Since the Iriver IMP series can be firmware upgraded to use OGG i thought i might get one and flash the firmware. Even though its limited to 160kbps (not sure if this means it will skip at any bitrate beyond that, but i thought, cool gotta be better than mp3)

Ive just done some testing and the first track i pick (Orbital Meltdown) sounds terrible (to me, and this is played back through an on-board soundcard and cheapo labtec phones - what i have handy at work), literally in the first second of playback with Aotuv tuning and QK tuning at q4 and at q5 (and tried 4.5). GT3b2 Sounds much better at Q5, but bitrates peak at huge levels (and probably not copable with by the iriver decoder) Interestingly the LAME -V5 --athaa-sensitivity 1 as used in the multiformat test fares much better. I have ABX results for the OGG encodes i tried. Not quite sure how to describe the effect other than as very obvious and high frequency mushiness/blurring. I also tried the new release 1.1 and the problem is still there, though slighly reduced. A quick trial with mpc (just radio setting i beleive - encoded quickly for comparison) showed no evidence of this problem.

Original VS Aotuv @ 4.5

ABX Results:
Original vs C:\Music\THEALTOGETHER_LPAC_\Original\TRACK11autuv4.5.wav
12 out of 12, pval < 0.001


---- Detailed ABX results ----
Original vs C:\Music\THEALTOGETHER_LPAC_\Original\TRACK11autuv4.5.wav
Playback Range: 00.000 to 10:17.760
11:20:20 AM p 1/1 pval = 0.5
11:20:24 AM p 2/2 pval = 0.25
11:20:30 AM p 3/3 pval = 0.125
11:20:33 AM p 4/4 pval = 0.062
11:20:36 AM p 5/5 pval = 0.031
11:20:39 AM p 6/6 pval = 0.015
11:20:41 AM p 7/7 pval = 0.0070
11:20:43 AM p 8/8 pval = 0.0030
11:20:46 AM p 9/9 pval = 0.0010
11:20:48 AM p 10/10 pval < 0.001
11:20:51 AM p 11/11 pval < 0.001
11:20:54 AM p 12/12 pval < 0.001

Original VS QKTune@Q4

ABX Results:
Original vs C:\Music\THEALTOGETHER_LPAC_\TRACK11.wav
13 out of 13, pval < 0.001


---- Detailed ABX results ----
Original vs C:\Music\THEALTOGETHER_LPAC_\TRACK11.wav
Playback Range: 00.000 to 10:17.760
10:17:29 AM p 1/1 pval = 0.5
10:17:33 AM p 2/2 pval = 0.25
10:17:35 AM p 3/3 pval = 0.125
10:17:37 AM p 4/4 pval = 0.062
10:17:44 AM p 5/5 pval = 0.031
10:17:46 AM p 6/6 pval = 0.015
10:17:53 AM p 7/7 pval = 0.0070
10:17:55 AM p 8/8 pval = 0.0030
10:18:03 AM p 9/9 pval = 0.0010
10:18:06 AM p 10/10 pval < 0.001
10:18:10 AM p 11/11 pval < 0.001
10:18:14 AM p 12/12 pval < 0.001
10:18:19 AM p 13/13 pval < 0.001

megamix @ Q5

ABX Results:
Original vs C:\Music\THEALTOGETHER_LPAC_\Original\TRACK11megamixQ5stillthere.wav
12 out of 12, pval < 0.001


---- Detailed ABX results ----
Original vs C:\Music\THEALTOGETHER_LPAC_\Original\TRACK11megamixQ5stillthere.wav
Playback Range: 00.000 to 10:17.760
10:49:10 AM p 1/1 pval = 0.5
10:49:12 AM p 2/2 pval = 0.25
10:49:14 AM p 3/3 pval = 0.125
10:49:16 AM p 4/4 pval = 0.062
10:49:20 AM p 5/5 pval = 0.031
10:49:22 AM p 6/6 pval = 0.015
10:49:24 AM p 7/7 pval = 0.0070
10:49:26 AM p 8/8 pval = 0.0030
10:49:28 AM p 9/9 pval = 0.0010
10:49:32 AM p 10/10 pval < 0.001
10:49:38 AM p 11/11 pval < 0.001
10:49:41 AM p 12/12 pval < 0.001


GT3b2@ Q5

ABX Results:
Original vs C:\Music\THEALTOGETHER_LPAC_\Original\TRACK11GTBQ5-stillsomeevidence.wav
10 out of 10, pval < 0.001


---- Detailed ABX results ----
Original vs C:\Music\THEALTOGETHER_LPAC_\Original\TRACK11GTBQ5-stillsomeevidence.wav
Playback Range: 00.000 to 10:17.760
10:46:40 AM p 1/1 pval = 0.5
10:46:46 AM p 2/2 pval = 0.25
10:46:54 AM p 3/3 pval = 0.125
10:47:06 AM p 4/4 pval = 0.062
10:47:29 AM p 5/5 pval = 0.031
10:47:39 AM p 6/6 pval = 0.015
10:47:46 AM p 7/7 pval = 0.0070
10:47:54 AM p 8/8 pval = 0.0030
10:48:04 AM p 9/9 pval = 0.0010
10:48:11 AM p 10/10 pval < 0.001


GT3b2 @ Q5 is by far the best of the ones i tested (i did test aotuv at 4 as well as Q4.5 but it was not much different and plainly obvious)

I havent got anywhere to upload stuff at the moment, if i can i will (if anybody is interested - perhaps interesting that this problem is very obvious on megamix q5 but not anywhere near as bad with GT3b2 - which im correct in thinking is only tuned from q5 and up?)

what about the upload section of this board ?
I'm interested in your problem sample.

Doesnt seem to want to let me. Ive tried it as a FLAC and zipped, gives an error. Its only an 18 sec sample so its hardly too big. Not sure if its OK to upload it anyways.

UPDATE: Ive just tried megamix at q6 and the problem is still there, but less so. So chucking bits at it isnt the answer. gt3b2 @ q5 and above has a good stab, but below is just as bad as the others. Obviously something in the tunings works to alleviate this problem at q5 and above for gt3b2, but something that wasnt incorporated into the megamix, as at q5 megamix still very much suffers from the same problem.

ABX Results:
Original vs C:\Music\THEALTOGETHER_LPAC_\Original\TRACK11megamixq6.wav
10 out of 10, pval < 0.001


---- Detailed ABX results ----
Original vs C:\Music\THEALTOGETHER_LPAC_\Original\TRACK11megamixq6.wav
Playback Range: 00.000 to 10:17.760
3:51:52 PM p 1/1 pval = 0.5
3:51:55 PM p 2/2 pval = 0.25
3:51:59 PM p 3/3 pval = 0.125
3:52:04 PM p 4/4 pval = 0.062
3:52:11 PM p 5/5 pval = 0.031
3:52:15 PM p 6/6 pval = 0.015
3:52:19 PM p 7/7 pval = 0.0070
3:52:25 PM p 8/8 pval = 0.0030
3:52:40 PM p 9/9 pval = 0.0010
3:52:46 PM p 10/10 pval < 0.001

I can upload the sample to my webspace for you. Check your PM for email details

I've uploaded the sample

http://www.rarewares.org/quantumknot/Originalsmall.flac