|
|
Post by stanleyb on Jun 10, 2015 13:58:24 GMT
Thanks Martin. It might be as early as tomorrow that I put something in the post to you.
If you do have some 192kHz files and a USB cable, can you also try the SFWII on the SB Touch with EDO? But it's not a pressing case.
|
|
|
|
Post by MartinT on Jun 10, 2015 14:47:08 GMT
Yes, I do have a few and I can try it on USB for you, compare it with streaming via the Raspberry Pi (I no longer have the Touch).
|
|
|
|
Post by MartinT on Jun 13, 2015 10:47:07 GMT
The SFWII CPU is now in my Caiman-II. It coincides nicely with enforced use for all disc sources since my Ayre will be away for a while with TonyC for some potential super-clock upgrades.
I had never thought my Pioneer LX-55 Blu-ray player much cop as a CD transport, but then I hadn't been particularly fair to it. So last night I placed it on the Electric Beach platform and fed it with regenerated mains. The C-II remains powered by battery. Since the Pioneer can play every disc type including CD, SACD, DVD-A and, of course, BDs, it's going to take time to go through every permutation. I'll also be experimenting with hi-res files via the RPi2 and direct USB playback.
I won't say much at the moment as it's a bit early to be making conclusions. However, what I am hearing above all else with every type of disc tried so far is a very organic, analogue-like presentation. This has always come about, from my own and others' observations, whenever the data clocking has been improved and that is exactly what Stan has been working on.
So far, so very good. Much more later.
|
|
|
|
Post by stanleyb on Jun 13, 2015 17:04:43 GMT
I have been experimenting with two versions of the SFWII, and I sent out samples of the various versions to different persons. But Lee up in Scotland has two pieces of the CMII. So he is going to test both versions.
The main technical difference between the two is that one uses a timer with clock recovery disabled, whilst the other uses the timer with clock recovery enabled. The good news is that I have managed to figure out how to combine the two. The chip I use has two user programmable timers. So I have added a user selectable switch on routine to switch between the two different timers. Unfortunately this swap over of the timers can only be done at switch on of the CMII. I haven't quite figured out how to do it whilst the music is playing. I am out of program memory space for one. I'll probably delete the AUTO function in a test chip and use that memory space to see if I can come up with an "on the fly" swap between the two timers.
Are there differences? Well, I can hear something and it also comes through on the floor board vibration test. Which one sounds best is a different matter. It's not chalk and cheese. It's more Chedar or Edam, or Prosecco or Champagne. Some of my tracks sound better with one timer, and visa versa.
Martin, it might be worth doing a TOSLINK versus Coax comparison. The coax from my CD player definitely sounds different from the Toslink.
I tend to use the Norah Jones: Come Away With Me CD for certain test. One thing that gets to me is her trouble in pronouncing the letter r when she is singing. This new firmware makes it even more excruciation to listen to.
|
|
|
|
Post by MartinT on Jun 13, 2015 18:13:54 GMT
I'll listen to that, Stan. I also find the cymbal shimmer most telling in that track, as well as the soft textured bass. As it's an SACD, the hi-res allows me to hear very clearly into the soundstage.
|
|
|
|
Post by stanleyb on Jun 13, 2015 18:42:16 GMT
I am just listening to modified FW with on the fly switching. I replaced the Auto function for the timer switch. I'll have to send you a copy to try next week. The ability to select the most appropriate timer for each track has be heard to be believed.
|
|
|
|
Post by MartinT on Jun 13, 2015 19:46:17 GMT
Excellent. Meanwhile, I'm still listening and assimilating.
|
|
|
|
Post by krzysztof on Jun 17, 2015 7:01:17 GMT
Hello guys,
Anyone using Beyers T90 with Caiman II? Curious how they sound with such pairing, because they have quite non linear impedance especially with high peak at the low bass range over 250Ohms. Any experience about it?
|
|
|
|
Post by John on Jun 18, 2015 8:27:22 GMT
Night be worth contacting Stan as he has quite a few headphones that he uses for test
|
|
|
|
Post by krzysztof on Jun 18, 2015 8:31:33 GMT
Night be worth contacting Stan as he has quite a few headphones that he uses for test I am not sure if he has Beyers T90 specifically? In general they're 250 Ohm, but with quite high impedance irregularity in low frequency with its increase from 250 Ohm to higher values in that region. That is why I am curious if CMII could have enough power to drive them properly. |
|
|
|
Post by stanleyb on Jun 19, 2015 6:45:37 GMT
Done some further reliability and bug testing on the triple option version of the super firmware. For convenience sake I shall refer to it as SFW3. It is basically the SFWII plus two other options that improve on the original firmware codes. It took me a while to come up with a method to switch between options, but I somehow managed.
