Latency on Mac vs Windows using Denon X1850 soundcard.
I recently switched from an older pc laptop with i7 to a mac mini with a m1.
I am surprised to see that the lowest latency i can achieve on mac is 64 while I was able to do 32 on my windows old laptop. There isn’t even an option for it.
PC has the special denon program that i used to use to select my latency, is there any way to achieve 32 on mac in traktor, maybe with a 3rd party program?
I think it will handle it if given the option as the load bar is only at 25-40% right now at 64…
Best Answers
-
Coming back to this one as I've checked a few soundcards/mixers that I have access to.
TLDR: If you want to scratch, maybe don't use a DJM … latency is massive!
What I currently have access to:
DJM-900NXS2 / DJM-V10 / Xone:PX5 / TA10mk2 / TA6 / RME Fireface UCX II / Apollo Arrow
Notes: I cannot scratch well … it is not a massive part of how I play. I do however try and single hit cue points on beat without headphones but with foldback at my gigs so overall latency for me is important … it's live so I get one shot at it!
I also generally set my buffer size at 256 samples at 96KHz which gives me very good stability and lower overall CPU usage. I have used Traktor since the release of the Audio 8 DJ back in 2007, and used it on my PowerBook G4 at the time, and have owned the A4DJ, Z2, Z1, S4 and TA10(mk1), and still currently own (2) TA6 and the TA10mk2. Below is what I normally use and also the lowest sample rate available.
All the following are with my MacBook Pro 13" that I use for Traktor:
DJM-900NXS2 (at current residency, might update 64 samples after gig tonight)
- 96k/256 samples = Processing 2.7ms / Output 6.9ms / Overall 9.6ms
DJM-V10 (borrowing)
- 96k/256 samples = Processing 2.7ms / Output 6.9ms / Overall 9.6ms
- 96k/64 samples = Processing 0.7ms / Output 6.9ms / Overall 7.6ms
Xone:PX5 (own)
- 96k/256 samples = Processing 2.7ms / Output 3.0ms / Overall 5.7ms
- 96k/64 samples = Processing 0.7ms / Output 3.0ms / Overall 3.7ms
TA10mk2 (own)
- 96k/256 samples = Processing 2.7ms / Output 3.0ms / Overall 5.7ms
- 96k/64 samples = Processing 0.7ms / Output 3.0ms / Overall 3.7ms
TA6 (own)
- 96k/256 samples = Processing 2.7ms / Output 3.0ms / Overall 5.7ms
- 96k/64 samples = Processing 0.7ms / Output 3.0ms / Overall 3.7ms
Fireface UCX II (own … in studio)
- 96k/256 samples = Processing 2.7ms / Output 0.5ms / Overall 3.2ms
- 96k/64 samples = Processing 0.7ms / Output 0.5ms / Overall 1.2ms
Apollo Arrow (Thunderbolt not USB … rarely used in studio)
- 96k/256 samples = Processing 2.7ms / Output 0.3ms / Overall 3.0ms
- 96k/64 samples = Processing 0.7ms / Output 0.3ms / Overall 1.0ms
For ALL of the above, 48k/64 samples processing goes to 5.3ms and at 44.1/64 goes to 5.8ms.
For ALL of the above, 48k/64 samples processing goes to 1.3ms and at 44.1/64 goes to 1.5ms.
Note that at 44.1 there are some slight increases to Output latency by 0.1ms on some units. It was something I noticed, I've never really used 44.1 so never been an issue for me.
