|
Post by kcatthedog on May 30, 2015 12:33:00 GMT -6
well remember that the experience of hearing the sound is visceral whereas reading a two dimensional graph on paper or a screen is not; simply intellectual.
the graphs represents elements of the sound ; it is not the sound and there in lies the difference.
|
|
|
Post by svart on May 30, 2015 12:50:11 GMT -6
well remember that the experience of hearing the sound is visceral whereas reading a two dimensional graph on paper or a screen is not; simply intellectual. the graphs represents elements of the sound ; it is not the sound and there in lies the difference. Yes, but measured data can tell you whether what you hear is a function of the circuitry, or a function of the setup, or something like a null in the room that wasn't apparent due to the attributes of a different piece of gear. The graph shows that electrically, the low end seems OK on my DAC. If he hears missing low end, then I have to say that according to the graph, there is something else happening here that accounts for the missing low end, I simply want to figure out what that is.
|
|
|
Post by tonycamphd on May 30, 2015 13:07:54 GMT -6
Could it be a voltage rail thing? Will be interesting when u find out. I will say it again though, identical frequency plots rarely translate to identical sounds, I screw with all kinds of shit in the mic's I build, and I hear real diffs from equal values/different manufacturers parts, a lot of times with a single swap!
|
|
|
Post by kcatthedog on May 30, 2015 13:33:24 GMT -6
I understand svart and while I am sure the symphony sounds superb I also believe that it has a designed tailored sound just like the 2192, burl b2 and your box; the difference is yours was tailored to be linear. So, I also understand that you are considering all this technologically in terms of the design effect, but but I feel we are back to where we started in the sense of do people want the large part of their final mix sound built on the the sound of their box's converter: enhancing or flat; both work it comes down to preference. The recorder/engineer/producer and audiophile listener preference are likely different
|
|
|
Post by Martin John Butler on May 30, 2015 13:43:53 GMT -6
Can one of you guys just make me something that sounds like a Burl A/D-D/A for $500? Uhh.. please.
|
|
|
Post by svart on May 30, 2015 14:53:39 GMT -6
Ok, Ive spent all morning running tests. This is feeding out from the DAC directly into my SSL alphalink. Doing loopbacks through the alphalink show it flat down to around 10hz up to beyond 50K, so it shouldn't affect the graphs much, although it might contribute to making the DAC look more rolled off below 10hz, as if anyone cared about anything below 10hz.. First, a frequency plot of 5hz to 30khz over a 10 second period: Since there seems to be questions over the low end, I ran some burst tests at -1dB out from 5hz to 100hz in 5hz/3 second intervals to see what the response was and if the level changed if the unit dwelled on a frequency for a few seconds, rather than simply sweeping through it. It's better than 1dB flat down to 15hz, with a -3dB point around 5hz: 5hz: -4dB 10hz: -2.3dB 15hz: -1.8dB 20hz: -1.6dB 25hz: -1.5dB 30hz: -1.4dB 35hz: -1.4dB 40hz: -1.4dB 45hz: -1.4dB 50hz: -1.4dB 55hz: -1.4dB 60hz: -1.4dB 65hz: -1.4dB 70hz: -1.4dB 75hz: -1.4dB 80hz: -1.4dB 85hz: -1.4dB 90hz: -1.4dB 95hz: -1.4dB 100hz: -1.4dB I also decided to look at the whole mid-low end, so I ran the same test from 100hz to 300hz, in 25hz/3 second intervals: 100hz: -1.4dB 125hz: -1.4dB 150hz: -1.4dB 175hz: -1.4dB 200hz: -1.4db 225hz: -1.4db 250hz: -1.4dB 275hz: -1.4dB 300hz: -1.4dB I honestly still don't see a problem with the low end. It's perfectly flat as far as I can tell.
|
|
|
Post by kcatthedog on May 30, 2015 15:17:58 GMT -6
Thx svart, ah correct me if I am wrong but if you are essentially flat within 1 - 1.4 db, isn't that a much tighter tolerance than most products post , i.e. we are used to hearing sonic foot prints with 3-5 db ranges ?
If people want a converter that is imparting a sound to their mix, then your converter is just not the right one for them and or run your signal through something before its hits your converter ?
|
|
|
Post by svart on May 30, 2015 15:23:23 GMT -6
Thx svart, ah correct me if I am wrong but if you are essentially flat within 1 - 1.4 db, isn't that a much tighter tolerance than most products post , i.e. we are used to hearing sonic foot prints with 3-5 db ranges ? If people want a converter that is imparting a sound to their mix, then your converter is just not the right one for them and or run your signal through something before its hits your converter ? Most products specify -3dB points. So if they say "20hz-20khz", they usually mean that it's within 3dB flatness in that range. Mine will be something more like 5hz-50Khz+ at -3dB. No, this converter will impart almost zero sound on something. I'd probably want something with transformers, tubes or class A after it if you want some kind of euphonic distortion. Martin John Butler, I'm thinking about the transformer out option. I think a couple of transformers with varying current through them might work as an adjustable harmonic box. I'll look into it, maybe I'll make it an add-on box that you plug into the ADC/DAC.
|
|
|
Post by NoFilterChuck on May 30, 2015 15:43:44 GMT -6
before you add transformers, can you add a volume knob?
|
|
|
Post by NoFilterChuck on May 30, 2015 15:51:37 GMT -6
Ok, Ive spent all morning running tests. This is feeding out from the DAC directly into my SSL alphalink. Doing loopbacks through the alphalink show it flat down to around 10hz up to beyond 50K, so it shouldn't affect the graphs much, although it might contribute to making the DAC look more rolled off below 10hz, as if anyone cared about anything below 10hz.. Why wouldn't you go from the DAC on yours into the ADC on yours and back into the computer via spdif on the Alphalink? Isn't that assuming the Alphalink is an archive quality converter? That's how I did the sweeps. Symphony DAC -> Symphony ADC, and Symphony SPDIF out -> Svart DAC -> Svart ADC -> Symph SPDIF in. Apogee wrote me back about the spdif clocking thing. Sunday i'll be testing it with the MOTU as the SPDIF connection. And then some time this week, we'll be doing a hang hopefully with teh other NYC guys. someone said they would bring a BLA Sparrow for comparison as well.
|
|
|
Post by Martin John Butler on May 30, 2015 15:54:21 GMT -6
Martin John Butler, I'm thinking about the transformer out option. I think a couple of transformers with varying current through them might work as an adjustable harmonic box. I'll look into it, maybe I'll make it an add-on box that you plug into the ADC/DAC. Whoa Svart, now you really got my attention. I'm sure I would be thrilled with "The Box" as is, but man, if there was a way to impart slight warmth and tone the way a Burl does, it would be fantastic, and kind of a no brainer. You could conceivably use it without transformers for some tracks and with for others.
|
|
|
Post by NoFilterChuck on May 30, 2015 16:22:42 GMT -6
I'll see what i can do. But on a side note, i hooked up the demo board from TI tonight and shot it out with my sound card inputs. Let's just say that the difference wasnt subtle in the least. Is the alphalink your sound card? If it is, how can you use it to measure your device when the TI Demo board sounds better (and is probably more accurate) than your sound card?
|
|
|
Post by NoFilterChuck on May 30, 2015 16:25:45 GMT -6
well remember that the experience of hearing the sound is visceral whereas reading a two dimensional graph on paper or a screen is not; simply intellectual. the graphs represents elements of the sound ; it is not the sound and there in lies the difference. Yes, but measured data can tell you whether what you hear is a function of the circuitry, or a function of the setup, or something like a null in the room that wasn't apparent due to the attributes of a different piece of gear. The graph shows that electrically, the low end seems OK on my DAC. If he hears missing low end, then I have to say that according to the graph, there is something else happening here that accounts for the missing low end, I simply want to figure out what that is. I'll pull out that REW software and sweep the room using both the symphony and your box and post the plots when I'm home. It'll be interesting to compare the two with that software.
|
|
|
Post by LesC on May 30, 2015 16:33:04 GMT -6
before you add transformers, can you add a volume knob? The last thing I would want on an AD-DA is a volume knob. Maybe I could see it on an interface, especially for a headphone out, but not on an AD-DA. I need a monitor controller anyway, with four sets of monitors that I use for various purposes, so an extra "volume knob" on DA would be less than useless. I realize we have different requirements, you need a volume knob, but a high-quality volume knob would probably be a relay-controlled resistor ladder, which would add enormously to the cost.
|
|
|
Post by Johnkenn on May 30, 2015 17:25:15 GMT -6
Yeah. Again, the TC Level Pilot is like $69.
|
|
|
Post by tonycamphd on May 30, 2015 17:46:25 GMT -6
before you add transformers, can you add a volume knob? The last thing I would want on an AD-DA is a volume knob. Maybe I could see it on an interface, especially for a headphone out, but not on an AD-DA. I need a monitor controller anyway, with four sets of monitors that I use for various purposes, so an extra "volume knob" on DA would be less than useless. I realize we have different requirements, you need a volume knob, but a high-quality volume knob would probably be a relay-controlled resistor ladder, which would add enormously to the cost. yeah, built in attenuators usually drive the cost up and sound like poop to boot.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 30, 2015 20:44:36 GMT -6
+1 against volume knobs in converters. A good volume knob is a rotary switch. People can decide how to route and control the volume, digitally before the converter- from a cheap passive pot volume controller like there are plenty up to a passive routing box with rotary switch like the better NOS or even a high end active volume control and routing matrix like SPL and up. Options from 50 to several hundred dollars and more. I can't see people getting excited about changing the pot in a high end converter chassis as soon as it's stereo tracking went off or when they feel like they need more steps in the rotary. Everyone can be fine with his own solution - after the converter. I read people complaining that otherwise pretty good converters HAVE a volume knob that degrades stereo tracking or sound and you try to debug what the problem is...odd stuff. Personally i use a RM in the bedroom with a cheaper volume control, a below thousand bucks superbeast with a 150$ passive rotary controller into an 8k$ amplifier...i don't know what kind of guy i am... 8-0 It's great that the last generation of converters now IS affordable to anyone, from the audio enthusiast to the bedroom producer up to mastering guys. *Less* options, that can go wrong mechanically or even degrade the sound successively over time (pots), are better.
|
|
|
Post by svart on May 30, 2015 20:53:06 GMT -6
I'll see what i can do. But on a side note, i hooked up the demo board from TI tonight and shot it out with my sound card inputs. Let's just say that the difference wasnt subtle in the least. Is the alphalink your sound card? If it is, how can you use it to measure your device when the TI Demo board sounds better (and is probably more accurate) than your sound card? The alphalink is SSL's flagship converter system.. www.sweetwater.com/store/detail/AlphaLinkMAXNo, my soundcard is an Echo MIA and WAS used as an analog mixdown input and output for monitoring. I have since ceased using the analog I/O on the MIA and only use the SPDIF I/O to my ADC/DAC. I used the alphalink as an input source because it has specs of 20-20K at +0/-0.5. It's at least as good as my ADC, but I wanted to use a "big name" device with excellent specs as a control for testing my DAC. This allows me to not worry about how my ADC would effect the DAC, and simply test the DAC only. I'll do the same thing with the ADC, by using the Alphalink output into the ADC. I'll finish by doing a round trip from the DAC to the ADC.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 30, 2015 21:12:10 GMT -6
