recommendations required for mic preamp and mic for use with Holmimpulse

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
so currently I have a umik-1 USB mic for using with rew and holmimpulse

I believe its only capable of measuring at 48khz samples

I am using holmimpulse in a car so first (and 2nd, and 3rd and 10th!) reflections are a major issue, now im slowly working my head around gating and removing reflections from my measurements, however I suspect if I could use a 96khz interface and a shorter initial pulse then I can see more clearly where my initial rise is

I am aligning at various crossover points and welcome suggestions for improving my measurements

subs to midbass 60-70hz
midbass to mids 300-500hz
mids to tweeters 3-5khz

a, I may well post some impulse responses as they look very bizarre and not clean and am open to any suggestions to make them more 'clean' so to speak but I suspect a shorter impulse may help with this

b, I am after recommendations for a mic and interface for use with holmimpulse which can be used with a loopback measurement rather than time locked as it can be seen drifting (for example if I lock it and select use for the time and repeat the very same measurement it then creeps back so the peak isn't then at zero, I suspect its something to do with my laptop/mic combination perhaps or a setting (I will ensure holmimpulse is set to my mics native 48k setting for measurements)

would setting the subwoofer in the corner of the van rather than in the centre help improve measurements of the sub as I have massive tuning capabilitys with a Helix DSP pro 2 so freq response isn't a massive issue for me although the centre of the van does exhibit better freq reponse to some degree but its not the end of the world if its not prefect and more eq has to be used, vehicles are way more awkward than rooms to get sounding nice sadly :(
 
If it helps I’m looking at a focusrite Scarlett solo and a Dayton Emma-6

I am open to suggestions if that’s not a decent combination for using for loopback measurements, however I am aware I will be needing patch leads etc, having never done this method of measurement before I would love some help with the leads and connections I will be required to use also
 
Unless your tweeters extend beyond 24 kHz using a higher sample rate will not make any difference to your measurements, and even then it will not affect the kinds of issues you describe. A car is a very difficult measuring environment, higher sample rates won't change that.

Thanks for the reply, the higher sample rates were more to enable me to use a shorter impulse (increasing the sample rate means the minimum length is shorter) as the reflective surfaces are closer than in a room for example, as you up the sample rate it enables a shorter impulse length initially so my theory which may be right or wrong was that the shorter the impulse the more able I would be to separate the initial rise from the reflections rise... whether it makes any odds in reality I don’t know, but yes a vehicle is horrible for reflections etc with glass in close proximity to the speakers
 
The theory is wrong, I'm afraid. Neither HolmImpulse nor REW use an impulse to make measurements, they derive an impulse response from the system's output when driven by a logarithmic sine sweep. The sharpness of features in that impulse response depends primarily on the bandwidth of the drivers being measured and the mic being used for measurement. Even your tweeters will likely not reproduce frequencies above 24 kHz at any meaningful level and the mic you are considering does not have the bandwidth or the calibration data to take advantage of a higher sample rate. Higher sample rates are not going to make your measurements any easier to interpret.
 
The theory is wrong, I'm afraid. Neither HolmImpulse nor REW use an impulse to make measurements, they derive an impulse response from the system's output when driven by a logarithmic sine sweep. The sharpness of features in that impulse response depends primarily on the bandwidth of the drivers being measured and the mic being used for measurement. Even your tweeters will likely not reproduce frequencies above 24 kHz at any meaningful level and the mic you are considering does not have the bandwidth or the calibration data to take advantage of a higher sample rate. Higher sample rates are not going to make your measurements any easier to interpret.

That does make sense on some parts...