My Razer Blade W10 LTSC laptop for some of the above, 256 samples and then as low as it will go, just to see differences in Windows vs Mac:
Fireface UCX II (ASIO - RME Drivers)
- 96k/256 samples = Processing 2.7ms / Output 0.8ms / Overall 3.5ms
- 96k/96 samples = Processing 1.0ms / Output 0.8ms / Overall 1.8ms
- 48k/256 samples = Processing 5.3ms / Output 0.9ms / Overall 6.2ms
- 48k/48 samples = Processing 1.0ms / Output 0.9ms / Overall 1.9ms
DJM-V10 (ASIO - Pioneer Drivers)
- 96k/256 samples = Processing 2.7ms / Output 5.8ms / Overall 8.5ms
- 96k/64 samples = Processing 0.7ms / Output 5.8ms / Overall 6.5ms
- 48k/256 samples = Processing 5.3ms / Output 9.3ms / Overall 14.6ms
- 48k/64 samples = Processing 1.3ms / Output 9.4ms / Overall 10.7ms
TA10mk2 (ASIO - USB Buffer @ 1ms in TA10 settings … can be upto 4ms)
- 96k/256 samples = Processing 2.7ms / Output 4.0ms / Overall 6.7ms
- 96k/64 samples = Processing 0.7ms / Output 4.0ms / Overall 4.7ms
- 48k/256 samples = Processing 5.3ms / Output 4.4ms / Overall 9.7ms
- 48k/32 samples = Processing 0.7ms / Output 4.4ms / Overall 5.1ms
It comes down to Denon using the class-compliant Core Audio Drivers, which result in your Output Latency changing from what you are used to. THIS makes the biggest difference. The CC Core Audio drivers are a good solid foundation to build upon with less variables. The TA10/6 also use this, as well as the Xone mixers. Very similar to how ASIO drivers work on Windows, but as you can see it comes down to how each manufacturer has implemented ASIO within their driver set as they vary widely.
Lastly, you can also see when a manufacturer develops their own driver set for macOS (Pioneer and RME), there are vastly different results with Output Latency! If only RME made a soundcard with multiple Phono Pre's … hmmm …
So … for you to reduce your latency I would recommend changing to 96k instead of the 44.1k that you are currently using. It should drop your latency by another 0.8ms as your Processing will go down to 0.7ms, that's the best you'll get with a combination of macOS and the Denon mixer.
1 -
wow thank you for this detailed analysis of your sound interface and their latency!
i ended up switch back to a windows for the time being. Would like to switch back to Mac in the future as it felt very smooth…
I’m surprised that that the higher sample rates are giving less latency. I would have thought the opposite…. I will have to play around with these settings and see if I can adjust. Probably tough in the CPU.Again thanks for this detailed info. I’m in a happy place right now with a windows desktop. Having fun experimenting with hdmi capture cards to output the UI to multiple mini screens:).
0
Answers
-
Hello!
Can you please tell us the operating system of your device and the version of Traktor you are using?
0 -
Also, 32 and 64 are not your "latency", it's your buffer size and that will affect your overall latency. Can you show the full round trip latency for each OS and how it compares? Your sample rate of 48k or 96k will also change the overall latency as well.
I understand that while "lower is better" from some technical aspects, is there a reason for wanting this to be lower?
1 -
Any update? @merve
0 -
"PC has the special denon program that i used to use to select my latency…"Doesn't your Mac have the same denon soundcard application to change the specs as you mentioned your windows PC has? I would think it installed with you Denon soundcard install package (for mac). Have you been to Denon website to download the latest drivers for your denon soundcard for MAC?
1 -
thanks everyone for the replies. My old laptop is unavailable for a couple of days. Will post pics of the settings and setup etc.
Apparently when plugged into Mac the drivers are automatically installed… but this denon program isn’t. There are no downloads for the Mac except for the firmware updater app. It is using the coreaudio Mac drivers. I’ve played around with a program called reaper to set the audio buffer to 32. Bit traktor seemed to override it.32 is for scratching. I was pleasantly surprised when I found it in the options of the Denon program. It feels just a touch tighter scratching… I think if offered would make many turntablists happy. I previously used the A10 on 64 buffer size. I switched to denon bc I wanted the internal sound card with external effects. 64 is great, but once you feel 32 it’s tough to go back. Like going from 120 fps back to 60 for gaming… I feel like a lot of computers could handle it and NI could add it to the options fairly easily? Maybe use at your own risk setting
Will post pics tomorrow but this weekend when I get the laptop back.
0 -
here are some pic of my older windows 10 laptop vs Mac mini. Traktor 4.0.2.
