Question about adaptive and Q1

[jarek]

Peter,

what is relation between adaptive mode and Q1 ? So, let's say, I have adaptive mode, and Q1=4. Is it working different when I change Q1 to 5 ? I feel it matters...

[PeterSt]

For now in brief : sure it does. But I don't even know myself anymore how exactly. I think it is so that Q1 = 1 uses the Device Buffer Size for the smallest latency. Set Q1 to 2 and it may be twice that size.
This said differently : Q1 is not much valuable for Adaptive Mode, except for finding the buffer size (from the device) itself. I may have defined Q1 = 2 etc. and it may do something, but it will be "worthless". Unless ... unless a higher Q1 for the same Device Buffer Size creates a better sound.

... And it can be, because things changed since Adaptive was created for the first time.

So you better tell me !
haha

[jarek]

Hi Peter,

I was using latency 32 adn Q1=4, but from time to time cracks are heard in speakers in Adaptive mode. When I moved to Q1=5, they dissapeared. I asked the question, because I thought that function of Adaptive mode was to find optimal Q1 ?

[PeterSt]

Almost true. It is there to find the buffer size of the sound device if you don't know it, and next it can use that without "buffer errors" (which means loosing samples on the way). This is, however, not related to other parts of the system being in stress, and therefore letting hear glitches etc.

If you'd ask me, I honestly would not know the difference between a formally deteced lost buffer, and a glitch (etc.) while the buffer formally is not lost, but it sure is not the same if you look/listen to the result.
Anyway, if you set Q1 higher than the "optimal 1" for the set device buffer size, you allow the system to do more in between, because you highered the latency (and used buffer size).
Something like that ...

Peter

[bgjohan]

Peter

I know you have outlined the procedure to set Adpative mode max sound quality from setting Q1=0, find lowest possible buffer to operate w/out ticks and then reset Q1=1.
I will call this minimize Adaptive buffers for max sound quality, given Q1=1.

Some time ago, I discovered that when changing from Adaptive to Special Q1 changed to a system specific value (in my case Q1=14).
Although Special did not play well on my system, I discovered that when returning to play with Adaptive and Q1=14, it played files wihout ticks at buffer sizes that caused ticks with Q1=1.
My conclusion is that playing with Special (even if not succesful and plays with ticks and noise) re-sets Q1 to a system specific value, where the stress on the PC (processor, memory) is minimized.
I have therefore startef using the following optimization strategy, for a given pre-selected buffer size (I choose 1024, but could be different as per user preference).
1. Set buffers to value to be tested (ex 1024)
2. Play a track in Special mode and let XX set the Q1 to a system specific value (in my case Q1=13).
3. Return to Adaptive mode and try your most demanding setting (in my case to play native 192-24 files or 4x AP upsampling of 48-24 files), with Q1 value as set by Special.
4. If plays OK, try lowering Q1 in stages; if you reach Q1=1, then Peter's target is met, if not yo have still optimized as far as your PC/ DAC combo allows. 


Bjorn