The ultimate low cost solution for room correction?

[Count Arthur]

Everything arrived safe and sound earlier this afternoon.

I had a bit of fun and games at first. I thought I could use the Focusrite Scarlett as the microphone input, but leave XLR speaker output connected to my DAC, but that didn't seem to work.

Luckily, I had bought some cheap XLR to TRS adapters just in case:

Using these to connect to the Scarlett's output, we were off to the races.

The ARC Studio software guides you through the process and I did the full 21 point measurement.

Once the measurement is taken, I disconnected the microphone and audio interface and connected the ARC Studio between my DAC and subwoofer input:

I started the ARC 4 software and messed around with some of the pre defined house curves, but then played with some custom curves. I set this one up to more or less replicate Toole's "all listener" preference curve:

I like how this sounds, so that's what I have stored in the box for now. But, using the software you can quite easily switch between different target curves and hear how they sound.

Once a target is stored in the box, you can close the software and disconnect it from the PC if you wish, then switch the correction on or off via a button on the front. As such the correction can work with anything from the PC: internet radio, YouTube, Spotify, Foobar, JRiver, etc..

It's a shame that you need a separate audio interface as, if you're like me, once you have taken your measurements, you're unlikely to use it very often. Given that the ARC box already has an ADC in it, it's a shame that they couldn't add a microphone input with 48v phantom power suitable for the included measurement microphone, even if it increased the price a little.

Lastly, the £8.60 microphone stand I bought was far better than expected. Given it's low cost, I expected something incredibly flimsy, but it's actually fairly decent, and totally up to the job of holding a measurement microphone for the 20 minutes or so that I used it.

All in all, I'm very pleased with the results and if it does cause any degradation to the sound, I can't hear it. If you've ever used room correction, you'll have an idea of what to expect, an increase in clarity, less boomy bass, better imaging and, on the whole, a general improvement.

 

[Overseas]

Yeap, more joy and less hifi.

 

[Golf]

Is it possible to mouse-tune the white target curve? And if so, can it be done separately for the left and the right channel?

The reason for asking is the area between 150 and 800 Hz. Would be nice if some tweaking of the target line(s) could lead to smoother after curves here.

 

[Timcognito]

Everything arrived safe and sound earlier this afternoon.

I had a bit of fun and games at first. I thought I could use the Focusrite Scarlett as the microphone input, but leave XLR speaker output connected to my DAC, but that didn't seem to work.

Luckily, I had bought some cheap XLR to TRS adapters just in case:

View attachment 354356

Using these to connect to the Scarlett's output, we were off to the races.

The ARC Studio software guides you through the process and I did the full 21 point measurement.

Once the measurement is taken, I disconnected the microphone and audio interface and connected the ARC Studio between my DAC and subwoofer input:

View attachment 354357

I started the ARC 4 software and messed around with some of the pre defined house curves, but then played with some custom curves. I set this one up to more or less replicate Toole's "all listener" preference curve:

View attachment 354361

View attachment 354362

I like how this sounds, so that's what I have stored in the box for now. But, using the software you can quite easily switch between different target curves and hear how they sound.

Once a target is stored in the box, you can close the software and disconnect it from the PC if you wish, then switch the correction on or off via a button on the front. As such the correction can work with anything from the PC: internet radio, YouTube, Spotify, Foobar, JRiver, etc..

It's a shame that you need a separate audio interface as, if you're like me, once you have taken your measurements, you're unlikely to use it very often. Given that the ARC box already has an ADC in it, it's a shame that they couldn't add a microphone input with 48v phantom power suitable for the included measurement microphone, even if it increased the price a little.

Lastly, the £8.60 microphone stand I bought was far better than expected. Given it's low cost, I expected something incredibly flimsy, but it's actually fairly decent, and totally up to the job of holding a measurement microphone for the 20 minutes or so that I used it.

All in all, I'm very pleased with the results and if it does cause any degradation to the sound, I can't hear it. If you've ever used room correction, you'll have an idea of what to expect, an increase in clarity, less boomy bass, better imaging and, on the whole, a general improvement.

Great and concise report Arthur. Do you think you will try to better or fine tune this iteration or try something else as an experiment or just live with it? Are there features that you did not try?

 

[Count Arthur]

Is it possible to mouse-tune the white target curve? And if so, can it be done separately for the left and the right channel?

The reason for asking is the area between 150 and 800 Hz. Would be nice if some tweaking of the target line(s) could lead to smoother after curves here.