I don’t use a logarithmic sine sweep like rew does, I use mls (it almost sounds like pink noise...) as I read a ‘how to’ that this is more accurate for at least subwoofer to midbass timing (not sure if this is correct or not as I am 100% new to using holmimpulse)

Any suggestions for settings are more than welcome, if I am doing something incorrect then put me right, I am very new to using impulse responses to set timing, any references or guides are more than welcome also
 

Attachments

  • 41944299-62F4-4BB7-AB5C-6205814A79D0.png
    41944299-62F4-4BB7-AB5C-6205814A79D0.png
    877 KB · Views: 76
Well as a new user and there being no handbook such I have to go by what I can find, and the only link to a how too was in

The Subwoofer DIY Page v1.1 - Time Alignment using HolmImpulse

So that’s why I used mls, rather than just telling me I’m doing wrong could you not suggest helpful hints and things to do as I’m sure the entire forum will benefit vs just telling me what not to do and why it’s incorrect?
 
I initially posted because it looked like you were about to spend money on things that wouldn't have achieved what you expected. I have commented on areas where I feel I can contribute, there are no doubt other forums on DIYAudio and elsewhere that can offer much better advice on setting up car audio systems than I can.

There are particular issues associated with using USB mics for measurements that will be used for timing alignment. When the input and output of a system are on different devices they do not stay in alignment because their clocks will be running at slightly different rates. The Holmimpulse option to keep streams active is really only suitable when input and output are on the same interface and so share a common clock. USB mics also do not provide any means of connecting a loopback from the output device to provide a timing reference. To get around that problem REW offers an "Acoustic timing reference", which means using one full range driver (or tweeter) in a system to reproduce a timing signal that the software generates and detects and uses to provide a consistent reference from measurement to measurement. Your USB mic could be used with REW, if you wish to use Holmimpulse you would need a different (analog) mic.
 
I initially posted because it looked like you were about to spend money on things that wouldn't have achieved what you expected. I have commented on areas where I feel I can contribute, there are no doubt other forums on DIYAudio and elsewhere that can offer much better advice on setting up car audio systems than I can.

There are particular issues associated with using USB mics for measurements that will be used for timing alignment. When the input and output of a system are on different devices they do not stay in alignment because their clocks will be running at slightly different rates. The Holmimpulse option to keep streams active is really only suitable when input and output are on the same interface and so share a common clock. USB mics also do not provide any means of connecting a loopback from the output device to provide a timing reference. To get around that problem REW offers an "Acoustic timing reference", which means using one full range driver (or tweeter) in a system to reproduce a timing signal that the software generates and detects and uses to provide a consistent reference from measurement to measurement. Your USB mic could be used with REW, if you wish to use Holmimpulse you would need a different (analog) mic.

And I appreciate the input. Do you use holmimpulse with loopback? If so, what mic and preamp do you use?

I am more interested in a mic/preamp setup to use with loopback in both holm and rew than anything... it also lets me calibrate the soundcard which should be more accurate than not doing so as you can’t loopback in from the usb mics output as you say, it just seems there are no disadvantages to a mic with loopback capability, but there are a few disadvantages with a usb mic so I don’t mind spending a few quid on a mic and preamp 👍🏼
 
Member
Joined 2004
Paid Member
A car is probably the worst case for windowed measurements. There are a lot of nearby reflective surfaces that will seriously confuse any measurement. In dash speakers are the most hopeless for example. Getting meaningful time alignment would be difficult at best. And with two ears separated the way human heads are you cannot be sure where to align to.

If you know the free field (anechoic) response of the drivers when together you can get some idea by looking at the response variations as you change the delays. When its closest to the free field response you should be pretty close to optimum. You can't really do that in a windowed measurement and you need more like a 1/6 octave RTA measurement.

The good part about a car usually is all the padded surfaces that damp the reflections, (except windows etc.) so an RTA is a pretty good starting point. Listening to recordings of familiar voices can also help spot issues that degrade the comprehension of complex words and sounds.

Something may have a pleasing balance on simple music but makes dialog or complex musical phrasing hard to follow because the components are not close enough in time. We do not have good tools that can handle those issues in a car for example so listening may be your best option. Even Detroit, which spends a whole lot on technology and instrumentation falls back to listening for the critical adjustments.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.