CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS
Lobby => Amp and DAC Measurements => Topic started by: Luckbad on August 12, 2015, 06:44:25 AM
-
I have a Focusrite Scarlett 2i4 and I'm starting to take measurements of amps and dacs.
I picked up a couple of Hosa Pro cables, one 3.5mm > Dual XLR, and one RCA > Dual TS.
I also have a Sound Blaster ZxR sound card with a discrete input card.
My issue is that theoretically the Scarlett should be really good for measurements, but I get far worse readings on the Scarlett than the Sound Blaster, and I don't know why.
For example, on my ODAC revB + O2, I get: Very good, Average x3, Good x3, Poor x1. On the Sound Blaster, I get Excellent x7 and Good x1.
Any advice for using the Scarlett to effectively take amp measurements? I tried Line/Inst/Pad combos and none seemed to make it much better. Should I give up on it and just use the ZxR?
Thanks!
-
I have a 2i2 but mostly use it for headphone measurements.
What do you get when you do a loopback into the 2i4?
-
The scarlett series is hardly good enough for unloaded DAC measurements. For amp measurements you need to give the amp a load(like 16 ohms) to have it distort while driving the load.
-
I have a 2i2 but mostly use it for headphone measurements.
What do you get when you do a loopback into the 2i4?
I'll try the loopback tonight.
The scarlett series is hardly good enough for unloaded DAC measurements. For amp measurements you need to give the amp a load(like 16 ohms) to have it distort while driving the load.
How do you give am amp load like that while connected to the Scarlett? Do I need an resistor adapter? I actually have one of those in my mailbox right now that gives 75 Ohms of load (in theory) that I'm going to try out with sensitive IEMs on a powerful amp.
Thanks for the replies, both of you. I'm still new to measuring amps/dacs and I'm about to figure out how to do full-size headphones as well. I already use the Scarlett with an electret mic to measure IEMs as best I can.
-
I could actually use a variety of resistor adapters myself for DAC and amp measurements. I'm too dumb to know how to make them myself unless I have a guide full of pretty pictures.
-
A nice guide for a dummy load http://tangentsoft.net/audio/hp-dummy.html. Personally I would use BA drivers for low ohm loads. essentially you have a split cable to the interface and to the load.
-
I would need a guide more like what Pete Millet did for the DIY Butte amp (here are your parts, here are your pretty step-by-step pictures, so plop these parts here and solder away, don't worry about anything else, because I've done it all for you). Yes, I'm that dumb in this area. Using test equipment and software, I can eventually figure that out and concepts behind it...mostly. Sometimes. Maybe. But anything that falls in the circuitry or electrical engineer area, I'm essentially a moron. Trying to learn what I can, slowly, but that area is not my brain's forte. So, if you know anything like that for both an RCA load for DACs and something with a 1/4" connector for amps, that would be absolutely wonderful for me. That link does a good job with the parts, but everything else is not laid out for total idiots like myself. Does that even have an output to send to a measurement device?
-
I would need a guide more like what Pete Millet did for the DIY Butte amp (here are your parts, here are your pretty step-by-step pictures, so plop these parts here and solder away, don't worry about anything else, because I've done it all for you). Yes, I'm that dumb in this area. Using test equipment and software, I can eventually figure that out and concepts behind it...mostly. Sometimes. Maybe. But anything that falls in the circuitry or electrical engineer area, I'm essentially a moron. Trying to learn what I can, slowly, but that area is not my brain's forte. So, if you know anything like that for both an RCA load for DACs and something with a 1/4" connector for amps, that would be absolutely wonderful for me. That link does a good job with the parts, but everything else is not laid out for total idiots like myself. Does that even have an output to send to a measurement device?
It looks to me like it doesn't have an output, but expects you to use a Y splitter and hook one of the ends up to this. Which means you could just use a Y splitter and hook one of the sides up to a headphone if you don't care about it surviving tests since the volume will be really loud. I'm wondering if my resistor adapter will do the job, though. 75 Ohms of extra load? Dunno. I'll try it out this weekend.
-
I use a 2i2 for headphone, amp and DAC measurements. The line in should be high impedance. Like most and any line input. You need indeed a Y spliter and loads. I made my own loads using two resistors per load going from + to GND and - to GND. The loads are soldered to ab RCA male connector. I made a pair w 4 33 ohm resistors, 300 ohm resistors, and 150 ohm resistors. I can't remember what type I used but I think I used Metal film. I'll make more to cover the ones w decide are stadard here.
I will qualify the 2i2 interface w the O2 n maybe the ODAC Rev B. I've already done measurements w the Sansa Zip n think I got very similar results than NwAvGuy. Except I don't think he pointed out some issues around the expected distortion values.