As for the different response curves shown... Yes, these are measurements. They say alot. Both fine boxes witch are for sure very useful. They have different qualities. As in rice and noodles. Not so obvious from the measurement, but this is not what we hear. Psychoacoustically, it makes a real difference if you go linear down with volume after a certain point. Bob Katz suggested a target response for a DRC system that goes linear down after 10k. Rhode&Schwarz used somehow similar curves. Even listening volume plays a huge role in the perception of a gears frequency response, too - nothing to do at all with the converter itself. (This alone makes me want to have a flat monitoring - so i can have at least something i can rely my ear on.) I.e. the lower the listening volume, the more the bass goes away - psychoacoustically! This already makes smaller differences you can SEE in a diagram pretty obviously different in perception. No voodoo, just science... (well, maybe I am mr. science...) So you may even measure out your rooms, and everything is flat and STILL you hear a lack of bass in this comparison for the one or the other psychoacoustical reason. AND we did not even talk about how people percieve different transient behaviour and sharpness. For me, a clinical converter is a good converter. Period. (Totalitarian approach.) You can always screw it up to your liking before or after. Saturation, transformers, roll-offs, whatever. But if it sounds like a wire, it's perfect. Like a good monitoring amp. (I leave the speakers out on purpose here so we don't end up in religious fights at this point. :-) )
|
|
|
Post by Johnkenn on May 30, 2015 22:49:08 GMT -6
We're not in a religious fight yet? lol
|
|
|
Post by tonycamphd on May 31, 2015 0:26:40 GMT -6
Ok, Ive spent all morning running tests. This is feeding out from the DAC directly into my SSL alphalink. Doing loopbacks through the alphalink show it flat down to around 10hz up to beyond 50K, so it shouldn't affect the graphs much, although it might contribute to making the DAC look more rolled off below 10hz, as if anyone cared about anything below 10hz.. First, a frequency plot of 5hz to 30khz over a 10 second period: Since there seems to be questions over the low end, I ran some burst tests at -1dB out from 5hz to 100hz in 5hz/3 second intervals to see what the response was and if the level changed if the unit dwelled on a frequency for a few seconds, rather than simply sweeping through it. It's better than 1dB flat down to 15hz, with a -3dB point around 5hz: 5hz: -4dB 10hz: -2.3dB 15hz: -1.8dB 20hz: -1.6dB 25hz: -1.5dB 30hz: -1.4dB 35hz: -1.4dB 40hz: -1.4dB 45hz: -1.4dB 50hz: -1.4dB 55hz: -1.4dB 60hz: -1.4dB 65hz: -1.4dB 70hz: -1.4dB 75hz: -1.4dB 80hz: -1.4dB 85hz: -1.4dB 90hz: -1.4dB 95hz: -1.4dB 100hz: -1.4dB I also decided to look at the whole mid-low end, so I ran the same test from 100hz to 300hz, in 25hz/3 second intervals: 100hz: -1.4dB 125hz: -1.4dB 150hz: -1.4dB 175hz: -1.4dB 200hz: -1.4db 225hz: -1.4db 250hz: -1.4dB 275hz: -1.4dB 300hz: -1.4dB I honestly still don't see a problem with the low end. It's perfectly flat as far as I can tell.If the symphony or SB is 1.4db up over the Svartacus at ALL those frequencies from 100-300, you could bet you'd hear it. Like I said, I haven't shot this out, but i did listen to some files, and i also read comments about "clinical" being paramount. What svarts doing here is awesome, and I respect him for having the balls to try it, and for being so far toward success. But I'm of the opinion that easy, natural, musical, and accurate are not words that are exclusive from one another, something does NOT have to sound clinical to be accurate, these analyzer readings mean nothing in that regard, trust your ears to guide your formidable EE chops, and try some things would be my suggestion, if this rig ends up being even remotely unsatisfying to listen to, who's gonna want to use it? Personally, I want something that sounds great to all that listen to it, represents accurately what's going on with the frequency range regarding amplitude, image accuracy concerning depth and stereo width, so i can confidently place tracks precisely within the soundstage, if ALL these essential aspects are not happening in a synergistic, musical way, I just won't dig it. I think it would behoove you to explore what is going on beyond the analyzer at this point, you're so close, you just need to do a little massaging to dial it in. Of course JMO.
