New sound from Creative in new OS from Microsoft
I decided to publish an article for which I received an invitation to Habr, nevertheless I wrote it not for invites, but to pay attention to the problem.
A long time ago, it seems last Friday, or rather a couple of years ago, I was tempted by the beauty of the aero interface and put vista on my computer, on which creative audigy 2 stood then. After a while, listening to music (as usual through foobar2000 ), I noticed that there is something wrong with the sound: the high frequencies sounded somehow too “clean”, but they lacked detail. First of all, I got to google on the topic of changing the sound in whist and did not find anything like it. Strange, I thought, or I have something with hearing or others. Or is it just that nobody uses creative, vista and good acoustics at the same time? :) I decided that I needed to check it somehow to find out the reason, but I didn’t have time for that, and I quickly returned to XP, mentally dumping everything on the raw drivers and the new sound subsystem in whist.
As time went on, I got the X-Fi Titanium (under PCI-E), and then the “seven” arrived in time - I decided to try it all together and ... again it's great - again something is wrong with the sound (despite the fact that the sound is different , although the same company, and the OS has been updated, although again only 6.1, and not 7). Once again it’s useful to google, I stumbled a couple of times on the messages that after installing the timbres in the drivers are set incorrectly. Climbed into the settings, turned down the high frequencies, it seemed to help, but ... it was that “like” - the reboot in XP showed that this was not the case, everything sounded right in it. I decided to get down to business more seriously: I put RMAA 6.2.3 , from the kolkhoz binder and sewing machine, I built a audio cable, one end of which was plugged into the front exit and the other into the line input; I spent my free weekend on a bunch of tests, and this is what came of it.
In the course of the test revealed some details.
Detail 1. Or the cable was completely bad (the binder is not the best material for it) or problems due to the sound output going to its input and there were some parasitic currents to check, otherwise I can’t, the second zvukovuhi no other cable I have. But at least the temporary removal of the cable between tests, in which the sampling rate was changed, greatly influenced the result.
')
Detail 2. Switching the recording source causes the playback settings to be reset, for example, turning on the microphone on the recording turns it on and on playback, although it is turned off in the mixer.
those. like this: if you knock on the microphone - the sound goes to the speakers, although it should not :).
Detail 3. Switching presets (rmaa 44.1, rmaa 48) in a mixer in audio creation mode does not change the settings of the mixer (visually), but it affects the results, especially on intermodulation and frequency response. In general, creative as usual is much better than their software (but even such software is better than their marketing;)).
Detail 4. RMAA in windows 7 behaves extremely unstable: for example, it continues to reproduce test sounds after it finished the recording test or the whole system hangs when changing the clock generator in the panel settings in acm.
Testing conditions:
First, let's see what XP shows us.
As you can see, different frequency response for different modes - the result of hardware oversampling. The choice of interface MME or DirectSound had no effect.
Intermodulation and harmonic distortions appear in “non-native” modes, i.e. when conversion from 44 to 48 kHz is required. This effect is only in acm, it does not appear in gm:
Now 48kHz mode:
The frequency response is impressive ... especially after a flat line of 44.1 kHz. Where is the catch?
Intermodulation distortion also looks weird. Obviously, I somewhere nakosyachil ... or not me. (if you think that having read to the end, you will understand exactly where - you are mistaken :)).
Now Windows 7.
The frequency response looks strange when using MME, the change in volume only managed to reduce vibrations, but the meaning did not change - the “saw” remained in place. A similar result was obtained in XP in the 48kHz mode; moreover, the intermodulation and harmonic distortions also clearly went off scale. I don’t know the reasons for this behavior - apparently there were some peculiarities of the RMAA, there wasn’t anything heard by hearing, and the volume controls didn’t fix it either.
BUT , this is not the strangest thing - look at the top of the range. I will give a comparison with XP right away:
As you can see, after 15 kHz, a gradual decrease in amplitude begins and this is repeated in ALL tests in windows 7, when the sound source gives 44.1 kHz.
Now look at 48 kHz:
It's ok. As well as with distortion - the problem is only in 44.1 kHz.
Foobar2000 uses the DirectSound interface. Before using X-Fi (in which there is hardware resampling from 44.1 kHz to 48), I used a plugin (resampling) built into the player. I checked it and this time, turned on 48kHz and ... everything with sound became normal in windows 7.
Total:
At a minimum, I hope that this scribbling will encourage someone to further investigate the problem, and at the maximum, will push creative to solve the problem with the drivers in the new OS.
Just in case, the archive of all tests.
PS
It turned out that the same problem exists on some embedded realtek, but there are no problems with ASUS in the new windows.
A long time ago, it seems last Friday, or rather a couple of years ago, I was tempted by the beauty of the aero interface and put vista on my computer, on which creative audigy 2 stood then. After a while, listening to music (as usual through foobar2000 ), I noticed that there is something wrong with the sound: the high frequencies sounded somehow too “clean”, but they lacked detail. First of all, I got to google on the topic of changing the sound in whist and did not find anything like it. Strange, I thought, or I have something with hearing or others. Or is it just that nobody uses creative, vista and good acoustics at the same time? :) I decided that I needed to check it somehow to find out the reason, but I didn’t have time for that, and I quickly returned to XP, mentally dumping everything on the raw drivers and the new sound subsystem in whist.
