[russdog]

One of the guys at Pacific Valve writes a blog under the name Vic Trola. In his 9/13 post, he reported on using a DAC-AH-Modded both with and without their upsampler. He reports that sending an upsampled 96kHz signal to the DAC-AH-M made it sound way, way better. This was counter to his expectations, as he had expected upsampling to magnify the DAC-AH's weaknesses, not help it improve. (He was equally surprised that upsampling hurt the DAC-AM-Modded, as he had expected it to help it. At least he's not shy about admitting when his expectations were wrong.) You can read what he says about it here.

I'm waiting on a DAC-AH-M, but I don't want to use their $570 upsampler with it. Instead, I want to achieve the same effect by using the 24/96 MAD mp3-decoder that's a free add-on for MediaMonkey. When I first used it with my Echo Indigo I/O card, the effect was immediate and dramatic (much to my surprise).

Does anybody here know if feeding an upsampled signal helps their DAC-AH-M? If so, great, I'm set. If not, he makes the DAC-AM-M sound pretty nice without upsampling. So, if you have a DAC-AH-M, and if you have any way to see what upsampling does to it (e.g, MediaMonkey with the MAD add-on), I'd sure like to know what you think about this. (I'd especially like to know *before* my DAC ships

 

[russdog]

Here's what he said about feeding an upsampled 96kHz signal to it...

Quote:

Originally Posted by cut and paste from Vic Trola's blog The DAC AH Modded I was not sure what to expect when the up sampler was used with a NOS DAC. My first thought was the upper midrange and high end would become glossy and have a metallic sheen to the sound. In other words, it would exacerbate the DAC AH’s weaknesses. The compete opposite happened. I chose the 96kHz up sample with the DAC AH.. What a difference! Gone was the aggressive constrictive high end. The upper midrange just sang. And here is the best part: the dynamics of the NOS method were kept in tact. This is definite winning combination that places the DAC AH / Up sampler combination above the pack. The DAC AH Modified without the Up sampler: Smooth High End 10 Mid Range Smoothness 12 Deep Tight Low End 17 Annoying Mid Bass 17 Resolution 12 Dynamics 19 Articulation 17 Total 104 The DAC AH Modified with Up sampler: Smooth High End 18 Mid Range Smoothness 18 Deep Tight Low End 18 Annoying Mid Bass 17 Resolution 14 Dynamics 19 Articulation 14 Total 118

FYI, higher scores are better. From his other comments in the blog, it seems that over-15 is very good, 10-15 is so-so, and under 10 is crap. In an email response to me, he sees no reason why any upsampling wouldn't do the same thing, he doesn't claim there's anything magical about the box he used. So, I am hopeful about the effect of using the MAD decoder add-on with MediaMonkey, set to 96kHz...

 

[evil-zen]

Maybe this will explain it a bit more. I recorded the Ack dAck 1 using tda1545 via RMAA and a RME digi96/8 pad soundcard.

normal 44.1khz

upsampled 88.2khz using foobar2000 SSRC

 

[russdog]

Well, as the saying goes, a picture is worth 1000 words. Thanks for the pictures. Do you hear differences that correspond to the pictures?

I'm neither a EE or an electronics hobbyist, so I have no clue as to why the upsampling would cause the difference shown in those graphs. Can someone explain it to me?

 

[warrior05]

i just tried this with foobar and confirmed via my juli@ display panel that i was comparing two different sample rates. i don't claim to have the most discerning ears but i did not hear a difference. my set-up is foobar playing flac > juli@ > dac ah-m > headfive > K701 w/xfeed on. i listened to a passage from beethoven's 6th symphony and a song sung by etta james.

i'll probably play with this a bit more but for now i'm keeping my sample rate at 44.1.

 

[ebrius35]

I've only got a stock DAC-AH, but in my chain of Audigy S/PDIF-> DAC-AH -> GLite -> HD650s, my initial reaction for the first 3 weeks was that the deeper bass notes (timpani, taiko, etc.) weren't quite "there" enough for me, and used EQ to nudge up the lowest 4 bands a tiny bit and I was happy.