|
|
|
Post by Johnkenn on May 31, 2015 9:56:31 GMT -6
This is what frustrates me. Who says there's a problem? Chuck? Dan said it sounded very, very similar to his un-modded Superbeast. There are people lining up for the 10 month wait for a Superbeast. No offense, but so far, Dan is the only person on the planet that I've heard of that felt the need to mod his Superbeast. And there is nothing wrong with the word "clinical," although most of the time it is used in a derogatory manner. Have you ever listened to a Lavry DA? The Lavry Gold has been one of the mastering standards for years - and it is prided on being "clinical."
I thought this DA was heads and shoulders better than the UA Apollo and I thought it was absolutely an improvement over the RM 1794 I use now. And I chose the 1794 DA over the Symphony. So - there is no consensus that there is a "problem" with the bass. I just want to make that clear.
|
|
|
Post by NoFilterChuck on May 31, 2015 10:33:56 GMT -6
How many people other than svart have actually heard the unit in their rooms besides you, dandeurloo and myself?
|
|
|
Post by Johnkenn on May 31, 2015 11:00:39 GMT -6
How many people other than svart have actually heard the unit in their rooms besides you, dandeurloo and myself? Chuck - you can be right...maybe this just isn't the box for you. "Dan made some suggestions and I greatly appreciate his testing work, but I've not heard anything from Jim as he hasn't tested a unit (not sure where that came from), but anyway, I have a couple people saying that it lacks bottom compared to known hyped units.. I also have some others saying it's perfect the way it is. I also have some audiophiles listening as well (you didn't think that you guys were the ONLY ones testing units did you?) and so far just some suggestions from them. I have frequency sweeps of my own showing that it's essentially flat down to almost 10hz. I've honestly found no electrical reason why anyone should think it's excessively light on the low end. In fact, the DAC is completely DC coupled from the DAC chips right through to the TRS connectors."
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 31, 2015 12:54:41 GMT -6
If this box sound *anything* like the superbeast...which is very likely, since it shares it's main topology and active components in the design and is designed to meet the same goal, and since Dan already confirmed (and i tend to believe that someone who feels the need to mod a superbeast has some very picky ears...), i would be *very* surprised if a lack of bass would be, what can be used to *characterize* it. The measurements also tell that there isn't. The superbeast, that i listened to yesterday the whole day again, is the most balanced and precise DAC i heard yet, throughout the sprectrum. If the converter is also shot out in audiophile circles, it is most likely it will be judged against emotivas or benchmarks or whatever, too. I know what it is made of and followed the threads from the beginning, and the only thing I would have liked to see would have been the hint of trying out different resistors in the critical audio path, as has been suggested by Jim, i think. But really - i have no clue if this would make a difference at all. If i needed another stereo quality converter, this box would be my bet, that's no secret, i wrote it already. Not because i am a fanboy, but because i know how it was designed and built, i know it's components from own listening, and because there is a consensus it sounds clinical like the superbeast - no matter if you like or don't like clinical. It must not sound *nice* or *good* but true to me. This said, i am a fan of K&H/Neumann monitors, because they sound so "boring", "ugly" and whatever is used to describe them, because they are so well translating. Sure, i could have bought some small genelecs instead, and everybody would go "yeah, this sounds nice", but i have heard quite some mixes that did not translate good from these sunshines... Clean flat devices are revealing, they may revealing your preferences in the rest of the listening chain, in your mixes and even in your personal listening perception. So you may like it or not. In the end everyone will have to make up his own mind on this. And, well - never i would try to mod the superbeast to make it sound more hifi, hehe. But what the heck, why not, if someone feels the need to squeeze the last bit of performance out of it's active components and design...
|
|