WuBai

Auke, I would love to tell you where it comes from, but I'm afraid I do not understand the question! The tweaks come from looking up the functions of ethtool on various Linux forums. Autoneg OFF for example gives a more open but less precise sound. Autoneg ON gives a more detailed but 'digital' sound, at least for me. As for the Target Diretta setting, I'm sorry I've been working and haven't had a chance to connect my Target to my router and check. You will find it if you look. It is in the same place as Diretta Host settings, but on your Target machine. There you can turn on and off DSP or various PCM support. You can also change your latency setting. I tried setting '3' but it did not work at all, Diretta wouldn't connect. I settled on '6' but try to find it and experiment with changing the setting. It is time consuming to listen to each one, but it is worth it. Also don't forget to turn OFF Web Interface on your Target before connecting back to your Host, this is very important for sound.

This is an EXCELLENT find Auke! For me, eth0 seems to be for the connection between my target and host. eth1 seems to be the connection from the my internet switch to the host. By changing both, I get a very clear improvement in sound. I hadn't thought to do this before until you mentioned it. Such a great tweak!

You are correct, each setting impacts every other setting. It is balancing many points to move towards a goal. Unfortunately there is no way for me to understand what each point does in relation to another. For example, my instinct is that turning on 'debug' in Diretta would add an unnecessary process and make it sound worse, but in my system it is the opposite. I've long ago stopped trying to understand and make very small tweaks over a period of many weeks or months. One day, possibly soon, I will write all of my settings for my target and host. This will include all of my memory and cpu settings. That will be a long post and unfortunately I would have difficulty formatting it in a way that makes it easy to follow, but I will try once I have time.

Please keep in mind that I have a LOT of other tweaks in my configurations for both Host and Target, so it is highly probable that my settings will either do nothing for you, or make this sound considerably worse. But I will share my current configuration, but don't expect anything positive! I don't run anything with Ramsystem enabled. Host RPI4 IO Board with CM4 (GP running on EMMC flashed memory) Kernel: 6.9.2-GentooPlayer-ST-CLTO-TEST+ Target is Allo USBridge Signature Kernel: 5.15.79-GentooPlayer-MIN-NOGRAPH-EUP100RCU Profile 1 My Target Diretta setting is to run at a latency of 6. I also turn off the Web Interface on my Target On the Host, I use Eth Tool with these settings: -s eth0 speed 10 duplex half autoneg off For my Diretta Settings on the Host they are below: TargetProfileLimitTime: 406 ThredMode: 2048 InfoCycle: 74500 FlexCycle: random CycleTime : 10000 Debug : enable periodMAX : 32 periodMIN : 12 periodSizeMax : 20000 periodSizeMin : 960 I used to post settings but then I realized that each system has far too many variables and differences where it would be pretty much impossible to just cut and paste my settings in any other system and have it sound right. Even if I changed my own speakers, my current settings would not work and I'd have to make adjustments. So keep that in mind!

Always scared updating since everything always sounds so good, but I can attest that the 8.8 on Target and Host sounds amazing. Better richer tonality, more authentic sounding bass and hearing a lot more details and spatial cues compared to before.

This sounds interesting, can you share the commands on the target when login in under SSH? I'd like to give this a try.

Write down all your settings for 8.60 if you've done tweaks. You will need to flash again from the beginning to access 8.70. Make sure you download the correct version. If you have CLTO then you must download the XTRME version. If not, then just the normal.

I cannot give you any scientific or technical reasons because truthfully I don't know them. What I've found is that for my target system I like to not 'squeeze performance' but rather slow it down considerably. I lower the ram voltage as well. For me, it sounds better when the Target is running fairly slow. The Host, on the other hand, can run faster and I do not use a profile for the host RPI4. I also run Ramsystem on the Target but not the host. This is just one combination, there are so many combinations that people can run where there is no particular 'rule' that can be applied to everyone. Now I see why Le Dom did not post his settings himself. Once a kernel changes, one must start over. The only constant thing in tuning is that there must be a balance and always there is a compromise. Something can sound better than it ever has before, and the limits can continue to increase, but even at a new limit and sound quality there are still tradeoffs. All I can do is tune for myself, with my system and to my ears. I pull levers and take notes, but I don't know exactly what the levers do. Even if somebody who was an expert programmer could come in and 'optimize' the system perfectly, even then it doesn't necessarily mean that the 'perfect system' will have a better sound. So the fun of this is just playing around and having fun as a hobby. Perfect hobby for the curious!

