Cayin N3 Hi-Res DAP with AKM4490 DAC, apt-X Bluetooth, and Line, USB & Coax Out for $150
Apr 11, 2017 at 7:43 AM Post #1,171 of 6,262
...AK4490EN chip and the 32-bit, 8x sampling SCDF (Sound Color Digital Filter) cutoff filter... ...SHARP was the worst filter of all, with the clean brassy resonance of the ride cymbal losing all harmonic resonance... ...The most accurate was the SUPER SLOW, which actually drops the SCDF's sampling to half... ...I also had a go with the keyboard and noticed that the recorded track did work best with SHARP... ...As long as I don't touch the EQ and/or set the EQ with nothing playing, power down and restart, the issue hasn't returned and all test tracks sound very close to how they play on other devices.

 
Hi, I've been lurking and trying to skim through the whole thread, and it's fascinating to read this kind of insight. Thanks for sharing.
 
Is there any website we can hear examples of the filter differences for ourselves on our own PCs or DAPs? or would someone have to make a recording of the output and upload it?
 
Apr 11, 2017 at 7:55 AM Post #1,172 of 6,262
I use mediamonkey to analyse the volume of the ripped tracks and tag them with the adjustment value.

OK, got it. I thought you could set the gain per album or song on the N3 itself. Thank you very much for the clarification, I don't think I will ever use it, seems like more work for something marginally more convenient but nice to have the option I guess. 
 
Apr 11, 2017 at 9:16 AM Post #1,173 of 6,262
  What do you get when you combine FedEx with UPS? ...Fed UP. 
biggrin.gif
  In Canada Purolator is the big player and they are even worse than Fed UP.  In Toronto the largest city in Canada they have only two pickup locations and the one closest to me is a 35 minute drive.  Trust me you can do worse than Fed UP. 

Hello fish 1050,
 
We have many service points in Toronto, for all the details pls don’t hesitate to contact us at customer.care@purolator.com.
 
Cheers,
 
Catherine
Customer CareTeam/ Purolator Inc.
 
Apr 11, 2017 at 9:29 AM Post #1,175 of 6,262
 
Wow! Purolator is on HeadFi?!


Well Catherine is obviously not a lurker (?) since you see it's her first post and just joined I'm just wondering how a post about them got picked up so quickly
blink.gif
I mean do the walls actually have ears
rolleyes.gif
 
 
Apr 11, 2017 at 9:37 AM Post #1,178 of 6,262
 
Well Catherine is obviously not a lurker (?) since you see it's her first post and just joined I'm just wondering how a post about them got picked up so quickly
blink.gif
I mean do the walls actually have ears
rolleyes.gif
 


I wonder how long it would take the secret service to post if I said something negative about Mr. Trump...
 
Apr 11, 2017 at 10:07 AM Post #1,180 of 6,262
  Under currently viewing I noticed Purolatorhelp listed...I guess we are always being watched.


No, not the POTUS
rolleyes.gif
He's about to put a major whoopazz and Putin and Kim Jung and I can't put everything on his shoulders since your kids, kids kids will be suffering from Mitchy boys ending of the SC fillibuster. I guess you just change the rules to "Make America Great Again"
frown.gif

 
How about a nice shot of an N3 to brighten your day
ph34r.gif


 
Apr 11, 2017 at 10:11 AM Post #1,181 of 6,262
 
No, not the POTUS
rolleyes.gif
He's about to put a major whoopazz and Putin and Kim Jung and I can't put everything on his shoulders since your kids, kids kids will be suffering from Mitchy boys ending of the SC fillibuster. I guess you just change the rules to "Make America Great Again"
frown.gif

 
How about a nice shot of an N3 to brighten your day
ph34r.gif



That's better than depressing politics and the prospect of WW3.
 
Apr 11, 2017 at 10:34 AM Post #1,182 of 6,262
  OK, got it. I thought you could set the gain per album or song on the N3 itself. Thank you very much for the clarification, I don't think I will ever use it, seems like more work for something marginally more convenient but nice to have the option I guess. 

