I MUST KNOW! D2 Measurements and Observations [LONG]

[Grandpa Jones]

Argh. Cables arrived. I guess it's time to look for cheap flights to Korea.

This is stupid. I emailed JetMall with a request to please change the description to have a disclaimer that it only works with TV Out mode. I'll go post cowonglobal and cowonamerica and see if I get a reply. (yea right) It obviously ties to the Wolfson inputs because the volume exhibits the same 34->35 and 40->41 switchover behavior. Hopefully it's not a design limitation but something that could be changed via firmware.

I'd send it back, but the shipping is more than the stupid cable is worth. I'll encode an all-black video with the test tone track as the audio and measure it, I guess. Not much else I can do about it.

[Tac21]

ahh shameful to say..

so the D2 doesn't do everything leave it be and use it what its used for. for now. While see what they update in future firmwires

[Grandpa Jones]

Heh, fair enough. I don't mean to imply I'm unhappy with the D2, it just seems like a design flaw; reading the datasheet for the Wolfson and it has those nice, shiny pins #21 & #22 dedicated for line-level outputs, not to utilize them for everything seems wrong. It's either they didn't think of it or that it might be used that way, or due to the way the video chip is integrated it cannot be done. It's worth asking about.

[PrincipalValiant]

Very good thread, thanks gramps!

[Lapis Molari]

Gramps,
Thanks for your tests! Much appreciated and very useful.
This has improved my D2's performance as a line-out device.
Keep 'm comin'!

[sigsegv0x0b]

Do you have an RMAA frequency sweep of the D2 trying to drive 16ohm load?

[wader]

Wow, interesting topic - I never saw this one (was away from the boards for a time) and was actually wondering about RMAA tests with various effects and EQ settings at different volumes. This scopes down the areas to check out nicely.

Guess I don't have to test out volume 50 for use in the car . . . and am now reading on how volume 34 may be a current/Class switching point in the codec implementation . . . neat!

- wader

[Captain Ødegård]

Good job, nice read. Though personally I prefer my ears over such graphs, as tests like this have more to do with placebo effect than actually proving what sounds best.

[TheDarkSide]

Thanks for putting the time into an excellent thread!!! Great info,........ I'm gonna' keep it bookmarked for reference!!!

[P8ntBal1551]

Originally Posted by grandpa jones

Or are you looking for more simplistic answers like... Uh... the d2 rocks but you have to experiment with settings? :-)

yes woooo the d2 rules man!!!!!
D2! D2! D2! D2! D2! D2!

[ben.]

Great thread. Makes me want to change career paths, follow my heart, and get stuck in with DSP! Really should have done that by now to be honest!

The 6% THD for +5 EQ is nuts! That's pretty awful. I'm guessing you don't want to fiddle with the eq unless you have to. Maybe the guys at BBE should make their own 20-band eq and give it to Cowon, that would be sweet. I'm very impressed with the BBE enhancement results. To my ears, BBE off -> BBE 1 is a huge difference, and anything above 1 doesn't do a great deal. Maybe the tilt in the spectrum even with BBE on just 1 is considerable. I can't really hear any serious difference between +1 and +5.

"I'd love it, love it if..." the D2 had a 20-band eq. Would be hard adjusting it by finger I'll admit.

I also think they could release a 'high end' D2 with integrated digital amplifier stage, like the Kenwood media players - http://www.anythingbutipod.com/archives/kenwood/
"Those who admired Kenwood's focus on high sound quality will be pleased to hear that the HD60GD9 still makes use of a dual stage clear digital amp EX and Supreme EX decoder technology." Sounds impressive anyway.

[paulr]

+5 at level 50 would be clipping of course. I cringe at ever setting the level to 50 because of the possibility of plugging in headphones (or iems!!) by accident.

I expect the d2 cpu is easily powerful enough to do 20 band eq in software if there was some use for it. It sounds from the description like BBE is a non-causal filter (i.e. it looks at the "future" samples to see if a sudden loud section is coming, and adjusts the gain ahead of time) and that's not too difficult either. So, maybe these features could be put into rockbox if they're not there already.

I agree, an audiophile d2 would be great

[paulr]

Another thing occurs to me: yeah it's weird to measure a 25 mw DAP like a power amp, but I wonder if a THD measurement is available for a low frequency (20 hz) tone into a low impedance (16 ohm I guess) load? That would show how well the power supply is holding up. Even better would be some measurements with reactive loads.

I notice that fancier headphone amplifiers have quite high power supply voltages (Core Sound Headline uses 9v but is happier with 12v, some USB dacs have 7.4 volt internal packs, the Headroom Total Bithead uses four AAA's but advises against rechargeables because of their lower voltage). Why is that, if 0db is 0.775 volts? Even if they're trying to operate completely class A, you'd think 3.6 volts was plenty. Should they really use more, i.e. is it a dirty secret of 3.6 volt DAP's that they would sound better with voltage boost circuits inside? It can't be that hard to do, since 1.5-volt amps like the Fiio supposedly sound good by DAP standards.

[Jingo Lingo]

Originally Posted by Captain Ødegård

Good job, nice read. Though personally I prefer my ears over such graphs, as tests like this have more to do with placebo effect than actually proving what sounds best.

How do measurements have anything to do with the placebo effect?