Ideally I’d like to be able to try 32 or even 16 samples! But this driver program doesn’t exist for Mac and I don’t think wine would work with drivers…
0 -
Coming back to this one as I've checked a few soundcards/mixers that I have access to.
TLDR: If you want to scratch, maybe don't use a DJM … latency is massive!
What I currently have access to:
DJM-900NXS2 / DJM-V10 / Xone:PX5 / TA10mk2 / TA6 / RME Fireface UCX II / Apollo Arrow
Notes: I cannot scratch well … it is not a massive part of how I play. I do however try and single hit cue points on beat without headphones but with foldback at my gigs so overall latency for me is important … it's live so I get one shot at it!
I also generally set my buffer size at 256 samples at 96KHz which gives me very good stability and lower overall CPU usage. I have used Traktor since the release of the Audio 8 DJ back in 2007, and used it on my PowerBook G4 at the time, and have owned the A4DJ, Z2, Z1, S4 and TA10(mk1), and still currently own (2) TA6 and the TA10mk2. Below is what I normally use and also the lowest sample rate available.
All the following are with my MacBook Pro 13" that I use for Traktor:
DJM-900NXS2 (at current residency, might update 64 samples after gig tonight)
- 96k/256 samples = Processing 2.7ms / Output 6.9ms / Overall 9.6ms
DJM-V10 (borrowing)
- 96k/256 samples = Processing 2.7ms / Output 6.9ms / Overall 9.6ms
- 96k/64 samples = Processing 0.7ms / Output 6.9ms / Overall 7.6ms
Xone:PX5 (own)
- 96k/256 samples = Processing 2.7ms / Output 3.0ms / Overall 5.7ms
- 96k/64 samples = Processing 0.7ms / Output 3.0ms / Overall 3.7ms
TA10mk2 (own)
- 96k/256 samples = Processing 2.7ms / Output 3.0ms / Overall 5.7ms
- 96k/64 samples = Processing 0.7ms / Output 3.0ms / Overall 3.7ms
TA6 (own)
- 96k/256 samples = Processing 2.7ms / Output 3.0ms / Overall 5.7ms
- 96k/64 samples = Processing 0.7ms / Output 3.0ms / Overall 3.7ms
Fireface UCX II (own … in studio)
- 96k/256 samples = Processing 2.7ms / Output 0.5ms / Overall 3.2ms
- 96k/64 samples = Processing 0.7ms / Output 0.5ms / Overall 1.2ms
Apollo Arrow (Thunderbolt not USB … rarely used in studio)
- 96k/256 samples = Processing 2.7ms / Output 0.3ms / Overall 3.0ms
- 96k/64 samples = Processing 0.7ms / Output 0.3ms / Overall 1.0ms
For ALL of the above, 48k/64 samples processing goes to 5.3ms and at 44.1/64 goes to 5.8ms.
For ALL of the above, 48k/64 samples processing goes to 1.3ms and at 44.1/64 goes to 1.5ms.
Note that at 44.1 there are some slight increases to Output latency by 0.1ms on some units. It was something I noticed, I've never really used 44.1 so never been an issue for me.