The files have to be scanned first so that the highest peak can be detected and values stored within the file that show the deviation from the reference volume. Now that you have a fixed target, it can level the volume correctly. The Track vs Album value are both stored. The Track is an absolute difference based on the reference (probably 89db). It ensures that the volume is leveled even if you play tracks from different albums randomnly. The Album value is the difference based on only songs in that tagged album so that quiet tracks are maintained if playing album by album. This is how ReplayGain works. 
 
There is also Volume Leveling, that some manufacturers use instead of RG. There isn't much of a standard there, so it's up to the manufacture to make up how they think it should work.
 
It's not hard at all, just highlight all your tracks and click "Analyze Volume", then go to bed. In the morning, copy the files over to your card and you're good to go. Foobar also does it via the right-click menu "ReplayGain->Scan Selection as Albums (by tag)".
 
It's more than convenient, as it keeps two tracks from playing greatly different in volume and blowing out your ears because you turned up the volume on the first (quieter) track to hear it.
 
Apr 11, 2017 at 10:52 AM Post #1,183 of 6,262
   
Hi, I've been lurking and trying to skim through the whole thread, and it's fascinating to read this kind of insight. Thanks for sharing.
 
Is there any website we can hear examples of the filter differences for ourselves on our own PCs or DAPs? or would someone have to make a recording of the output and upload it?

There's no way to accurately guarantee uploading samples and making it heard. There are still a lot of variables involved, such as the resolution of the phones playing back and whether the device recording is capable of picking up the values (remember, it's involving an analog-to-digital converter again). 
 
You'd have to be in a recording studio with very accurate monitors to be able to play back and hear it. And even then, it's much harder to hear via speakers than phones since they are further away from your ears and begin to involve the "acoustic chamber" known as the ear canal.
 
Odds are that if you haven't noticed it already, your phones aren't capable of reproducing those frequencies, your ears can't hear them, or your source media doesn't contain anything with the nuances to be noticed. There's quite a few tracks that I have where the original lossless 16/44 CD wasn't noticeable, but the high-res 24/96 remastering or SACD/DSD is noticeable. If you only listen to lossy material or poorly recorded/mixed tracks, you'll probably never notice it either as most of the data is already lost before it got to you (or in the lossy conversion).
 
The other thing is that without hearing original (real) instruments as a reference, it's hard to tell what's been changed in the recording process. If you were a drummer, you'd pay more attention to and notice that cymbals don't sound right or snares sounded dry. If you played piano, you'd probably notice the string resonance not ringing right. Engineers would notice that sounds decay too quickly or cut off suddenly, like a gate was installed. I originally noticed it after listening to a track I'd just recorded and thought I screwed up the mix. So I went on the other side of the glass, listened, and followed the sound path through until I noticed the difference. That difference occured as I was playing back a rough daily on the N3 and I thought it was a bad gate in the mix. Once I played it back in studio, the edge wasn't there. It was then that I started looking into the filters and how they affected the playback of that track.
 
Apr 11, 2017 at 11:00 AM Post #1,184 of 6,262
To clarify my recommendation more, I'll offer the following in best-to-worst:

Super Slow: the most accurate and clean sounding. Shows instruments as they should sound in the studio without digital effects. Runs the 8x sampling of the AK4490 filter at 4x.Will cost a bit of detail in high res files, but some of those details are overrated and unnatural. Piano, guitar, bass, drums, cymbals sound dead accurate. Not the most sub-bass/bass reverb. Listening to actual tracks recorded off a sound board are decently accurate to the sound board itself during recording. Reminiscent of analog playback from a good source, but with better highs. Smooth, rich.
SEE: Phil Collins "The Singles", Eric Clapton "Complete Clapton", Fleetwood Mac (HDTracks remasters, especially Tusk), Alison Krauss (Live), most classical works
EQ: You could add +1db to 31, 63, 125 and 8k, 16k to give a bit more definition back.

