manueljenkin
Member
- Joined
- Nov 4, 2018
- Messages
- 43
- Likes
- 10
Hi, I would like to have some guidance on making my measurement rig for measuring DACs, amps and headphones.
Of first priority is DACs because I believe I have most of the necessary gear and I have been able to personally design the tones needed to measure the same. I personally feel that audio measurements should go beyond just single tone sine sweeps and IMD burst to capture the different things that we perceive as auditory pleasure. My intention is not to create a super precise measurement rig, Amir has an excellent one and his measurements on those parameters can be taken as the gold standard. My intention is to convince Amir to measure more parameters for DACs (and media players and USB cables) that I find audible and worthy and design test tones for the same, and hopefully get a recommendation letter from him for my University admissions. Since the admission deadlines are close by I'm trying to get it done quick with what I have.
I have personally designed test tones and scenarios and I have been able to hear the difference audibly. I just need to measure it to validate it. A friend has already measured the difference between winyl (an audio player based on BASS audio library) and foobar2000, both in Asio configuration. I currently have a focusrite Scarlett 2i2 interface and I ordered a 50$ usb oscilloscope which will reach me soon. But I guess I need other plugs.
Please recommend me a way to use my focusrite 2i2 to record output from my dac (Apogee groove). The focusrite has two 3 pin xlr supporting phantom power. Would connecting the two grounds of the interface together cause any short/issues? Maybe measuring even a single channel is enough for me. Less worried about it passing through the preamp section of the interface since I'm only trying to get some of the highly audible parameters measured and I'm only looking at the delta. Also should I connect it directly to the interface or should I connect it to a resistor and connect the inteface connections in parallel to it?
For the oscilloscope, connecting the dac lines to a resistor and measuring it using the oscilloscope in parallel would do fine right?
Another area I'd like advice for would be in Microphones for measuring headphone transients. It is not of immediate interest since I'm still in the process of coding my test scenarios but it is better that I procure the necessary gear in advance. My interface would preferably still be the 2i2, but I would be happy to upgrade/change if the benefits are worthwhile.
I'll be building my diy rig and will consider a lot of aspects like ear matching (it'll be a modified flatplate accounting for ear variations), rig stability (not get rattled by tactility of headphones) etc. I've come to realize that a single type of mic may not be good enough to capture all parameters of a headphone. I have a couple of Presonus PRM1 measurement mics that I used for my University project on sound source localization (Training computer to localise sound source like humans do). It performed well but it had limitations. It worked well in a fairly enclosed space (and constrained noise conditions) though front back localization suffered a little both due to dynamic range compression and also due to geometry of my dummy setup (dint have proper earlobe structures). I restricted the space to a 2d plane so I couldn't analyse vertical accuracy. While I think it will suffice to measure fr, phase and overall power handling, it is specifically designed for low dynamic range (probably for better noise rejection) which would leave the dynamic range measurement out of the equation. It works most ideal when the input is between 70db and 90db I guess. I also have a lewitt lct440 pure that i use for my personal applications. Would be nice if i can dual purpose it for some of my other science pursuits.
Ideally I'd want 3-4 different setups for measuring different parameters. I'd like to have guidance on what are all the things I need to know about specific mics. I have only some basic knowledge of condenser, dynamic and ribbon mics and very few knowledge of pickup pattern, self noise and dynamic range choices of different mics. I do know that universally ideal mics/materials don't exist for all purposes and a mic meant for capturing voices may not be effective in capturing drum kits (might clip or get destroyed) and vice versa (might get compressed).
Of particular interest for me is to try and measure the transients (response to sharp stimuli), low level detail (responding to loud input followed by fairly quiet input), and a few other things that I don't believe is captured in a frequency response chart calculated through sine sweeps. I wouldn't want the mic to add its own aberration like - transient lag, proximity effect, dynamic range compression etc into it. The setups will initially be used for characterizing headphones using my test tones but eventually I'll use them for spectral and spatial analysis of other sounds (bike engine, electrical whine etc) some in close proximity, some with lots of distance. I am working on a few papers on headphones and a few other papers on sound in general. I plan to measure and quantify the effects of different amplification topologies, design choices in headphones etc. I think measuring with a R load is not fully representative of how it will drive a headphone since headphones exhibit other phenomenon like back-emf which can destabilise an otherwise stable amplification system. Trying to fit everything into output impedance also seems far fetched since output impedance may not just be output resistance but it could be more parameters. I'm trying to characterize all non-linearities, and behavior of pairing. I intend to measure artefacts like phase-skew as well.
I wouldn't want a pre built system like mini dsp ears since I'm indifferent to pre built plastic structures on their interaction to headphones that are tactile enough to rattle the rig.
I am trying to restrict my budget to 2-3 mics each of 200-250 dollars but would be happy to stretch the budget if improvements are worthwhile. As an alternative of one mic does all, I also checked at a few mics that were around 500$ ish. One of them is earthworks sr25. It seems to have a power handling and detail capability and since it's meant to mic loud sources close I think they'd have optimised it to minimise proximity effect. They advertise that it doesn't exhibit proximity effect but I'm quite skeptical of marketing speak from most companies.