Sort of, when you are in "Edit" mode, you can move around the six points to adjust the shape of the curve:

However, you cannot add extra points with Q settings to create peaks or low shelves, so in that regard it's quite limited and you can't create separate left and right curves. I think the idea is that you create a target and it automatically calculates the filters to match the curve as closely as practical. If I change to a different target, you can see the "After" lines moving around briefly until they settle to match the new target; you can also hear the change.

Also, bear in mind that we're talking about small deviations from the target in the region of +/- 1 to 1.5dB - that's pretty good, if it's prediction is correct. At some point, I may try playing some sweeps, so that I can actually measure the output with the correction in place. Although I'm not sure I can use the included microphone for that, as I don't think it has a calibration file; presumably the ARC software accounts for that. However, I could try with my UMIK-1 and REW.

Great and concise report Arthur. Do you think you will try to better or fine tune this iteration or try something else as an experiment or just live with it? Are there features that you did not try?

I have tried picking alternative targets and playing with the virtual monitoring:

Here's what Smartphone does; basically boosts the midrange and lops off the top and bottom end, to make your monitors sound like a crummy smartphone speaker:

I guess that might be useful for an engineer that wants to check how their mix sounds on a variety of devices, but for my purposes, it's just a novelty.

For a more detailed view and to see it in action, take a look at some of the videos on YouTube: https://www.youtube.com/results?search_query=ik+multimedia+arc+studio+4

 

[Golf]

However, you cannot add extra points with Q settings to create peaks or low shelves, so in that regard it's quite limited and you can't create separate left and right curves.

Especially the latter would render it nearly useless for »serious« room correction, I’d say :-(

 

[Count Arthur]

Especially the latter would render it nearly useless for »serious« room correction, I’d say :-(

Bear in mind, that it does measure the left and right channels separately; I could hear it playing the funny noises from each channel individually as I took a measurement in each of the 21 positions it indicated.

So, it is applying its EQ settings to the left and right channels separately to match your curve at the listening position. It's just that these EQ settings aren't exposed to you to adjust manually.

 

[Golf]

Bear in mind, that it does measure the left and right channels separately; I could hear it playing the funny noises from each channel individually as I took a measurement in each of the 21 positions it indicated.

So, it is applying its EQ settings to the left and right channels separately to match your curve at the listening position. It's just that these EQ settings aren't exposed to you to adjust manually.

Sure. But in real »room eq life« there’s almost always reason for channel wise fine-tuning.

See the area between 150 and 800 Hz again: the »after curve« of the right channel shows a much worse result than that of the left channel. To improve this »by hand«, you’d need the possibility of channel wise fine-tuning.

 

[Count Arthur]

Sure. But in real »room eq life« there’s almost always reason for channel wise fine-tuning.

See the area between 150 and 800 Hz again: the »after curve« of the right channel shows a much worse result than that of the left channel. To improve this »by hand«, you’d need the possibility of channel wise fine-tuning.

I see what you mean, but I'm not sure how important that might be, we are talking about fairly small deviations and a significant improvement on the uncorrected output; and likely better than an awful lot of unmeasured, uncorrected set-ups.

We should also bear in mind that this is a prediction, how close it is to the actual output is uncertain - but I might look into that. Also, how discerning is our hearing; even if we fiddled around using hundreds of filters to get it absolutely perfect (maybe not possible anyway), chances are that the difference when listening might be very subtle. Large boosts and cuts over broad frequency range will be pretty obvious, but +/- 1 or 2dB over a narrow frequency range, will likely be very subtle.

Regardless, enabling and disabling the whole correction, is very obvious and, to my ears at least, the effect of the correction is very positive.

 

[Eric Natural]

I have the ARC v.4 software installed in my DAW's master and really like the corrections it's done - particularly of the bass response. It seems considerably more detailed.
I've emulated the EQ correction curve that it's calibration mic calculated for playback of Apple Music and like the result there as well. In the context of only happening in my little studio room. My stereo room with the CDs and vinyl are going to have to wait for their correction - home theater(Sony ES) and the hobby studio are dialed-in and sound great.
Best of luck

 

[Count Arthur]

I did some testing to measure the before and after correction frequency response using REW with both a UMIK-1 and the MEMS microphone included with ARC Studio.

Incidentally, the IK Multimedia MEMS microphone did not come with a calibration file, and I could find no link to one on the web site. I contacted IK Multimedia support via their web site and within 3 hours they had replied and sent me a link to the calibration file - top marks for customer support.

