kradcliffe
Agreed. 1ms is crazy and there's no need for that setting. Probably crashing as your system is not capable of achieving that from the driver.
I use WASAPI now but when I had the ASIO drivers on the MC6000 I seem to remember 15ms worked fine, including using the jogs for scratching.
I use WASAPI now but when I had the ASIO drivers on the MC6000 I seem to remember 15ms worked fine, including using the jogs for scratching.
Mensajes Tue 09 Feb 16 @ 3:39 pm
PressNPlayDJ
kradcliffe wrote :
Agreed. 1ms is crazy and there's no need for that setting. Probably crashing as your system is not capable of achieving that from the driver.
I use WASAPI now but when I had the ASIO drivers on the MC6000 I seem to remember 15ms worked fine, including using the jogs for scratching.
I use WASAPI now but when I had the ASIO drivers on the MC6000 I seem to remember 15ms worked fine, including using the jogs for scratching.
Depending the machine...i have been using forever 1ms with no issues. If anybody is getting Blue Screen Of Death its due to nvidia drivers with windows 10...Simply install an older nvidia driver. I had the same issue on my 970m nvidia card and driver 361...rolled back to 347....
Mensajes Tue 09 Feb 16 @ 8:10 pm
PhantomDeejay
@PressNPlayDJ: You seem to be a helpful member with some nice knowledge. However 1ms latency was simply unrealistic a few years ago, and it is a very high risk now even for the most powerful systems.
Let's do some theory:
1) What do we call latency ?
We call latency the time it takes for the sound to travel from your audio software to the soundcard's DAC (Digital-to-Analog Converter)
So, in order to understand latency, we need to understand how the audio travels from the software to the DAC's which is quite simple actually:
The software writes the data to be played as a sound on RAM, the processor reads them, processes them and finally sends them on the sound interface. So, what's effectively moving the sound from the software to the sound interface is the CPU.
Let's take a brake here: Modern computers are designed to execute multiple tasks at once. That's what makes playing music with a computer possible in the first place since it's the CPU that reads the data from your hard disk, it's the CPU that synthesizes the sound within the software, apply effects, e.t.c and it's the CPU that finally makes the software to write the data to be played into RAM. All these actions are different tasks for the CPU. If the CPU didn't had to do other tasks but just read the data from RAM and push them to your sound interface then you wouldn't need buffers and we would not had any latency at all.
2) Buffers ? What's this ?
Well as we said the CPU has multiple tasks to execute. And since it cannot execute everything at once it has a queue. Each task waits on the queue until the CPU has free time to deal with it. That means that the CPU cannot constantly process the data of our software. It has to do it in parts. So the software writes small chunks of data that the CPU processes one at a time and then it keeps doing it's other tasks. That's what we call buffers: These small chunks of data.
3) How does the size of buffer affects sound ?
Well, let's take an example:
You and your friend are trying to fill a pool with water but you only have two buckets. You stand in the faucet and you fill the buckets with water while your friend has to carry the buckets from the faucet to the pool which is 50 feet away, empty them and come back in time before you finish filling the next bucket.
The bigger the bucket is, the more time you need to fill it and thus the more time your friend has to go in the pool and come back. He may even start chit-chating with the girl next door while he's going back and forth. So, you can assume that your friend will always be back in time to bring you the empty bucket and receive the full one.
Now if we replace the buckets with something smaller like a pitcher, you will need far less time to fill them, but your friend also will have much less time to go to the pull empty the pitcher and come back in time. Therefore he starts running.
If we replace the pitchers with glasses then you can fill them really quick. But your friend despite the fact that he's running as fast as he can, he can't catch up with you and you start loosing water.
Finally while he tries to make as fast as he can he stumbles, falls down and brakes the glass.
As you can imagine, you are the software, the bucket/pitcher/glass is the buffer, and your friend is the CPU.
If it wasn't for the fact that big buffers cause bigger latency (delays) when we try to manipulate the sound in real time since we are dj's then we would all use big buffers the same way media players do.
