It would be great if in a future release of Equalify Pro I could select jplay ASIO device and allow playback in core mode with Audiophile optimizer.
Support for ASIO devices is something that isn’t in huge demand.
But it might be possible and i’ll try to do some experiments over the next few months and see if its possible to make spotify work with it.
The audio quality will be identical though as it will receive the exact same raw audio as the normal devices. So there isn’t any gain, what so ever, except being able to use an asio device for playback.
ASIO support due Equalify Pro?
I have to respectfully disagree with your assessment here. While you may be correct that there isn’t enough demand to warrant your development time (this is difficult for me to assess myself). I can assure your there is demand (I just bought this product and am dissapointed there isn’t ASIO support as it forces me to set my DAC’s audio settings down to 44.1KHz to ensure there isn’t upsampling going on). Being able to output bitperfect audio over ASIO to an external DAC is something that more and more people are either doing or interested in. Right now spotify doesn’t offer lossless audio (there are API based projects like fidelify which are interesting) but with competing services like Tidal and Deezer heading that direction I am hopeful there will be pressure for Spotify to follow suit. When you say “it will receive the exact same raw audio as the normal devices” that may or may not be true. By bypassing the Windows mixer and streaming at the native rate it makes it easier for the enthusiast to ensure that what’s coming out of spotify (plus your EQ in this case) is as untouched as possible
As an example, I force 44.1KHz in shared mode when not using another application such as Foobar to direct stream via ASIO (Which will send the native rate to the DAC by definition) - I do this because it’s my assumption that Spotify and this EQ are generally providing output at 44.1KHz and I want to mess with that as little as possible. Also, I set the default windows sounds to go through my monitor’s speakers so the Windows audio mixer isn’t combining other sounds with those of Spotify / Equalify. If that’s not true (for example if Equalify is upsampling before applying its EQ) that would be good to know so one can force shared mode to output at that rate (e.g. 88.2KHz, 176.4KHz, or whatever the case may be)
You are correct, the output from Spotify (and the EQ) is 16bit/44.1kHz.
Remember that i can not change how Spotify handles the sound before its sent to the audio driver.
Spotify uses 320kbps OGG (as their highest setting for premium users), they decompress it to 16bit/44.1kHz PCM, they do their own internal filtering if enabled (normalize/crossfade), then they pass it on and thats where i intercept it. I can then do what i want with the audio before passing it on. In theory i could then have started up another audio output via ASIO in addition to the normal output, and send the audio there, but the audio would be identical and it would have higher latency and i could still not shut down the normal output, since Spotify needs that to function.
ASIO being bit perfect does not matter as long as the input is lossy. As far as i know there are no plans for lossless audio in Spotify in the foreseeable future.
Fidelify may have ASIO support, but it still uses the same lossy ogg audio, so the result will be the same.
All this said, i’m afraid that Spotify just isn’t an enthusiast player. The OGG compression/decompression does alter the audio, but at the highest bitrate its actually very good. ASIO or windows mixer output, noone would ever be able to hear any difference in a blind test, it will sound exactly the same.
Hope that explains why i’m hesitant when it comes to implementing ASIO “support”. There is no way to do it without doing it in a very “hacky” way and it would be less than optimal in every way. And with the very low demand i’m afraid i just do not see the point.
This isn´t exactly True. If i View a Youtube Video with Google Chrome, If i Download the Video and now view it with a “Bitperfect” able Videoplayer like JRiver, I can tell the difference. And it is not a very tiny Difference, I Talk about Youtube Videos with AAC 128-kbit/s . You never get Bitperfect Audio with Windows Direct Audio. This is because The internal Windows Resampler will always convert up to 32 bit float and then Dither down to the bitrate you choose. This means with 44.1khz 16bit set in Windows, you at distortion quite a bit, even more than with 24bit, but with 24bit the added Quantisation Noise is very Low. You at Distortion ether way. But it is Significantly worse if the Windows Audio Kernel does Resampling, lerts say I play a Audio-file with 44.1khz 16bit and set the Windows Resampler to 96khz 24bit. The Windows Resampler Creates a Big amount of Intermodulation Distortion. The IMD+N are -60db at 6000hz / -45db at 10000hz and -30db at 20KHz for this exact Circumstance. The IMD+N Distribution climbs linear to Frequency. With a decent Audio String this is audible. The IMD+N of a 320kbs CBR Lame 3.995 MP3 are Better ( a average Value of -65db IMD + N and more constant/Flat/(the same value) over the Frequency range). The Windows Resampler also use a Lowcut Filter with -2db at15Khz and -5db at20khz. The Windows Audio kernel also has a Buffer who ads about 30ms of delay. I personally would like the integration of ASIO or Wasapi. Equalify would be the only way, that provides bit perfect audio with free Spotify on a Windows platform. Because Fidelify requires a Spotify Premium Account, and its User-Interface is less User Friendly as standard Spotify, so some Premium Spotify Members that use Fidelify will likely change to using Equalify. Sorry for my bad English i am German.
So if I understood correctly the point there is no improvement in the sound quality using Jplay with Fidelity or Equalify?
So better use Equalify using directly the DAC ad output?
Just to know what is worth or not…
I am using Jplay with Foobar2000 and I think it sounds better but honestly I am not sure I would guess it correctly in a blind test…
ASIO support will be nice.