P amp gain up to +6dB.
I'm not sure FIIO BTR1 AK4376 H/P amp gain configuration, but as I check the specification,
http://www.fiio.net/en/products/77
it seems that ES100 provides slightly more power with no distinguishable difference, I guess.

ES100 2.5mm output provides much higher power,
but DT770 doesn't support the balanced cable replacement.

Thanks and Regards,
WS

 

[wslee]

Thank you. I understand that decreasing preamp and increasing volume will increase noise floor. But anything potentially bad with dynamic range cut due to bit loss?
I used -3dB before and unfortunately, it was making weird increase of treble. I'm just not sure if EQ is done in 24bit and 12dB would cut audible details.

@SubMash:
The Kalimba DSP on CSR8675 is a 24-bit fixed point DSP, not a floating point.
It has 24 x 24-bit MAC with a 56-bit accumulator.
That is, the mathematical errors propagation on IIR filter would be quite larger than 64-point double precision floating math.
But, again, since audio track does have very little power in the high-frequency bands, 24-bit math would be enough.
Anyway, I don't think it's a good idea to boost up the high-frequency bands from a lossless codec, with that much of excessive gain.
Thanks and Regards,
WS

 

[HD800NL]

FIIO BTR1 has AK4376, while ES100 has AK4375a.

Both DACs support the same output power, and H/P amp gain up to +6dB.
I'm not sure FIIO BTR1 AK4376 H/P amp gain configuration, but as I check the specification,
http://www.fiio.net/en/products/77
it seems that ES100 provides slightly more power with no distinguishable difference, I guess.

ES100 2.5mm output provides much higher power,
but DT770 doesn't support the balanced cable replacement.

Thanks and Regards,
WS

Thanks a lot! When the ES100 is available in Europe I will buy one. Seems the best option for my wireless setup. BTW I have a DT770 from Custom Cans with a detachable cable mod and a balanced 4 pin mini XLR

 

[SubMash]

@SubMash:
The Kalimba DSP on CSR8675 is a 24-bit fixed point DSP, not a floating point.
It has 24 x 24-bit MAC with a 56-bit accumulator.
That is, the mathematical errors propagation on IIR filter would be quite larger than 64-point double precision floating math.
But, again, since audio track does have very little power in the high-frequency bands, 24-bit math would be enough.
Anyway, I don't think it's a good idea to boost up the high-frequency bands from a lossless codec, with that much of excessive gain.
Thanks and Regards,
WS

I appreciate detailed insight. Problem with se846 is that treble goes down after 8khz quite significantly. I don't see much of the problems with AAC so I'm assuming that issues should not be significantly amplified. -12dB preamp should not by itself go out of such math engine, probably. I will try to conduct thd test to compare.

 

[m4rkw]

I appreciate detailed insight. Problem with se846 is that treble goes down after 8khz quite significantly. I don't see much of the problems with AAC so I'm assuming that issues should not be significantly amplified. -12dB preamp should not by itself go out of such math engine, probably. I will try to conduct thd test to compare.

The FR graph posted here suggests it goes down after 10k rather than 8k: https://www.innerfidelity.com/images/ShureSE846WhiteFilterSample2.pdf

 

[m4rkw]

Actually the curve doesn't look that different from the SE535 one: https://www.innerfidelity.com/images/ShureSE535.pdf

In fact the 846 seems to have additional response waves in the higher end that aren't present on the 535 graph.

 

[SubMash]

I think you post in the wrong topic. While both Shure suck at treble - you can see 535 is more flat in response, impedance phase and ability to play squares. Anyway, they both need any help available to boost treble.

 

[m4rkw]

You really think? I find the 535LTD is quite bright, any more treble is just fatiguing. Never heard the 846 though.

 

[SubMash]

Then it means you don't like trebles. It's ok. To make them close to reality I have to make this

But I can clearly hear harmonics in such case.

Ideal solution would be to wait when sonar works will make 24 bit truefi for Spotify android and run that through LDAP or APTX HD (if its not as noisy as aptx).

 

[m4rkw]

How are you measuring "reality" for this?

 

[SubMash]

How are you measuring "reality" for this?

1. I can see charts. 2. I know some of the tracks quite well. 3. I can quickly switch with sonarworks reference calibrated hd650.

I don't have headphones calibration rig. You can get used to some sound and call 535 bright. But try to listen half an hour with something I made above and then return back and you will hear how much you lost. Again, it's a matter of personal preference - I prefer to hear same stuff that producer heard.

 

[m4rkw]

Your settings are for 846 though aren't they? Not 535

 

[m4rkw]

Hmm ok, with your settings I can definitely hear a lot more detail. It's also quite bright and in-your-face though, fatiguing to listen to. I'll experiment later and see if I can find a happy point somewhere in between.

 

[SubMash]

Your settings are for 846 though aren't they? Not 535

Adjust somewhat to have balance. 535 doesn't need to fix anything below 2k. Trebles, though have same issue. 846 bass is grossly overpowers music and makes whole thing booming.

 

[SubMash]

Few bugs:
Sometimes switching back to the app crashing it and eventually, app comes to the state where it refuses to work and connect to any device. The only solution is to reconnect the device - app restart doesn't help, device search doesn't work.
Another problem - Avantree Priva III ignores aptx turned off and even after re-pair it forces to connect aptx no matter what I do.
Avantree has APTX LL, but it never connects to LL. And there are no APTX HD USB transmitters.