Loopback shows the 2i2 is decent as a single ended DAC n very good as a balanced interface.
When I get back from vacations I'll try to take some pics.
-
Ultra, any chance you'd be able to make me some loads w/ RCA? Removes the idiot, me, from the equation. I'd pay for parts, labor, and shipping, of course. PM me if you want.
-
Ill see what I can do Hands. Will be sometime next week. Im still out of town. But it's relatively easy to do.
-
I'm in no hurry. Can work with a good guide for dummies, too.
-
I would suggest a build using XLR connectors? You would be able to test with a bunch of other stuff this way with the appropriate adapters.
-
Been re-reading on the 2i2:
http://d3se566zfvnmhf.cloudfront.net/sites/default/files/focusrite/downloads/7317/scarlett-2i2-user-guide-v2.pdf
The Line-In option offers much better specs for measurement than the Instruments option (which at -72dB THD seems quite horrible for low distortion DAC measurements), proly related to being lower gain. According to specs the best the Line-In interface can do is about -91 dB of THD+N which is about 0.0028%, which I think is not too bad if correct. But need to check how to tweak things to achieve that and if it's practical to do so.
I'm currently checking results with the ODAC and the interface seems to like 0.319 Vrms. I will see if I can relate that to the gain setting in the 2i2 spec for line-in, but it could be that this corresponds to 0 dB of gain (i.e. min gain) and if so, will run a 2i2 loopback test at this setting to verify claims.
I will also try to see separately what the 2i2 can do when configured for the tests performed here: https://www.quantasylum.com/content/Products/QA400/QA400Specs.aspx
It seems to me that most interfaces have a sweet spot for measuring. But I'm still getting up to speed.
For measuring things like a Clip+, I think the 2i2 can do pretty good, but not sure if it's upto snuff for certain low distortion equipment... yet.
If feasable, hopefully out of this a tutorial or whatever can be eventually put together to use the 2i2 for measurements, again, if feasable or at least to get the most out of it. Hopefully too, some of the things found can be applied to other interfaces.
-
OK. So I get far better results if I use loopback in balanced (differential mode) than in single ended mode. I knew this before but I didn't look too much further into it. I'll see what I can do to exploit this. Perhaps by using a differential prove like the QA400 guyses are suggesting in their app notes here and there. Will report back...
-
Recently I posted some ODAC measurements using the 2i2 here:
http://www.changstar.com/index.php/topic,442.msg78440.html#msg78440
I do not believe these measurements are accurate, specially the 2.0 (0 dBFS) ones. I believe the problem comes down to the 2i2 (and other entry level pro interfaces) unbalanced line-in performance.
Consider the following loopback measurements done at 0 dBFS (1.14Vrms for the 2i2):
Unbalanced
THD 1kHz
(http://www.changstar.com/index.php?action=dlattach;topic=2693.0;attach=10685;image)
SMPTE
(http://www.changstar.com/index.php?action=dlattach;topic=2693.0;attach=10687;image)
Balanced
THD 1kHz
(http://www.changstar.com/index.php?action=dlattach;topic=2693.0;attach=10689;image)
SMPTE
(http://www.changstar.com/index.php?action=dlattach;topic=2693.0;attach=10691;image)
Clearly balanced shows superior performance. Stuff like this has been discussed elsewhere in the context of the EMU0404:
http://www.diyaudio.com/forums/equipment-tools/276823-usb-vs-pcie-sound-card-audio-analysis-14.html#post4415566
However, in my setup, it seems the different connections make a sizable difference, as was discussed here:
http://www.changstar.com/index.php/topic,1163.msg31210.html#msg31210
In the Vali thread the unbalanced measurement was at -7 dBFS, while here it is at 0 dBFS. Unbalanced performance improves as the input levels hit somewhere around 0.3 Vrms, but are worse as voltage levels increase. For the ODAC measurement, 0 dBFS is about 2.0 Vrms and measurements would be completely off IMO. To get the most of this interface w/o modifications I would recommend to measure using levels close to 0.3 Vrms which is of course a severe limitation (and even then performance will not be all that great).
So I do not currently recommend the 2i2 for accurate measurement of low distortion unbalanced DACs and Amps. Can help get an idea though.
A QA400 is likely to yield better results. Getting an AP, SR or D-Scope is proly even better. But we are going from a $150 do-it-all interface, to a $300 dedicated setup, to kilobuck gear. And my audio piggy bank is currently drained. Will have to wait.
-
I have a Focusrite Scarlett 2i4 and I'm starting to take measurements of amps and dacs.