Then on a whim I set my audigy to output at 96khz digital out, and foobar resampled to 96khz to match to see what would happen.Yyeah the audigy handles internal sound at 48khz but 96 was available in the settings box so I used it.

Amazingly my EQ settings started distorting the low notes. When I took that out, the bass I was missing was actually noticably (to my newbie ears) more "there" than before, around the same levels I had EQed things to. Other things might've changed too, but if they're there I'm not listening hard enough to notice them.

 

[Superpredator]

Setup:

Foobar2000 0.83 (32bit fixed-point, kernal streaming)
192kbit CBR MP3 EAC; 320kbit CBR MP3 EAC ~200kbit VBR MP3 EAC; ~50% FLAC
M-Audio Revolution 7.1
Belden '1694A' Coax
DAC-AH.M
Signal Cable 'Silver Resolution Reference' RCA
Darkvoice 336i
Headphiled K340
Beyer DT880

Warning: ~10 hours on DAC-AH.M and interconnects. I neither have golden nor silver ears. Even copper might be a stretch. Initial impressions only.

Alrighty. I've done some quick switching between 96000Hz and 44100Hz by way of Foobar's PPHS resampler in ultra mode. No other DSPs running. Specific differences between the two did not jump out at first, but after replaying selected portions of several (15 or so) different tracks of various genres, female/male vocalists, etc, primarily with the K340 but with the DT880 thrown in for good measure, there are pretty distinct, but not dramatic, differences.

IMO, the 96000Hz clearly has more openness and better separation. Instruments/vocals sound more 3D, with much better depth, and a negligible boost in outer limit soundstage width. Vocals have more weight and texture. Bass has noticably deeper extension, and I would definitely agree that annoying mid-bass is reduced as a result, almost like it just moves out of the way, taking with it low-end congestion. I'm pretty terrible at analyzing highs, but I'll give it a shot. With females and some Radiohead, the vocal highs had a slight rough quality to them (almost grainy) at 44100Hz. This was smoothed out at 96000.

These are just first impressions, but I will most likely be resampling to 96000 with my current setup. 44100 vs. 96000 is not the difference between screeching cats mating and a chorus of angels orgasming on clouds, but it is noticable and worth using IMO (at least with the AH).

 

[regal]

I tried it . Resampled to 96k with foobar ultra mode. You lose bitperfect ouput this way. It does sound better especially the upper midrange and treble. The grain is gone.

Kinda makes me wonder if the whole non-OS DAC theory is hogwash.

 

[russdog]

Quote:

Originally Posted by regal I tried it . Resampled to 96k with foobar ultra mode. You lose bitperfect ouput this way. It does sound better especially the upper midrange and treble. The grain is gone. Kinda makes me wonder if the whole non-OS DAC theory is hogwash.

Why? I'm not trying to give you a hard time, I just don't understand what you mean. Based on what you just said, I'd think it would make you wonder if the concern for bit-perfect output is hogwash.

 

[regal]

By feeding a nonOS DAC an oversampled 96k signal it is no longer functioning as a non-OS DAC. Its still filterless which may be the benefit.

 

[lan]

Quote:

Originally Posted by regal By feeding a nonOS DAC an oversampled 96k signal it is no longer functioning as a non-OS DAC. Its still filterless which may be the benefit.

The signal you send the DAC is not "oversampled". It's just a signal plain and simple. The DAC non-oversampling DAC is still non-oversampling with whatever incoming signal. Oversampling is a function of hardware ONLY.

 

[regal]

no a 44.1k resampled to 96k is oversampling, its just done by the computer instead of the dac digital filter

 

[lan]

I know what you are saying though since it's "similar". But if it's done before the DAC, it should be called upsampling or resampling. Sorry just a matter of semantics