Short Delay Slow: the best compromise between detailed and accurate. Cymbal hits don't ring as long as they should, but still sound natural and clear. Drum vibrations ring as they should. High-res details appear, but subtle and visible. You can still tell the difference between a keyboard and piano, as well as real drums vs digital (midi) drumpads. Good for more "powerful" music with real instruments where you would want more detail to appear.
SEE: Sierra Hull (Daybreak), Joan Osborne (Relish, 20th Anniv Ed), Doobie Brothers (HDTracks Best of the Doobies), Enya (Best of), Amelia Brightman (self-titled), Jewel (Best of), Dar Williams (most albums)
EQ: flat works great

Short Delay Sharp: details start to sound digital, drums 'tap' instead of 'thwap'. Cymbals get 'ting' instead of 'taaannng'. Vocals get rich detail and bass nuances appear. Keyboards show harmonics in the highs. Pianos sound like keyboards. Digital sourced music (electronica, trance, ambience) sounds pretty good here.
SEE: NIN (HDTracks The Downward Spiral), Conjure One (any album), Depeche Mode (Violator DSD), Lindsey Stirling (Shatter Me), Dido (most albums)
EQ: had to take the 8K down -1db to cut off a bit of harshness in some vocals. A de-esser would have been nice.

Sharp: Everything sounds digital. Drums sound more like drumpad digital, including cymbals. Everything has more punch, but less nuances. Very little of the natural details of instruments remain. MP3 (lossy) files start to show that harmonic anti-aliasing edge where the definition is lost (around 12-16k). Electronics stuff sounds crisp and powerful (Conjure One sounds good here too, but can get some harsh treble). Very reminiscent of early days of CD.
SEE: Better to not use this, but use Short Delay Sharp as the result will be cleaner and less harsh. I did find that the HDTracks version of Megadeth Symphony of Destruction sounded really good with this though. So if you want everything as sharp, high, and crisp as possible, this is for you. Don't do it if you want any instrument to not sound synthesized though.
EQ: To taste, as there just ain't no right here.

Slow: Don't do it, just don't. Everything sounds fake and muddy. Digital sounds wrong, real instruments get warbly and sound like a gate/compressor has been applied. All the other filters should cover what you need. This filter is just not right.
SEE: n/a

* Note that EQ should be used to equalize a listening environment FIRST. In the case of a room and speakers, it should neutralize any reflections or absorbtions. For headphones, it's to correct dips and other artifacting to make them neutral. THEN, you would add/subtract what you want to taste. And, the typical rule of preferring subtractive EQ before additive doesn't work as well with the N3 for some reason. Probably gain related. In the case of the UERR, EQ was off or flat. For the studio monitors, well they are already in a properly treated room with unity gain and no effects rack inline.

* The "SEE" items are tracks that I've really noticed the difference in. Though my analysis of the filters does NOT originally come from pre-packaged music, but rather tracks that I created for testing in the studio recorded off the mixing console where I listened to them via UERR and monitors. I also listened from the other side of the glass to study the details first.

The filter difference isn't as subtle as it's made out to be by AKM or others. It just takes a good ear and accurate monitors, good reference material, a *dead* quiet & accurate environment, and a lot of damn time! The difference between Super Slow and Short Delay Slow can be noticed in the first 15 seconds of Phil Collins remastered "True Colors" from The Singles as well as in the fine detail of "In the Air Tonight". Listen for the snare decay, cymbal splash decay, as well as the tiny keyboard pan left.


Thanks for this, it's very nice to have a detailed description on what exactly these filters do.
 
Apr 11, 2017 at 11:09 AM Post #1,185 of 6,262
No, not the POTUS:rolleyes: He's about to put a major whoopazz and Putin and Kim Jung and I can't put everything on his shoulders since your kids, kids kids will be suffering from Mitchy boys ending of the SC fillibuster. I guess you just change the rules to "Make America Great Again":frowning2:

How about a nice shot of an N3 to brighten your day:ph34r:



lol, but hey that pic is making me regret ordering the black N3 :frowning2: after having a titanium x3 2 and a titanium x5 3 I thought it was time to go back to black. Can't wait to get mine, I always like a smaller dap- hence the plenue D. Unfortunately it didn't have enough punch and sound stage for full sized cans.

I can't believe how much Cayin has put into this little unit for 150 bucks. Unless the build quality is utter carp or crap lol.
 

Users who are viewing this thread

Back
Top