Power Technology debuts iDFX iTunes add-on | iLounge News


Power Technology debuts iDFX iTunes add-on


Power Technology has introduced its new iDFX Audio Enhancer software add-on for iTunes. According to the company, iDFX re-encodes users’ current MP3 and AAC files via a “patent-pending method that repairs the damage and lost harmonics that occurred during the original encoding process.” In addition, the company also claims that the software sets all songs to a standard volume level, similar to Apple’s SoundCheck feature, and increases the audio clarity and undistorted maximum output level when used with iPod docks and in-car connections. iDFX requires Windows XP, Vista or higher and iTunes 8 or higher, and is available now for $40. A free trial version is also available.

Related Stories



And yet a majority of people can’t even ABX between lossless and 128(VBR) AAC or LAME mp3, and less than 1% can manage to do so between 192(VBR) AAC or mp3…

It’s amazing how shameless and downright evil some corporations are by exploiting the general ignorance in the populace.

$40 to automate applying an EQ to your files while incurring the standard transcoding artifacts from a lossy to lossy encode, genius.

Posted by Code Monkey on October 8, 2009 at 10:10 PM (CDT)


This seems like a 40 dollar waste of time. Trans coding a file is not the best thing to to if you are so concerned about how your files sound you should encode them the way you want them the first time. Don’t spend money on this it more than likely will introduce more artifacts into the file.

Posted by David on October 9, 2009 at 10:02 AM (CDT)


Crashes. So much for that….

Posted by beetsnotbeats on October 9, 2009 at 10:47 AM (CDT)


“you should encode them the way you want them the first time”

Right, but that’s the rub: this isn’t about encoding them right, it’s about applying EQ to the files. There is apparently an option to apply these EQ tricks to the files at the time you rip, but there’s no proof this is done losslessly either.

I understand that some sort of intelligent sound analysis and dynamic application of EQ will make files subjectively sound better to a demographic that’s used to unnatural distortion of the bass and treble, but if that’s the goal, there are better ways of achieving it than this $40 proprietary iTunes add-on.

Just as a quick example, dBpoweramp will give you a truly secure ripper, absolute fine tuned control of your lossy encoder settings, volume normalization, and embedded EQ applicaton if you want for $36 and it even comes without the dung smelling outright lies about restoring harmonics and repairing damage from encoders.

Posted by Code Monkey on October 9, 2009 at 10:48 AM (CDT)


Hi, this is Paul Titchener, I’m with Power Technology and I’m the developer of the audio processing in iDFX.

I’ve been trying to post a response but it keeps failing, maybe its too long so I’ll try posting it in parts.

Thanks for the comments on iDFX. Regarding any crashes or similar problems, if you could please report them we’ll chase them down right away. This is a new product and although we tested it heavily, with any new product you always hit some scenerios that are unexpected. We just today located and are fixing one such problem and that likely will eliminate the crash that was mentioned above.
(to be continued in next post)

Posted by Paul Titchener on October 9, 2009 at 7:49 PM (CDT)


(post above continued)

Regarding the processing iDFX performs, I can’t tell you specifically what we are doing as its subject to several patent applications, but I can state that its definitely not a simple (or complex for that matter) EQ or single or multi-channel compression operation. To prove this to yourselves, try iDFX on some acoustic piano or similar material with a lot of dynamic range. You’ll find iDFX really improves the audible detail of the instrument without limiting the dynamic range in any way. This isn’t the case with the multi-band compression and EQ based systems.
(to be continued)

Posted by Paul Titchener on October 9, 2009 at 7:51 PM (CDT)


Were not telling in “lies” about what iDFX does, iDFX really does make most mp3/aac files sound better. Please try it out before passing judgement, the trial version lets you fully enhance a limited number of songs. For a quick evaluation, try the youtube demo available at the link below, and note how much more detail you can hear in the vocal and instruments when iDFX is switched on.
(to be continued in next post)

Posted by Paul Titchener on October 9, 2009 at 7:58 PM (CDT)


The link below isn’t some canned demo, its the exact sound difference you would hear when using iDFX on this AAC encoded track.


In summary, thanks for the input, but please give it try before passing judgement (it appears none of the commenters here actually tried it) and I apologize again for the crash, we’ll definitely get that fixed.

Posted by Paul Titchener on October 9, 2009 at 8:02 PM (CDT)


@Paul, thanks for taking the time to post, but it’s not going to change my opinion. It’s not a question of trying it, it’s that the notion there’s anything audibly destructive about moderate lossy compression is complete and utter hogwash on the face of it. Whatever you’re doing to the audio doesn’t matter, you start from a BS premise: mp3s sound worse than the original CD. Well, ABX testing out the yin-yang proves that’s complete nonsense.

