Unclear what is meant by "mediocre" ... other than a personal, subjective opinion of the sound charc. of noise-shaping DACs. Like OS vs NOS, perhaps. Apples and oranges.
Excluded from your description here are the other Philips datasheet-suggested "combo" devices such as various NPC df's and TDA1307 that appeared in the mid 1990s.
Recall that dual diff. 1547s, in combo with the SAA7350 (used as up-sampler) and a DF, like 1307, upstream of that , was Philips' high-end strategy for the 1990s including the LHH series, and that $7500 Marantz SACD from late 2000.
Not as familiar with chip ladders and older R2Rs such as AD1862 and PCM63 --- but do have several in my parts bin awaiting experiments.
The Philips CC is so-so. Ditto opinions about with discrete R2R (Soekris, etc) -- dull and unexciting.
The real surprise is the mid-1990s TDA1305 (see my Naim thread). Somehow, this hybrid Bitstream can be made quite effective w/o much modding or tweaking.
As far as modern dacs ... To my ears, AKM and ESS are very good. The Wolfson was also v. good. The TI and Cirrus, less so.
Personal bottom line: If the sound output of the final DAC project is boring -- regardless of clarity or "warmth" or smoothness or depth -- the magic is gone and dream has derezzed.
With modern DAC chips that are up to the latest developments and all currently available DSD modes/high res PCM one has both excellent measurements and good sound, good availability (AKM seems to be in full action again), lowest costs, lowest part count. Often a newer chip is pin compatible with the predecessor too.
Investing/loosing time on decades old chips that cost too much, were and certainly are not best of class and are available (refurbished/desoldered) on TaoBao or from tacky sources etc. is a choice
More often than one such recommendations are done by people that have rails full of these otherwise it makes not much sense to do so for a DIYer. The golden 3 worth the hassle were TDA1541/1543/1545 provided one still spins red book disks on cd players.
Investing/loosing time on decades old chips that cost too much, were and certainly are not best of class and are available (refurbished/desoldered) on TaoBao or from tacky sources etc. is a choice
More often than one such recommendations are done by people that have rails full of these otherwise it makes not much sense to do so for a DIYer. The golden 3 worth the hassle were TDA1541/1543/1545 provided one still spins red book disks on cd players.
Last edited:
Hmmm ... are you suggesting those ICs were designed from the bottom up -- both objectively; and subjectively [say, in-house listening tests in the Philips R&D labs] -- for Red Book optical disk spinners? And not, say, fed computer data (music via I2S), from a PC hard-drive. Or whatever Philips/Sony were using in late 70's and 80s' during product development?
It's easy to look at just the metrics and finite number or tracked specs (say, freq. response, IM, S/N, dynamic range, THD, etc) and conclude that's what's fundamental.
But the whole CD player, spinning a disc inside a box scenario, and out comes music, is magical, dreamy stuff. A cesspool of "fundamental" specs ... probably the vast majority of which are still "hidden". And that might explain the subjective vs objective debates. Or why DIYers and certain audiophiles still select ICs like the 1541 or even 1540, over the latest AKM or ESS... even though those dinosaurs don't spec out too well and are fussy to implement.
In those days such a thing would have been by far the most revolutionary thing in the room.
Imagine our teacher telling the class that if we would be grownups and have a career a song would be stored on a ultra high tech chip (as in 1 song on 1 chip)
Some went “ooh”, others “aah”’and some laughed in disbelief. The ease of plugging in a chip instead of playing a single. How wonderful that would be (CD was just introduced).I am now looking at a 256 gb USB stick. No motor, no laser, no moving parts at all. Enough albums for years to playback.
BTW I am quite sure the guys at the Philips/Sony labs played the first CDs for testing. There was no computer involved AFAIK. Not much hidden stuff as Philips sold CD players but also the technology to competitors.
Last edited:
I've got one sitting in my cupboard.
It's a Studer A727, basically a Revox but with balanced outs on XLRs.
Okay, thank you. What published controlled listening tests are there of TDA1541 dacs? And what published controlled listening tests are there of ESS dacs?
Well, there is no "romance" there --- no effort.
"Nothing ventured, nothing gained".
The physical action of getting up off the couch to change a record or CD does influence perception (quality?) -- tho' I'll be damned if I can point to any psycho-acoustic papers that have published such results. We're getting into the weeds here: The "Hard Problem" of consciousness, etc.
Also, it might just be case that if the mind-brain system does not encounter much effort in sound retrieval, the perception of fidelity is compromised (enjoying what you are listening to, or being engaged in it). That is, tubes measure worse than solid state, but can sound more musical . Or the latest / greatest Chinese DAC that Amir measures perfectly and gets a golfing panther rating, actually sounds boring after a while.
OK. You take a sentence out of context but OK. Some take a lot of effort in finding that romance in either old or modern stuff. Did that ever occur to you? Nothing lazy there and one is out of the boring circle of "older is better" and also only the stuff one thinks to know. And even if one could/would find the "perfect sound", that is also boring after a while It seems like coupling emotions to cold devices.
Coincidentally I recently compared an old "better" CD player with a few current streamers and DACs. The perception was like a getting a cold shower. The memory is a tricky thing. That mysterious Sony PLM has been long superseeded in terms of measured performance and sonical qualities. I think I have kept a few of those ICs and you may have them if they are still here. I modded literally hundreds of CD players and especially those with 45.xx clocks by Sony were very susceptible to clock jitter.
It seems like coupling emotions to cold devices.Coincidentally I recently compared an old "better" CD player with a few current streamers and DACs. The perception was like a getting a cold shower. The memory is a tricky thing. That mysterious Sony PLM has been long superseeded in terms of measured performance and sonical qualities. I think I have kept a few of those ICs and you may have them if they are still here. I modded literally hundreds of CD players and especially those with 45.xx clocks by Sony were very susceptible to clock jitter.
Last edited:
If you look for a DIY DAC that takes lots of money and effort, I recommend my valve DAC. It's expensive, requires a four-layer board with a very specific stack-up, requires dealing with 0.5 mm pitch SMDs (TQFP-48's and two 80-pin connectors) as well as potentially lethal voltages, you need to wind your own filter inductors and its noise floor is far higher than that of any modern audio DAC.
https://linearaudio.nl/valve-dac-submicron-silicon-meets-submillimetre-vacuum-0
https://linearaudio.net/design-data-package-marcel-van-de-gevels-valve-dac-vol-13
Last edited:
When you put it like that it does sound very romantic
Last edited by a moderator: