REW For Simple Desktop Speakers?

[wacomme]

Can you drive the speakers directly from the Mac? This could eliminate the DAC as the source of the 5kHz drop.

Also, nothing you’re doing with respect to audio is taxing your computer. Do not be concerned.

I'm pretty sure I can bypass the dac to make sure that is not the source of the 5kHz drop.

 

[wacomme]

REW won't allow different interface for ADC and DAC, but that is possible to work around in Windows with ASIO4ALL. For MacOS I don't know.

The sharp 5K drop should be quite audible though so your suggestion of another DAC ought to work for just confirmation.

I have a Qudelex I could use as an alternate dac if I can't bypass the SMSL dac.

 

[dc655321]

REW won't allow different interface for ADC and DAC, but that is possible to work around in Windows with ASIO4ALL. For MacOS I don't know.

Not sure what you mean here - I can easily select which mic and outputs to use in REW (linux or mac).

 

[Trell]

I have a Qudelex I could use as an alternate dac if I can't bypass the SMSL dac.

Yeah, with a tone sweep from REW you should hear the 5 kHz drop on your right monitor or channel.

 

[dc655321]

I have a Qudelex I could use as an alternate dac if I can't bypass the SMSL dac.

That's worth a shot, just to narrow down the possibilities.

FWIW, I have a similar desk setup and also see large suckouts in a similar region. Rooms suck

 

[HarmonicTHD]

I'm pretty sure I can bypass the dac to make sure that is not the source of the 5kHz drop.

If you can hear the drop it’s the DAC or speaker. If you can’t hear it, it is your computer / UMIK not ramping up fast enough.

You can look at the non smoothed / RAW data. If the data shows an abrupt drop it is a glitch during the capturing (CPU). If it is a roll off (more gentle) than it is most likely speaker or DAC.

My laptop i7 (2021) didn’t like it when in Energy saving / economic mode and glitched during the capturing phase.

 

[wacomme]

That's worth a shot, just to narrow down the possibilities.

FWIW, I have a similar desk setup and also see large suckouts in a similar region. Rooms suck

I presume the problem is not the dac, but it's good to rule it out. I did get the Qudelix to work with the monitors and REW, so bypassing my desktop dac should work. I also presume the 5k drop is the the R monitor, but I'll switch monitors to make sure that's not the issue either. I suspect the issue is the room and the R monitor location. I'll test later to to find out!

 

[Trell]

I suspect the issue is the room and the R monitor location.

Not for a drop like that at 5 kHz that you measured.

 

[wacomme]

Dang! It is that dac that's the problem. I still need to switch monitors to make sure, but the green and red lines are quick measurements with the Qudelix dac. The blue line is the SMSL DO100 dac. It seems pretty obvious.

Unfortunately, I'm outside the Amazon return window for the SMSL DO100. Should I contact the manufacturer for a replacement? It's the last thing I thought would be the problem. So, am I right in that the dac is the reason for the R monitor dropoff at 5kHz? Could it be my computer?

If it IS the dac, that's pretty bad. First the Topping D50 with a similar issue, and now SMSL? Really?

 

[Trell]

Dang! It is that dac that's the problem. I still need to switch monitors to make sure, but the green and red lines are quick measurements with the Qudelix dac. The blue line is the SMSL DO100 dac. It seems pretty obvious.

Unfortunately, I'm outside the Amazon return window for the SMSL DO100. Should I contact the manufacturer for a replacement? It's the last thing I thought would be the problem. So, am I right in that the dac is the reason for the R monitor dropoff at 5kHz? Could it be my computer?

If it IS the dac, that's pretty bad. First the Topping D50 with a similar issue, and now SMSL? Really?

View attachment 218639

I would return it as defective, which it clearly is. Unless the manufacturer has a firmware fix.

I don’t know where you live but even Amazon in USA has some period for defects?

Edit: As you wrote below something else can be going on as well.