So, whatever you’re doing isn’t about restoring the original sound, it’s about spiking and doping the original audio to make it sound better to the same idiots who keep Bose in business. That and you can’t seriously expect anybody who’s knows squat about audio to ignore the fact that you’re doing a lossy to lossy transcode on people’s libraries - pretty much one of the cardinal sins of maintaining the integrity of your lossy audio.

I like my audio to actually sound like the CD, which it already does without paying you $40.

Posted by Code Monkey on October 9, 2009 at 10:52 PM (CDT)


Still crashes. The “initializing” box appears, iTunes launches, then Windows XP SP3 says iDFX has crashed. Is the currently available version the new one? My second download was 3 KB smaller than the first. How about some version numbers?

Posted by beetsnotbeats on October 9, 2009 at 11:56 PM (CDT)


As for the demo video linked above, whatever else is going on, you clearly didn’t match output levels, which makes the video one more great big deception on the face of it. Testing has shown repeatedly that when two otherwise identical audio samples are played back, people overwhelmingly go with louder = better, and the DFX portions are clearly louder than the non-DFX samples. People are going to hear it as more clear even if there is literally nothing more going on than a volume boost.

Take your snake oil somewhere else.

Posted by Code Monkey on October 10, 2009 at 7:19 AM (CDT)


Hi, this is Paul Titchener from Power Technology again.

Beetsnotbeats, we haven’t updated our .exe yet, were still testing a release that will go up later today. However, I’d like to verify that we fix the problem on your particular PC, so if will test this new release for us I’ll give you a complementary serial number to the iDFX Plus version. Please send me an email at pt AT fxsound DOT com if you would be willing to test the new release on your PC.

Code Monkey, to keep these posts a reasonable size, I’ll respond in a separate post to yours.

Posted by Paul Titchener on October 10, 2009 at 1:45 PM (CDT)


Code Monkey, thanks for the comments. What you are hearing in the demo, both the youtube one and if you try the product, are comparisons to the unaltered MP3 file and then to the file with our enhancement. For a fair comparison we don’t modify the volume level of the original audio in any way, and then we do a direct comparison to our enhanced version.

Doing an output level matched comparison in this case doesn’t make sense, as we really are adding signal components to the audio that were either flat not present or were highly diminished in the original version, so it will naturally sound louder to you because of these added components.

But if we tried to do a output level matching, to make the overall output levels match, now in the enhanced version we would be making some frequency ranges quieter than they were in the original version, so for example perhaps a bass instrument wouldn’t be at the right level now, so doing that just doesn’t make sense, and its not the way a user would want to listen to the song.

Another thing that iDFX accomplishes with its processing is allowing a much higher undistorted output level than you can acheive with the original mp3s. To hear this, try playing back the original and then one of our enhanced files on an iPod speaker dock or laptop PC or any system with limited audio playback capabilities. You’ll find the iDFX processed version give you much higher undistorted playback levels.

We have been providing MP3 enhancement for many years in the non-iPod world, we have millions of satisfied users of our DFX plug-in for the Windows Media Player and Winamp, to a search on “DFX MP3” to get a feeling for our DFX plug-in user base.

I know audio is very personal thing, and if our product isn’t for you, at least for full time use, I can certainly understand that. But it still may be a useful tool for you in some cases. Here’s an example where it really did me some good. I’ve been on a “Taj Mahal” (a sixties era blues band) kick for the last week, listening to his first three albums recorded in ‘67 - ‘68. But they unfortunately weren’t recorded very well, on a lot of the tracks the vocals or instruments sound pretty flat. This is case were iDFX really helps even directly on the CD versions of the songs, it really helps compensate for the limitations of the original recording and makes listening to these CD’s much more enjoyable. Try it also on any live recording .mp3’s you may have, it can really make a big difference on those.

Code Monkey, I also need to apologize on one issue, I saw your comment (and a couple more that were pretty positive) in the comments range on our youtube demo from the link in the post above, but when I was trying to update our information page on the video, somehow both that information page and all the comments that had been placed got blasted, and I haven’t figured out if there is a way to get those comments back, so bear with me on that until I figure it out.

Posted by Paul Titchener on October 10, 2009 at 2:18 PM (CDT)


Level matching is the *only* valid comparison. Anything else is pure bollocks.

Posted by Code Monkey on October 10, 2009 at 2:45 PM (CDT)


Just some more info: I saved the youtube video as an .mp4 and exported the audio to wave so no further compression or anything else will affect it. Imported it into Audacity and had some fun…

There is a 6 dB difference in the volume between the DFX off portions and the DFX on portions. That’s a significant difference and responsible for the “improvement” in the video.

Not at all surprisingly, once I amplified the “off” portions by 6 dB there’s no noticeable difference between the two alternating sections.

I’ll give you the benefit of the doubt that there might be *some* audible difference outside of the volume boost since audio in a youtube video is highly compressed beyond what most people would have in their libraries, but anybody with freeware audio software can see for themselves the sort of deception and trickery you’re engaged in.

Posted by Code Monkey on October 10, 2009 at 3:19 PM (CDT)


(Paul Titchener from Power Techology again)

Code Monkey, clearly I would dispute your opinion that there is no audible difference in the two versions, but like I said, audio is pretty personal and users reactions will vary, so I’d just recommend for people to try it for themselves.

I’ll stand on our customers opinions, check out http://www.fxsound.com/dfx/pages/overview/reactions.php
to see some of these. We have hundreds of thousands of customers for DFX who really like how it improves the sound of their MP3’s, so I’d again just ask potential users to try out the trial version and decide for themselves, your ears won’t lie to you.

One more time on the volume difference, a big advantage of iDFX is that is allows you to acheive much higher undistorted output level on portable audio systems like iPod docks. Basically, with a small system you can get much more usable audio output from the system before distortion occurs. This is a big advantage of iDFX and it wouldn’t make much sense to try to artificially remove that capability in the demo when its part of what people are getting when they buy the product.

Posted by Paul Titchener on October 10, 2009 at 4:34 PM (CDT)


Actually, ears do lie to you. That’s why there are things like level matching and ABX comparisons and RMAA output readings, so we can see what’s actually going on, and I’ve *seen* what’s actually going on with your product.

Posted by Code Monkey on October 10, 2009 at 4:55 PM (CDT)


Wanting to give you some fair shakes, I downloaded the trial. It installed fine and runs on my computer without any obvious issue.

What I wanted to do was compare two files, one DFX processed, one native, that were level matched in an ABX comparison to see if there actually is an audible difference once the deceptive volume boost aspect is eliminated. Encoder, bitrates and number of transcoding steps would be identical so the only difference between the files would be the DFX processing.

For my first attempt, I started with an Apple lossless file figuring it would involve the least “contamination” for later testing. Unfortunately, the resulting DFX generated file had chunk errors that both caused mp3gain to crash and prevented me from converting it to a wave file with dBpoweramp for further analysis. So, hoping that the corrupted header part of the file was due to me starting from Apple lossless, I opted to start with a 320 kbps AAC file, nope, same bunk header that prevents me from using the file for anything but playback. Simply put, that’s not cool.

Fortunately, Foobar2000 has a very forgiving wave conversion, so I went back to my lossless sourced files and converted them to wave there. Interestingly (and disturbingly), Foobar2000 could not read the bunk header at all, it treated the file as though it were completely tagless.

Anyhoo, this time the difference between the processed and non processed file was 5.1 dB, so I applied that gain to the unprocessed file’s wave so that both initial waves will have the same amplification consequences. Converted these to FLAC, and ABX compared them in Foobar2000 with ReplayGain level’s matched precisely.

So, here’s the good news. I was able to ABX between the original unprocessed audio and the DFX processed audio 100% of the time on 10 trials, a 99.9% probability that there is a distinctly audible difference between the two.

The bad news is that when the levels are matched, even in blind testing, I subjectively prefer the unprocessed audio (it was easy to match up to the known samples afterward based on their clear cut audible differences).

Now, on the off chance that me using Foobar2K to force a wave conversion had some adverse effect, I let iTunes handle the wave conversion since I figured I could assume that whatever bunk headers you’re using are, at the least, iTunes compatible. Same ABX results, same preference for the original sound.

And, because I am a fair guy, I even repeated this trial with another track, only skipping the trouble shooting steps this time ;-). Audio differences were more subtle, but I was able to ABX the two tracks 9/10 times, for a 98.9% probability there were audible differences between them, which is as good as certain. My subjective preference was, again, for the non-DFX processed track.

Spectral analysis of the files shows what I already guessed, lots of high freqency noise is added, this is your so-called “restoration”. Unfortunately, for a reasonably trained ear, it just gives the audio the so-called “tinny” sound character, which is why I can ABX it from the non-processed audio and also why I find myself preferring the non-processed audio.

So here’s my review after giving the product the benefit of the doubt:

One, the product is unbelievably slow. Based on the performance of these test tracks, I estimate the doing a total library conversion on a large library like mine (more than 35K items) would take about six weeks running my computer 24/7. Times might vary based on your computer, but mine, while not blazing fast, is a 3.2GHz Dual Core Pentium, and each conversion was around 90 seconds.