My Razer Blade W10 LTSC laptop for some of the above, 256 samples and then as low as it will go, just to see differences in Windows vs Mac:
Fireface UCX II (ASIO - RME Drivers)
- 96k/256 samples = Processing 2.7ms / Output 0.8ms / Overall 3.5ms
- 96k/96 samples = Processing 1.0ms / Output 0.8ms / Overall 1.8ms
- 48k/256 samples = Processing 5.3ms / Output 0.9ms / Overall 6.2ms
- 48k/48 samples = Processing 1.0ms / Output 0.9ms / Overall 1.9ms
DJM-V10 (ASIO - Pioneer Drivers)
- 96k/256 samples = Processing 2.7ms / Output 5.8ms / Overall 8.5ms
- 96k/64 samples = Processing 0.7ms / Output 5.8ms / Overall 6.5ms
- 48k/256 samples = Processing 5.3ms / Output 9.3ms / Overall 14.6ms
- 48k/64 samples = Processing 1.3ms / Output 9.4ms / Overall 10.7ms
TA10mk2 (ASIO - USB Buffer @ 1ms in TA10 settings … can be upto 4ms)
- 96k/256 samples = Processing 2.7ms / Output 4.0ms / Overall 6.7ms
- 96k/64 samples = Processing 0.7ms / Output 4.0ms / Overall 4.7ms
- 48k/256 samples = Processing 5.3ms / Output 4.4ms / Overall 9.7ms
- 48k/32 samples = Processing 0.7ms / Output 4.4ms / Overall 5.1ms
It comes down to Denon using the class-compliant Core Audio Drivers, which result in your Output Latency changing from what you are used to. THIS makes the biggest difference. The CC Core Audio drivers are a good solid foundation to build upon with less variables. The TA10/6 also use this, as well as the Xone mixers. Very similar to how ASIO drivers work on Windows, but as you can see it comes down to how each manufacturer has implemented ASIO within their driver set as they vary widely.
Lastly, you can also see when a manufacturer develops their own driver set for macOS (Pioneer and RME), there are vastly different results with Output Latency! If only RME made a soundcard with multiple Phono Pre's … hmmm …
So … for you to reduce your latency I would recommend changing to 96k instead of the 44.1k that you are currently using. It should drop your latency by another 0.8ms as your Processing will go down to 0.7ms, that's the best you'll get with a combination of macOS and the Denon mixer.
1 -
wow thank you for this detailed analysis of your sound interface and their latency!
i ended up switch back to a windows for the time being. Would like to switch back to Mac in the future as it felt very smooth…
I’m surprised that that the higher sample rates are giving less latency. I would have thought the opposite…. I will have to play around with these settings and see if I can adjust. Probably tough in the CPU.Again thanks for this detailed info. I’m in a happy place right now with a windows desktop. Having fun experimenting with hdmi capture cards to output the UI to multiple mini screens:).
0 -
If only RME made a soundcard with multiple Phono Pre's … hmmm …
RME makes several interfaces with two independent Phone outputs. Or have you speaked about inputs?
I may add values for Babyface Pro FS on Win10, that I use with Traktor Pro 3.
Babyface Pro FS (ASIO - RME Drivers)
- 96k/256 samples = Processing 2.7ms / Output 0.8ms / Overall 3.5ms
- 96k/96 samples = Processing 1.0ms / Output 0.8ms / Overall 1.8ms
- 48k/256 samples = Processing 5.3ms / Output 0.8ms / Overall 6.1ms
- 48k/48 samples = Processing 1.0ms / Output 0.8ms / Overall 1.8ms
PS.
It seems to me that your values for UCX II for 48k/256 samples are wrong (the first value should be the double of 96k/256).
0 -
Hmmm … I’ve just finished my gig deep in suburban Melbourne so gotta drive home and stop for a kebab on the way so I’ll check in a bit, lol
0 -
Oh and I’m talking about multiple phono inputs (with real phono preamps) as in can connect a pair of turntables to the RME so that you can use time code vinyl for scratch.
Technically you can connect 3x RME sound cards and they will all be processed through the one instance of TotalMix so it’s actually possible if your pockets are deep …0 -
Fixed up the RME value for 48/256 on the Razer Blade … it was a long afternoon of plugging and un-plugging, I was bound to get something wrong!
0
Categories
- All Categories
- 19 Welcome
- 1.5K Hangout
- 62 NI News
- 785 Tech Talks
- 4.1K Native Access
- 16.5K Komplete
- 2K Komplete General
- 4.3K Komplete Kontrol
- 5.7K Kontakt
- 1.6K Reaktor
- 378 Battery 4
- 845 Guitar Rig & FX
- 429 Massive X & Synths
- 1.3K Other Software & Hardware
- 5.8K Maschine
- 7.3K Traktor
- 7.3K Traktor Software & Hardware
- Check out everything you can do
- Create an account
- See member benefits
- Answer questions
- Ask the community
- See product news
- Connect with creators