What is the mic setup that zeos used to have (the older one)? I remember that being fairly good sounding even when passed through YouTube compression algorithm, so can be nice to have as a backup in case I encounter issues with my other mics.
Of first priority is DACs because I believe I have most of the necessary gear and I have been able to personally design the tones needed to measure the same. I personally feel that audio measurements should go beyond just single tone sine sweeps and IMD burst to capture the different things that we perceive as auditory pleasure. My intention is not to create a super precise measurement rig, Amir has an excellent one and his measurements on those parameters can be taken as the gold standard. My intention is to convince Amir to measure more parameters for DACs (and media players and USB cables) that I find audible and worthy and design test tones for the same, and hopefully get a recommendation letter from him for my University admissions. Since the admission deadlines are close by I'm trying to get it done quick with what I have.
I have personally designed test tones and scenarios and I have been able to hear the difference audibly. I just need to measure it to validate it. A friend has already measured the difference between winyl (an audio player based on BASS audio library) and foobar2000, both in Asio configuration. I currently have a focusrite Scarlett 2i2 interface and I ordered a 50$ usb oscilloscope which will reach me soon. But I guess I need other plugs.
Please recommend me a way to use my focusrite 2i2 to record output from my dac (Apogee groove). The focusrite has two 3 pin xlr supporting phantom power. Would connecting the two grounds of the interface together cause any short/issues? Maybe measuring even a single channel is enough for me. Less worried about it passing through the preamp section of the interface since I'm only trying to get some of the highly audible parameters measured and I'm only looking at the delta. Also should I connect it directly to the interface or should I connect it to a resistor and connect the inteface connections in parallel to it?
For the oscilloscope, connecting the dac lines to a resistor and measuring it using the oscilloscope in parallel would do fine right?
Another area I'd like advice for would be in Microphones for measuring headphone transients. It is not of immediate interest since I'm still in the process of coding my test scenarios but it is better that I procure the necessary gear in advance. My interface would preferably still be the 2i2, but I would be happy to upgrade/change if the benefits are worthwhile.
I'll be building my diy rig and will consider a lot of aspects like ear matching (it'll be a modified flatplate accounting for ear variations), rig stability (not get rattled by tactility of headphones) etc. I've come to realize that a single type of mic may not be good enough to capture all parameters of a headphone. I have a couple of Presonus PRM1 measurement mics that I used for my University project on sound source localization (Training computer to localise sound source like humans do). It performed well but it had limitations. It worked well in a fairly enclosed space (and constrained noise conditions) though front back localization suffered a little both due to dynamic range compression and also due to geometry of my dummy setup (dint have proper earlobe structures). I restricted the space to a 2d plane so I couldn't analyse vertical accuracy. While I think it will suffice to measure fr, phase and overall power handling, it is specifically designed for low dynamic range (probably for better noise rejection) which would leave the dynamic range measurement out of the equation. It works most ideal when the input is between 70db and 90db I guess. I also have a lewitt lct440 pure that i use for my personal applications. Would be nice if i can dual purpose it for some of my other science pursuits.
Ideally I'd want 3-4 different setups for measuring different parameters. I'd like to have guidance on what are all the things I need to know about specific mics. I have only some basic knowledge of condenser, dynamic and ribbon mics and very few knowledge of pickup pattern, self noise and dynamic range choices of different mics. I do know that universally ideal mics/materials don't exist for all purposes and a mic meant for capturing voices may not be effective in capturing drum kits (might clip or get destroyed) and vice versa (might get compressed).
Of particular interest for me is to try and measure the transients (response to sharp stimuli), low level detail (responding to loud input followed by fairly quiet input), and a few other things that I don't believe is captured in a frequency response chart calculated through sine sweeps. I wouldn't want the mic to add its own aberration like - transient lag, proximity effect, dynamic range compression etc into it. The setups will initially be used for characterizing headphones using my test tones but eventually I'll use them for spectral and spatial analysis of other sounds (bike engine, electrical whine etc) some in close proximity, some with lots of distance. I am working on a few papers on headphones and a few other papers on sound in general. I plan to measure and quantify the effects of different amplification topologies, design choices in headphones etc. I think measuring with a R load is not fully representative of how it will drive a headphone since headphones exhibit other phenomenon like back-emf which can destabilise an otherwise stable amplification system. Trying to fit everything into output impedance also seems far fetched since output impedance may not just be output resistance but it could be more parameters. I'm trying to characterize all non-linearities, and behavior of pairing. I intend to measure artefacts like phase-skew as well.
I wouldn't want a pre built system like mini dsp ears since I'm indifferent to pre built plastic structures on their interaction to headphones that are tactile enough to rattle the rig.
I am trying to restrict my budget to 2-3 mics each of 200-250 dollars but would be happy to stretch the budget if improvements are worthwhile. As an alternative of one mic does all, I also checked at a few mics that were around 500$ ish. One of them is earthworks sr25. It seems to have a power handling and detail capability and since it's meant to mic loud sources close I think they'd have optimised it to minimise proximity effect. They advertise that it doesn't exhibit proximity effect but I'm quite skeptical of marketing speak from most companies.
What is the mic setup that zeos used to have (the older one)? I remember that being fairly good sounding even when passed through YouTube compression algorithm, so can be nice to have as a backup in case I encounter issues with my other mics.
Last edited:
