http://src.infinitewave.ca/
Somewhat amazing to see just how bad (in theory) some of the popular converters are. Naturally as with all gear testing it against real music is the only way to know.
The winners:
* r8brain
* iZotope
* Weiss Sarcon (except for one oddity)
Some odd results:
* Pro Tools HD does fairly well, except that is has a significant noise floor
* Pyramix has an uneven frequence response down to DC (down about .1dB at 10k, but it oscilates by ~.03dB in a sine wave of ~800hz)
* Several start dropping off at only 16k
* Several have really odd phasing (some just reverse, but others...)
- It is currently Mon Jul 27, 2026 2:40 pm • All times are UTC + 10 hours [ DST ]
Sample Rate Converter comparison site
Moderators: rick, Mark Bassett
8 posts
• Page 1 of 1
Sample Rate Converter comparison site
by JulienG » Wed Jan 31, 2007 2:16 am
- JulienG
- Regular Contributor
- Posts: 263
- Joined: Sat May 07, 2005 3:02 pm
- Location: Sydney, Australia
by wez » Thu Feb 01, 2007 3:54 pm
going by those results you'd buy wavelab over the weiss sfc and spend the difference on quite a decent small car... yet the weiss is commonly regarded (by people who should know) as the primo SRC. so as interesting as they are, i don't know how much use those test results are in the real world.
what i do know is that src is the devil's work, avoid it whenever you can!
what i do know is that src is the devil's work, avoid it whenever you can!
-
wez - Valued Contributor
- Posts: 1259
- Joined: Fri May 06, 2005 10:38 pm
- Location: Slightly to the left.
by smash » Mon Feb 05, 2007 5:36 pm
I worked out how to interpret the 1kHz tone graph...
Someone enlighten me, but when the difference between programs is, for example, a 110db Noise floor (Logic Pro 7.2.3) and a 165db Noise floor (Protools HD) doesn't this become purely a theoretically thing?
I am only aware of a very small number of D/A converters that can resolve a greater than 110 dB dynamic range...
Prism ADA-8XR
Prism Dream DA/2
Lynx L22
Benchmark DAC
Apogee Rosetta 200 - Just (only 114 db Dynamic Range, but -103db Noisefloor)
Apooge Rosetta 800 - - Just (only 114 db Dynamic Range, but -103db Noisefloor)
Of cpurse this is only 1 aspect of the test, but to me the difference between an induced noise of -110 dB versus an induced noise of - 150db seems largely theoretical - especially since the medium that we eventually playing back on - CD - comes nowhere near this point.
Someone enlighten me, but when the difference between programs is, for example, a 110db Noise floor (Logic Pro 7.2.3) and a 165db Noise floor (Protools HD) doesn't this become purely a theoretically thing?
I am only aware of a very small number of D/A converters that can resolve a greater than 110 dB dynamic range...
Prism ADA-8XR
Prism Dream DA/2
Lynx L22
Benchmark DAC
Apogee Rosetta 200 - Just (only 114 db Dynamic Range, but -103db Noisefloor)
Apooge Rosetta 800 - - Just (only 114 db Dynamic Range, but -103db Noisefloor)
Of cpurse this is only 1 aspect of the test, but to me the difference between an induced noise of -110 dB versus an induced noise of - 150db seems largely theoretical - especially since the medium that we eventually playing back on - CD - comes nowhere near this point.
- smash
- Regular Contributor
- Posts: 443
- Joined: Tue Jun 07, 2005 12:49 am
- Location: Northern Beaches, Sydney
by smash » Mon Feb 05, 2007 5:53 pm
Just to quantify
a 110 DB noise floor is equivalent to an hearing leaves rustling in the wind...
... when you are standing 1m from a functioning Air Raid siren.
a 110 DB noise floor is equivalent to an hearing leaves rustling in the wind...
... when you are standing 1m from a functioning Air Raid siren.
- smash
- Regular Contributor
- Posts: 443
- Joined: Tue Jun 07, 2005 12:49 am
- Location: Northern Beaches, Sydney
by JulienG » Mon Feb 05, 2007 7:43 pm
smash wrote:... when you are standing 1m from a functioning Air Raid siren.
Ouch! Actually did that this weekend, was in a room in a school in NZ where the alarm went off a few times...
- JulienG
- Regular Contributor
- Posts: 263
- Joined: Sat May 07, 2005 3:02 pm
- Location: Sydney, Australia
by heathen » Tue Feb 06, 2007 8:26 am
[quote="smash"]Just to quantify
a 110 DB noise floor is equivalent to an hearing leaves rustling in the wind...
... when you are standing 1m from a functioning Air Raid siren.[/quote]
Also especially once on CD after dithering (depending on which type), I cant remember what the noise floor would be then, but probably way higher than -110 noisefloor. I'm not sure what all this says (the graphsH) in a real world application but I reckon quantization errors would be way more of a worry than an extremely low noise floor, the quantization errors would sound heaps worse than a little noise at such a low noise floor.
If I have to do an src I'll usually do it with the Rosetta 200 as it's src sounds ok but there is a difference, though very very small.
All interesting though.
a 110 DB noise floor is equivalent to an hearing leaves rustling in the wind...
... when you are standing 1m from a functioning Air Raid siren.[/quote]
Also especially once on CD after dithering (depending on which type), I cant remember what the noise floor would be then, but probably way higher than -110 noisefloor. I'm not sure what all this says (the graphsH) in a real world application but I reckon quantization errors would be way more of a worry than an extremely low noise floor, the quantization errors would sound heaps worse than a little noise at such a low noise floor.
If I have to do an src I'll usually do it with the Rosetta 200 as it's src sounds ok but there is a difference, though very very small.
All interesting though.
-
heathen - Valued Contributor
- Posts: 1745
- Joined: Sun May 08, 2005 11:15 pm
- Location: Sydney
8 posts
• Page 1 of 1
Return to The Turtlerock Forum
Who is online
Users browsing this forum: No registered users and 19 guests