As time went on, I got the X-Fi Titanium (under PCI-E), and then the “seven” arrived in time - I decided to try it all together and ... again it's great - again something is wrong with the sound (despite the fact that the sound is different , although the same company, and the OS has been updated, although again only 6.1, and not 7). Once again it’s useful to google, I stumbled a couple of times on the messages that after installing the timbres in the drivers are set incorrectly. Climbed into the settings, turned down the high frequencies, it seemed to help, but ... it was that “like” - the reboot in XP showed that this was not the case, everything sounded right in it. I decided to get down to business more seriously: I put RMAA 6.2.3 , from the kolkhoz binder and sewing machine, I built a audio cable, one end of which was plugged into the front exit and the other into the line input; I spent my free weekend on a bunch of tests, and this is what came of it.
In the course of the test revealed some details.
Detail 1. Or the cable was completely bad (the binder is not the best material for it) or problems due to the sound output going to its input and there were some parasitic currents to check, otherwise I can’t, the second zvukovuhi no other cable I have. But at least the temporary removal of the cable between tests, in which the sampling rate was changed, greatly influenced the result.
')
Detail 2. Switching the recording source causes the playback settings to be reset, for example, turning on the microphone on the recording turns it on and on playback, although it is turned off in the mixer.
those. like this: if you knock on the microphone - the sound goes to the speakers, although it should not :).
Detail 3. Switching presets (rmaa 44.1, rmaa 48) in a mixer in audio creation mode does not change the settings of the mixer (visually), but it affects the results, especially on intermodulation and frequency response. In general, creative as usual is much better than their software (but even such software is better than their marketing;)).
Detail 4. RMAA in windows 7 behaves extremely unstable: for example, it continues to reproduce test sounds after it finished the recording test or the whole system hangs when changing the clock generator in the panel settings in acm.
Testing conditions:
- The recording level is 50% (+ 0db), the wave level is 100% / 75% (+ 0db) for acm (audio creation mode) and 100% wave for gm (game mode), the other sources are 0. The test recording level was adjusted by the overall system volume ( 90-96%). The front output was looped (loopback) to the linear (aka mic) input. All questionable results were checked with different levels of recording and reproduction, but this did not bring fundamental differences in the results.
- Windows XP sp3, driver 2. 17. 7
- Windows 7 RC 7106, driver 2. 17. 7
First, let's see what XP shows us.
As you can see, different frequency response for different modes - the result of hardware oversampling. The choice of interface MME or DirectSound had no effect.
Intermodulation and harmonic distortions appear in “non-native” modes, i.e. when conversion from 44 to 48 kHz is required. This effect is only in acm, it does not appear in gm:
Now 48kHz mode:
The frequency response is impressive ... especially after a flat line of 44.1 kHz. Where is the catch?
Intermodulation distortion also looks weird. Obviously, I somewhere nakosyachil ... or not me. (if you think that having read to the end, you will understand exactly where - you are mistaken :)).
Now Windows 7.
The frequency response looks strange when using MME, the change in volume only managed to reduce vibrations, but the meaning did not change - the “saw” remained in place. A similar result was obtained in XP in the 48kHz mode; moreover, the intermodulation and harmonic distortions also clearly went off scale. I don’t know the reasons for this behavior - apparently there were some peculiarities of the RMAA, there wasn’t anything heard by hearing, and the volume controls didn’t fix it either.
BUT , this is not the strangest thing - look at the top of the range. I will give a comparison with XP right away:
As you can see, after 15 kHz, a gradual decrease in amplitude begins and this is repeated in ALL tests in windows 7, when the sound source gives 44.1 kHz.
Now look at 48 kHz:
It's ok. As well as with distortion - the problem is only in 44.1 kHz.
Foobar2000 uses the DirectSound interface. Before using X-Fi (in which there is hardware resampling from 44.1 kHz to 48), I used a plugin (resampling) built into the player. I checked it and this time, turned on 48kHz and ... everything with sound became normal in windows 7.
Total:
- The reason for the deterioration of sound in Windows 7 and Vista in the cards creative is obvious and visible on the penultimate graph, but who is to blame for this behavior zvukovuhi? In any case, if you are using windows vista or 7 and soundblaster, use a software resampler for 44.1 kHz sources whenever possible. Or use XP, in it and with EAX problems less.
- I didn’t find out the reasons for “drinking” and off-scale distortions - I hope there are people who will find out. I repeat, they are not audible, and they appear only in tests.
At a minimum, I hope that this scribbling will encourage someone to further investigate the problem, and at the maximum, will push creative to solve the problem with the drivers in the new OS.
Just in case, the archive of all tests.
PS
It turned out that the same problem exists on some embedded realtek, but there are no problems with ASUS in the new windows.
Source: https://habr.com/ru/post/73391/
All Articles