Author Topic: What causes the 250msec delay in Spectrum visualizer and on the Waveform cursor?  (Read 4053 times)

peewee678

  • Full Member
  • ***
  • Posts: 142
There's a delay of the waveform cursor and the spectum visualizer relative to the music I'm hearing.
The delay in the visualizer is about 250msec while the waveform cursor has a delay of about a second or so.

Am I the only one experiencing this? What can I do to prevent this?
Last Edit: April 06, 2016, 01:47:13 PM by peewee678

peewee678

  • Full Member
  • ***
  • Posts: 142
Bump ....because earlier I thought this problem was due to the buffer but now I think something else is wrong.. (I changed the topic title and my question)
Last Edit: April 06, 2016, 05:03:05 PM by peewee678

Steven

  • Administrator
  • Hero Member
  • *****
  • Posts: 33483
if you are using DirectSound output then there is a delay between when sound is generated and when you hear it. I guess i could make an adjustment but something like that should be on the wishlist

peewee678

  • Full Member
  • ***
  • Posts: 142
Thanks Steven.

I already tried WASAPI earlier with the same results so there's not much left except ASIO (but I'm afraid with ASIO I will have other problems with other sound output like my browser's or media player classic; I once used this PC also for my music production (I now use a laptop for that) and I remember I had to constantly switch between ASIO and "default audio").

I guess most of you guys are using MotherBoard sound or a sound card... I'm using a DAC so I guess that's why my choices are limited.

So yeah, I will put this on the wishlist but if you or someone else have other tips, please feel free to share :-)
Last Edit: April 06, 2016, 06:02:03 PM by peewee678

Steven

  • Administrator
  • Hero Member
  • *****
  • Posts: 33483
i have done a bit of analysis on this. It turns out with WASPI there is a different issue coming into play, where for a reason i dont understand yet, the same data is being returned 3 times in a row. So that would explain the slowness that people report if they use WASPI.
DirectSound doesnt have that problem.
For the discrepency between hearing and seeing, according to the bass documentation the data should represent whats being header i guess that means its when the data has been sent by the bass library to the output device.
So can you confirm the delay you see is the visualiser is ahead of what you hear or is behind what you hear?
The waveform is probably a different issue but i havent looked at that yet

peewee678

  • Full Member
  • ***
  • Posts: 142
Thanks.

I checked it again and I see my earlier description was actually wrong; it's the other way around: first I see the the visualizer peaking and a few hundred msec later I hear that peek in the audio (to be sure about the order I tried this at the start of some tracks ofcourse). There's no difference between Direct Sound and WASAPI however. This has never been the case.

FYI: with regards to the waveform: I see the cursor passing a peek in the music and somewhat later I hear this peak.


hiccup

  • Hero Member
  • *****
  • Posts: 6382
i have done a bit of analysis on this. It turns out with WASPI there is a different issue coming into play, where for a reason i dont understand yet, the same data is being returned 3 times in a row. So that would explain the slowness that people report if they use WASPI.
DirectSound doesnt have that problem.
For the discrepency between hearing and seeing, according to the bass documentation the data should represent whats being header i guess that means its when the data has been sent by the bass library to the output device.
So can you confirm the delay you see is the visualiser is ahead of what you hear or is behind what you hear?
The waveform is probably a different issue but i havent looked at that yet

F.w.i.w., on my system:
- ASIO:  spectrum is slightly (negligible) behind the sound.
- DirectSound:  spectrum is slightly (negligible) ahead of the sound.
- Wasapi:  spectrum is clearly ahead of the sound. (rough guess: around 200 ms) Also at some occasions the whole spectrum appears to get slower and less responsive (convulsed, oversaturated?) with very dynamic content.

Steven

  • Administrator
  • Hero Member
  • *****
  • Posts: 33483
i have done a bit of analysis on this. It turns out with WASPI there is a different issue coming into play, where for a reason i dont understand yet, the same data is being returned 3 times in a row. So that would explain the slowness that people report if they use WASPI.
DirectSound doesnt have that problem.
For the discrepency between hearing and seeing, according to the bass documentation the data should represent whats being header i guess that means its when the data has been sent by the bass library to the output device.
So can you confirm the delay you see is the visualiser is ahead of what you hear or is behind what you hear?
The waveform is probably a different issue but i havent looked at that yet

F.w.i.w., on my system:
- ASIO:  spectrum is slightly (negligible) behind the sound.
- DirectSound:  spectrum is slightly (negligible) ahead of the sound.
- Wasapi:  spectrum is clearly ahead of the sound. (rough guess: around 200 ms) Also at some occasions the whole spectrum appears to get slower and less responsive (convulsed, oversaturated?) with very dynamic content.
that is mostly consistent with what i have found. For WASAPI, its because the bass library is returning the same data 3 cycles and that makes the visualiser look slow to react. As to why the data is behaving that way, i dont know just yet.

hiccup

  • Hero Member
  • *****
  • Posts: 6382
I just noticed that on the bass website basswasapi v. 2.4.1.2 is offered (mb includes 2.4.1.0), but just to let you know, trying that one out, it makes no difference in this respect.

Steven

  • Administrator
  • Hero Member
  • *****
  • Posts: 33483
the issue i raised with WASAPI refreshing only every 3 cycles was actually related to using WASAPI in "event mode" which is not available in v2.x or v3.0