4) So what ? I have a 4th generation i7 processor which is extremely powerful and fast and therefore I can use very small buffers!
Hmmm. Maybe, or maybe not!
In the above example it was only you and your friend. If we want to make the above example more realistic then your friend would have to carry water to the pool from 10 different persons, each person could possibly use a different size bucket, and finally he would also have to do any chores for the ladies of the company.
What I'm trying to say here is that CPU is not focused only on serving one particular task in due time.
CPU has a list of tasks to do and it goes by doing them one by one. Of course each task on the queue has a priority flag. So tasks with "high priority" flag are served first. However, we can't assume that our task (to move the audio) will be the only one with the "high priority" flag all the time. If several tasks exist with the "high priority" flag then these tasks are waiting in queue as well. But even if our task is the only one with "high priority" flag, there's always a chance that the previous task take too much time to complete and therefore execution of our task remains on hold.
That's why a fast CPU can improve things, but is not the only thing that matters. Other things we need to consider are the OS itself and the drivers of our hardware. All these demand CPU power as well and all their tasks are market as "critical priority" This means that most of the time they will get ahead of our own task in the queue of the CPU, and if one of them takes too much time to complete it will make our own task fail.
5) What happens if the task waits for too long before it gets executed by the CPU ?
In simple words the sound interface will run out of data at some point before the next packet arrives and that will be heard as an audio drop out (sound stops and then resumes) or audible clicks / pops
6) General Remarks:
While CPU is responsible for everything, a fast CPU is not enough for a system to be capable to play back properly with low latency. Most of the time it's the drivers of various components (like Wireless cards, GPU's, even ACPI power management) that are responsible for audio drop outs on systems with small buffers. All the systems of a PC/Laptop consume CPU time and some of them can start a heavy task at the wrong time. Also I avoid to add USB kernels on the equation because it would get too complicate to explain, but that's another component that could affect how small buffers you can set on a system without risking. Also USB kernels have another bad thing: They can make the system crash!
So finally back to your case, if you understand how it all works then you are basically using a 64 samples buffer (which actually is 1.6ms @44.100Hz or 0.75ms @48KHz) and you ask from your entire system to be fast enough without any "bad" or unpredictable drivers to be able to compensate with that.
All researches agree that most people can't understand audio latency if it's lower than 12ms. A very few people claim to be able to understand it down to 6ms. From that point on no-one can understand a difference. Therefore even if you belong on that special case of people who can understand latency down to 6ms, from 6ms to 1 ms is an unnecessary risk.
As a last word: As Dj's we seek to achieve the lowest latency possible only when using DVS. That's because in that case the latency is double. You have the "input" latency (time it takes from sound-interface to software) and "output latency" (time from software to sound interface). So in that case we want to have 6ms input and 6ms output latency (or better) in order to be lower than the 12ms that's noticeable for most people. As a best case scenario we would be trying to achieve 3+3ms if possible and of course we would not try to go lower than that.
Let's do some theory:
1) What do we call latency ?
We call latency the time it takes for the sound to travel from your audio software to the soundcard's DAC (Digital-to-Analog Converter)
So, in order to understand latency, we need to understand how the audio travels from the software to the DAC's which is quite simple actually:
The software writes the data to be played as a sound on RAM, the processor reads them, processes them and finally sends them on the sound interface. So, what's effectively moving the sound from the software to the sound interface is the CPU.
Let's take a brake here: Modern computers are designed to execute multiple tasks at once. That's what makes playing music with a computer possible in the first place since it's the CPU that reads the data from your hard disk, it's the CPU that synthesizes the sound within the software, apply effects, e.t.c and it's the CPU that finally makes the software to write the data to be played into RAM. All these actions are different tasks for the CPU. If the CPU didn't had to do other tasks but just read the data from RAM and push them to your sound interface then you wouldn't need buffers and we would not had any latency at all.
2) Buffers ? What's this ?
Well as we said the CPU has multiple tasks to execute. And since it cannot execute everything at once it has a queue. Each task waits on the queue until the CPU has free time to deal with it. That means that the CPU cannot constantly process the data of our software. It has to do it in parts. So the software writes small chunks of data that the CPU processes one at a time and then it keeps doing it's other tasks. That's what we call buffers: These small chunks of data.
3) How does the size of buffer affects sound ?
Well, let's take an example:
You and your friend are trying to fill a pool with water but you only have two buckets. You stand in the faucet and you fill the buckets with water while your friend has to carry the buckets from the faucet to the pool which is 50 feet away, empty them and come back in time before you finish filling the next bucket.
The bigger the bucket is, the more time you need to fill it and thus the more time your friend has to go in the pool and come back. He may even start chit-chating with the girl next door while he's going back and forth. So, you can assume that your friend will always be back in time to bring you the empty bucket and receive the full one.
Now if we replace the buckets with something smaller like a pitcher, you will need far less time to fill them, but your friend also will have much less time to go to the pull empty the pitcher and come back in time. Therefore he starts running.
If we replace the pitchers with glasses then you can fill them really quick. But your friend despite the fact that he's running as fast as he can, he can't catch up with you and you start loosing water.
Finally while he tries to make as fast as he can he stumbles, falls down and brakes the glass.
As you can imagine, you are the software, the bucket/pitcher/glass is the buffer, and your friend is the CPU.
If it wasn't for the fact that big buffers cause bigger latency (delays) when we try to manipulate the sound in real time since we are dj's then we would all use big buffers the same way media players do.
4) So what ? I have a 4th generation i7 processor which is extremely powerful and fast and therefore I can use very small buffers!
Hmmm. Maybe, or maybe not!
In the above example it was only you and your friend. If we want to make the above example more realistic then your friend would have to carry water to the pool from 10 different persons, each person could possibly use a different size bucket, and finally he would also have to do any chores for the ladies of the company.
What I'm trying to say here is that CPU is not focused only on serving one particular task in due time.
CPU has a list of tasks to do and it goes by doing them one by one. Of course each task on the queue has a priority flag. So tasks with "high priority" flag are served first. However, we can't assume that our task (to move the audio) will be the only one with the "high priority" flag all the time. If several tasks exist with the "high priority" flag then these tasks are waiting in queue as well. But even if our task is the only one with "high priority" flag, there's always a chance that the previous task take too much time to complete and therefore execution of our task remains on hold.
That's why a fast CPU can improve things, but is not the only thing that matters. Other things we need to consider are the OS itself and the drivers of our hardware. All these demand CPU power as well and all their tasks are market as "critical priority" This means that most of the time they will get ahead of our own task in the queue of the CPU, and if one of them takes too much time to complete it will make our own task fail.
5) What happens if the task waits for too long before it gets executed by the CPU ?
In simple words the sound interface will run out of data at some point before the next packet arrives and that will be heard as an audio drop out (sound stops and then resumes) or audible clicks / pops
6) General Remarks:
While CPU is responsible for everything, a fast CPU is not enough for a system to be capable to play back properly with low latency. Most of the time it's the drivers of various components (like Wireless cards, GPU's, even ACPI power management) that are responsible for audio drop outs on systems with small buffers. All the systems of a PC/Laptop consume CPU time and some of them can start a heavy task at the wrong time. Also I avoid to add USB kernels on the equation because it would get too complicate to explain, but that's another component that could affect how small buffers you can set on a system without risking. Also USB kernels have another bad thing: They can make the system crash!
So finally back to your case, if you understand how it all works then you are basically using a 64 samples buffer (which actually is 1.6ms @44.100Hz or 0.75ms @48KHz) and you ask from your entire system to be fast enough without any "bad" or unpredictable drivers to be able to compensate with that.
All researches agree that most people can't understand audio latency if it's lower than 12ms. A very few people claim to be able to understand it down to 6ms. From that point on no-one can understand a difference. Therefore even if you belong on that special case of people who can understand latency down to 6ms, from 6ms to 1 ms is an unnecessary risk.
As a last word: As Dj's we seek to achieve the lowest latency possible only when using DVS. That's because in that case the latency is double. You have the "input" latency (time it takes from sound-interface to software) and "output latency" (time from software to sound interface). So in that case we want to have 6ms input and 6ms output latency (or better) in order to be lower than the 12ms that's noticeable for most people. As a best case scenario we would be trying to achieve 3+3ms if possible and of course we would not try to go lower than that.
Mensajes Tue 09 Feb 16 @ 10:57 pm
Adion
I remember 64 samples being possible over 10 years ago with the RME sound cards, but with most computers/audio drivers it's indeed not easy to achieve without any drop outs.
Mensajes Wed 10 Feb 16 @ 4:45 am
PressNPlayDJ
@Phantom
I understand how latency works. Really cool that you took some time to write all this. I recommend you pin it somewhere on the forum so everybody can understand this concept. However, in the case of my friend bringing water to fill up the pool, I have the Flash! And The Flash isn't filling up a pool with water but trying to save a building that is on fire with people in it LOL. My point, If you are going to use a computer for DJing, 1 ms latency is the way to go. Its not about if the crowd will notice it but you and how you feel overall. I used to use my old Laptop at 5ms until a friend who uses serato suggested to put at 1 ms. Personally, it really did make a difference. When scratching and using effects you want the latency tight. I heard that they are trying to make the latency even tighter which really crazy but will not be noticeable. If your PC is optimized for DJing you won't have issues with latency being at 1ms. Optimized meaning you have installed only one program! Virtual DJ! Optimzed meaning, nothing else is running or installed on that pc. Optimized meaning, not even hard drives are connected and all music is on a internal hard drive. This is the mistake a lot of people make. They want to DJ and want to use a laptop with a software (doesn't matter which) but they buy a cheap pc and use for dropbox, netflix, facebook, gaming.... It doesn't work like that...Would you use your cdj nexus to play music at home with your home system? No...You would use an ipad or any tablet...Same things goes for laptops. It should be strictly optimized to run your one and only desired dj software.
@Adion it is achievable as long as DJS invest more money into more powerful machines. I see Djs using 4gb ram and i3 cores and expect to have max performance...LOL. I updated from a i7 2,2 Windows 7 SONY VAIO 8GB RAM and 2GB NVIDIA to a Dell Alienware i7 2,6 Windows 10 16GB RAM 3GB NVIDIA and SSD hard drive running the OS and ONLY Virtual DJ just to get max performance and which i do. I record video at HQ with single video deck effects, milkdrop effects and tellyvisuals with the CPU hitting max at 40% without even a glitch.
ALSO to the OP and the rest: doing a DPC latency check helps a lot to understand how and what is using your CPU http://www.thesycon.de/deu/latency_check.shtml
I understand how latency works. Really cool that you took some time to write all this. I recommend you pin it somewhere on the forum so everybody can understand this concept. However, in the case of my friend bringing water to fill up the pool, I have the Flash! And The Flash isn't filling up a pool with water but trying to save a building that is on fire with people in it LOL. My point, If you are going to use a computer for DJing, 1 ms latency is the way to go. Its not about if the crowd will notice it but you and how you feel overall. I used to use my old Laptop at 5ms until a friend who uses serato suggested to put at 1 ms. Personally, it really did make a difference. When scratching and using effects you want the latency tight. I heard that they are trying to make the latency even tighter which really crazy but will not be noticeable. If your PC is optimized for DJing you won't have issues with latency being at 1ms. Optimized meaning you have installed only one program! Virtual DJ! Optimzed meaning, nothing else is running or installed on that pc. Optimized meaning, not even hard drives are connected and all music is on a internal hard drive. This is the mistake a lot of people make. They want to DJ and want to use a laptop with a software (doesn't matter which) but they buy a cheap pc and use for dropbox, netflix, facebook, gaming.... It doesn't work like that...Would you use your cdj nexus to play music at home with your home system? No...You would use an ipad or any tablet...Same things goes for laptops. It should be strictly optimized to run your one and only desired dj software.
@Adion it is achievable as long as DJS invest more money into more powerful machines. I see Djs using 4gb ram and i3 cores and expect to have max performance...LOL. I updated from a i7 2,2 Windows 7 SONY VAIO 8GB RAM and 2GB NVIDIA to a Dell Alienware i7 2,6 Windows 10 16GB RAM 3GB NVIDIA and SSD hard drive running the OS and ONLY Virtual DJ just to get max performance and which i do. I record video at HQ with single video deck effects, milkdrop effects and tellyvisuals with the CPU hitting max at 40% without even a glitch.
ALSO to the OP and the rest: doing a DPC latency check helps a lot to understand how and what is using your CPU http://www.thesycon.de/deu/latency_check.shtml
Mensajes Wed 10 Feb 16 @ 6:52 am
PhantomDeejay
Actually you should use LatencyMon ( http://www.resplendence.com/latencymon ) instead of DPC.
Also if you take a good brief at what LatencyMon yells out you would understand that it's not about power or having "Flash" doing the job. It's about all the components on your system playing nice with each other.
I'm happy you have such a powerful laptop and the risk is smaller for you. But you don't need a Ferrari in order to go for a ride in the city. You can go with a Ferrari, but it's not mandatory.
Same rule goes for PC's and Dj'ing. You don't need an i7 to Dj with software. You can, but it's not mandatory.
Also tweaking a system and especially a laptop is not always an easy task. It needs time, patience and knowledge. And even then results are not guaranteed. Three years ago I had a Toshiba Qosmio laptop. It was an i7 with 8GB RAM, 1GB dedicated nVidia GPU e.t.c. It was a "Ferrari" back at it's time. However the latency on that laptop could not go lower than 7ms. No matter how much I tweaked it, not matter how much I stripped down the OS, no matter what. It was because of the various components of the laptop itself that the DPC latency remained high enough to prevent the audio latency go lower than 7ms. For instance the GPU would create a spike every 40 seconds with 5000ms (5sec) execution time. There was nothing I could do about that, and you can't replace GPU on a laptop just like that! However that laptop served me perfectly at 7ms latency. Then I got my ASUS ROG. Without any tweaks the system components play nice among each other and I can lower the latency to unrealistic values like 1-2ms. However I choose to run at 5ms.
As for latency, at least we agree that we disagree. NO HUMAN being can understand a difference between 1ms and 5ms latency. And I'm not talking about your crowd, they wouldn't understand latency even if it was 10 seconds since they don't manipulate the audio. I'm talking about you who manipulate the audio. Dj's are demanding people (LOL) but musicians are even more. When musicians can't tell the difference when using soft synths or drum pads in 6ms versus 1ms, I tend to trust them more.
As someone who works constantly with various hardware and as someone who talks on behalf of Atomix I cannot advise anyone go lower than 4-5ms latency and especially with DENON ASIO drivers. It's just asking for troubles. No matter how fast or how powerful their system is.
As for you, you are free to use whatever value suits you. Honestly I think you are just experiencing a placebo effect, but even if you are not and I'm wrong, 1ms is not a value most systems can cope with and therefore you should not advise others to use this value, at least not without telling them the risks it involves. :)
@Adion: Yes, RME could technically go very low but still you needed a very fine tweaked system, especially when you started building up a track with a lot of soft synths and effects e.t.c. Despite being able to go as low as 64 samples most studios I know usually used them at 256 or 128 samples and honestly those cards (among with a few others like some Apogee's or MOTU's) were just a few bright exceptions on the basic rule!
Also if you take a good brief at what LatencyMon yells out you would understand that it's not about power or having "Flash" doing the job. It's about all the components on your system playing nice with each other.
I'm happy you have such a powerful laptop and the risk is smaller for you. But you don't need a Ferrari in order to go for a ride in the city. You can go with a Ferrari, but it's not mandatory.
Same rule goes for PC's and Dj'ing. You don't need an i7 to Dj with software. You can, but it's not mandatory.
Also tweaking a system and especially a laptop is not always an easy task. It needs time, patience and knowledge. And even then results are not guaranteed. Three years ago I had a Toshiba Qosmio laptop. It was an i7 with 8GB RAM, 1GB dedicated nVidia GPU e.t.c. It was a "Ferrari" back at it's time. However the latency on that laptop could not go lower than 7ms. No matter how much I tweaked it, not matter how much I stripped down the OS, no matter what. It was because of the various components of the laptop itself that the DPC latency remained high enough to prevent the audio latency go lower than 7ms. For instance the GPU would create a spike every 40 seconds with 5000ms (5sec) execution time. There was nothing I could do about that, and you can't replace GPU on a laptop just like that! However that laptop served me perfectly at 7ms latency. Then I got my ASUS ROG. Without any tweaks the system components play nice among each other and I can lower the latency to unrealistic values like 1-2ms. However I choose to run at 5ms.
As for latency, at least we agree that we disagree. NO HUMAN being can understand a difference between 1ms and 5ms latency. And I'm not talking about your crowd, they wouldn't understand latency even if it was 10 seconds since they don't manipulate the audio. I'm talking about you who manipulate the audio. Dj's are demanding people (LOL) but musicians are even more. When musicians can't tell the difference when using soft synths or drum pads in 6ms versus 1ms, I tend to trust them more.
As someone who works constantly with various hardware and as someone who talks on behalf of Atomix I cannot advise anyone go lower than 4-5ms latency and especially with DENON ASIO drivers. It's just asking for troubles. No matter how fast or how powerful their system is.
As for you, you are free to use whatever value suits you. Honestly I think you are just experiencing a placebo effect, but even if you are not and I'm wrong, 1ms is not a value most systems can cope with and therefore you should not advise others to use this value, at least not without telling them the risks it involves. :)
@Adion: Yes, RME could technically go very low but still you needed a very fine tweaked system, especially when you started building up a track with a lot of soft synths and effects e.t.c. Despite being able to go as low as 64 samples most studios I know usually used them at 256 or 128 samples and honestly those cards (among with a few others like some Apogee's or MOTU's) were just a few bright exceptions on the basic rule!
Mensajes Wed 10 Feb 16 @ 8:02 am
PressNPlayDJ
I used to play the piano and i have a midi synth which i use with FL for doing some production and remixes. There i feel the difference. I don't have my synth at 1ms but at 4, but as far as djing goes the big question is, why haven't i had any issues with 1 MS in the last 4 years that i have been using VDJ? (Le and Pro) And why hasn't anybody else that that I've helped had any issues when i recommended 1 ms? @ErieDj and me are the living proof on this post that 1 ms doesn't cause any issues. I'm sorry but i will see the difference between 1ms and 6ms especially whem scratching a song. I can't have that delay.
Mensajes Wed 10 Feb 16 @ 3:35 pm
djxhalx
Guys, I play live instruments with a band, I love polyrythmic structures of bands such as Meshuggah, I use vdj sampler as a liveact drum maschine - believe me if I tell you that there IS a difference between such values as 1ms or say 10ms. Probably, the real latency with settings as described in my previous post (asio4all v2 - 64samples, 3 kernel buffers, output processing only) is around 2.9ms (checked on Ableton). Still working smooth, 60fps skin all the time, great scratching response, playable sample activation (akai mpd218), no glitches whatsoever. Tried DENON asio drivers (both older and newer) - sucks big time, and gives BSOD on lowest conf. Tried WASAPI - it's unconfigurable, and slower than said configuration. Weird? Dunno, asio4all v2 works great :)
Mensajes Thu 11 Feb 16 @ 5:17 pm
PressNPlayDJ
djxhalx wrote :
Guys, I play live instruments with a band, I love polyrythmic structures of bands such as Meshuggah, I use vdj sampler as a liveact drum maschine - believe me if I tell you that there IS a difference between such values as 1ms or say 10ms. Probably, the real latency with settings as described in my previous post (asio4all v2 - 64samples, 3 kernel buffers, output processing only) is around 2.9ms (checked on Ableton). Still working smooth, 60fps skin all the time, great scratching response, playable sample activation (akai mpd218), no glitches whatsoever. Tried DENON asio drivers (both older and newer) - sucks big time, and gives BSOD on lowest conf. Tried WASAPI - it's unconfigurable, and slower than said configuration. Weird? Dunno, asio4all v2 works great :)
Don't know why or how but i using thr denon asio for controller drivers at 1ms with no issues....