I picked up a couple of Hosa Pro cables, one 3.5mm > Dual XLR, and one RCA > Dual TS.
I also have a Sound Blaster ZxR sound card with a discrete input card.
My issue is that theoretically the Scarlett should be really good for measurements, but I get far worse readings on the Scarlett than the Sound Blaster, and I don't know why.
For example, on my ODAC revB + O2, I get: Very good, Average x3, Good x3, Poor x1. On the Sound Blaster, I get Excellent x7 and Good x1.
Any advice for using the Scarlett to effectively take amp measurements? I tried Line/Inst/Pad combos and none seemed to make it much better. Should I give up on it and just use the ZxR?
Thanks!
Read your post a little more carefully.
Dude, The Sound Blaster ZxR is a little more expensive than a 2i2 and perhaps even the 2i4. Futhermore, it's a PCIe card, not a USB powered interface. I think the ADC interface there is optimized for unbalanced input instead of pro-gear balanced input.
I actually expect the ZxR to kick the Scarlett line's ass in Line-in DAC measurements. What the ZxR lacks however is better mic pre-amps for speaker/headphone measurments and artistic recording. It's a bit of a different beast that sits inside a PC.
If you want something similar I read great things about this PCI alternative (they also have a PCIe equivalent):
http://www.esi-audio.com/products/julia/
Or you can just buy a QA400 if you don't want it to be sitting inside your PC.
If sticking to what you got, by all means do your DAC/Amp measurements with your ZxR and speaker/headphone measurements with your Focusrite.
BTW, I would love to see your results using your ZxR interface.
-
Derp, I forgot to post my measurements last night. I have measurements for the ODAC revB both on the Scarlett and ZxR. The latter looks much better. All of my measurements on the ZxR have kinda low dynamic range compared to what they claim, so I think I'm doing something a bit wrong there.
I'm also planning on bringing my gear home from work to verify my measurements.
What software are you using for your measurements? I've just been using REW (Room EQ Wizard) for earphones and RMAA (RightMark Audio Analyzer) for amps/dacs.
-
Derp, I forgot to post my measurements last night. I have measurements for the ODAC revB both on the Scarlett and ZxR. The latter looks much better. All of my measurements on the ZxR have kinda low dynamic range compared to what they claim, so I think I'm doing something a bit wrong there.
I'm also planning on bringing my gear home from work to verify my measurements.
What software are you using for your measurements? I've just been using REW (Room EQ Wizard) for earphones and RMAA (RightMark Audio Analyzer) for amps/dacs.
I strongly dislike RMAA. Almost completely useless. Lots of information left out.
Here are some initial measurement standards that we've agreed upon: http://www.changstar.com/index.php/topic,2642.0.html
-
Derp, I forgot to post my measurements last night. I have measurements for the ODAC revB both on the Scarlett and ZxR. The latter looks much better. All of my measurements on the ZxR have kinda low dynamic range compared to what they claim, so I think I'm doing something a bit wrong there.
I'm also planning on bringing my gear home from work to verify my measurements.
What software are you using for your measurements? I've just been using REW (Room EQ Wizard) for earphones and RMAA (RightMark Audio Analyzer) for amps/dacs.
I acutally use REW for Amps/DACs also. One could use ARTA as well, but it was easier for me to get things going with REW. I basically have two crap laptops (one over 8 years old). One runs REW as a source and the other one runs REW as a line-in receive end. The latest REW has distortion numbrs on their Real Time Analyzer, but you have to set the source and the receiving end generators the same, even if the receiving end is not being used as a source. I'm not certain if the distortion numbers work with non-standard exitation (SMPTE and so on), but you can certainly get a plot and do a dB to % conversion by hand.
Like Marv said, be very careful about RMAA. NwAvGuy was right to question it. You can get it to work but it's a bit of a pain and there is little flexibility in the freeware version. Not sure if there is a paid version that is more flexible. But I can tell you that I had a terrible time getting it to work with the 2i2 using ASIO at high rates. REW should run ASIO fine at all rates on the 2i2 and likely the 2i4. Furthermore, you want to use ASIO all around. It makes a difference in your results. For the ODAC, you can use ASIO4All or something like that. It took a little work, but I got that to work in the end. ASIO makes a difference in the ODAC performance.
You should be able to do all of the measurements that Marv linked. The agreement was to use 0.55 Vrms for Amps, but for DACs you may want to do 0 dBFS (full-scale) or maybe -1 dBFS (if DAC is clipping badly for x or y reasons, or even less if your interface has issues at those levels like my 2i2). Usually 0 dBFS this is about 2.0 Vrms or so. But some other DACs may go lower or higher. For example, the 2i2 is about 1.14 Vrms I think at 0 dBFS. The reasons for using 0 dBFS is that in many cases the DAC does not have volume control and are indeed run full scale (digital volume control all the way up) while the amp in the chain takes care of gain (pretty, shinny, sexy volume control knob).