Two, the inability to produce AAC files with standard, universally compatible headers is simply unacceptable.

Three, the audio is decidedly different than what you start with. Unfortunately, to my ears, it is not an improvement at all and is actually a degrading of the original sound (a.k.a tinny).

Four, spectral analysis confirms the main thing you’re doing outside of the volume boost is spiking the audio with high frequency noise. This might offer some subjective improvement to low bitrate files with a fairly low cut-off ceiling, but for higher bitrate files it just leaves them sounding artificial and processed.

Oh, and thanks for overriding my iTunes import settings by the way, at least I noticed before it screwed up later work.

Once more, people don’t have to take my word for it, they can confirm any of this themselves with your trial.

The product is snake oil in the finest tradition of snake oil.

Posted by Code Monkey on October 11, 2009 at 8:28 AM (CDT)


(Paul Titchener again)
Code Monkey, thanks for taking such a hard look at our product. Audio is a very personal thing, and clearly you don’t like the processing we’re doing, and that’s fine, as I’ve mentioned we expect when people try our product some will like it, some won’t, but the people that like it is definitely a clear majority. In addition to the hundreds of thousands of plug-ins we’ve sold, one of the major in-store suppliers of audio systems puts DFX on every system they install for two big reasons. They really like the sonic improvement and as I also stated above, it also lets them acheive a much higher usable sound level without distortion. So clearly one of things you don’t like about it (it lets you acheive a much higher undistorted output level) is a feature a lot of people really do like. Its not easy to do that, we’re not just “turning up the volume” which as you know would cause distortion, we’re processing the audio so that without losing any dynamic range you can get a lot more usable output level out of your audio system.

Your analysis clearly showed that we certainly are modifying the audio although we of course would differ with the categorization of “noise”. But I would view your accessment as “snake oil” as being invalid and unfair, that implies we are somehow not doing anything to the audio, which is clearly not the case as you yourself showed. You don’t like our processing and that’s fine, but many customers and oem users have very carefully evaluated DFX with critical listening and they really like the improved sound level and undistorted increased output level they get with DFX.

Posted by titchener on October 11, 2009 at 7:54 PM (CDT)


Paul, this isn’t subjective, this is looking at precisely what the program is doing objectively: it’s adding a non-reversible volume level gain, a bad idea on the face of it, and something people can do for free (and reversibly) with mp3gain or iVolume, and it’s doping the audio with high frequency noise.

That’s it. I looked for something else - a dynamic application of EQ, something, anything, but that’s all you seem to be doing.

As for your claims about allowing higher non-distorted listening: You’ve amplified the audio while minimizing clipping, whoopee, not exactly tricky. It’s a basic checkbox in freeware audio software out the yin-yang (and doesn’t accomplish diddly with many modern CDs that begin brickwalled). At best, you are applying the same sort of gain compression that plagues recent CD mastering and actually destroys dynamics in music, though I found no evidence of that in the samples I examined. Given that you don’t appear to be engaged in gain compression, it’s a basic gain amplification on the audio, and that’s hardly what I call a feature given how many other ways there are to accomplish this.

I stand by my claim that it’s total snake oil and I ask anyone who disbelieves my claim to get themselves a copy of Foobar2000, Audacity, and iTunes (and, if they have access to it, Audition for spectral analysis) and try out objective comparisons themselves.

You keep trying to make it sound like I’m suffering from sort of subjective bias, but all I’m dealing in here is the facts about audio. Spin it how you want, people like me obviously challenge your business model, but objective facts speak for themselves. Number of consumers and testimonials mean bupkiss, finding a million stupid people in the world isn’t exactly hard or an accomplishment.

Don’t worry, this will be off the front page by lunch tomorrow most likely and since there’s plenty of folks who wouldn’t know how to do a matched level comparison if their life depended on it, I’m sure you’ll find plenty of fools to part with their money. I just hope the facts will speak for themselves and at leas some people will think before falling for such trickery.

Posted by Code Monkey on October 11, 2009 at 10:41 PM (CDT)

Subscribe to iLounge Weekly

Sign up for the iLounge Weekly Newsletter

iLounge is an independent resource for all things iPod, iPhone, iPad, and beyond.
iPod, iPhone, iPad, iTunes, Apple TV, Mac, and the Apple logo are trademarks of Apple Inc.
iLounge is © 2001 - 2018 iLounge, Inc. All Rights Reserved. Terms of Use | Privacy Policy