Option 1 has a noise discriminator built in that is designed to reject some of the input changing etc. noises you can get with the likes of airport express etc.
Option 2 uses the main part of the code of the original super firmware, plus a few enhancement.
Option 3 uses a completely new method. It is basically an active clocking/reclocking circuit with clock recovery and PLL based on the bitrate and sampling frequency of the incoming audio data. It has a dramatic effect on high quality material, but can also expose poorly recorded or remastered material. A few people have been testing the option 3 code in the SFWII.
|
|
|
|
Post by krzysztof on Jun 19, 2015 7:18:12 GMT
Done some further reliability and bug testing on the triple option version of the super firmware. For convenience sake I shall refer to it as SFW3. It is basically the SFWII plus two other options that improve on the original firmware codes. It took me a while to come up with a method to switch between options, but I somehow managed. Option 1 has a noise discriminator built in that is designed to reject some of the input changing etc. noises you can get with the likes of airport express etc. Option 2 uses the main part of the code of the original super firmware, plus a few enhancement. Option 3 uses a completely new method. It is basically an active clocking/reclocking circuit with clock recovery and PLL based on the bitrate and sampling frequency of the incoming audio data. It has a dramatic effect on high quality material, but can also expose poorly recorded or remastered material. A few people have been testing the option 3 code in the SFWII. Very good to know Stan. I have been using SFW for quite some time and it's amazing, also fixes my previous problems with clicks heard without music playing. Now it's been completely quite. What kind of enhancements you have done to Option 2? I think that CMII is revealing enough already to be picky with poor quality material (quite common nowadays), but on the other hand, the DAC should be the most transparent audio gear in the chain IMHO. If you can fix the harsh sound when changing the sampling rates time to time that would be also welcome. Nice job and waiting for testers feedback on SFWII and Option 3. I would be happy to test it as well, but guess that's since I am out of the UK it's impossible. |
|
|
|
Post by stanleyb on Jun 19, 2015 9:34:03 GMT
Option-2 now has a better rejection of the sample rate or CD track switch over noise that some people have noticed.
Option-3 can show an increase in dynamic range, where that dynamic range is available, and also produces a better signal separation. The clock recovery part of option-3 seems to be recovering data that would otherwise get rejected by the input receiver clock. I am not yet clear in my own mind how that comes about. I am just highlighting the important parts that I noticed.
|
|
|
|
Post by krzysztof on Jun 19, 2015 10:06:19 GMT
Option-2 now has a better rejection of the sample rate or CD track switch over noise that some people have noticed. Option-3 can show an increase in dynamic range, where that dynamic range is available, and also produces a better signal separation. The clock recovery part of option-3 seems to be recovering data that would otherwise get rejected by the input receiver clock. I am not yet clear in my own mind how that comes about. I am just highlighting the important parts that I noticed. That's interesting, especially about Option-3, since in the normal situation nothing should have been rejected. How do you noticed that, by ear or by debugging the firmware? Dynamic range is to be persistent every time, though depends on S/N ratio and then bit resolution the DAC is able to present. BTW, you observation are related to the particular input (COAX/TOSLINK/USB) or generally to how the DAC is working despite the input flavor? |
|
|
|
Post by stanleyb on Jun 19, 2015 10:18:37 GMT
If I couldn't hear it I wouldn't mention it.
|
|
|
|
Post by krzysztof on Jun 19, 2015 10:26:28 GMT
If I couldn't hear it I wouldn't mention it. That's fine. Waiting for the conclusion. BTW, do you have Beyers T90? |
|
|
|
Post by stanleyb on Jun 19, 2015 19:51:56 GMT
I don't have any Beyer headphones. Sorry.
|
|
|
|
Post by krzysztof on Jun 19, 2015 20:09:19 GMT
I don't have any Beyer headphones. Sorry. No problemo, just asking. |
|
|
|
Post by krzysztof on Jun 21, 2015 17:20:37 GMT
Stan, please put me on the list for new firmware version II in the future.
|
|
|
|
Post by stanleyb on Jun 21, 2015 19:41:51 GMT
Anyone interested should email me. I am not good with remembering requests on forums for some reason.
|
|