Furthermore, don't use loads with DACs. A DAC will usually see a high impedance anyway. That loading stuff applies to Amps which have to deal with headphone/speaker crazyness.
-
These were my original measurements in RMAA, one with the ZxR's dBPro, and the other with the Scarlett 2i4.
If I wasn't so busy at work that I forgot to bring it home, I would have measured with other software today. Instead, this is what you get for now.
http://www.basshead.club/dac/ODAC_revB_ZxR.html
http://www.basshead.club/dac/ODAC_revB_Scarlett.html
The ZxR measurements were RCA -> RCA, and the Scarlett measurements were RCA -> Dual 1/4 Inch.
-
The Scarlett measurements look sort a weird. Try REW better and check your voltage levels.
-
Okay... Just measured the Vorzuge VorzAMP Duo with a 75 Ohm load using a simple resistor adapter (http://www.ebay.com/itm/291329605627?_trksid=p2050601.m570.l5999&_trkparms=gh1g%3DI291329605627.N36.S2.R3.TR3). The measurements are FAR better than without load.
http://www.basshead.club/measurements/vorzuge_vorzamp_duo/
I need to figure out how to use REW for amp measurements. I'm going to try the Scarlett next.
Update: Scarlett can't friggin' measure my amps to save its life. It seems to work fine for headphones via an electret mic, but it just doesn't cut it for amp measurements.
Which is fine for me, because the DBpro is working great for it. Now I have a couple of nice Hosa Pro cables that I have no need for, though... anyone need a 3.5mm -> Dual XLR or RCA -> Dual TS?
-
Well, the 2i2 48V phantom power mic interface is balanced, so that helps. Pretty much every other Pro audio interface I've found is balanced, and seem to do well balanced. Unbalanced... Maybe the EMU doesn't suck that bad, but still...
The DBpro is designed to work well unbalanced.
As far as 75 Ohms doing awesome vs no-load, that's weird. Because for the same amount of voltage a 75 ohm load will demand more current than a 10k ohm load (which is "no-load"). So far I've not seen a situation where a no-load (open) will do worse than a loaded measurement.
-
I'll re-run the tests with 75 Ohm load and no load.
Here's a fun little comparison between volume levels (I always see debates of 100% PC volume vs. 99% vs. 80%):
http://www.basshead.club/measurements/vorzuge_vorzamp_duo/vol_comparison.html
-
Seems to roll a bit too soon in the lows. This is not loaded or 75 ohms? BTW, I maybe reading things wrong, but you are getting about 17 something bits of resolution, assuming I'm doing the math right and the ADC is not limiting readings. Not too bad if that's the case.
-
Just did loaded vs. unloaded and the results are virtually identical. I realized that previously, I was running from an ODAC to the amp, and now I'm running from a ZxR... apparently the ZxR has a better DAC implementation than the ODAC.
As for the lows rolling off too soon, that is really there. I could hear it rolling off before I even measured, and all of my results with the DBpro and Scarlett both in REW and RMAA show the rolloff.
It's strange that it does that, but it definitely does. It has a crazy switch that pumps the bass up +15dB and a treble switch that pushes treble up ~5dB.
http://i0.wp.com/www.basshead.club/wp-content/uploads/2015/08/VorzAMP_duo_eqduo.png
-
LOL! I wouldn't be surprised that much if the ZxR > ODAC. I would proly joint the Massdrop (https://www.massdrop.com/buy/creative-sound-blaster-zxr-sound-card)deal on it if I had a desktop. As it is I only have 2 laptops.
As far as the roll off, I was wondering about the roll off and the loading because sometimes there are coupling caps that reduce the low frequency extension depending on how small the load is. I would expect the lows to roll less w/o load (1k ohms) vs load (1k || 75 ~ 75 ohm).
-
The ZxR measures really well.
http://www.basshead.club/measurements/zxr/zxr.html
I've always felt like the ACM module (volume control pot) added a little noise to the signal, so I tested it:
http://www.basshead.club/measurements/zxr/zxr_acm.html
Yep.
http://www.basshead.club/measurements/zxr/acm_comparison.html
-
Last measurement of the night. For grins, I checked my onboard sound.
It's less of a train wreck than I expected, probably because it's about the best onboard sound you can get (SupremeFX 2014 aka Realtek 1150).
http://www.basshead.club/measurements/maximus_vii_hero/