I could be mistaken, but I'm not sure XTRM is limited to specific kernels. There is no XTRM kernel that I can select like I can with a CLTO kernel for example. I think XTRM is more having to do with the build of GP for a specific CPU, but I could be mistaken and if so please somebody correct me. There is not a lot of information posted about the specifics, I don't know if this is intentional. You have to pay money to try it without really knowing its benefits, and if it doesn't sound good then too bad. For my Allo USBridge Target I am running 6.1.54-GentooPlayer-ULTRA-EUP-NOR+ with Profile 3

Just updated my RPI4 HOST to Extreme, but I'm still running the 6.1.77 CLTO Kernal. The latest kernels sound way too bright for my system. I will say that running the same old 6.1.77 kernel but now with EXTRM I notice that the sound is smoother with a better soundstage. I am running my original Diretta settings (Classic) and not the last one (Vivid). The Classis is very similar to Le Dom's configuration since that is what I used originally, and I think overall I prefer something closer to that. I will update my RPI3 Allo USBridge TARGET to EXTRM but still run the old kernal and then begin tweaking. Whenever there is a major version change it always takes me almost a month to get everything sounding right again. I always say I will leave it alone, but now that EXTRM came out for RPI I have to go mix things up again like usual. 😁

Thanks for testing, I'm looking forward to trying these settings after work! I saw that you had something filled in under Interface where mine is blank. Was this a setting you changed and added yourself? I am trying to figure out why mine is blank.

I always had a suspicion that we have similar sounding equipment on the (warm-neutral-bright) scale. I thought about this based on the feedback from the previous years. Usually you and Auke will try some tweaks and you will say they area good for your system but also help me with some feedback. Auke usually says they are too much, but I really appreciate him trying it out! So without ever hearing your systems, I developed an idea of what they might sound like. Then last week I found some settings which were very soft and detailed, but too warm for me on my already warm system. Surprisingly this was the first time Auke liked my setting! 😂 So that is when I said to myself "If it is too warm for me, then it might be good for Auke, but also too warm for PML." I'm using a Luxman 509X which is known for detail and warmth and my DAC is the NewOnTech DAC07 which is also on the neutral-warm side. I still do get a sense of 'air' in recordings, but the air feels like it has a lot of humidity if that makes sense. From everything I've read and heard about Holo, it is exactly my cup of tea as well. I recently installed two media converters and some Cisco fiber cable right where my internet connection comes into the house and it has made a huge difference. Together with the tweaks this past week, I can't imagine how it can sound more realistic. But also whenever I say that, there is always another layer and my mind is blown again, so I'm hoping there is still a long way to go!

It is trial and error. I start with one setting and increase it very high and listen, then decrease it very low and listen. I want to learn what that setting does to the sound at the extremes. Typically the higher some values, the softer the sound will be but will lose detail. It is very much like cooking. Sometimes what tastes good for the first bite might become too much for the whole meal! I would start with using Le Dom's original settings like I did, then find what ThredMode you like. I usually take notes and go one by one up the line (I went to 50 for ThredMode) and then the best ones will fight it out together until I find the one that works. The main thing I focus on is the physical feeling I am getting from the sound. There is an 'ease' or natural lack of 'digititus' when the setting is correct. It is like when you listen to an album compared to streaming, the air feels empty because that digital ghost is gone. I really try to focus on that sensation when I tune my system so that it is exciting but also never has that 'digital feeling' in the air that makes my ears and body uncomfortable. So I'm sorry that there is no data based approach to it. I am sure that if you do one variable a day then you will have a system sounding very dialed in for your tastes in a short amount of time. It is a big upgrade and the worst thing that can happen is that a setting sounds bad and we change it. But at least with every bad sounding setting we learn more about its impact on the sound and whether we should increase or decrease its value in relation to the rest. Good luck and have fun!

Thank you for trying and giving feedback! My system is warm, so even on a neutral system these settings might be too much, but I am at the end for me (for now!) But of course if I change some gear, I will have to come back and tweak more. 😅 I've been looking at the TeraDak T-S212 and have been trying to resist the urge. Looking forward to what the next big software/tweak improvement will be. Over the years whenever I think it can't get better, it always does somehow.

I'm just happy there are fellow audiophiles around the world who truly understand this. 😅 I got a little lost in the weeds with the previous settings, but so happy you mentioned the sound being colored. I approached with fresh ears and more patience. If it wasn't for you and Le Dom, my system and musical enjoyment wouldn't be as good. Great community here!