 

[wacomme]

OK. The story gets stranger. I swamped monitors - had to swap cable L/R connections and switch L to R and R to L in REW - and I don't get the dropoff from EITHER speaker. I tested it several time and cannot replicate the dropoff, though I haven't swapped the speakers back to their original positions. Now, along the way I'm getting output selection errors in REW (REW V5.20.9 running Azul Systems, Inc. JRE 1.8.0_332 64-bit on OS X 12.4 Locale: language en, keyboard US, UTF-8 at 81 DPI) when I select the SMSL dac, and (and one low volume issue despite not adjusting the volume). If I turn on/off the dac several times I can get the dac to "stick" within REW, but most of the time I get that error message.

Now, is that error message due to the computer or the dac?

 

[Trell]

On Windows I’ve to restart REW if the DAC is not powered on at startup. Some changes like sample rate or ASIO buffer size can require the same.

 

[wacomme]

Ok. I moved the monitors back to their original positions - L/R. I did a quick REW measurement (with the SMSL dac) and no 5kHz dropoff. What's happening?

Anyway, for what it's worth, I am getting a replacement dac from Amazon - despite being outside the 30 day return window. Thanks Amazon!!!!

 

[wacomme]

More accurate sweep tests will be made tonight - umik-1 vertical with 90 degree setting and hopefully without the 5kHz dropoff. Is there anything else I need to know about measuring before I assume the measurements are good and start playing with speaker position (I might wait for the new dac - in a week).

From the measurements I'm assuming that the monitors are fine; I bought them used.

 

[wacomme]

If you can hear the drop it’s the DAC or speaker. If you can’t hear it, it is your computer / UMIK not ramping up fast enough.

You can look at the non smoothed / RAW data. If the data shows an abrupt drop it is a glitch during the capturing (CPU). If it is a roll off (more gentle) than it is most likely speaker or DAC.

My laptop i7 (2021) didn’t like it when in Energy saving / economic mode and glitched during the capturing phase.

I can't really hear it, though I don't know what I'm listening for.

Here is yesterday's R speaker measurement with the 5kHz dropoff without smoothing. What do you think? Computer/mic or dac/speaker? If it's the UMIK-1, is there any way to speed up the ramping? If it's the computer, is there anything that can be done?

 

[wacomme]

I measured again (sweep, 90degree UMIK-1 calibration and mic pointing straight up in listening ear position). I took two measurements for both L and R monitors. No 5kHz dip!!!! Speakers are in the same location as yesterday. The R speaker is nearly a perfect match with the measurement above 35Hz. The L speaker is close with a bobble between measurements at 1200Hz. I should repeat the R side, but haven't.

 

[wacomme]

I'm a little surprised the L speaker has a double dip at 90 and 110Hz, and larger than the single 100Hz of the R speaker. I thought the R speaker would be worse since it's more tangental to the wall (30 degrees) vs 10 degree from perpendicular for the L speaker. However, the L speaker has a partially drawn window shade behind it (the monitor port is at the top of the monitor). What do you make of this?

 

[HarmonicTHD]

I can't really hear it, though I don't know what I'm listening for.

Here is yesterday's R speaker measurement with the 5kHz dropoff without smoothing. What do you think? Computer/mic or dac/speaker? If it's the UMIK-1, is there any way to speed up the ramping? If it's the computer, is there anything that can be done?

View attachment 218662View attachment 218665

Yes something is glitching during capturing. Strange that it only occurs intermittently. Did you try to increase the buffer size? Does it get better?

 

[HarmonicTHD]

I'm a little surprised the L speaker has a double dip at 90 and 110Hz, and larger than the single 100Hz of the R speaker. I thought the R speaker would be worse since it's more tangental to the wall (30 degrees) vs 10 degree from perpendicular for the L speaker. However, the L speaker has a partially drawn window shade behind it (the monitor port is at the top of the monitor). What do you make of this?View attachment 218683

Happy to see you make progress.

At these frequencies sound is not directional and angle does not matter. What matter is positioning in space and proximity to walls.

REW has a room simulator built in. It can give you hints what dip is caused by which position. However in the end there will always be a compromise. Again only sub(s) positioned strategically can give you a near perfect low frequency response.

 

[wacomme]

Yes something is glitching during capturing. Strange that it only occurs intermittently. Did you try to increase the buffer size? Does it get better?

I did not increase the buffer size since I'm no longer seeing the drop - a conundrum.