Mensajes Thu 11 Feb 16 @ 5:34 pm
kradcliffe
The Denon Asio driver is an enigma. It works on some computers and not others.
It's badly written so who knows if the latency it supposedly achieves is even a true figure.
It's badly written so who knows if the latency it supposedly achieves is even a true figure.
Mensajes Thu 11 Feb 16 @ 6:08 pm
PressNPlayDJ
Yeah i agree on that. I've seen a lot of users have issues in general...maybe its configuration thing...who knows. Me on 2 laptops, 1 Sony Vaio 2012 and Alienware 2015 i have at 1 ms and never ever had issue. The craziest thing is that my old sony vaio was being used for everything! Gaming, Fl and Sony Acid, Photoshop and even movies...my new alienware is strictly for vdj. Not even a simple program like mp3gain or drop box is installed on that pc
Mensajes Thu 11 Feb 16 @ 6:57 pm
PressNPlayDJ
I totally forgot about this post... On another similar post i wrote that there is a solution for Asio drivers to work flawless. You would have to totally disable input lines everywhere.
1)Within Virtual DJ
2) Windows Sound settings.
Having only output settings and no input lines and there results have been great for the past 6 months. No BSOD, no stutter no crash and still on 1MS latency
1)Within Virtual DJ
2) Windows Sound settings.
Having only output settings and no input lines and there results have been great for the past 6 months. No BSOD, no stutter no crash and still on 1MS latency
Mensajes Tue 09 May 17 @ 5:53 pm
Don Moir
I work with audio a lot but for listening and reaction time to audio it is not like it is for audio engineers and DJs. Audio engineers and DJ's are probably some of the best at this. I used to work with audio engineers and they would point out the problems and ruin the music for me... haha
latency was defined above as the amount of time the software processes it and it gets to your sound card but that is only part of it.
You can find studies on reaction time to audio all you want... Seems at least some studies show 125ms or more on average. dogs can hear a lot better than people and one study showed about 16 ms for a dog.
Who or what might have better reaction times than audio engineers or DJs?
1) electronic monitoring
2) dogs
3) blind people
This is an educated guess so correct if you think it is not right.
When you set some software to 1 or 5 MS for latency there is no way it is getting to your ears and you can process it that fast. All it probably means is your telling the software/hardware to process things at a different rate and I believe you when you say you can tell the difference when you change the setting. It is not like you tell the difference between 1 or 5 ms though for real but the change has possibly made a difference between maybe 120ms and 160ms or something you can actually measure with your ears and do something about it.
Sound engineers use the rule of thumb that it takes audio 3ms to travel one meter in air based on the speed of sound in dry air etc. You can't hear something faster than it can travel to your ears. Head phones could help reduce air travel time.
latency was defined above as the amount of time the software processes it and it gets to your sound card but that is only part of it.
You can find studies on reaction time to audio all you want... Seems at least some studies show 125ms or more on average. dogs can hear a lot better than people and one study showed about 16 ms for a dog.
Who or what might have better reaction times than audio engineers or DJs?
1) electronic monitoring
2) dogs
3) blind people
This is an educated guess so correct if you think it is not right.
When you set some software to 1 or 5 MS for latency there is no way it is getting to your ears and you can process it that fast. All it probably means is your telling the software/hardware to process things at a different rate and I believe you when you say you can tell the difference when you change the setting. It is not like you tell the difference between 1 or 5 ms though for real but the change has possibly made a difference between maybe 120ms and 160ms or something you can actually measure with your ears and do something about it.
Sound engineers use the rule of thumb that it takes audio 3ms to travel one meter in air based on the speed of sound in dry air etc. You can't hear something faster than it can travel to your ears. Head phones could help reduce air travel time.
Mensajes Tue 09 May 17 @ 8:48 pm