Here are the measurement results:

Observations:

1. The microphone measurements, are similar, peaks and dips are in the same place, but not identical. They are quite different in the ~15K region, but consistent with themselves when the corrected and uncorrected measurements are compared, the MEMS microphone shows a broad dip, while UMIK-1 is flatter. I would need a third microphone to determine which one is likely the more accurate.

2. The correction clearly does something. However, it is in no way as accurate as the predictions indicated within the software. It has boosted the response from 400Hz up and reduced some peaks and troughs in the bass region:

3. Regardless of the measured effect, the sonic effect is very positive, to my ears at least. I definitely prefer the sound with correction.


Things to bear in mind:

1. I am in no way an audio professional, or a trained listener; just some random guy on the internet that enjoys listening to music and occasionally messing about with stuff.

2. This is a desk set-up in my office (spare bedroom), not a purpose built professional studio.

 

[Timcognito]

I did some testing to measure the before and after correction frequency response using REW with both a UMIK-1 and the MEMS microphone included with ARC Studio.

Incidentally, the IK Multimedia MEMS microphone did not come with a calibration file, and I could find no link to one on the web site. I contacted IK Multimedia support via their web site and within 3 hours they had replied and sent me a link to the calibration file - top marks for customer support.

Here are the measurement results:

View attachment 354807

Observations:

1. The microphone measurements, are similar, peaks and dips are in the same place, but not identical. They are quite different in the ~15K region, but consistent with themselves when the corrected and uncorrected measurements are compared, the MEMS microphone shows a broad dip, while UMIK-1 is flatter. I would need a third microphone to determine which one is likely the more accurate.

2. The correction clearly does something. However, it is in no way as accurate as the predictions indicated within the software. It has boosted the response from 400Hz up and reduced some peaks and troughs in the bass region:

View attachment 354808

3. Regardless of the measured effect, the sonic effect is very positive, to my ears at least. I definitely prefer the sound with correction.


Things to bear in mind:

1. I am in no way an audio professional, or a trained listener; just some random guy on the internet that enjoys listening to music and occasionally messing about with stuff.

2. This is a desk set-up in my office (spare bedroom), not a purpose built professional studio.

Keep posting you're saving me a lot of time and helping me decide what to do with my system.
Things to bear in mind: I'm lazy

 

[GXAlan]

This is a quick video comparing the Trinnov Nova and the ARC Studio. They didn’t do real measurements afterwards and just went with what the software reports the correction is, but it’s a nice look at the interfaces.

 

[Golf]

Hmm ... Is this chart showing the right channel only?

 

[Count Arthur]

Hmm ... Is this chart showing the right channel only?

This was a single measurement with each microphone in the same position, with and without correction, left and right combined.

This is a quick video comparing the Trinnov Nova and the ARC Studio.

I saw that. It should be noted that the Trinnov system is about 10 x the price of the ARC system.

 

[Golf]

left and right combined.

What sense could that make?

 

[Count Arthur]

What sense could that make?

1. I refer you to my previous post: "I am in no way an audio professional, or a trained listener; just some random guy on the internet that enjoys listening to music and occasionally messing about with stuff."

2. This is just for fun - I don't work here.

 

[Golf]

This is a quick video comparing the Trinnov Nova and the ARC Studio. They didn’t do real measurements afterwards and just went with what the software reports the correction is, but it’s a nice look at the interfaces.

At around minute 7:50, some »34-band« editing of the target curve is shown. Valid for the £ 4,000 solution »Trinnov Nova« only ...

For comparison: The software solution »MathAudio Room EQ« is only $ 99.95, and it features an even better possibility of editing the target curves (left and/or right), as you can literarily mouse-draw them.

 

[Timcognito]

At around minute 7:50, some »34-band« editing of the target curve is shown. Valid for the £ 4,000 solution »Trinnov Nova« only ...

For comparison: The software solution »MathAudio Room EQ« is only $ 99.95, and it features an even better possibility of editing the target curves (left and/or right), as you can literarily mouse-draw them.

$99.95 + $400 computer always in loop. IK Arc Studio is for those of us that don't want to use a computer in audio playback. That has been stated by several posters and YouTubers. If one routinely uses a computer in audio playback, MA may be a better deal. IK at $300 seems to be the lowest no computer (other than setup) real DSP deal.

In the video on page 3 the presenter reshapes curves by drawing but its probably not as sophisticated as MathAudio.

 

[Golf]

$99.95 + $400 computer always in loop.

True.

Nevertheless, it’s hard to imagine to make good use of »IK Arc Studio« or even more »Trinnov Nova« without a computer.