Curious:
Did you verify that the levels of all outputs were equal ?
if yes, how close were they ?
And how did you perform the "switching" between the devices being compared ?
was it "instantaneous A/B" ?
blind ?
(I get fooled a lot by those two aspects, when I perform comparisons )
^ Good point.
In that case, this is just another listening impression post which is only useful for the listeners who were there.
Why, because Jam's audio business sells DACs?
What test method revealed that?
Well, the DAC module doesnt output current at the outputs does it? I havent investigated the board in any detail, as it doesnt particularly interest me for various reasons and discrete dacs are a rabbithole a multichannel enthusiast should probably avoid; unless designing and building it yourself; or have a few dB more disposable income to waste than I do, then fully characterise, match, model, compensate and trim out each discrete network.
I dont personally see the point. I have other things i'd rather be doing. there is a reason laser trimming was deemed the appropriate manufacture technology moving forward a few decades ago. but whatever the exact situation you cannot easily separate it like the others. My point being, that taking care of all of that and employing an identical voltage follower/buffer would be as close as you might get, and even that isnt really close enough.
I have said all I need to say already on this forum about the DIYINHK board. Not a fair fight at all and ultimately worthy of curiosity only. Fun for all involved on the day, but no use to anyone else really IMHO.
Agreed. All listening tests are, by definition, subjective, and a consensus over which component is 'best' even in a properly controlled and executed test (which this one does not appear to be) does make it objectively so.
My apologies for the word salad above. I clearly got a bit excited with the paste command before bed. Edit window is closed. it should have read.
Well, the DAC module doesnt output current at the outputs does it? I havent investigated the board in any detail, as it doesnt particularly interest me for various reasons and discrete dacs are a rabbithole a multichannel enthusiast should probably avoid; unless designing and building it yourself; or have a few dB more disposable income to spend than I do. Fully characterising, matching, modelling, compensating and trimming out each discrete network is a lot of work and I just struggle with why? Seems to me there is tastier and lower hanging fruit.
Well, the DAC module doesnt output current at the outputs does it? I havent investigated the board in any detail, as it doesnt particularly interest me for various reasons and discrete dacs are a rabbithole a multichannel enthusiast should probably avoid; unless designing and building it yourself; or have a few dB more disposable income to spend than I do. Fully characterising, matching, modelling, compensating and trimming out each discrete network is a lot of work and I just struggle with why? Seems to me there is tastier and lower hanging fruit.
Last edited:
Just for the sake of general interest, we have listened to DAC_Lite with and without buffering. Have also compared it to Topping D90, and to a custom AK4499 DAC with hardware conversion to DSD256. Maybe the stiffest competition DAC_Lite has gone up against so far, not sure. One reason I would like to hold off describing the sound for now is because I'm not sure how the DAC_Lite Vref supply here compares to the Vref supplies used for previous listening comparisons (Vref characteristics can have an influence on the sound). Also as it happens another reason for waiting is because Andrea is traveling though California on vacation and it works out he can stop by Auburn for a few short hours on one day. Would like Andrea to have a chance to form his own impressions using our Vref supply and his impression of the other DACs here before we then compare notes. When there is more to share I will post here.
Thanks for checking this out. It would be cool if there are locals with similar R2R discrete compensated DACs for a direct comparison. A Lavry Gold or Holo would be interesting.
Is the DAC_Lite expensive by the time all the modules, power supplies and case are added up?
Meaning that they claim? There are tons of expensive audio electronics tagged with all sorts of claims. They are also known as snake oil.
That's something I would like to do eventually. For now I have agreed to not to get into detailed specifics. In general terms I designed a board to do clocking, DSP, and to accept I2S/SPDIF/TOSLINK inputs. The I2S input of that board is fed by an externally clocked I2SoverUSB. My board's I2S output feeds a modified AK4499 evaluation board which includes a custom discrete Vref supply, and custom discrete output stage (except IC opamps are still used for IV, as there didn't seem to be a better solution). For asynchronous USB input, all the clocked circuitry is synchronous. For now at least SPDIF/TOSLINK use ASRC for low latency. AK4499 operates in DSD volume bypass mode. Other than that, I have written about some other things to consider (say for example, layout, bypassing, regulator loading, etc), in threads scattered around in the forum. I will stop here for now.
Measurements don't contain any IP so what's keeping you from showing some measurements of your 4499 dac?
As you may have noticed I was not depicted on the far right side of the graphic below. IME focusing on measurements can be a distraction for listening dacs. For those who want to focus on measurement dacs, there are other threads.
Attachments
Last edited:
His boss may get upset if he does. If you are developing a product for future release, you wouldn't want to spoil it beforehand, would you?💡
We must have some pre-defined set of specifications that a particular audio device should meet to be deemed a good sounding. How else would you get an assurance that a device will sound good and be of great quality irrelevant of who is listing to it, what the quality of other sound system components are, and what the particular room acoustics are?
I encourage research, development, and design. However, the aim should be to achieve exceptional specifications set before starting to tune the sound by ear.
- Status
- Not open for further replies.